Renal-specific loss of ferroportin disrupts iron homeostasis and attenuates recovery from acute kidney injury

Abstract

Chronic kidney disease is increasing at an alarming rate and correlates with the increase in diabetes, obesity, and hypertension that disproportionately impact socioeconomically disadvantaged communities. Iron plays essential roles in many biological processes including oxygen transport, mitochondrial function, cell proliferation, and regeneration. However, excess iron induces the generation and propagation of reactive oxygen species, which lead to oxidative stress, cellular damage, and ferroptosis. Iron homeostasis is regulated in part by the kidney through iron resorption from the glomerular filtrate and exports into the plasma by ferroportin (FPN). Yet, the impact of iron overload in the kidney has not been addressed. To test more directly whether excess iron accumulation is toxic to kidneys, we generated a kidney proximal tubule-specific knockout of FPN. Despite significant intracellular iron accumulation in FPN mutant tubules, basal kidney function was not measurably different from wild type kidneys. However, upon induction of acute kidney injury (AKI), FPN mutant kidneys exhibited significantly more damage and failed recovery, evidence for ferroptosis, and increased fibrosis. Thus, disruption of iron export in proximal tubules, leading to iron overload, can significantly impair recovery from AKI and can contribute to progressive renal damage indicative of chronic kidney disease. Understanding the mechanisms that regulate iron homeostasis in the kidney may provide new therapeutic strategies for progressive kidney disease and other ferroptosis-associated disorders.

NEW & NOTEWORTHY Physiological iron homeostasis depends in part on renal resorption and export into the plasma. We show that specific deletion of iron exporters in the proximal tubules sensitizes cells to injury and inhibits recovery. This can promote a chronic kidney disease phenotype. Our paper demonstrates the need for iron balance in the proximal tubules to maintain and promote healthy recovery after acute kidney injury.

INTRODUCTION

Iron is an essential mineral used by living organisms for a variety of biological processes including oxygen transport, enzyme catalysis, and bioenergetic reactions. The majority of iron is found intracellularly, where it is bound to hemoglobin and myoglobin, the major oxygen transport and storage proteins. Much of the remainder is bound to transferrin in circulation. Iron levels must be regulated, as iron overload leads to oxidative stress and cell damage (1), whereas iron deficiencies lead to chronic anemia. The versatility of iron is due in part to its ability for redox recycling, as the oxidation state can vary greatly from +2 to −7, although different biological systems use mostly the ferrous Fe(II) and ferric Fe(III) forms for electron transfer. Iron is also linked to a novel cell death pathway termed ferroptosis, which is driven by iron overload and a redox imbalance that results in the robust oxidation of polyunsaturated fatty acid-containing phospholipids (2). Implicated in both acute kidney injury (AKI) and chronic kidney disease (CKD), the mechanisms and causes of ferroptosis in the kidney remain to be fully characterized. Whether iron overload is sufficient to cause renal injury or exacerbate existing chronic or acute conditions remains to be fully characterized.

The kidney is a critical regulator of systemic iron homeostasis. Under normal physiological conditions, transferrin-bound iron is filtered by the glomerulus and reabsorbed in kidney tubules by binding to apical membrane transferrin receptors (3). Once taken up by renal epithelial cells, iron can be exported into the plasma by basolateral ferroportin (FPN), the only known iron export protein. FPN function is inhibited following degradation by hepcidin, a peptide expressed primarily in the liver that limits iron export into the circulation to prevent systemic iron overload (4). A kidney-specific deletion of FPN in whole nephrons resulted in excess intracellular iron retention, increased ferritin heavy chain expression, and a reduction in plasma and hepatic iron levels, whereas deletion of FPN in more distal tubules had little impact on iron homeostasis (5). Thus, proximal tubules appear to be the major site of iron resorption and export into the plasma. Proximal tubules are also the only part of the nephron capable of regeneration after acute injury. Despite their regenerative capacity, repeated bouts of AKI can lead to a more chronic progressive kidney disease (6, 7). Iron deposition has been observed in about one-third of CKD biopsies (8), yet how iron imbalance impacts regeneration after AKI and progression to CKD has not been investigated.

In this report, we generated a proximal tubule-specific deletion of the ferroportin gene (Fpn1 or Slc40A1) and protein. Using a Pepck-Cre driver that is active in the S1, S2, and S3 segments of the renal proximal tubules, we characterized the loss of FPN in mice before and after experimentally induced AKI. The data show significant accumulation of iron in the proximal tubules before injury, but little effect on physiology or pathology. However, the loss of FPN attenuated the ability of injured proximal tubules to fully recover in large part due to an increase in ferroptosis. Thus, an imbalance in iron homeostasis in the kidney can accentuate renal injury, inhibit recovery, and indirectly promote a chronic renal phenotype.

METHODS

Animals

All animal protocols complied with the National Institutes of Health Guide for the Care of Use of Laboratory Animals and were approved by the University Committee on Use and Care of Animals at the University of Michigan. Mice were housed in a specific pathogen-free facility with a 12:12-h light-dark cycle and given free access to food and water. Fpn^f/f; mT/mG; Pepck-cre mutated mice (FPN−) were obtained by crossing the Fpn^f/f mouse strain (9, 10) with the reporter Gt (Rosa)26Sor^{tm4(ACTB-tdTomato,-EGFP)luo} mouse strain (Jackson Labs, Stock No. 007676) and the phosphoenolpyruvate carboxykinase-Cre (Pepck-cre) transgenic mice. The Pepck-cre transgenic mice were generated by targeted single-copy transgenesis in embryonic stem cells, a gift from Volker H. Haase (11). For genotyping, genomic DNA was prepared by a standard method and amplified by PCR using the following primer pair sets:

eGFP Fwd- GGGCGATGCCACCTACGGCAAGCTGACCCT

eGFP Rev- CCGTCCTCCTTGAAGTCGATGCCCTTCAGC

FPNf/f Fwd- GGCATTCCCAACACTTTAGC

FPNf/f Rev- CCCATAGGTTAAACTGCTTCAA

Cre Fwd- CGAGTGATGAGGTTCGCAAG

Cre Rev- TGAGTGAACGAACCTGGTCG

TdTomato WTF wd- CTCTGCTGCCTCCTGGCTTCT

WT Rev- CGAGGCGGATCACAAGCAATA

Floxed Rev- TCAATGGGC GGGGGTCGTT

AKI Models

For folic acid nephrotoxicity, FPN mutant mice Fpn^f/f; mT/mG; Pepck-cre (FPN−) and control mice Fpn^f/f or FPN+ mT/mG; Pepck-cre (FPN+) with similar genetic background and ages were injected intraperitoneally with a single dose of 200 mg/kg of folic acid in a total volume of 500 µL of 0.15 M NaHCO₃. Because of the high percentage of death in the FPN mutant mice with 250 mg/kg of body weight, the dose was adjusted to 200 mg/kg and mice were put under constant observation for the first 48 h after injection. Mice were then euthanized at 2, 7, and 28 days after injury and blood samples were obtained from the abdominal aorta for renal function and serum analyses. Mice were immediately perfused with cold PBS to improve fixation and tissue processing as described (12, 13). At each time point, 0, 2, 7, and 28 days, three to four age-matched male mice were taken of each genotype and kidneys were harvested; one kidney of each was used for histology, and the other kidney was freshly frozen and used for protein and RNA analysis. The Pepck-Cre transgene was X chromosome linked; thus, we used male mice for all experiments because females were mosaic for Cre unless bred to homozygosity. Females are also more resistant to AKI and can exhibit greater variability.

Histology, Perl’s Staining, and Immunofluorescence

Kidney samples were fixed overnight in 4% paraformaldehyde/1× PBS and processed for paraffin embedding. Sections of 5 μm were used for histology with hematoxylin and eosin (H&E) staining following standard protocols and slides were scanned using Vectra Polaris whole slide scanner (Akoya). Sections were visualized, processed, and quantified using QuPath (v. 0.3.2) in combination with ImageJ. Detailed step-by-step instructions for each domain as well as ImageJ and QuPath macros for automated processing are available online as described (14). Acute injury was scored by calculating the percentage of tubules that display cell necrosis, loss of the brush border, cast formation, or tubular dilatation as described (14, 15) using the following cutoffs: 0: No injury, 1: 1–10%, 2: 11–25%, 3: 26–45%,4: 46–75%, and 5: 76–100%.

Tissue iron detection was performed in paraffin-embedded sections; tissue sections were deparaffinized using xylene, and then rehydrated in ethanol. Slides were then stained for 1 h with Prussian blue and counterstained with hematoxylin (NovalUltra Special Stain Kits). Immunofluorescence detection was done using a modified version of published protocols (16). Paraffin sections were rehydrated and stained with mouse monoclonal anti-α-smooth muscle actin (SMA) (Invitrogen, Cat. No. ma1-06110), rabbit anti-collagen I (Abcam Cat. No. ab34710), goat anti-kidney injury molecule-1 (Kim1) (R&D Systems, anti-mTIM1, Cat. No. AF1817), chicken anti-green fluorescent protein (GFP) (Life technology, Cat. No. A10262), goat anti-villin (Santa Cruz Biotech, Cat. No. SC7672), rabbit anti-laminin (Sigma Cat. No. L9393), rabbit anti-SLC4OA1 (Novus, Cat. No. NBP1-21502), rat anti-Ki67 (Biolegend, Cat. No. 562402), rabbit anti-Vcam (Abcam, Cat. No. ab134047), rabbit anti-4HNE9 (Abcam, Cat. No. ab46545), rabbit anti-ACSL4 (AbCam Cat. No. ab155282), and rabbit anti-NCOA4 (Bethyl, Cat. No. A302-272A). Images were collected using an Olympus DP80 camera (Olympus America), scanned, and quantified as described (12).

Western Blots

Half of each fresh frozen kidney was used for protein analysis, and kidney extracts were prepared as briefly described in Ref. 16. Tissue was homogenized in the RIPA Lysis Buffer System (Santa Cruz Sc-24948); the kit includes lysis buffer, PMSF, sodium orthovanadate, and protease inhibitor cocktail. Crude lysates were incubated at 4°C on the rotisserie for 2 h, and centrifuged at 14,000 g for 10 min. Supernatant was transferred to fresh tubes and the protein concentration was measured using the BCA Protein Assay Kit (23227, Pierce). Samples were mixed in 2× SDS/PAGE sample buffer, separated on gradient precast polyacrylamide gels, and transferred to PVDF membranes (Millipore IPFL00010). Specific proteins were detected with the following primary antibodies: monoclonal mouse anti-α-smooth muscle actin (SMA) (Invitrogen, Cat. No. ma1-06110), goat anti-Kim1 (R&D Systems, Cat. No. AF1817), rabbit anti-SLC4OA1 (Novus, Cat. No. NBP1-21502), rabbit anti-Vcam (Acam, Cat. No. ab134047), rabbit anti-4HNE9 (Abcam, Cat. No. ab46545), rabbit anti-NCOA4 (Bethyl, Cat. No. A302-272A), goat anti-lipocalin-2/NGAL (R&D Systems, Cat. No. AF1867), rabbit anti-ACSL4 (Abcam, Cat. No. ab55282), rabbit anti-GPX4 (Abclonal, Cat. No. A1933), rabbit anti-FTH1 (Cell Signaling, Cat. No. 3998), rabbit anti-Slc7A11 (Abcam, Cat. No. ab37185), rabbit anti-β-actin (Sigma-Aldrich, Cat. No. A2066), and horseradish peroxidase-conjugated secondary antibodies (GE Healthcare) using Western Lightning Enhanced Chemiluminescence (Pierce). Blots were exposed to autoradiography films (Lab Scientific, Cat. No. XARALF1318). Films were scanned and band intensities were analyzed with ImageJ software.

Renal Function

For quantification of blood urea nitrogen (BUN) and serum creatinine, a LIASY 330 Chemistry Analyzer with standard protocols for mouse serum samples was used. Services were provided by the In-Vivo Animal Core (IVAC) Animal Diagnostic Laboratory, University of Michigan. Specific product parameters and procedures are found in the LIASY 330 instrument software, with a normal range of 5.15–30.70 mg/dL for BUN and a normal range of 0.09–0.40 mg/dL for creatinine.

Gene Expression and Quantification by qPCR

Total RNA was extracted from the kidney cortex using RNeasy Mini Kits (QIAGEN), and cDNA was generated using High-Capacity cDNA Reverse Transcription Kits (Thermo Fisher), according to the manufacturer’s instructions. RNA (1–2 μg) in final volume of 10 μL was added into PCR tubes and 10 μL of the prepared master mix were added to each reaction. Samples were mixed gently and then placed in BioRad T100 Thermal cycler. qPCR runs for SYBR Green primers were carried out using Quant Studio 3 (Thermo Fisher) in triplicate using a master mix that was prepared. Per reaction, the mix contained 5 μL of Maxima SYBR Green/ROX qPCR Master Mix (2X) (Thermo Fisher), 0.25 μL of SYBR Green forward primer, 0.25 μL of reverse primer, 2.5 μL of water, and 2 μL of cDNA template. The qPCR cycling conditions for SYBR Green primers were as previously described (17). All samples were amplified in triplicates. The average threshold cycle (Ct) values for the differential genes and housekeeping gene (GAPDH) were obtained from each reaction. The gene expression in the tested samples was quantified using the 2(−ΔΔCt) relative method and the following primer pairs:

Zeb1 Fwd- TCAAGTACAAACACCACCTG and Zeb1 Rev- TGGCGAGGAACACTGAGA, Pal1 Fwd- ACATGTTTAGTGCAACCCTG and Pal1 Rev- GGTCTATAACCATCTCCGTG, and Snail1 Fwd GGAAGCCCAACTATAGCGA and Snail1 Rev- AGCGAGGTCAGCTCTACG.

Statistical Analyses

Each experiment was performed using at least three independent samples as described in methods. Averages and standard deviations were calculated using GraphPad Prism 8.0.0 and significance was assessed by two-way ANOVA multiple comparisons or the Student’s two-tailed t test for independent variables. P values less than 0.05 were considered to be significant.

RESULTS

To specifically delete FPN in renal proximal tubules, we used the X chromosome-linked Pepck-Cre transgene in male mice. Using the mouse mT/mG fluorescent reporter strain, robust and efficient excision of the GFP reporter allele was observed throughout the cortex in villin-positive proximal tubules (Fig. 1, A and B). FPN^+/+;Pepck-Cre or FPN^f/f mice, hence referred to as FPN+, and FPN^f/f;Pepck-Cre mice, referred to as FPN−, were immunostained for FPN (red), villin (white), and GFP (green) and showed a near complete loss of FPN in villin-positive cells of mice carrying both FPN^f/f alleles and the Pepck-Cre driver (Fig. 1, C and D). As expected, the deletion of FPN resulted in iron accumulation in proximal tubules as exhibited by increased Prussian blue staining for intracellular ferric iron (Fig. 1, E and F). Western blotting for FPN (Fig. 1G) and quantification of protein and mRNA (Fig. 1H) confirmed the reduction of FPN in the FPN− strain of mice, whereas the residual FPN protein is most likely from distal tubules, collecting ducts, and glomeruli that do not delete with the Pepck-Cre driver. Despite the renal iron overload, we did not see a noticeable difference in renal function when comparing serum BUN or creatinine between FPN+ and FPN− mice. Histology appeared normal in all controls and mutants.

Figure 1.

A renal proximal tubule-specific deletion of FPN. A: mice carrying the mT/mG reporter allele (Rosa)26Sor^{tm4(ACTB-tdTomato,-EGFP)luo} show efficient Cre-mediated excision (GFP) in cortical proximal tubules, but not in collecting ducts or renal medulla (tomato red, scale bar is 2 mm). B: high magnification of A, stained for villin (white), laminin (red), and GFP (green), shows GFP in villin-positive (white) proximal tubules. Scale bar is 50 µm. C: control FPN+ kidney section stained for FPN (red), villin (white), and GFP (green). Scale bar is 25 µm. D: Pepck-Cre FPN− mutant kidney stained for FPN (red), villin (white), and GFP (green). Scale bar is 25 µm. E: Perls Prussian blue staining of a control FPN kidney section. Scale bar is 200 µm. F: Perls Prussian blue staining of a Pepck-Cre FPN− kidney section. Scale bar is 200 µm. G: Western blot of kidney lysates from FPN+ and FPN− kidneys. H: quantification of FPN protein levels and FPN mRNA levels in control FPN+ and mutant FPN− kidneys, **P < 0.01. I: survival curves of WT and FPN− mice after AKI at 200 mg/kg or 250 mg/kg folic acid injection. J: Western blot of FPN after AKI in 3 independent samples. K: quantitation of J shows twofold increase of FPN in wild-type kidneys at 7 days after AKI (***P < 0.001). L: Western blot for Kim1 and FTH1 show increases after AKI, with persistent expression in FPN− kidneys at 7 days after folic acid. M: quantitation of L (*P < 0.05, ***P < 0.001). AKI, acute kidney injury; FPN, ferroportin; WT, wild-type.

To test whether renal iron overload can impact injury and recovery, we used an established folic acid nephrotoxicity model of AKI (12, 13, 18). We examined the response to folic acid-induced AKI and subsequent regeneration of injured proximal tubules in FPN+ and FPN− mice. Initial experiments used a standard dose of 250 mg/kg folic acid. However, we noted that FPN− mice were significantly more sensitive to folic acid-induced AKI than the FPN+ cohorts, with no survivors at 250 mg/kg (Fig. 1I). Thus, we used a single dose of 200 mg/kg body weight to induce nephrotoxicity and an acute injury. At this lower dose, all FPN+ mice recovered, whereas 58% of FPN− mice survived to 28 days post-AKI. Also, males were used exclusively as they are more sensitive to AKI and are not subject to inactivation of the X-linked Pepck-Cre driver. After AKI, upregulation of FPN was seen in controls (Fig. 1, J and K), whereas the injury marker KIM1 (19) and ferritin heavy chain FTH1 (5) were upregulated in both FPN+ and FPN− kidneys (Fig. 1, L and M), as predicted.

Mice were euthanized at 2, 7, and 28 days post-AKI, and kidneys were isolated, decapsulated, halved for fixation, histology (Fig. 2A), and immunostaining. Protein lysates were made from halved kidneys for Western blotting. Folic acid was used because it induces rapid AKI as measured by increased serum BUN and creatinine. FPN+ and FPN− mice exhibited similar increases in BUN 2 days after folic acid (Fig. 2C), whereas FPN− mice had significantly more serum creatinine when compared with FPN+ mice after AKI (Fig. 2B). H&E-stained sections from kidneys were scored for AKI severity according to previously described methods (14, 15). At 2 days post folic acid, FPN− mice showed a significant increase in AKI severity, with a median score of 5 compared with 3 for FPN+ mice (Fig. 2D). By 7 and 28 days after AKI, FPN+ mice recovered well and showed few signs of prolonged injury, whereas FPN− mice continued to exhibit renal injury with AKI scores above 3 even at 28 days after folic acid when recovery is usually complete (Fig. 2, A and D).

Figure 2.

Loss of PFN inhibits recovery from AKI. A: representative images of trichrome-stained kidney sections from uninjured (UN) or post AKI (as indicated) FPN+ and FPN− mice. Scale bar is 100 µm. B: serum creatinine taken 2 days after folic acid (FA) from FPN+ and FPN− mice. C: serum blood urea nitrogen (BUN) taken at 2 days after folic acid from FPN+ and FPN− mice. D: compiled AKI histology scores over time as assessed by standard scale. Significance used the Student’s t test for two independent variables. **P < 0.01, ***P < 0.001. AKI, acute kidney injury; FPN, ferroportin.

Failed or incomplete repair of proximal tubules after AKI can lead to interstitial fibrosis, scarring, and CKD (7, 20). Thus, we examined FPN+ and FPN− kidneys for fibrosis and markers that indicate severe chronic injury. At 7 and 28 days after folic acid-induced AKI, FPN− kidneys exhibited persistent expression of SMA, Kim1, and Vcam, all markers of failed repair after AKI (Fig. 3A). Quantitative Western blotting also confirmed a rise in injury markers Vcam, Ngal, SMA, and Kim1 in the acute phase of injury, which persisted to a greater extent in FPN− kidneys at 7 and 28 days (Fig. 3, B and C). Regeneration of proximal tubules after AKI requires surviving epithelial cells to reenter the mitotic cycle and proliferate to repair damaged tubules (21). Proliferation of surviving cells peaks at ∼2 days after AKI and can be quantified with Ki67 staining. Consistent with the inability to recover, the FPN− kidneys showed significantly fewer Ki67-positive nuclei at 2 days after AKI (Fig. 4, A and B). Targets of the profibrotic cytokine TGF-β are often increased in fibrotic kidneys and progression to CKD (20, 22, 23). Among these, we examined Pal1, Snail1, and Zeb1, all of which were significantly increased in FPN− mice 7 days after folic acid (Fig. 4C). These data suggest that excess intracellular iron in proximal tubules compromises their ability to regenerate after injury and can lead to chronic renal pathologies.

Figure 3.

Loss of FPN and iron overload accentuates chronic renal damage. A: immunostaining for αSMA (green), Vcam (red), and Kim1 (white) in sections from uninjured or 2, 7, or 28 days after folic acid (FA) of FPN+ and FPN− kidneys. Scale bar is 100 µm. B: Western blotting of kidney lysates from three independent FPN+ or FPN− mice taken at various times after folic acid and probed with antibodies against the indicated proteins. C: quantitation of protein levels shows persistent higher levels of fibrotic and injury markers in FPN− mice after AKI. Significance used the Student’s t test for two independent variables. *P < 0.05, **P < 0.01, ***P < 0.001. AKI, acute kidney injury; FPN, ferroportin.

Figure 4.

Reduced proliferation after AKI in FPN mutant kidneys. A: immunostaining for the proliferation marker Ki67 in uninjured and 2 days after folic acid (FA) kidneys from FPN+ and FPN− mice. B: quantitation from multiple images taken from FPN+ and FPN− kidneys. C: quantitative RT-PCR for the TGF-β target genes as indicated. Significance used the Student’s t test for two independent variables. *P < 0.05, **P < 0.01, ***P < 0.001. AKI, acute kidney injury; FPN, ferroportin.

Excess iron accumulation can promote ferroptosis, an iron-dependent programmed cell death mechanism associated with lipid peroxidation and the loss of cell membrane integrity (24). Given the iron overload in FPN− kidneys, we examined potential changes in the expression of proteins linked to iron homeostasis, including both inhibitors and inducers of ferroptosis (Fig. 5). At 7 and 28 days after folic acid, FPN− kidneys had higher levels of ferritin heavy chain (FTH1) but reduced levels of the ferroptosis repressor Slc7A11 (25) compared with FPN+ kidneys. The acyl-CoA synthetase long-chain family protein (ACSL4) is known to promote apoptosis (26, 27) and remains highly expressed in the FPN− kidneys even at 28 days after AKI. Loss of GPX4 is thought to promote ferroptosis (28, 29) and is decreased in FPN− kidneys at 7 and 28 days after injury. The nuclear receptor coactivator 4 (NCOA4) is responsive to intracellular iron levels and also mediates ferritinophagy (30, 31). When iron is abundant or in excess, NCOA4 is degraded through ubiquitination by the E3 ubiquitin ligase Herc2. Consistent with this mechanism, NCOA4 protein expression was lower in FPN− mice and remained so after AKI. Especially at 28 days after AKI, NCOA4 levels were less than half that of WT controls. Lastly, persistent NGAL expression (32) also indicated chronic renal injury in the FPN− kidneys even after 28 days.

Figure 5.

Expression of iron homeostasis genes in FPN+ and FPN− kidneys before and after AKI. A: Western blotting of lysates from six independent FPN+ and FPN− kidneys were probed for the indicated proteins. Samples were from uninjured, 2 days, 7 days, and 28 days after folic acid (FA) as indicated. B: quantitation of relative protein levels from A. Relative protein levels are determined from single blots and are presented in two graphs because protein gels are limited to 12 samples plus markers. Significance used the Student’s t test for two independent variables. *P < 0.05, **P < 0.01, ***P < 0.001. AKI, acute kidney injury; FPN, ferroportin.

To confirm the high levels of ACSL4 expression in tubules, we immunostained sections for ACSL4 at various times after injury (Fig. 6). Although FPN+ kidneys showed ACSL4 staining at 2 days after injury, there was little overlap between ACSL4 and GFP that marked the Pepck-Cre-expressing proximal tubules. However, in FPN− kidneys, high levels of ACSL4 protein were detected in multiple epithelial cells including GFP-positive proximal tubules at 2 and 7 days after injury and remained high at 28 days. We also stained sections for NCOA4 (Fig. 7A), which confirmed the significant decrease in FPN− kidney as observed by Western blotting. α,β-Unsaturated hydroxyalkenal (4HNE) is produced by lipid peroxidation and increases with oxidative stress leading to ferroptosis. Western blotting and immunostaining show increased 4HNE levels in FPN− kidneys at 28 days after injury, consistent with ferroptosis. In total, these findings suggest that ferroptotic cell death underlies the failure to fully recover and regenerate after AKI in the FPN− mice that exhibit iron overload in the proximal tubules.

Figure 6.

Expression of ACSL4 in FPN+ and FPN− kidneys after injury. Sections were immunostained with anti-ACSL4 (red), anti-Kim1 (white), and anti-GFP (green) at the indicated times after injury. Control FPN+ kidneys show little overlap between GFP and ACSL4, whereas strong ACSL4 expression in GFP+ proximal tubules is seen in FPN− kidneys at 2 and 7 days after injury. Note examples of overlap between GFP and ACSL4 (arrows). Scale bar is 50 µm. FA, folic acid; FPN, ferroportin.

Figure 7.

Reduced NCOA4 and increased 4HNE in FPN− kidneys after AKI. A: immunostaining of sections for NCOA4 (red) in uninjured and after folic acid (FA) kidneys from mice as indicated. Pepck-Cre mice carrying the mT/GFP reporter were costained for GFP (green). Scale bar is 50 µm. B: Western blot of proteins from kidneys taken at 28 days after folic acid from FPN+ and FPN− mice and probed with anti-4HNE antibodies. C: quantitation of 4HNE-reacting proteins as probed in B. *P < 0.05. D: sections from 28 days after folic acid were stained for 4HNE (red), villin (white), and GFP (green) as indicated. Scale bar is 50 µm. AKI, acute kidney injury; FPN, ferroportin; 4HNE, α,β-unsaturated hydroxyalkenal; NCOA4, nuclear receptor coactivator 4.

DISCUSSION

Iron homeostasis is tightly regulated, as either low or excess iron levels can be detrimental to cellular functions. The kidney is central to the control of iron levels, through the reabsorption of iron from the glomerular filtrate by proximal tubule cells that then actively transport iron back into the plasma. FPN is the only known iron export protein that is responsible for iron recycling. We used a specific Pepck-Cre driver to delete FPN primarily in the S1 to S3 segments of the proximal tubules. Although this results in measurable iron overload in specific cell types, we did not observe significant effects on renal function or pathology in the absence of secondary insults.

Given that the proximal tubules are the primary cells injured in nephrotoxicity, we used a folic acid model of AKI and recovery to test the effects of iron overload on proximal tubule cells and their ability to regenerate. Before injury, there were no gross pathological effects of FPN deletion in the proximal tubules nor did deletion alone impact renal functions, as measured by serum creatinine and BUN. However, FPN mutants were sensitive to nephrotoxicity with no surviving male mice at a standard dose of 250 mg/kg, suggesting that cells with increased intracellular iron levels due to FPN deletion were already stressed or primed for cell death. By using a lower dose of folic acid, we could examine a window where significant acute injury occurred in the FPN mutants with enough survival so that we can address the ability of injured tubules to repair. Our findings show that excess iron in proximal tubule epithelial cells impairs proliferation, regeneration, and recovery from AKI and promotes ferroptosis.

The kidney is highly metabolic and rich in mitochondria, with 20–25% of cardiac output going directly to the kidneys (33). As mitochondria are also rich in iron and the main source of reactive oxygen species, this makes the proximal tubules particularly sensitive to iron overload and oxidative damage. Increased iron concentrations in the kidney and urine are also associated with the onset and progression of CKD (34). However, anemia is a common complication of CKD and often requires iron supplementation, especially in hemodialysis patients (35, 36). Given that excess iron can exacerbate oxidative stress leading to ferroptosis and tissue damage, optimizing intracellular iron levels in the nephron is essential to maintain homeostasis. Iron chelators have been shown to protect against AKI (37), suggesting that iron overload can contribute to AKI to increase damage and prevent recovery. Furthermore, serum hepcidin levels correlate with the severity of CKD and may affect iron transport in the proximal tubules and vulnerability to AKI (38). That interpretation is consistent with our observations that FPN mutations leading to proximal tubule iron overload is an impediment to recovery and repair after folic acid-induced nephrotoxicity. With increased cell death after AKI, the ability to repair becomes limited and results in loss of nephrons, interstitial fibrosis, and progression toward CKD.

Among the many forms of cell death, ferroptosis has become increasingly relevant due to its association with a variety of pathological conditions including AKI (28). With their high level of metabolic activity and abundant mitochondria, proximal tubules are particularly sensitive to redox imbalance and the oxidation of polyunsaturated fatty acids, which promote ferroptosis. Among the regulators of ferroptosis is NCOA4, which binds iron via ferritin and delivers it to lysosomes (30). Glutathione peroxidases (GPX1 and GPX4) catalyze the reduction of lipid peroxides and work to protect cells from oxidative stress and ferroptosis (39). Genetic ablation of GPX4 is sufficient to drive ferroptosis in the kidney and causes acute renal failure and death (28). Gpx4 and NCOA4 are downregulated in our AKI model after 28 days; this together with high levels of 4HNE all suggest a proferroptotic environment that can account for the reduced recovery and regeneration. Although the deletion of FPN and the high levels of intracellular iron in mutant proximal tubules may not be sufficient to induce AKI, the additional stress of nephrotoxic cell death can exacerbate damage due to the release of free radicals from dead epithelial cells. Regeneration of proximal tubules after AKI is driven by surviving cells that dedifferentiate and reenter mitosis (21). Thus, if FPN mutant-surviving cells find themselves in an environment where free radicals are more abundant, their propensity for ferroptosis is increased and thus recovery may be impaired.

Perspectives and Significance

The renal proximal tubules are among the most metabolically active cells in the body and thus are particularly sensitive to oxidative stress. These kidney epithelial cells also regulate iron levels through resorption from the glomerular filtrate and transport into the plasma. We demonstrate that disruption of iron homeostasis by deletion of FPN in renal proximal tubule cells does not directly impact kidney function but sensitizes the kidney to acute injury and cell death due to ferroptosis. Given that iron supplementation is often used to treat anemia associated with CKD, the implication of our study points to the need to precisely control iron levels so that iron overload does not exacerbate the progression toward end-stage renal failure.

DATA AVAILABILITY

Data will be made available upon reasonable request.

GRANTS

This work was supported in part by National Institutes of Health (NIH) Grants DK054740 and DK073722 to G.R.D., DK125776 to J.A.B., and CA148828, CA245546, and DK095201 to Y.M.S. and a Fulbright scholarship to A.S. This work was also supported by a generous gift from Audrey and Josh Rumsey in memory of Isaiah.

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

A.S. and G.R.D. conceived and designed research; A.S., V.L., J.A.B., S.A., M.H., and N.K.D. performed experiments; A.S., V.L., J.A.B., S.A., M.H., N.K.D., Y.M.S., and G.R.D. analyzed data; A.S., J.A.B., M.H., N.K.D., Y.M.S., and G.R.D. interpreted results of experiments; A.S., V.L., and G.R.D. prepared figures; A.S. and G.R.D. drafted manuscript; M.H., Y.M.S., and G.R.D. edited and revised manuscript; N.K.D., Y.M.S., and G.R.D. approved final version of manuscript.