Immunopathogenesis of Acute Kidney Injury

Sepsis is one of the most common causes of AKI (Swaminathan, Rosner, and Okusa 2015). In sepsis, there is a systemic inflammatory response, triggered by the innate immune system, to attack foreign bacterial pathogens (LPS) and toxins to limit their spread (Mårtensson and Bellomo 2016). Such systemic inflammatory response leads to the release of PAMPs and DAMPs into the circulation (Gomez et al. 2014). PAMPs are released from pathogens, and DAMPs are released from damaged and injured tissues at sites of infection, and both PAMPs and DAMPs prime, and signal the immune system to fight back the invading pathogens and toxins. Interactions of PAMPs and DAMPs with sensors of the innate immune system called pattern recognition receptors (PRRs; e.g., TLRs, nucleotide-binding oligomerization domain–like receptors, and retinoic-acid-inducible-gene-I-like receptors) amplify the systemic inflammatory response (Gomez et al. 2014; Mårtensson and Bellomo 2016). For example, there is TLR4-dependent LPS recognition mechanism in the RTECs in the S1 segment of the proximal tubules (Kalakeche et al. 2011). Kidney-specific DAMPs (e.g., uromodulin [Tamm–Horsfall protein]) are synthesized by distal epithelial tubules and selectively released into the tubular lumen and leak into the interstitium after renal tubular injury. This activates renal DCs via TLR4, and NACHT (NAIP, CIITA, HET-E, TP-1), Leucine-rich repeat (LRR), and possessing pyrin domain (PYD) domains-containing protein 3 inflammasome potentially leading to sterile renal inflammation (Darisipudi et al. 2012; Kurts et al. 2013). The kidney parenchyma is especially vulnerable to the pro-inflammatory response to infection and subsequent renal tissue damage because it receives approximately 20% to 25% of the cardiac output (Khan, Hard, and Radi 2012). This is because the toxin-rich blood, PAMPs, DAMPs, and pro-inflammatory mediators gain access to kidney parenchyma and its various compartments via glomerular filtration or peritubular capillaries microcirculation (Gomez et al. 2014; Mårtensson and Bellomo 2016). For example, LPS is filtered into the tubular fluid and can directly interact with RTECs via a TLR4-dependent mechanism (Gomez et al. 2014; Kalakeche et al. 2011). As a consequence, various compartments of the kidney, including RTEC, glomerular, and endothelial, are injured. This is because the filtered PAMPs and DAMPs, also called “danger signal,” bind to and activate RTECs, renal endothelial, and glomerular cells and can lead to endothelial cell activation and structural changes in the glomerular and peritubular capillaries (Mårtensson and Bellomo 2016). Collectively, exogenous and endogenous PAMPs and DAMPs are integral to AKI pathogenesis.

In the kidneys, various TLRs (TLR1, TLR2, TLR3, TLR4, and TLR6) are expressed by various renal compartments and contribute to immune system inflammatory responses and renal tissue damage (Tsuboi et al. 2002; Wolfs et al. 2002; Anders, Banas, and Schlondorff 2004; Zhang et al. 2008). For example, in kidneys obtained from healthy humans, TLR2 is expressed on the endothelium and Bowman’s capsule, while TLR4 is expressed on the endothelium and there is no tubular expression of TLR2 or TLR4 (Stribos et al. 2015). Similarly, in normal kidneys of rats and mice, no TLR2 and TLR4 expression was noted on tubular epithelial cells (Stribos et al. 2015). TLRs contribute to the renal innate and adaptive immune defense mechanisms by recognizing PAMPs as well as endogenous signals of tissue injury. TLRs induce the expression of inflammatory cytokines via myeloid differentiation primary response 88 (MyD88)-dependent and MyD88-independent signaling pathways (Rothfuchs et al. 2004; West, Koblansky, and Ghosh 2006). In sepsis and ischemia–reperfusion (I/R) injury, renal tubular cells become necrotic and the released DAMPs activate TLR2 and TLR4 on renal cells (Kurts et al. 2013). In a renal I/R mouse model, ischemia-induced renal inflammation significantly enhanced TLR2 (messenger RNA [mRNA]) and TLR4 (mRNA and protein) expression in the distal tubular epithelium, the thin limb of Henle’s loop, and collecting ducts (Wolfs et al. 2002). TLR4 recognizes Gram-negative bacteria via surface LPS moiety (Chow et al. 1999).

After glomerular filtration, renal tubular cells are sensitive to the freely filtered toxins. Accumulation of toxins and LPS can induce renal tubular cell necrosis and subsequently lead to TLR4-mediated renal inflammation (Zhang et al. 2008). Proximal tubular epithelial cells in the S1 region recognize the filtered LPS via their TLR4, and LPS is then internalized via endocytosis (Gomez et al. 2014). Binding of LPS to TLR4 is dependent on its interaction with LPS-binding protein (LBP) and the subsequent interaction with the membrane-bound CD14 (Wright et al. 1990). Within the kidney, this LPS LBP-CD14-TLR4 complex is located at the apical brush border of proximal tubules and initiates a downstream signaling cascade (El-Achkar et al. 2006; Morrell et al. 2014). There is an LPS-induced expression of CD14 gene in human kidney proximal tubular epithelial cells (Bussolati et al. 2002) and in mouse renal interstitial and tubular epithelial cells (Fearns et al. 1997). The downstream activation includes TLR4 activation via the MyD88 pathway leading ultimately to phosphorylation of inhibitor of kappa B (IκB) by IκB-kinase-β (El-Achkar et al. 2006; Morrell et al. 2014). Once phosphorylated, IκB is then released to facilitate nuclear migration and subsequent transcription of various inflammatory cytokines (e.g., interleukin [IL]-1, IL-6, IL-8), oxidative outburst, and inflammatory response (Morrell et al. 2014). It is interesting to note that the oxidative outburst takes place in the S2 segment of the proximal tubular epithelial cells (Gomez et al. 2014). Such inflammatory response generated by TLRs can activate renal DCs and T lymphocytes. Collectively, there is a strong interplay between renal TLRs, CD14, LPS, and LBP signaling pathways in AKI.

The primary site for mitochondrial dysfunction in AKI is the proximal convoluted tubules (PCTs), and mitochondrial biogenesis represents a crucial step in the recovery phase of renal function (Emma et al. 2016). In the kidney of rats subjected to I/R injury, necrosis occurs in PCTs accompanied by medullary congestion, which is most prominent in the outer stripe (Khan, Hard, and Radi 2012). In AKI, the morphology, function, and expression of mitochondria-related genes and proteins are altered when there is mitochondrial injury (Tábara et al. 2014). This is not surprising since proximal tubular cells are rich in mitochondria to perform oxidative phosphorylation (Khan, Hard, and Radi 2012). During cellular oxidative phosphorylation, reactive oxygen species (ROS) are generated as by-products (Tábara et al. 2014), and renal dysfunction can be related to mitochondrial abnormalities and generation of ROS. Mitochondria generate superoxide radical and ROS such as hydrogen peroxide and hydroxyl radicals, and these ROS can lead to irreversible mitochondrial DNA (mtDNA) damage, dysfunction, and death (Kowaltowski et al. 1999). Thus, mitochondria are targets of oxidative stress, and subsequent mitochondrial damage and associated mitochondrial membrane permeability changes, mitochondrial swelling, rupture of renal tubules, and release of mitochondrial ROS contents are cellular and molecular events that take place in AKI (Tábara et al. 2014). Drug-induced proximal tubular injury and renal failure leads to a clinical syndrome called Fanconi syndrome (FS; Heidari et al. 2018). In a rat model, valproic acid–induced FS was caused by mitochondrial dysfunction and oxidative stress (Heidari et al. 2018). Sulfasalazine-induced renal injury in rats was associated with decreased mitochondrial succinate dehydrogenase activity, mitochondrial depolarization, mitochondrial glutathione (GSH) depletion, increased in mitochondrial ROS, and mitochondrial swelling, suggesting mitochondrial dysfunction as a mechanism associated with renal injury (Niknahad et al. 2017).

In Balkan nephropathy caused by aristolochic acid in rats, ultrastructural examination of the degenerative renal proximal tubular epithelia cells showed extremely enlarged and dysmorphic mitochondria with loss and disorientation of cristae, and kidney mtDNA content, as assessed by polymerase chain reaction, was reduced (Jiang et al. 2013). In mice, cisplatin-induced nephrotoxicity is associated with structural and functional damage to the mitochondria (Zsengeller et al. 2012). In rats with myoglobinuric AKI, induced by glycerol injection, and mice with ischemic AKI, mitochondrial respiratory proteins NADH dehydrogenase (ubiquinone) 1 β subcomplex subunit 8, adenosine tripohsphate synthase β, cytochrome c oxidase subunit I (cyclooxygenase [COX] I), and COX IV were decreased, while mitochondrial biogenesis, such as peroxisome proliferator–activated receptor γ coactivator 1-α (PGC-1α) and PGC-1-related coactivator (PRC) were increased in both models and did not recover by 144 hr (Funk et al. 2012). In a rat model of rhabdomyolysis and myoglobinuric AKI, induced by intramuscular injection of glycerol, the expression of a key mitochondrial fission protein, dynamin-related protein (Drp1), and mitochondrial division inhibitor 1 (Mdivi-1) was examined (Cho et al. 2010; Tang et al. 2013). After AKI, Mdivi-1 suppressed the accumulation of Drp1, inhibited the insertion of proapoptotic Bax in mitochondria, and inhibited the release of cytochrome c, and this suppression of Drp1 in the mitochondria further supported mitochondrial function and reduced apoptosis of renal tubular cells (Tang et al. 2013).

As far as drug-induced renal toxicities, cyclosporine A nephrotoxicity has been linked to mitochondrial oxidative stress and ROS production. Cyclosporine increased the expression of Drp1 and decreased mitofuson 2 and optic atrophy protein 1, protein involved in the fusion process in an in vitro renal tubule cell culture system (de et al. 2013). Drp1 is upregulated in both animals and cell culture models of cisplatin and ischemia–reperfusion kidney injury (Cho et al. 2010; Tang et al. 2013). In another renal tubular cell culture system, cisplatin caused Drp1translocation to mitochondria early during tubular injury and mitochondrial fragmentation prior to cytochrome c release and apoptosis (Brooks et al. 2009). Mitochondrial fragmentation occurred in proximal tubular cells in mice during renal I/R injury and cisplatin-induced nephrotoxicity (Brooks et al. 2009). Tubular cell apoptosis and AKI were attenuated by Mdivi-1, a potential pharmacological inhibitor of Drp1 (Brooks et al. 2009). An in vivo multiphoton approach can be used to evaluate renal mitochondrial structure and function in anesthetized experimental animals such as rodents using multiphoton excitation of endogenous and exogenous fluorophores (Hall et al. 2013). Such approach can be used to demonstrate real-time changes in renal mitochondrial redox, structure, and function by imaging rodent kidneys in vivo in response to I/R injury or drug-induced toxicity (Hall et al. 2013). For example, using such approach, gentamicin-induced AKI causes cell death and mitochondria swelling and dysmorphia in the proximal tubules, while the morphology of mitochondria was intact in distal tubules and collecting ducts (Hall et al. 2013). Interestingly, it has been recently reported, using a rat AKI model, that renal dysfunction (i.e., elevated serum creatinine with a concurrent decrease in renal lactate clearance) in sepsis can occur without structural tubular injury, decreases in RBF and oxygen delivery, and hemodynamic instability (Arulkumaran et al. 2018). It is suggested that mitochondrial dysfunction mediated by circulating mediators induces local oxidative stress as a potential pathophysiologic mechanism for the abnormal renal function (Arulkumaran et al. 2018). In fact, antioxidants (e.g., iron chelators) are used as potential therapies in AKI (Benoit and Devarajan 2018). Collectively, mitochondrial pathologies, especially in the proximal tubular segments, play a key role in AKI.

To maintain a homeostatic steady state, cells are constantly adapting to physiological needs and responding to extracellular stimuli (Radi, Stewart, and O’Neil 2018). However, once the limits of adaptive cellular physiologic capabilities are exceeded or following exposure to lethal extracellular stimuli, cell injury progresses into cell death (Figure 2). The Nomenclature Committee on Cell Death suggested that cell death can be classified into three major groups: unregulated/accidental cell death (ACD), regulated cell death (RCD), and programmed cell death (PCD; Davila, Levin, and Radi 2018; Radi, Stewart, and O’Neil 2018). ACD (unregulated necrosis) is nongenetically and/or non-PCD which can be triggered by chemical, physical, or mechanical stress and occurs in an uncontrollable manner (e.g., oncosis; Davila, Levin, and Radi 2018; Radi et al. 2018). Necrosis is the presence of dead tissues or cells in a living organism regardless of the initiating process and can be observed in infectious and noninfectious diseases and toxicities (Davila, Levin, and Radi 2018; Radi et al. 2018). RCD describes cell death triggered by a genetically, or not, encoded machinery and regulated (induced and/or inhibited) by specific pharmacologic or genetic interventions (e.g., necroptosis, mitotic catastrophe; Davila, Levin, and Radi 2018; Radi, Stewart, and O’Neil 2018). Key components of the necroptotic pathway include receptor-interacting protein 1 (RIP1), RIP3, or mixed lineage kinase domain–like protein (MLKL; Davila, Levin, and Radi 2018). In cisplatin-induced AKI, necroptotic cell death contributes to various inflammatory cytokines induction, and cisplatin-induced necroptosis of renal tubular cells was shown to be RIP1-, RIP3-, and MLKL-dependent (Xu et al. 2015). In fact, necroptosis was abrogated in kidneys obtained from RIP3 knockout (KO) animals after ischemia and recipients receiving kidneys from RIPK3^−/− animals following transplantation had longer survival and improved renal function when compared with controls (Lau et al. 2013).

Cellular development and homeostasis is regulated via PCD (apoptosis and pyroptosis), which is a genetically regulated cellular process (Davila, Levin, and Radi 2018; Radi, Stewart, and O’Neil 2018). Receptors in the tumor necrosis factor (TNF) superfamily trigger PCD (extrinsic pathways and activation of death receptors such as Fas) or PCD can be triggered directly via the activation of mitochondrial effectors (intrinsic pathway and mitochondrial or endoplasmic reticulum [ER] stress; Davila, Levin, and Radi 2018). Tubular cells constitutively express fas ligand (FasL), and cyclosporine increased the expression of Fas (Justo et al. 2013). During renal I/R injury and in AKI, TNF-α and TNF-related weak inducer of apoptosis TNF-related weak inducer of apoptosis (TWEAK) are increased in renal proximal tubular cells (Dong et al. 2007; Sanz et al. 2016). Caspases (derived from cysteine-aspartic proteases or cysteine-dependent aspartate-directed proteases) are a family of enzymes that play essential roles in PCD. Mitochondria are key regulators of apoptotic, necrotic, and autophagic cell death and changes in the expression of multiple mitochondria-related antiapoptotic (e.g., B-cell lymphoma 2 extra large [Bcl-xL]) and proapoptotic (e.g., p53, Bcl-2-associated X [Bax]) regulators have been observed in AKI (Lorz et al. 2005; Sanz et al. 2008; Xu and Han 2016). Cyclosporine-induced apoptosis of renal tubular cells is associated with Bax aggregation, and translocation of Bax to the mitochondria and cyclosporine also leads to a caspase-dependent loss of mitochondrial membrane potential and mitochondrial injury (Justo et al. 2003). Aminoglycosides such as gentamicin accumulate in lysosomes and ultimately causes renal proximal tubular injury and nephrotoxicity by the release of gentamicin to the cytosol which triggers a Bax-mediated mitochondrial pathway of apoptosis and activation of caspase-3 (Servais et al. 2005). DNA injury activates p53-mediated apoptosis in cisplatin nephrotoxicity (Sanz et al. 2008). Paracetamol induces features of ER stress in tubular epithelium and Bcl-xL plays a protective role in paracetamol-induced tubular cell injury (Lorz et al. 2005). Apoptosis inhibitors (e.g., QPI-1002) are used as potential therapies for AKI (Benoit et al. 2018). Collectively, the molecular machinery in various forms of cellular death operates to protect the kidneys from AKI.

Oxygen tensions in the kidney are heterogeneous (Rosenberger et al. 2002). Renal tissue hypoxia by itself can lead to pathologic changes, independent of other known risk factors for kidney disease (Ow et al. 2018). HIF is a key gene and transcription factor that regulates adaptive responses against hypoxic conditions to control oxygen delivery, vascularization and neoangiogenesis, vascular tone, and glucose metabolism and anaerobic glycolysis in the kidney (Table 1; Rosenberger et al. 2002; Nangaku et al. 2013; Andringa and Agarwal 2014). HIF accomplishes this by targeting genes involved in these cellular processes such as erythropoietin, vascular endothelial growth factor (VEGF), and glucose transporters (Rosenberger et al. 2002; Nangaku et al. 2013). There are two active isoforms of HIF (HIF-1 and HIF-2) and in normoxic rats, no significant HIF-1 expression, using immunohistochemistry, was noted in the renal cortex or medulla, while weak cytoplasmic staining of HIF-2 was occasionally demonstrated in some tubular cells in the cortex and medulla and in medullary interstitial cells (Rosenberger et al. 2002). However, in hypoxic rats, HIF-1 was expressed in tubular cells, with strongest expression in the S2 segment tubular cells, and HIF-2 was expressed in peritubular interstitial cells in the cortex and medulla, capillary endothelial cells of the vasa recta, and in some glomerular cells (Rosenberger et al. 2002). In a rat model of regional hypoxia (induced segmental renal infarcts), upregulation of HIF was colocalized with regional upregulation of VEGF (Rosenberger et al. 2003). In a rat model of AKI, injection of HIF-1 improved survival and significantly alleviated I/R injury of renal tubules (H. Wang et al. 2018). Pimonidazole, a marker of deep hypoxia, has been observed in many AKI animal models (Tanaka, Tanaka, and Nangaku 2015). In AKI, both ischemic and nonischemic events lead to renal hypoxia and activation of HIF to protect the kidney against hypoxia and induce adaptive responses (Nangaku et al. 2013). For example, hypoxia can be shown in human renal biopsies after transplantation, and HIF accumulates immediately after engraftment and during the hypoxia phase and is also widely expressed during the recovery phase of AKI at 2 weeks postimplantation, irrespective of renal histology (Rosenberger et al. 2007). Intrinsic activation of HIF is submaximal in AKI, and additional augmentation of HIF can ameliorate ischemic renal injury (Nangaku et al. 2013). There are several factors that can predispose to AKI and activate renal HIF, including volume depletion, aging, diabetes, radiocontrast media, inhibition of cyclooxygenase, or vasoconstriction (Nangaku et al. 2013). In a sheep model of septic AKI, by continuous intravenous infusion of Escherichia coli, no evidence of renal tubular necrosis or increased apoptosis was noted, the cortical expression of all nitric oxide synthase (NOS) subtypes increased significantly, and cortical HIF-1α expression also increased (Langenberg et al. 2014). Activation of the innate immunity during hypoxia might be dependent on hypoxia-induced stabilization of HIF in neutrophils (Walmsley et al. 2005) and might lead to increased intravascular leukocyte stasis and mechanical obstruction of the vasculature by immune cells, which might further compromise blood flow (Ow et al. 2018). Interestingly, the tubular cells’ intrinsic capacity to mount an HIF response during acute hypoxia is most pronounced in collecting ducts, less pronounced in proximal tubules, and very limited in medullary thick limbs (Nangaku et al. 2013). Finally, under renal hypoxic conditions, various inflammatory cytokines (e.g., interferon gamma [IFNγ], IL-2, IL-10, and granulocyte–macrophage colony-stimulating factor [M-CSF]) were noted in renal tubular cells, and inflammatory cell infiltrates such as neutrophils, DCs, macrophages, and T cells (Goes et al. 1995; L. Li and Okusa 2010). In summary, HIF is a key regulator of adaptive responses in renal hypoxic conditions to regulate renal oxygen homeostasis, especially in I/R injury of AKI.

Both TLRs and complement act as early sensors of tissue damage (McCullough, Renner, and Thurman 2013). Complement-mediated immune function controls homeostatic process and eliminates infectious agents (Zipfel and Skerka 2009). In the kidney of humans and/or rodents, cell surface and soluble complement regulatory proteins (e.g., decay accelerating factor [DAF; CD55], membrane cofactor protein [MCP; CD47], complement receptor type 1 [CR1], CD59, clusterin, and factor H) are expressed by various microanatomic renal compartments such as glomerular capillaries, peritubular capillaries, proximal tubules, collecting ducts, medullary interstitium, and glomerular cells (endothelial, epithelial, and mesangial; Ichida et al. 1994; Nangaku 1998; Thurman et al. 2006; Zipfel and Skerka 2009). As part of the innate immunity, the complement system is a major player in the pathogenesis of I/R injury, and AKI and blocking complement in I/R injury can be effective for the prevention of renal fibrosis after ischemic AKI (McCullough, Renner, and Thurman 2013; Danobeitia et al. 2017). For example, increased renal I/R injury was noted in CD55 deficient (CD55^−/−), but not in CD59-deficient (CD59^−/−; Yamada et al. 2004). Mice deficient in complement factor B, an essential component of the alternative, not classic, pathway, were protected from ischemic AKI (Thurman et al. 2003; McCullough, Renner, and Thurman 2013). There is production of complement proteins by RTECs, which in turn promotes local complement activation and injury (McCullough, Renner, and Thurman 2013). In morphologically normal human kidney biopsy samples, small amounts of activated C3d were noted in the tubular basement membrane, while in biopsy samples obtained from patients with acute tubular necrosis (ATN), significant C3d complement deposition was noted in tubular basement membrane with no evidence for C4d deposition (Thurman et al. 2005). This indicates complement activation taking place primarily via the alternative pathway activation in AKI. Similarly, in rats with I/R injury, C3 was localized within necrotic tubular epithelial cells after 1 hr of total ischemia (Stein et al. 1985). In mice with I/R injury, C3 deposition was noted in the tubulointerstitial tissue along with neutrophil infiltration in the outer medulla (Thurman et al. 2003). In addition, activation of the complement system via dialysis or infections may exacerbate AKI (McCullough, Renner, and Thurman 2013). In mice, deficiency of C3aR and C5aR reduces renal production of inflammatory mediators after I/R injury and both C3a and C5a contribute to renal I/R injury, with predominant pathologic role for C5a (Peng et al. 2012). In rodent models of renal I/R injury, treatment with C5a-receptor (C5aR) antagonists prevented renal injury (Arumugam et al. 2003; De Vries et al. 2003). In rats, postischemic treatment with C5aR antagonist showed significant reduction in renal tissue damage compared to drug-free I/R injury animals (Arumugam et al. 2003). Interestingly, expression of C3aR and C5aR can be noted in circulating leukocytes contributes to the pathogenesis of renal I/R injury. There is downregulation of mRNA expression of various adhesion molecules (intercellular adhesion molecule [ICAM]-1, CD146, von Willebrand factor) in kidney tissues obtained from mice with deficiency in C3aR and C5aR and postischemia (Peng et al. 2012). In conclusion, the complement system should not be ignored because it is an important component of the innate arm of the immune system and can play a key role in the pathogenesis of AKI. Such understanding would help in finding new therapies for kidney injury.

The adherence of leukocytes and immune cells to the vascular endothelium and subsequent transmigration is mediated via various adhesion molecules (e.g., selectins and integrins) expressed on leukocytes and vascular endothelial cells. Leukocytes of humans and animals express L-, P-, and E-selectin, and β1 through β8 integrins (Radi, Kehrli, and Ackermann 2001). In the initial step of slowing of leukocytes on the vascular endothelium, selectins play a key role (Takada et al. 1997; Radi, Kehrli, and Ackermann 2001). This step is initiated after vasodilation and neutrophils flow through the vessel to allow for selectin-selectin receptor binding events (Radi, Kehrli, and Ackermann 2001). Following activation, neutrophils shed their selectin molecules and the expression of β2 integrins that bind ICAMs expressed on activated endothelial cells takes place to allow for the migration using intercellular junctions of endothelial cells, and this process is partly mediated by other adhesion molecules (Radi, Kehrli, and Ackermann 2001). Very late activation integrin-4, α4β1 integrin, is an important ligand for vascular cell adhesion moledule-1 (VCAM-1) as well as various extracellular matrix components such as fibronectin and vitronectin (Radi, Kehrli, and Ackermann 2001). Such adhesion molecules are upregulated after renal I/R injury; and in patients with sepsis-induced AKI, there is impairment of rolling and migration of neutrophils which is due, in part, to the decreased phosphorylation of selectin-dependent intracellular signaling pathways (Rossaint et al. 2011). Interestingly, resistin, a pro-inflammatory cytokine and uremic toxin, was significantly elevated in septic AKI patients and inhibited neutrophilic migration and intracellular actin polymerization (Singbartl et al. 2016). Mutant mice that are genetically deficient in ICAM-1 were protected from AKI (Kelly et al. 1996). In a mouse model of septic AKI, β₂-integrins, E-selectin, and P-selectin were shown to be involved in neutrophil recruitment into renal tissue (Herter et al. 2014). E-selectin (CD62E) is synthesized and expressed by endothelial cells, and its expression is induced by cytokines (Radi, Kehrli, and Ackermann 2001). E-selectin-deficient mice have a 75% reduction in myeloperoxidase activity (an indicator of neutrophil infiltration) in the postischemic kidneys at 24 hr (Singbartl et al. 2000). In mice, basigin (Bsg)/CD147 is a ligand for E-selectin that promotes renal inflammation in I/R injury, and Bsg-deficient (Bsg−/−) mice have significant suppression of neutrophil infiltration in the kidney after renal I/R injury (Kato et al. 2009). There is an overall good correlation for the increased levels of soluble adhesion molecules, with the presence of sepsis in neonates, children, and adults (Zonneveld et al. 2014). For example, in nonsurgical adult patients with severe sepsis AKI, a 1 ng/ml increase in serum E-selectin level increased the risk of septic AKI by 1%, and the E-selectin levels in the septic AKI group were significantly higher than those in the non-AKI group at two different times (days 1 and 4; Su et al. 2016). The serum levels of E-selectin, ICAM-1, and VCAM-1 were significantly higher in septic AKI patients (Su et al. 2016). Anti-inflammatory agents (e.g., recombinant alkaline phosphatase) are used as potential therapies for septic AKI (Benoit et al. 2018). Collectively, immune cells adhesion and subsequent infiltration into site of renal tissue injury comprises one of the first steps in trafficking of cells of the innate and adaptive arms of the immune system.

Interestingly, renal-resident CX3CR1+ DCs represent an extensive, contiguous network within normal kidney in mice (Soos et al. 2006). Such renal DCs are negative for CD4 and CD8 and differ from splenic conventional DCs by their prominent expression of F4/80, a surface marker of monocytes and macrophages (Soos et al. 2006). DCs present antigens to T cells and contribute to renal homeostasis and regulation of immune responses (Jang and Rabb 2015). Renal DCs are present in the tubulointerstitium and have a protective role in cisplatin-induced AKI (Tadagavadi and Reeves 2010). It is thought that renal DCs are the predominant secretors of TNF within 24 hr of I/R injury (Dong et al. 2007). In addition, kidney DCs might contribute to the recovery after I/R injury by phenotypic change from pro-inflammatory to anti-inflammatory with modulation of immune response (Kim et al. 2010). Depletion of DCs in mice before or coincident with cisplatin treatment exacerbated nephrotoxicity and resulted in more severe renal dysfunction, tubular injury, neutrophil infiltration, and mortality compared with mice with an intact renal DCs population (Tadagavadi and Reeves 2010). In addition, in cisplatin-induced AKI in mice, major histocompatibility complex class II expression decreased slightly and expression of inducible costimulator ligand (ICOS-L) increased significantly in DCs which suggests that modulation of the inflammatory response by DCs might have contributed to the observed protective effects in cisplatin-induced nephrotoxicity (Soos et al. 2006). In a mouse model of nephrotoxic nephritis, renoprotective effect was noted through DCs expressing ICOS-L and/or by inducing IL-10 in infiltrating CD4+ Th1 cells (Scholz et al. 2008). Depletion of DCs using clodronate prevented aggravation of I/R injury in single Ig IL-1-related receptor–deficient mice (Lech et al. 2009). In addition, resident macrophages and DCs are activated by DAMPs, which are released by activated renal epithelial and necrotic cells. Together, the activated cells recruit further leukocytes and initiate an immune response to clear debris and necrotic tissue, before tissue repair can take place (Kinsey and Okusa 2012). Collectively, the kidney DCs network is extensive, and contiguous, and AKI pathogenesis involves modulating such cellular pathway.

In renal I/R injury, macrophages play a key role to facilitate renal repair and interstitial fibrosis. Systemic depletion of monocytes and macrophages in a rat I/R injury model, using liposomal clodronate, attenuated renal tubular necrosis, markedly reduced inflammation in the corticomedullary junction and outer medullary area by 24 hr, and significantly decreased apoptosis of RTECs by 24 hr (Jo et al. 2006). In another study in osteopontin (a macrophage chemoattractant) KO mice, macrophage infiltration significantly decreased in cortex and the outer stripe of the outer medulla as well as there was significantly less expression of renal fibrosis markers (collagen I and IV) in the postischemic AKI phase (Persy et al. 2003). However, in a cisplatin-induced AKI in mice, liposome-encapsulated clodronate decreased renal CD11b-positive macrophages on day 3 but did not protect from AKI on day 3 (Lu et al. 2008). It has been demonstrated in mice that macrophages are not the source of injurious IL-18 in ischemic AKI (He et al. 2009), but the injured RTECs release DAMPs and PAMPs which act as a signal to resident DCs and macrophages (Han et al. 2018). Macrophages recognize such initial injury signal through PRRs, and this leads to stimulation of macrophage phagocytosis, phagolysosome maturation, antigen presentation, and production of the pro-inflammatory cytokine (e.g., TNF-α; Han et al. 2018). After this initial injury response, resident macrophages can further prolong the inflammatory process by recruiting neutrophils, monocytes, and lymphocytes (Han et al. 2018). In a uninephrectomized I/R injury rat model, the numbers of kidney monocytes/macrophages, determined using ED-1 staining, significantly increased at 24 hr post-I/R injury and became prominent by day 5 in proximal tubular cells in the outer stripe of outer medulla (Ysebaert et al. 2000). In the inner stripe of the outer renal medulla in a uninephrectomized I/R injury rat model, macrophages adhered/accumulated in the vasa recta by 2 hr after reperfusion (De Greef et al. 2001). In addition, there is loss of endothelial cell expression of basement membrane heparin sulfate proteoglycans in renal I/R injury, and this induces the expression of l-selectin and chemokines such as monocyte chemoattractant protein 1 which leads to monocytes/macrophages influx in human renal allograft rejection (Celie et al. 2007).

Macrophages exhibit two, biphasic phenotypes during renal injury and repair: pro-inflammatory (classically activated/M1) when they first respond to tissue injury within the first 24 to 48 hr, and reparative and pro-regenerative and anti-inflammatory (M2) phenotype which occurs toward the resolution of injury (i.e., increase in tubular cell proliferation that peaks on day 3; S. Lee et al. 2011; Han et al. 2018). In the kidney, increased M1 cytokines (e.g., IL-1, IL-6, IL-12, inducible NOS, TNF) expression can be noted within the first 48 hr after I/R injury in AKI and significantly decrease by 3 days postinjury (S. Lee et al. 2011; Han et al. 2018). M1 macrophages participate in innate immunity and are activated by the binding of DAMPs and PAMPs to PRRs and by IFNγ released from T helper cells and NKT cells (Han et al. 2018). The goal of the M1 inflammatory response is to remove damaged and pathogenic particles (Han et al. 2018). M2-activated macrophages are comprised of wound healing and immunoregulatory macrophages (Shapouri-Moghaddam et al. 2018). M2 macrophages secrete resolvins, lipoxins, transforming growth factor beta, and matrix metalloproteinases to cleave chemokines and chemoattractants and suppress the inflammatory response, contribute to tissue repair, remodeling, vasculogenesis, and retain homeostasis (Han et al. 2018; Shapouri-Moghaddam et al. 2018). M-CSF is a hematopoietic growth factor responsible for the survival and proliferation of monocytes and the differentiation of monocytes into macrophages (Radi et al. 2011). After AKI, M-CSF is produced by proximal tubular cells (Menke et al. 2009; Y. Wang et al. 2015). Tubular M-CSF can limit TECs apoptosis and promote tubular repair (Lenda et al. 2003). It has been shown in mice that the injection of CSF-1 postnatally enhanced kidney weight and was associated with increased infiltration of tissue macrophages, and thus CSF-1 facilitated renal repair and attenuated renal fibrosis by enhancing the macrophage expression of insulin-like growth factor 1 and anti-inflammatory genes (Alikhan et al. 2011). In conclusion, the polarization and biphasic phenotypic nature of macrophages (pro-inflammatory [classically activated/M1], anti-inflammatory [M2]) contribute to AKI pathophysiologic responses.

Neutrophils produce IL-17 to regulate IFNγ-mediated neutrophil migration in mouse I/R injury model (L. Li et al. 2010) and ischemia promotes neutrophil trafficking into the postischemic kidney in mice (Fukuzawa et al. 2009). In a mouse model of cisplatin-induced AKI, neutrophil infiltration into the kidney and the expression of IL-1β, IL-18, and IL-6 were increased; however, blocking either neutrophil infiltration or these cytokines did not prevent cisplatin-induced AKI (Faubel et al. 2007). Peptidylarginine deiminase-4 (PAD4) plays a critical role in ischemic AKI in mice by promoting renal tubular inflammation and neutrophil infiltration (H. Li et al. 2018). Renal proximal tubular cell in PAD4 deficient mice had profoundly reduced renal tubular apoptosis (H. Li et al. 2018). Although neutrophil infiltration can be noted in the kidney following ischemic injury in I/R injury animal models (Nemoto et al. 2001; Tadagavadi et al. 2015) and human biopsies of AKI patients (Solez, Morel–Maroger, and Sraer 1979; Friedewald and Rabb 2004), neutrophil depletion and neutropenia in rats did not protect against ischemic injury (Paller et al. 1989). Despite such conflicting results in the literature, neutrophils might participate in the induction of renal injury, by obstructing renal microvasculature and releasing oxygen-free radicals and proteases (Jang and Rabb 2015). During tissue inflammation and injury, leukotriene B₄ (LTB₄), a potent lipid chemoattractant, and its receptor Leukotriene B₄ receptor 1 (BLT1) are released at the site of inflammation (Deng et al. 2017). Although innate immune cells such as neutrophils are a major source of LTB₄ after inflammatory stimulation, RTECs act as the major cellular source of the increased LTB₄ in the kidney (Deng et al. 2017). BLT1^−/− mice exhibited significant reduction in the infiltration of neutrophils, inflammatory cytokine expression, and cell apoptosis in cisplatin-induced AKI (Deng et al. 2017). In these BLT1^−/− mice, reduced levels of oxidants and apoptosis might be a result of decreased neutrophils infiltration (Deng et al. 2017). Such data suggest a pro-inflammatory and proapoptotic role of the LTB₄-BLT1 axis in pathogenesis of cisplatin-induced AKI by recruitment of neutrophils to sites of renal injury (Deng et al. 2017). In summary, the role of neutrophils as an innate component of the immune system is AKI appears to be more of an indirect effect in AKI related to intravascular leukocytostasis leading to sluggish blood flow in renal microvasculature and peritubular capillary system with subsequent ROS release in their microenvironment and contributing to AKI pathogenesis.

In mouse kidneys, the degree of lymphocyte infiltration at 3 and 24 hr after renal I/R injury was characterized using immunohistochemistry (Ascon et al. 2006). Although CD3+ T lymphocytes were noted in normal mouse kidneys, a slight increase in CD3+ T cells was noted 3 hr in the peritubular capillaries, particularly in the medullary outer stripe in kidneys, followed by a decrease 24 hr after renal I/R injury (Ascon et al. 2006). Increased CD19+ B cells was noted at 3 hr, and increased infiltration of NK1.1+ and CD4+ NK1.1+ cells were noted at 3 and 24 hr after renal I/R injury (Ascon et al. 2006). By 24 hr of renal I/R injury, decreased percentages of CD3+, CD19+, and NK1.1+ cells were noted in I/R injury mouse model (Ascon et al. 2006). T cells contribute to increased vascular permeability after I/R injury (Martina et al. 2014). Peritubular T lymphocytes have been identified in experimental I/R injury, and adhesion between T cells and RTECs underlies the pathophysiological role of T cells in renal I/R injury (Rabb et al. 2000). In genetically engineered CD4/CD8 double KO mice, marked improvement in renal function was noted at 48 hr postischemia, with a two-fold increase in adherence of T lymphocytes to RTECs in vitro after hypoxia and reoxygenation. Mice deficient in T cells (nu/nu mice) are both functionally and structurally protected from postischemic renal injury (Burne et al. 2001). The pathogenic role for T lymphocytes in AKI was examined in cisplatin-induced AKI model in T cell–deficient mice, and it was found that T lymphocytes play a role in experimental cisplatin nephrotoxicity (Liu et al. 2006). For example, the number of T cells was significantly increased as early as 1 hr after cisplatin administration in kidneys from wild-type mice, peaked by 12 hr, and declined by 24 hr. Renal protective effects of adenosine 2A receptor agonists were shown to be mediated by actions on CD4+ T cells in an I/R injury mouse model (Day et al. 2006). Because CD4 T cells can functionally differentiate to either a pathogenic Th1 (IFNγ producing) phenotype or protective Th2 (IL-4) phenotype, the renal protective effects of CD4+ T cell subsets were studied in a mouse model of renal I/R injury (Yokota et al. 2003). The KO mice used in this study had deletions of signal transducer and activator of transcription (STAT) 4 and STAT6, enzymes that regulate differentiation of T cells into Th1 or Th2 phenotypes, respectively (Yokota et al. 2003). STAT4-deficient mice were partially protected from renal I/R injury, whereas STAT6-deficient mice had markedly impaired renal function and tubular injury postischemia, which suggests an important renal protective role for STAT6 pathway in I/R injury (Yokota et al. 2003). Interestingly, the phenotype of renal lymphocytes and the expression of cytokines in the normal kidney and after I/R injury in mice are influenced by microbial stimuli (Jang et al. 2009). Furthermore, aging, at least in mice, affects the intrarenal immunologic micromilieu with small effects on the initial renal injury following I/R injury (Jang et al. 2015). For example, normal kidneys of aged mice (12 months old) contain more leukocytes, activated T cells, and effector memory T cells than younger (9 weeks and 6 months) mice, and higher numbers of T cells and less numbers of B cells were noted in the postischemic kidneys of old mice (Jang et al. 2015). CD4-deficient (and, to a lesser degree, CD8-deficient) mice showed decreased cisplatin-induced mortality and renal dysfunction compared with wild-type mice (Liu et al. 2006). In cisplatin-induced AKI in mice that lack mature T cells, it was found that adoptive transfer of CD4+CD25+ Treg cells attenuated renal dysfunction and tubular injury, renal TNF-α, and IL-1β cytokine levels (H. Lee et al. 2010). Cisplatin-treated mice with severe combined immunodeficiency/beige mice (B-, T-, and natural killer cell–deficient) showed significant survival, with only 55% mortality, but exhibited significant renal failure (Linkermann et al. 2011). Cisplatin-induced AKI is related to Fas-mediated apoptosis, which is driven by FasL that is expressed on renal tubular cells and infiltrating T cells (Linkermann et al. 2011). Furthermore, immunosuppressive medications that affect lymphocyte trafficking such as tacrolimus and mycophenolate mofetil substantially attenuated early renal injury in rats following I/R injury (Sakr et al. 1992; Jones et al. 2000). In uninephrectomized rats, blockade of the T cell CD28-B7 costimulatory pathway with CTLA 4–Ig significantly inhibited T cell and macrophage infiltration and activation in situ and reduced early renal injury after I/R injury (Takada et al. 1997). In uninephrectomized rat model of chronic progressive proteinuria, CTLA 4–Ig treatment on the day of and during the first week after I/R injury was effective in ameliorating development of progressive proteinuria (Chandraker et al. 1997).

Th17 cells play an important role in sepsis-induced AKI pathogenesis, and data suggest that Th17 cells in the circulation of patients with septic shock are hyperactivated as evidenced by the ex vivo over-release of IL-17 in patients with septic AKI (Maravitsa et al. 2016). In a mouse model of multiple organ dysfunction complicated with AKI and bacterial gut translocation, IL-17 was the only cytokine produced at greater quantities from CD4 lymphocytes (Maravitsa et al. 2016). In IL-17A KO mice with sepsis-associated AKI, decreased serum creatinine and blood urea nitrogen, improved ATN score, and significantly decreased apoptosis of tubular epithelial cells, including decreased TUNEL-positive tubular cell number and cleaved caspase-3 level, were noted (Luo et al. 2016). Collectively, inhibiting lymphocytes trafficking and activation of these immune cells with the subsequent modulation of downstream signaling pathways involved in a release of pro-inflammatory cytokines and chemokines can offer protective therapeutic benefits against renal pathology and AKI.

NKT cells are a unique lymphocyte population that serves as a bridge between innate and adaptive immunity (Hu, Zhang, and Yang 2017). In renal I/R injury in mice, NKT cells are increased by 3 hr return to normal levels by 24 hr after I/R injury and NKT cells contribute to the induction of early renal injury by mediating neutrophil IFNγ production (Ascon et al. 2006; L. Li et al. 2007). Three hours after reperfusion in mice, kidney CD1d-restricted NKT cells were noted (L. Li et al. 2007). Renal I/R injury in mice administered antibodies to block CD1d, or deplete NKT cells or in mice deficient of NKT cells (Jalpha18[−/−]), was markedly attenuated (L. Li et al. 2007). Interestingly, protective effects of isoflurane anesthetic treatment were mediated, in part, by a significant reduction in NK1.1+ lymphocyte cell infiltration, which was demonstrated in a mouse model of renal I/R injury (H. T. Lee et al. 2007). However, in another experimental I/R mouse model, it was shown that the absence of NKT cells (especially type II NKT cells) accentuated the severity of kidney injury, whereas repletion of NKT cells attenuated kidney injury (Yang et al. 2011). In this study, an endogenous glycolipid, 3 sulfated galactosylceramide (sulfatide), was used to induce activation of type II NKT cells, which suggested renoprotective role for NKT cells, especially sulfatide-reactive type II NKT cells, after I/R injury (Yang et al. 2011). Interestingly, although rare numbers of NKT cells can be found in normal human kidney tissue, higher/increased number of cells were noted in kidney tissue from patients with ATN, and the number of NKT cells varied according to the severity of ATN and the timing of kidney biopsy (Yang et al. 2011). In conclusion, NKT cells bridges the two arms of the immune system to bring renoprotective properties in AKI.