Heme Oxygenase-1 Inhibits Renal Tubular Macroautophagy in Acute Kidney Injury

Presence of Autophagic Vesicles in HO-1^−/− and HO-1^+/+ Kidneys during Cisplatin Injury In Vivo

We have previously shown that HO-1^−/− mice have worse structural (apoptosis and necrosis) and functional evidence of kidney injury after cisplatin treatment compared with HO-1^+/+ mice.¹¹ In this study, we examined the effects of cisplatin on autophagy in HO-1–deficient mice to evaluate if dysregulated autophagic responses contributed to increased cell death. Renal tissue examined by transmission electron microscopy (TEM) at 72 hours after saline or cisplatin administration showed that HO-1^+/+ and HO-1^−/− mice administered cisplatin had a substantial number of autophagosomes (double-membraned vesicles with cytoplasmic content) and autolysosomes (autophagosomes that have fused with lysosomes and appear as vacuoles) in the kidney cortex, specifically in the proximal tubule segments (Figure 1). Interestingly, HO-1^−/− mice that were administered saline had a higher number of autophagic vesicles (autophagosomes and autolysosomes) compared with their saline-administered HO-1^+/+ littermates (Figure 1).

During autophagy, soluble LC3-I is lipidated with the addition of phosphatidyl-ethanolamine to LC3-II, an insoluble form. LC3-II binds to the inner and outer membranes of the autophagosome, appearing as puncta in the cells. Renal sections from HO-1^+/+ mice administered saline exhibited few tubules containing punctated LC3 expression, whereas renal sections from HO-1^+/+ administered cisplatin exhibited a modest increase in the number of tubules containing punctated LC3 expression, consistent with previous studies using a similar model of renal injury.¹³ Similar to our TEM findings, HO-1^−/− mice administered saline had a significantly higher level of punctated LC3-II expression compared with saline-treated HO-1^+/+ mice (Figure 2; Supplemental Figure 1). However, after cisplatin administration, no further increases in LC3 puncta were noted in HO-1^−/− mice.

Progression of Autophagy in PTC during Cisplatin Injury

PTC were isolated from the HO-1^+/+ and HO-1^−/− mice as described in the methods. The PTC were characterized by staining for γ-glutamyl transpeptidase (proximal tubule specific marker) and alkaline phosphatase (epithelial marker) (Supplemental Figure 2). PTC from HO-1^+/+ and HO-1^−/− mice were treated with 50 μM cisplatin for varying times and gene or protein expression was analyzed by real-time PCR or Western blot analysis, respectively. In the HO-1^+/+ PTC, expression of autophagy genes, Atg5 and Atg7, was induced after cisplatin. However, HO-1^−/− PTC failed to illicit a significant induction of either Atg5 or Atg7 after cisplatin treatment (Figure 3, A, and B). Similarly, HO-1^+/+ PTC responded to cisplatin with an increase in Atg5, beclin, and LC3-II protein levels, whereas HO-1^−/− PTC did not exhibit an increase above basal levels (Figure 4, A and B). Also, basal levels of LC3-II and beclin were higher in HO-1^−/− PTC compared with HO-1^+/+ PTC (Supplemental Figure 2B). HO-1^−/− PTC treated with cisplatin had significantly higher levels of cleaved caspase-3 compared with HO-1^+/+ PTC, indicating increased apoptosis in HO-1^−/− PTC (Figure 4C).

To further characterize the role of HO-1 in regulating autophagy, we used PTC cultured from HO-1^−/− mice in which HO-1 expression was restored (HBAC mice). These mice were modified to express the human HO-1 gene so that they could be easily distinguished from the HO-1^+/+ mice (Kim J, Agarwal A, in preparation). As expected, HBAC PTC did not express mouse HO-1 (Supplemental Figure 3A) and, furthermore, had a significant increase (>45-fold) in human HO-1 gene expression (Supplemental Figure 3B). HBAC PTC also had an increase (>3-fold) in HO-1 protein levels compared with wild-type PTC (Supplemental Figure 3, C and D). PTC from HBAC mice responded to cisplatin in a similar manner to the HO-1^+/+ PTC, with induction of Atg5 and LC3-II (Figure 5A), accompanied by a decrease in cleaved caspase-3 levels (Figure 5B; Figure 6B). In addition, the high basal level of autophagy seen in the HO-1^−/− PTC was attenuated. Figure 6C shows a correlative analysis of the time-dependent shift from autophagy to apoptosis in HO-1^−/−, HO-1^+/+, and HBAC PTC treated with cisplatin. HO-1^−/− PTC show no increase in autophagy and instead exhibit higher levels of cleaved caspase-3 and undergo apoptosis more readily compared with HO-1^+/+ and HBAC PTC. These results suggest that expression of HO-1 is capable of rescuing the HO-1^−/− PTC from impaired autophagic responses and apoptosis, further validating our finding that HO-1 is a key regulator of autophagy and apoptosis.

HO-1^+/+ and HO-1^−/− PTC were transfected with a green fluorescence protein (GFP)-LC3 fusion plasmid and treated with 50 μM cisplatin and examined by fluorescence microscopy for LC3 expression. Change in LC3 tagged with GFP (GFP-LC3) expression from diffuse to a punctated form is indicative of autophagosomes (Supplemental Figure 4). A time-dependent increase in punctated expression of GFP-LC3 in HO-1^+/+ cells treated with cisplatin was observed (Supplemental Figure 4). HO-1^−/− PTC had a significantly higher number of cells with punctated expression basally that did not increase further upon cisplatin treatment (Supplemental Figure 4).

Inducible HO-1 Overexpression Protects Cells from Cisplatin-Mediated Injury

To further validate the protective role of HO-1 in cisplatin-mediated cell injury, HO-1 overexpressing (pIND/HO-1) and receptor control (RXR) cells were generated using the ecdysone inducible system. pIND/HO-1 cells responded to inducer, ponasterone A (Pon A), with an increase in HO-1 mRNA (data not shown), and protein, in a dose- and time-dependent manner (Figure 7, A and B). As expected, RXR cells did not have an increase in HO-1 protein expression above basal levels upon Pon A treatment and served as a control. Pon A–induced pIND/HO-1 cells also demonstrated a significant increase in HO activity (>3-fold) compared with RXR cells, confirming that the overexpressed HO-1 protein was functional (Figure 7C).

RXR and pIND/HO-1 cells were induced with Pon A (5 μM), exposed to hemin (50 μM) or cisplatin (100 μM) for 24 hours, and cytotoxicity was assessed by trypan blue exclusion assay. HO-1 overexpressing cells had a significant increase in cell viability compared with RXR cells that were exposed to hemin or cisplatin (Figure 7D). These results, based on the use of an inducible expression system, confirm our previous data with constitutive HO-1 overexpression demonstrating a cytoprotective role for HO-1 in cisplatin-mediated cell death.¹¹

Delayed Progression of Autophagy in HO-1 Overexpressing Cells

To evaluate the role of HO-1 in cisplatin-mediated autophagy, RXR and pIND/HO-1 cells were transfected with a GFP-LC3 fusion plasmid, induced with Pon A and treated with 50 μM cisplatin, and examined by fluorescence microscopy for change in LC3 expression pattern. GFP-LC3 puncta formation was evident in both RXR and pIND/HO-1 cells treated with cisplatin (Figure 8A). Quantitation demonstrated a time-dependent increase in punctated expression of GFP-LC3 in RXR cells treated with cisplatin. In contrast, pIND/HO-1 cells had a significant delay in puncta formation (autophagosome) upon cisplatin treatment compared with RXR cells (Figure 8B).

Inhibition of Reactive Oxygen Species Generation in HO-1 Overexpressing Cells during Cisplatin Injury

Cisplatin mediates cytotoxicity through the generation of reactive oxygen species (ROS). To evaluate the role of HO-1 overexpression on ROS generation, RXR and pIND/HO-1 cells were induced with Pon A (5 μM) and treated with cisplatin (50 μM). RXR cells had an increase in ROS generation after cisplatin treatment, as evident by an increase in 2′,7′-dichlorodihydrofluorescein (DCF) fluorescence (Figure 9A). However, in addition to reduced cell death (Figure 7D), pIND/HO-1 cells had a significant inhibition of ROS generation after cisplatin treatment compared with RXR cells (Figure 9B).

Animals

HO-1^−/− and HO-1^+/+ mice [8 to 12 weeks of age, (C57BL/6xFVB)F_n] were used for both in vitro and in vivo studies. A humanized HO-1 transgenic mouse line (HBAC), expressing only the human HO-1 gene, were generated using a human bacterial artificial chromosome clone (Kim J, Agarwal A, in preparation). Briefly, an 87-Kb bacterial artificial chromosome DNA from chromosome 22 (accession no. z82244) containing the human HO-1 gene was microinjected into oocytes from C57BL/6 mice, which were then implanted into psuedopregnant female mice. After birth, potential founders were screened for the presence of the transgene using PCR with human HO-1 specific primers. Animals positive for the transgene were then mated with HO-1^−/− mice to generate human HO-1–expressing HBAC mice that were screened using PCR for the presence of the human HO-1 gene and the absence of the mouse HO-1 gene. HBAC mice were characterized for the expression of mouse and human HO-1 mRNA transcripts in the kidney by real-time PCR. HBAC mice (8 to 14 weeks of age) were used to generate primary renal proximal tubular cells.

Generation of Proximal Tubule Epithelial Cells from Mice

Primary PTC cultures were generated from the kidneys of HBAC, HO-1^−/−, and their wild-type littermates using a previously described procedure with minor modifications.^47,48 Kidney cortices from mice were dissected from the medulla, sliced, minced, and filtered through a 70-μm cell strainer over a 50-ml conical tube with media (Renal Epithelial Cell Growth Medium; PromoCell). The medium containing the tubules was centrifuged and plated on collagen-coated culture plates and then incubated for 72 hours at 37°C in 5% CO₂. Each experiment required PTC generated from a single animal and cells were not passaged. Cells were examined for expression of γ-glutamyltranspeptidase and alkaline phosphatase, markers for proximal tubule cells, as described previously.⁴⁷

Cisplatin Injury Model

HO-1^−/− and HO-1^+/+ mice were administered 20 mg/kg body wt cisplatin (1.0 mg/ml solution in sterile normal saline) or vehicle (normal saline) by a single intraperitoneal injection. Mice were sacrificed 24 and 72 hours after cisplatin administration. Kidneys were harvested, cut transversely, and fixed in 10% neutral buffered formalin for immunohistochemistry. The animal protocol was approved by the Institutional Animal Care and Use Committee at the University of Alabama at Birmingham. For in vitro experiments, 50 μM cisplatin was added to PTC or HEK293 cells in culture medium and analyzed at indicated time points by real-time PCR or Western blot analysis.

Electron Microscopy

Renal tissues were fixed at room temperature with a fixative containing 2.5% glutaraldehyde, 1% formaldehyde, 100 mM sodium phosphate, pH 7.2. Samples were rinsed in 0.1 M Na cacodylate (pH 7.4), then postfixed for 1 hour in 2% osmium tetroxide in 0.1 M Na cacodylate, and washed in 0.1 M Na cacodylate. The samples were dehydrated in a graded series of ethanols and propylene oxide and embedded in epoxy resin (Taab 812 Resin; Marivac Industries, Montreal, CA). Ultrathin (60 to 70 nm) sections were counterstained with uranyl acetate and lead citrate and then observed using an Hitachi 7600 transmission electron microscope (Hitachi High-Technologies America, Schaumburg, IL) equipped with a Macrofire monochrome progressive scan CCD camera (Optronics, Goleta, CA) and AMTv image capture software (Advanced Microscopy Techniques, Danvers, MA).

Immunofluorescence

Paraffin-embedded kidney sections were cut at 4-μm thickness and deparaffinized using CitriSolv. Citrate antigen retrieval was performed followed by incubation in 3% hydrogen peroxide solution to block endogenous peroxidase activity. Sections were incubated for 1 hour at room temperature in blocking buffer (1% BSA, 0.2% nonfat dry milk, and 0.3% Triton X-100 in PBS) and then exposed to primary antibody (Ab), rabbit anti-LC3B (1:5000), overnight at 4°C. Sections were washed and incubated for 1 hour at room temperature with secondary Ab, SuperPicTure goat anti-rabbit (1:20 dilution; Invitrogen, Carlsbad, CA). Ab binding was detected using TSA Plus Cyanine 3 System and visualized by fluorescence microscopy. Quantitation was performed on slides that were blinded to the reviewer using at least 3 to 6 different fields per time point with an average of 25 renal proximal tubules per field from at least 3 to 6 animals per group.

Real-Time PCR Analysis

RNA was isolated from cells using TRIzol (Invitrogen) according to the manufacturer's protocol. Three micrograms of RNA was used to convert to cDNA using the First Strand cDNA Synthesis Superscript kit (Invitrogen) according to the manufacturer's protocol. Quantitative real-time PCR was performed with SYBR Green Mastermix (Invitrogen) and 15 pmol primers for mouse HO-1, human HO-1, Atg5, Atg7, and GAPDH. Reactions were performed in triplicate and specificity was monitored using melting curve analysis after cycling. Primers used were as follows (5′-3′): mouse HO-1, 5′-GGTGATGGCTTCCTTGTACC-3′ and 5′-AGTGAGGCCCATACCAGAAG-3′; human HO-1, 5′-CATGACACCAAGGACCAGAG-3′ and 5′-AGTGTAAGGACCCATCGGAG-3′; Atg5, 5′-GACGTTGGTAACTGACAAAGTGAAAAAGCA-3′ and 5′-CCAAGGAAGAGCTGAACTTGATGCAAG-3′; Atg7, 5′-CAATCTGGGCTAAATGCCATTTCTGGAAG-3′ and 5′-AGCCCAGTACCCTGGATGG-3′; GAPDH, 5′-TCCCACTCTTCCACCTTCGA-3′ and 5′-AGTTGGGATAGGGCCTCTCTTG-3′. Relative mRNA expression was quantified using the ΔΔ Ct method and GAPDH was used as an internal control. Results were expressed as fold change over untreated cells.

Western Blot Analysis

Immunoblot analysis was performed as described previously.⁴⁷ Briefly, cell cultures were lysed in RIPA buffer, electrophoresed in a 15% SDS-polyacrylamide gel, and transferred onto a Hybond C Extra membrane (Amersham Biosciences). Membranes were incubated with anti–HO-1 (1:5000 dilution; Stressgen), anti-LC3 (1:2000 dilution, Sigma-Aldrich), anti–cleaved caspase-3 (1:2000 dilution; Cell Signaling), anti-beclin (1:2000 dilution; Santa Cruz Biotechnology), or anti-Atg5 (1:5000 dilution; Millipore) antibodies, followed by a peroxidase-conjugated goat anti-rabbit (or mouse) IgG antibody (1:10000 dilution; Jackson ImmunoResearch Laboratories, West Grove, PA). Horseradish peroxidase activity was detected using enhanced chemiluminescence. The membrane was stripped and probed with anti-actin antibody (1:5000 dilution; Sigma-Aldrich) to confirm loading and transfer. Results were normalized to actin or LC3-I and expressed as fold change over untreated cells.

Plasmids and Generation of Inducible HO-1 Overexpressing HEK293 cells

pVgRXR regulatory vector and pIND-inducible vector (ecdysone-inducible mammalian expression system; Invitrogen) were used to generate inducible HO-1 overexpressing HEK293 cells. The entire protein-coding region of the human HO-1 gene was cloned into the EcoRI and XbaI sites of the pIND vector to generate the pIND/HO-1 inducible vector. Pon A (Invitrogen) was used to induce expression of genes under the regulation of the ecdysone-inducible expression system.

HEK293 cells were transfected with pVgRXR regulatory vector using lipofectamine 2000 (Invitrogen) and stable clones were selected with the addition of zeocin (400 μg/ml; Invitrogen). Individual clones were isolated and maintained with 100 μg/ml zeocin. Inducible HO-1 overexpressing HEK293 cells were generated by stably transfecting pVgRXR cells with the pIND/HO-1 plasmid and selected with the addition of geneticin (600 μg/ml). Individual clones were isolated and maintained with 100 μg/ml zeocin and 200 μg/ml geneticin. HEK293 cells stably transfected with pVgRXR are referred to as RXR cells and serve as control. HEK293 cells stably transfected with pVgRXR and pIND/HO-1 vector are referred to as pIND/HO-1 cells. Stable cells were maintained in Dulbecco's Minimum Essential Medium supplemented with fetal bovine serum (5%), glutamine, antibiotic-antimycotic (1×), and HEPES (25 mM) at 37°C in 90% air–10% carbon dioxide. To overexpress HO-1, pIND/HO-1 cells were plated to 70% confluency, serum-starved, and induced with 5 μM Pon A for 24 hours. RXR cells treated with 5 μM Pon A served as controls.

The GFP-LC3 plasmid was a generous gift from Dr. Gur Kaushal (University of Arkansas for Medical Sciences, Little Rock, Arkansas). Transfections were performed with lipofectamine 2000 (Invitrogen) according to the manufacturer's protocol.

HO Enzyme Activity Measurement

HO activity was measured by bilirubin generation as described previously.^3,49 Briefly, cell lysates were centrifuged at 10,000g and supernatant was incubated with rat liver cytosol (3 mg), a source of biliverdin reductase, hemin (20 μM), glucose-6-phosphate (2 mM), glucose-6-phosphate dehydrogenase (0.2 μ), and NADPH (0.8 mM) for 1 hour at 37°C in the dark. The bilirubin formed was extracted with chloroform; Δoptical density of 464 to 530 nm was measured and enzyme activity was calculated as picomoles of bilirubin formed per 60 minutes per milligram of protein. Fold change in HO activity was determined by comparing the activity of HO-1 overexpressing cells to RXR control cells.

Cell Viability Assay

Cells were plated in 24-well plates and allowed to reach 70% confluency, serum-starved, and treated with 5 μM Pon A for 24 hours. Cells were incubated with 50 μM hemin or 100 μM cisplatin for 24 hours. Adherent cells were lifted gently by treatment with diluted trypsin and floating cells collected from the media. Cell suspensions were incubated with equal volumes of trypan blue for 5 minutes and observed under a microscope. The viable cells that excluded the dye and dead cells that did not were counted and cell viability was expressed as % live cells/total number of cells.

Detection of Autophagosomes by GFP-LC3

Primary mouse PTC cultures and HO-1 overexpressing or RXR control HEK293 cells were plated on chamber slides, grown to 60% confluency, and transfected with pGFP-LC3 (0.5 μg/ml) using lipofectamine 2000 (Invitrogen) according to manufacturer's protocol. PTC were allowed to reach 80% confluency and treated with 50 μM cisplatin for the times indicated. pIND/HO-1 and RXR cells were induced with Pon A for 24 hours, after which they were treated with cisplatin (50 μM) for the detection of autophagosomes. Cells were fixed at the end of incubation in cold methanol, mounted, and observed under an inverted fluorescence microscope. Cells with diffuse and punctated GFP-LC3 were counted and autophagosome formation was expressed as % punctated GFP-LC3 expression/total number of cells expressing GFP-LC3. Quantitation was performed in a blinded fashion on at least 12 different fields per time point with an average of 50 cells from at least three independent experiments. Confocal imaging was performed to emphasize the expression pattern of GFP-LC3 after cisplatin treatment.

Measurement of ROS

ROS generation was measured as described previously.⁵⁰ Briefly, RXR and pIND/HO-1 cells were plated on collagen-coated chamber slides, induced with Pon A, and treated with 50 μM cisplatin for the times indicated. Cells were loaded with 2.5 μM 2′,7′-2′,7′-dichlorodihydrofluorescein diacetate [H₂DCF-DA (Invitrogen)] for 1 hour before imaging. Cells were imaged with a laser-scanning confocal scan head (Olympus FV300) mounted on an inverted fluorescence microscope (Olympus IX71) equipped with oil immersion objectives (Olympus 60× 1.2NA). DCF fluorescence was excited with an Ar laser (488 nm) and its emission filtered at 510 ± 20 nm. Single images were scanned at 512 × 512 pixel resolution (1.96 seconds per frame), avoiding PMT saturation (400 to 500 V PMT voltage, 1× gain, zero offset), saved as TIFF files, and analyzed using ImageJ (National Institutes of Health). The average pixel intensity within a single region of interest defined around individual cells in DCF dye fluorescence images was used for quantitative estimation of ROS levels. Fluorescence intensity was measured in 8 fields and at least 40 different regions of interest per field and averaged from four independent experiments. Results are expressed as fold increase in DCF fluorescence (ROS) in cisplatin-treated cells compared with untreated cells.

Statistical Analysis

Data are presented as mean ± SEM. The t test was used for comparisons between two groups. For the comparisons that involved more than two groups, ANOVA and the Newman-Keuls test were used for analysis, with statistical significance considered at P < 0.05.