Meprin-β activity modulates the β-catalytic subunit of protein kinase A in ischemia-reperfusion-induced acute kidney injury

Abstract

Meprin metalloproteases have been implicated in the progression of kidney injury. Previous work from our group has shown that meprins proteolytically process the catalytic subunit of protein kinase A (PKA-C), resulting in decreased PKA-C kinase activity. The goal of the present study was to determine the PKA-C isoforms impacted by meprin-β and whether meprin-β expression affects downstream mediators of the PKA signaling pathway in ischemia-reperfusion (IR)-induced kidney injury. IR was induced in 12-wk-old male wild-type (WT) and meprin-β knockout (βKO) mice. Madin-Darby canine kidney cells transfected with meprin-β cDNA were also subjected to 2 h of hypoxia. Western blot analysis was used to evaluate levels of total PKA-C, PKA-Cα, PKA-Cβ, phosphorylated (p-)PKA-C, and p-ERK1/2. Meprin-β expression enhanced kidney injury as indicated by levels of neutrophil gelatinase-associated lipocalin and cystatin C. IR-associated decreases were observed in levels of p-PKA-C in kidney tissue from WT mice but not βKO mice, suggesting that meprin-β expression/activity is responsible for the in vivo reduction in kinase activity. Significant increases in levels of PKA-Cβ were observed in kidney lysates for WT mice but not βKO mice at 6 h post-IR. Proximal tubule PKA-Cβ increases in WT but not βKO kidneys were demonstrated by fluorescent microscopy. Furthermore, IR-induced injury was associated with significant increases in p-ERK levels for both genotypes. The present data demonstrate that meprin-β enhances IR-induced kidney injury in part by modulating mediators of the PKA-Cβ signaling pathway.

INTRODUCTION

Ischemia-reperfusion (IR)-induced injury is a leading cause of acute kidney injury (AKI) and is associated with high morbidity and mortality rates (39). The pathophysiology of AKI includes multiple cellular events that lead to damage to renal proximal tubules. This results in disorganization of the apical actin cytoskeleton and redistribution of actin and other proteins, collapse of brush-border microvilli, breakdown of tight and adherent junctions, loss of cell polarity, and loss of cell matrix adhesions, ultimately causing cell detachment into the tubular lumen (32). These alterations may lead to cell death of proximal tubular epithelial cells by apoptosis and necrosis (32). The molecular mechanisms of kidney injury are not fully understood. However, several causal factors, such as reactive oxygen species, neutrophil infiltration, vasoactive peptides, and ATP depletion, have been reported to contribute to the pathogenesis of IR-induced AKI (17). Meprins are zinc metalloproteinases that are abundantly expressed in brush-border membranes (BBMs) of kidney proximal tubules and the small intestine (52). Meprins are made up of two subunits, α and β, expressed by two distinct genes. The expression of these two subunits results in two protein isoforms: meprin A, which exists as a homodimer (α-α) or heterodimer (α-β), and meprin B, which exists as a homodimer (β-β) (52). Several studies have shown that meprins play a role in the pathogenesis of AKI and chronic kidney injury (9, 11, 53). Meprin-β expression is associated with enhanced IR-induced AKI, while meprin inhibitors or disruption of the meprin-β gene were shown to provide some protection against IR-induced kidney injury (9, 26). While the mechanisms by which meprins modulate IR-induce kidney injury are not fully understood, meprins are redistributed from the BBM to the cytosolic and basolateral compartments in IR-induced kidney injury (9, 43). This redistribution would bring meprins in close proximity with proteins in these cell compartments. Several studies have shown that meprin proteolytically processes and/or degrades several proteins present in kidney tissue. These include extracellular matrix (ECM) proteins (e.g., collagen type IV, laminin, nidogen-1, and fibronectin) (33), cytoskeletal proteins (actin and villin) (43), tight junction proteins (E-cadherin and occludin) (5, 27), and several inflammation modulators (IL-18, IL-1β, IL-6, and macrophage chemoattractant protein-1) (3, 24, 25, 34). Meprins also interact with proteins involved in cell signaling pathways, e.g., the catalytic subunit of protein kinase A (PKA-C) (14, 42) and protein kinase C (8). In vitro experiments have shown proteolytic processing of PKA-C by meprins is isoform specific and produces PKA-C fragments with decreased kinase activity (1, 14, 42). Furthermore, levels of total PKA-C in proximal kidney tubules were shown to dramatically decrease in kidney tissue from meprin-expressing mice at 3 h post-IR. In contrast, PKA-C levels in distal tubules, which do not express meprins, increased. However, the PKA-C isoforms mediating IR-induced kidney injury are not known. Furthermore, the effects of meprin-PKA-C interactions on downstream targets of the PKA signaling pathway are not known. The PKA signaling pathway plays a role in the hypoxia response, but it is not known whether meprin activity modulates the expression of hypoxia response genes. The objective of the present study was to determine the PKA-C isoform(s) modulated by meprin-β in IR and how meprin-β expression affects the levels of downstream targets of the PKA signaling pathway on kidneys subjected to IR injury and in in vitro cultured kidney cell lines subjected to hypoxia.

MATERIALS AND METHODS

Experimental animals.

Two genotypes of mice on a C57BL/6 background were used: 1) wild-type (WT) mice, which express normal levels of both meprin A and meprin B, and 2) meprin-β knockout (βKO) mice, in which the meprin-β gene was disrupted. βKO mice are deficient in meprin B (β-β) and the heterodimeric form of meprin A (α-β). βKO mice were initially generated in the laboratory of Dr. Judith Bond (Pennsylvania State University) and bred at the Laboratory Animal Resource Unit of North Carolina A&T State University. Age-matched WT mice were purchased from Charles River Laboratories (Wilmington, MA). Mice were housed in standard cages with a maximum of 5 mice/cage with a 12:12-h light-dark cycle. Mice were fed standard rat chow (Purina, St. Louis, MO) and water ad libitum. All animal protocols were approved by the North Carolina A&T State University Institutional Animal Care and Use Committee.

Induction of AKI.

AKI was induced in 12-wk-old male mice using surgical procedures to perform unilateral nephrectomy and followed by clamping of the renal pedicle of the contralateral kidney for 27 min as previously described (42). After reperfusion for 6 h, animals were euthanized by CO₂ asphyxiation, and the remaining kidney was excised. The contralateral kidneys harvested at 0 h served as the control for proteomics and immunohistochemical analysis.

Processing of blood and kidney tissues.

Blood samples were collected into lithium/heparin prechilled tubes (Sarstedt, Newton, NC). Before surgery, baseline blood sample (0 h) were obtained by tail nicking. At the time of mouse euthanization, blood was collected by cardiac puncture. Blood samples were centrifuged at 10,000 g for 10 min at 4°C to obtain plasma, which was aliquoted and stored at −80°C until used for biochemical assessment of kidney injury. The kidneys were excised and decapsulated, and sections of each kidney processed for protein extraction, RNA extraction, and immunohistochemistry. Kidney tissue samples for protein extraction were wrapped in aluminum foil, snap frozen in liquid nitrogen, and stored at −80°C. A 2-mm midsection tissue sample for immunohistochemistry were stored in Carnoy’s fixative (60% ethanol-30% chloroform-10% acetic acid) overnight at 4°C and then transferred to 70% ethanol until processed for paraffin embedding. Paraffin embedding and cutting tissue sections onto slides was performed at the Wake Forest University Pathology Laboratory.

Biochemical assessment of kidney function.

To confirm kidney injury, ELISAs were used to determine plasma levels of creatinine, cystatin C, and neutrophil gelatinase-associated lipocalin (NGAL). Cystain C and NGAL have been shown to be sensitive biomarkers of kidney injury and are suitable for use in the early time points (12, 20, 22, 29–31, 41, 46). The ELISA used commercial kits specific for mouse proteins; Diazyme Laboratories (Poway, CA) for creatinine and R&D Systems (Minneapolis, MN) for cystatin C and NGAL. All assays followed the manufacturers’ instructions. Absorbance was read using an ELISA plate reader (Tecan Infinite 200 PRO, Tecan Austria). GraphPad Prism 7.0 software was used to generate standard curves and extrapolate values and for data analysis.

Immunohistochemical analysis.

Immunohistochemistry used previously described protocols (42, 43). In brief, slide sections of kidney tissue were deparaffinized through xylene, 100% ethanol, 95% ethanol, and water. Antigen unmasking was achieved by boiling in 10 mM sodium citrate buffer (pH 6.0) for 10 min. Nonspecific binding sites were blocked by incubation in 5% normal goat serum with 0.3% Triton X-100 at room temperature for 1 h in a humidified chamber. Primary antibodies were diluted in PBS with 1% BSA and 0.3% Triton X-100. Sections were incubated in primary antibodies overnight at 4°C or at room temperature for 1 h. Rabbit polyclonal anti-meprin B antibodies (HMC77) were a gift from Dr. Judith Bond (Penn State Hershey Medical Center) and were diluted at 1:200. Mouse monoclonal anti-PKA-C antibodies (BD BioSciences, San Jose, CA) were diluted 1:200. Rabbit polyclonal anti-PKA-Cα (Cell Signaling, Danvers, MA) and anti-PKA-Cβ (Abcam, Cambridge, MA) antibodies were diluted 1:200. Mouse monoclonal anti-villin antibodies (Santa Cruz Biotechnology, Dallas, TX) were diluted 1:200. Slides were then rinsed three times in PBS for 10 min each and incubated in fluorophore-conjugated secondary antibodies: goat anti-rabbit Alexa Fluor 488 for meprin B, PKA-Cα, and PKA-Cβ and goat anti-mouse Alexa Fluor 555 for PKA-C and villin (Cell Signaling) diluted 1:1000 for 1 h at room temperature. We used goat anti-mouse antibodies to avoid nonspecific signals on mouse tissues. DAPI was used for nuclear staining. Coverslips with Prolong antifade reagent (Life Technologies, Carlsbad, CA) were mounted and allowed to dry at room temperature overnight. Tissue sections were evaluated for meprins and PKA-C expression and localization using a BZ-X700 Series all-in-one fluorescence microscope (KEYENCE Corporation of America, Elmwood, NJ) and imaged using BZ-X700 analysis software.

Protein extraction from kidney tissues.

Kidney tissues were thawed on ice and homogenized in 9 volumes of homogenization buffer [0.02 mM HEPES (pH 7.9), 0.015 mM NaCl, 0.1 mM Triton X-100, 0.01 mM SDS, and 1 mM Na₃VO₄) with protease and phosphatase inhibitors. Sequential centrifugation was used to fractionate protein lysates into cytosolic-, nuclear-, and BBM-enriched protein fractions as previously described (42, 43). RIPA buffer was used to obtain nonfractionated total protein lysates from sections of kidney tissue. Protein concentrations were determined by the Bradford protein assay method using Bio-Rad’s protein assay reagent (Bio-Rad, Hercules, CA) and stored in aliquots at −80°C until analyzed by Western blot.

In vitro cell culture.

Madin-Darby canine kidney (MDCK) cells were originally purchased from the American Type Culture Collection (Manassas, VA). MDCK cells were stably transfected with meprin-β cDNA as previously described (40). MDCK cells are not a mixture of proximal and distal tubules; however, MDCK cells transfected with meprin-β cDNA are expected to behave as proximal tubules as far as meprin expression is concerned, whereas mock-transfected MDCK cells behave more like distal tubules. This cell line has been extensively used for in vitro studies pertaining to meprins (5, 40). MDCK cells were propagated in DMEM supplemented with 10% FBS and antibiotics, subcultured, and preserved in liquid nitrogen.

Induction of hypoxia.

MDCK cells stably transfected with meprin-β cDNA and sham-transfected control cells were cultured to 90–95% confluence. MDCK cells were subjected to hypoxia (1% O₂) using hypoxia chambers (Biospherix, Parish, NY) for 2 h or for 2 h under hypoxic conditions followed by 2 h of recovery at 21% O₂. Control cells were cultured at 21% O₂. Proteins were extracted from the cells and fractionated into cytosolic- and nuclear-enriched fractions as previously described (40). Proteins concentrations were determined by the Bradford protein assay method using Bio-Rad’s protein assay reagent and stored in aliquots at −80°C until analyzed by Western blot. Western blot analysis was used to determine levels of hypoxia-inducible factor (HIF)-1α in nuclear-enriched protein lysates to confirm the induction of hypoxia. When we commenced the in vitro experiments, we reviewed the existing literature and could not find in vitro studies that could exactly mimic the in vivo time points that we were using, i.e., 27-min ischemia followed by 6-h reperfusion. Furthermore, cells cultured in vitro take time to equilibrate to the 1% O₂ levels. We therefore first obtained preliminary data for various time points. The data we obtained for HIF-1α Western blots led us to select 2 h for exposure. It may not perfectly mimic the in vivo data, but we believe that the time points reflect the early time points associated with AKI in vivo.

Western blot analysis.

Western blot analysis was used to evaluate protein levels of total PKA-C, PKA-Cα, PKA-Cβ, phosphorylated (p-)PKA-C, p-ERK1/2, and HIF-1α from kidney lysates and cell lysates. Equal amounts (60−90 µg) of kidney protein lysates or lysates from in vitro cultured cells were loaded onto a 12% SDS-PAGE gels and separated by electrophoresis. The separated proteins were transferred into nitrocellulose membranes (Bio-Rad Laboratories). To block nonspecific binding sites, membranes were incubated in 5% fat-free milk in Tris-buffered saline with 0.1% Tween 20 (TBS-T) for 1 h at room temperature or overnight at 4°C. Primary antibodies were added, and the membranes were incubated for 1 h at room temperature or overnight at 4°C. The antibodies used were mouse monoclonal anti-PKA-C (BD BioSciences, San Jose, CA), rabbit monoclonal anti-PKA-Cα (Cell Signaling), rabbit monoclonal anti-PKA-Cβ (Abcam), rabbit monoclonal anti-p-PKA-C (Cell Signaling), rabbit monoclonal anti-p-ERK1/2 (Cell Signaling), and mouse polyclonal anti-HIF-1α (Abcam); all were diluted 1:1,000 in 5% milk in TBS-T. Anti-α-tubulin antibodies (Abcam) and anti-TATA box-binding protein (TBP; Cell Signaling) were used for loading controls. While using total proteins is a more accurate way of obtaining relative intensities for the phosphorylated proteins, we used α-tubulin and TBP because the sizes for p-PKA-C and PKA-C are pretty close, making it difficult to distinguish the residual signal from the p-signal. Following the primary antibody incubation, membranes were washed three times for 10 min each in TBS-T. Secondary antibodies, either mouse or rabbit IgG (diluted 1:10,000 in TBS-T) were added onto the membrane and incubated for 1 h at room temperature or overnight at 4°C. Membranes were then washed three times for 15 min each in TBS-T. To detect bands, membranes were exposed to chemiluminescence substrates (Thermo Scientific, Waltham, MA) and developed on X-ray film. Protein band intensities were determined by densitometry using a GS800 calibrated densitometer (Bio-Rad) and Quantity One software (Bio-Rad). Optical densities (ODs) for each protein band were normalized to the ODs of the housekeeping gene products (α-tubulin or TBP) to obtain the relative OD.

Statistical analysis.

Data were analyzed by two-way ANOVA using GraphPad 7.0 Prism software (GraphPad, La Jolla, CA). The densitometric data for Western blots were first normalized to the loading controls (α-tubulin and TBP). Data are presented as means ± SE. P values of ≤0.05 were considered statistically significant.

RESULTS

Meprin-β expression was associated with increased levels of kidney injury biomarkers at 6 h post-IR.

To confirm injury in kidneys subjected to IR, ELISAs were used to determine plasma levels of known kidney injury markers, namely, creatinine, cystatin C, and NGAL. There were no significant differences in plasma creatinine levels for either genotype at 6 h post-IR (Fig. 1A). However, this was not surprising since creatinine levels typically rise at later time points. Interestingly, baseline creatinine levels were significantly lower in βKO mice compared with their WT counterparts (P = 0.002). In contrast, significant increases were observed in cystatin C levels at 6 h post-IR in WT mice (P = 0.05; Fig. 1B). Similarly, NGAL levels were significantly higher at 6 h post-IR in both WT (P = 0.0001) and βKO mice (P = 0.0022; Fig. 1C). While baseline NGAL levels were higher in βKO mice, the fold change was greater in WT mice (9-fold vs. 2-fold). Compared with cystatin C, NGAL levels had the most differentiating power for mice subjected to IR (P = 0.0001 in WT mice and P = 0.002 in βKO mice). For in vitro cultured MDCK cells, induction of hypoxia was confirmed by evaluating the levels of HIF-1α in nuclear-enriched protein lysates. Hypoxia is known to cause stabilization of HIF-1α, resulting in higher nuclear levels. Western blot data demonstrated an increase in the levels of HIF-1α in MDCK cells exposed to 1% O₂ for 2 h (Fig. 1D). HIF-1α levels returned to prehypoxia levels after 2 h of recovery at 21% O₂.

Fig. 1.

Biomarkers of kidney injury and hypoxia. Blood samples (n = 4 per group) were obtained at 0 h [before ischemia-reperfusion (IR)] and 6 h post-IR injury, and ELISA was used to determine the levels of creatinine (A), cystatin C (B), and neutrophil gelatinase-associated lipocalin (NGAL; C). Data are means ± SE from 4 mice/group and were analyzed by two-way ANOVA (GraphPad Prism). Levels of hypoxia-inducible factor (HIF)-1α in nuclear-enriched Madin-Darby canine kidney (MDCK) cells (D) were used to confirm the induction of hypoxia. The relative optic densities (ODs) were calculated by normalizing the ODs of HIF-1α to the ODs for TATA box-binding protein (TBP) in the same blot. Data are means ± SE from at least 3 independent experiments. Data were analyzed by two-way ANOVA (GraphPad Prism). *P = 0.01; ****P = 0.0001.

Meprin-β expression was associated with isoform-specific alterations in the levels of PKA-Cβ in kidneys after IR injury.

To determine the impact of meprin-β expression/activity on PKA-C levels during IR-induced kidney injury, cytosolic-enriched kidney extracts were evaluated by Western blot analysis. First, anti-PKA-C antibodies, which are not isoform specific, were used and thus total PKA-C levels were detected. Immunoblot analysis detected one PKA-C protein band, running at 43 kDa. The relative intensities of PKA-C in cytosolic-enriched kidney proteins were significantly lower in WT mice subjected to IR (P = 0.03) but not in their βKO counterparts (Fig. 2A). However, in in vitro cultured MDCK cell lysates, levels of PKA-C in cytosolic- and nuclear-enriched fractions showed no significant differences following exposure to hypoxic conditions (Fig. 2B). A similar pattern was observed in levels of p-PKA-C in kidney proteins with significant decreases in cytosolic-enriched kidney proteins from WT mice (P = 0.03) but not in βKO mice (Fig. 3A). Anti-PKA-Cα- and anti-PKA-Cβ-specific antibodies were then used to determine whether levels of PKA-Cα and PKA-Cβ isoforms are changed in IR-induced kidney injury and hypoxia. There were no significant differences in the levels of PKA-Cα in either cytosolic- or nuclear-enriched kidney protein fractions or in cell lysates from MDCK cells (Fig. 4). In contrast, levels of PKA-Cβ were significantly increased in both cytosolic- and nuclear-enriched kidney protein fractions at 6 h post-IR injury (Fig. 5A). Two PKA-Cβ protein bands running at 41 and 37 kDa were detected in cytosolic-enriched fractions, with the 41-kDa protein species being significantly higher in WT kidney lysates at 6 h post-IR (P = 0.0001) but not in βKO kidney lysates. In contrast, the 37-kDa protein band decreased (P = 0.002) in βKO kidney lysates following IR. Nuclear-enriched fractions of kidney lysate showed a significant increase (P = 0.0001) in WT kidneys following IR. For in vitro cultured MDCK cells, immunoblots for the PKA-Cβ isoform in cytosolic- and nuclear-enriched fractions detected two protein bands running at 41 and 37 kDa (Fig. 5B). Interestingly, cytosolic levels of the 41-kDa PKA-Cβ protein species were significantly higher after the induction of hypoxia (P = 0.05) and after recovery from hypoxia (P = 0.0003). In sham-transfected MDCK cells, the 41-kDa species of the PKA-Cβ isoform was significantly increased (P = 0.003) after 2 h of exposure to hypoxic conditions and stayed elevated even after 2 h of recovery from hypoxia (P = 0.0007). In contrast, the 37-kDa protein species of the PKA-Cβ isoform showed no significant difference in both genotypes of MDCK cells. PKA-Cβ has been shown to have over 10 splice variants. We believe that the 37- and 41-kDa bands represent different PKA-Cβ splice variants. At this time, PKA-Cβ splice variant-specific antibodies are not available to enable us to determine the splice variants represented by these protein bands.

Fig. 2.

Representative immunoblots of the catalytic subunit of protein kinase A (PKA-C) in cytosolic-enriched kidney proteins from wild-type (WT) and meprin-β-deficient mice (n = 3 per group) at 0 h and 6 h post-ischemia-reperfusion (IR)-induced kidney injury (A) and Madin-Darby canine kidney (MDCK) cells subjected to 0 or 2 h of hypoxia or 2 h of hypoxia with 2 h of recovery (n = 3 per group; B). The protein bands represent samples from individual mice. The relative optic densities (ODs) shown in the dot plots were calculated by normalizing the ODs of PKA-C to the OD for α-tubulin and TATA box-binding protein (TBP) in the same blot. Data are means ± SE from 3 mic /group for the in vivo study and from 3 independent experiments for the in vitro study. Protein bands in the blots were run on the same gel, but the order was rearranged for final presentation. Data were analyzed by two-way ANOVA (GraphPad Prism). *P = 0.05.

Fig. 3.

Representative immunoblots of phosphorylated catalytic subunit of protein kinase A (p-PKA-C) in cytosolic- and nuclear-enriched kidney protein fractions from wild-type (WT) and meprin-β-deficient mice (n = 3 per group) at 0 h and 6 h post-ischemia-reperfusion (IR; A) and Madin-Darby canine kidney (MDCK) cells subjected to 0 h or 2 h of hypoxia or 2 h of hypoxia with 2 h of recovery (n = 3 per group; B). The optical density (OD) for the 42-kDa p-PKA-C protein bands were normalized to the ODs for α-tubulin and TATA box-binding protein (TBP) to obtain the relative ODs shown in the dot plot charts. Data are means ± SE from 3 mice/group for the in vivo study and from at least 3 independent experiments for the in vitro study. Data were analyzed by two-way ANOVA (GraphPad Prism). *P = 0.03.

Fig. 4.

Representative immunoblots of the catalytic subunit of protein kinase A (PKA-Cα) in cytosolic- and nuclear-enriched kidney protein fractions from wild-type (WT) and meprin-β-deficient mice (n = 3 per group) at 0 h and 6 h post-ischemia-reperfusion (IR; A) and Madin-Darby canine kidney (MDCK) cells subjected to 0 h or 2 h of hypoxia or 2 h of hypoxia with 2 h of recovery (n = 3 per group; B). The optical density (OD) for the PKA-Cα protein bands were normalized to the ODs for α-tubulin and TATA box-binding protein (TBP) to obtain the relative ODs shown in the dot plot charts. Data are means ± SE from 3 mice/group for the in vivo study and from at least 3 independent experiments for the in vitro study. Data were analyzed by two-way ANOVA (GraphPad Prism).

Fig. 5.

Representative immunoblots of the catalytic subunit of protein kinase A (PKA-Cβ) in cytosolic- and nuclear-enriched kidney protein fractions from wild-type (WT) and meprin β-deficient mice (n = 3 per group) at 0 h and 6 h post-ischemia-reperfusion (IR; A) and Madin-Darby canine kidney (MDCK) cells subjected to 0 h or 2 h of hypoxia or 2 h of hypoxia with 2 h of recovery (n = 3 per group; B). The optical density (OD) for the PKA-Cβ protein bands were normalized to the ODs for α-tubulin and TATA box-binding protein (TBP) to obtain the relative ODs (dot plot charts). Data are means ± SE from 3 mice/group for the in vivo study and from at least 3 independent experiments for the in vitro study. Protein bands in the blot were run on the same gel, but the order was rearranged for final presentation. Data were analyzed by two-way ANOVA (GraphPad Prism). *P = 0.01; **P = 0.002; ***P = 0.0007; ****P = 0.0001.

Levels of PKA-Cβ increased in meprin-expressing kidney tubules at 6 h post-IR.

Immunofluorescence staining of kidney tissues with anti-meprin B and anti-PKA-C antibodies showed a positive correlation between expression of meprins and tubular PKA-C levels in kidneys subjected to 6 h of IR-induced kidney injury. A decrease was confirmed in total PKA-C levels in meprin expressing tubules at 3 h post-IR (Fig. 6A), as previously reported (42). In contrast to the observations at 3 h post-IR, levels of total PKA-C were higher in the proximal tubules, which express meprins at 6 h post-IR, compared with the distal kidney tubules, which lack meprins (Fig. 7, A and B), suggesting a biphasic response. PKA-C levels and expression patterns were comparable in proximal and distal tubules in control kidneys. For reasons of antibody compatibility, the PKA-Cβ immunostaining analysis counterstained with villin, a well-known proximal tubule marker in kidney sections. Meprins and villin have been previously shown to colocalize to the BBMs of proximal tubules (43). Our data show that while total levels of total PKA-C and the PKA-Cβ isoform increased at 6 h post-IR-induced kidney injury, levels of PKA-Cα were comparable in proximal and distal tubules. Furthermore, levels of PKA-Cα and PKA-Cβ isoforms were comparable in proximal and distal tubules for kidney sections from βKO mice (Figs. 7B and 8B, respectively). A closer histological analysis of WT kidney sections showed that the levels of damage to the brush borders were not uniform in all proximal tubules. These changes were not observed in proximal tubules from βKO kidneys. Individual interstitial cells (presumed to be resident macrophages) also stained positively for PKA-Cβ in WT kidney sections but not in βKO kidney sections.

Fig. 6.

Immunolocalization of meprins (green) and the catalytic subunit of protein kinase A (PKA-C; red) in tubules of kidney tissue from wild-type (WT; A) and meprin-β knockout (B) mice with ischemia-reperfusion (IR)-induced kidney injury (n = 3 per group). Images were captured using a BZ-X700 Series all-in-one fluorescence microscope. DAPI was used to stain the nuclei (blue).

Fig. 7.

Immunolocalization of villin (red) and the catalytic subunit of protein kinase A (PKA-Cα; green) in tubules of kidney tissue from wild-type (WT; A) and meprin-β knockout (B) mice with ischemia-reperfusion (IR)-induced kidney injury (n = 3 per group). Villin was used as a proximal tubule marker. DAPI was used to stain the nuclei (blue).

Fig. 8.

Immunolocalization of villin (red) and the catalytic subunit of protein kinase A (PKA-Cβ; green) in tubules of kidney tissue from wild-type (WT; A) and meprin-β knockout (B) mice (n = 3 per group) with ischemia-reperfusion (IR)-induced kidney injury (n = 3 per group). Villin was used as a proximal tubule marker. DAPI was used to stain the nuclei (blue).

Meprin-β expression was associated with increased levels of p-ERK in kidneys after IR injury.

To determine whether meprin-β expression impacts downstream modulators of the PKA signaling pathway, levels of ERK in kidney lysates and lysates from in vitro cultured MDCK cells were evaluated. Western blot analysis was used to compare changes in the levels of p-ERK1/2 in cytosolic- and nuclear-enriched kidney extracts at 6 h post-IR (Fig. 9A) and in MDCK lysates subjected to hypoxic conditions for 2 h (Fig. 9B). Levels of 44-kDa p-ERK1/2 in cytosolic-enriched kidney lysates significantly increased in βKO mice (P = 0.009) at 6 h post-IR but not in WT mice. In contrast, levels of 42-kDa p-ERK1/2 significantly increased in cytosolic-enriched fractions from both WT and βKO kidneys (P = 0.03) compared with levels in control kidneys. For nuclear-enriched kidney lysates, levels of 44-kDa p-ERK1/2 were significantly higher in lysates from both WT (P = 0.01) and βKO (P = 0.005) kidneys at 6 h post-IR. Similarly, levels of 42-kDa p-ERK1/2 in nuclear-enriched fractions significantly increased in both WT (P = 0.03) and βKO (P = 0.0007) mice (Fig. 9B). For in vitro cultured MDCK cells, cytosolic levels of a 44-kDa p-ERK1/2 species showed a significant decrease at 2 h after the induction of hypoxia (P = 0.05) but significantly increased after 2 h of recovery from hypoxia (P = 0.0007) in both meprin-β-transfected and nontransfected control cells (Fig. 9B). Interestingly, the p-ERK1/2 band running at 42 kDa increased (P = 0.05) only after 2 h of recovery from hypoxia in meprin-β-transfected and nontransfected control cells. For the Western blots, we used p-p44/42 MAPK (Erk1/2) (Thr²⁰²/Tyr²⁰⁴) antibody (no. 4370S, Cell Signaling), which detects endogenous levels of p44 and p42 MAPK (Erk1 and Erk2) when phosphorylated either individually or dually at Thr²⁰² and Tyr²⁰⁴ of Erk1 or Thr¹⁸⁵ and Tyr¹⁸⁷ of Erk2. The p-ERK bands detected correspond to p44 and p42 MAPK.

Fig. 9.

Representative immunoblots for phosphorylated (p-)ERK1/2 in cytosolic- and nuclear-enriched kidney protein fractions for wild-type (WT) and meprin-β knockout mice (n = 3 per group) at 0 h and 6 h post-ischemia-reperfusion (IR; A) and Madin-Darby canine kidney (MDCK) cells subjected to 0 h or 2 h of hypoxia or 2 h of hypoxia with 2 h of recovery (n = 3 per group; B). The relative optical densities (ODs) were obtained by normalizing the ODs for p-ERK to the ODs for tubulin and TATA box-binding protein (TBP). Two p-ERK bands were detected: a dominant 42-kDa band and a less abundant 44-kDa band. Data are means ± SE from 3 mice/group for the in vivo study and from 3 independent experiments for the in vitro study. Data were analyzed by two-way ANOVA (GraphPad Prism). *P = 0.01; **P = 0.009; ***P = 0.0007.

DISCUSSION

Meprins are zinc metalloproteinases abundantly expressed in the BBMs of proximal kidney tubules and the small intestine (52). Levels of meprin expression and localization are associated with IR-induced kidney injury and diabetic nephropathy (10, 47). Tubular epithelial cells are the primary target for hypoxia-mediated AKI and subsequent chronic kidney disease (7, 36). Dysfunctions in the hypoxia response in tubular cells is associated with the production of profibrotic genes and ECM buildup (38). The meprin-expressing S3 segments of proximal kidney tubules are more susceptible to IR-induced kidney injury compared with distal tubular cells lacking meprins (54). The PKA signaling pathway plays an important role in the hypoxia response and ECM metabolism (21). Furthermore, cytokine-induced PKA-mediated signaling has been implicated in oxidative stress in mesangial cells (6). In vitro experiments by our group and others have demonstrated that PKA-C is a meprin substrate (13, 14, 42) and that meprin interactions with PKA-C are isoform specific, with meprin A only cleaving PKA-Cα while meprin B cleaved PKA-Cα, PKA-Cβ1, and PKA-Cβ2 (42). Furthermore, meprin cleavage of PKA-C reduced PKA-C kinase activity (42). The objective of the present study was to determine the PKA-C isoforms impacted by meprin-β expression in response to IR-induced injury. To this end, levels of total PKA-C and the two PKA-C isoforms that were previously shown to interact with meprin B, namely, PKA-Cα and PKA-Cβ, were evaluated. The study further determined whether meprin-β impacts the levels of p-ERK, a downstream modulator of the PKA signaling pathway. The study used mice deficient in meprin-β (βKO mice) and WT mice, which express both isoforms of meprin (A and B). βKO mice only express the homomeric form of meprin A (α-α) and are deficient in meprins B (β-β) and the heterodimeric isoform of meprin A (α-β). Additional experiments used in vitro cultured MDCK cells transfected with meprin-β cDNA. The data from the present study confirm previous reports showing that meprin-β exacerbates AKI (9, 43). Two biomarkers of kidney injury, namely, cystatin C and NGAL, were significantly elevated in WT mice but not in βKO mice at 6 h post-IR. Cystatin C is a member of the cystatin superfamily and is produced by all nucleated cells. It is filtered by the glomerulus and then reabsorbed by proximal tubules but is not secreted by renal tubules. Thus, cystatin C is a sensitive biomarker for the detection of early kidney dysfunction (23, 55). NGAL, an iron-transporting protein, plays a critical role in the protection of proximal tubules after induction of IR injury (4). NGAL is upregulated in mouse models of renal IR-induced kidney injury with significant changes detected as early as 2 h post-IR (48). The modest changes in NGAL levels in meprin-β-null mice confirms that meprin-β deficiency protects mice from AKI. Potential mechanisms could involve direct processing of protein targets involved in the response to injury or changes in cell metabolism impacted by meprin targets. Data from mice with streptozotocin-induced type 1 diabetes had significant differences in metabolite profiles between meprin-β-expressing and meprin-β-deficient mice at 8 wk after streptozotocin (19).

Previous studies from our group and others have demonstrated that meprin-β is capable of proteolytically processing PKA-C in vitro and in vivo (14, 42). Furthermore, in vitro cleavage of PKA-C by meprin-β resulted in decreased PKA-C kinase activity (42). The present study showed a significant decrease in total levels of p-PKA-C in cytosolic protein lysates from WT but not βKO kidneys, suggesting that meprin-β expression/activity is responsible for the in vivo reduction in kinase activity. Additionally, confocal microscopy showed that at 3 h post-IR, levels of PKA-C significantly decreased in meprin-expressing proximal tubule cells but increased in distal tubules, which do not express meprins (38). However, the previous study could not determine the PKA-C isoforms impacted because isoform-specific antibodies were not available.

Data from the present study using anti-PKA-Cα- and anti-PKA-Cβ-specific antibodies show significant increases in the levels of PKA-Cβ in kidney protein lysates for WT mice, which express meprins at 6 h post-IR, but not in βKO mice, which are deficient in meprin B. There were no significant changes in the levels of PKA-Cα, indicating that it is the PKA-Cβ isoform that is involved in mediating the response to IR-induced renal injury. The increase in PKA-C at 6 h was surprising because we previously demonstrated a decrease in the levels of total PKA-C at 3h post-IR, an observation confirmed in the present study. This leads us to conclude that initial processing of PKA-Cβ in meprin-β-expressing tubules at the onset on injury induces a PKA-Cβ-mediated response that could be important in tubular repair. Bimodal regulation of the PKA signaling pathway by the endogenous PKA-C antagonist ARHGAP36 was also reported in human embryonic kidney cell lines (16). However, the increased levels of PKA-Cβ only occurred in select proximal tubules and may correlate to the level of injury and reparative mechanisms that should be further investigated.

The first evidence that meprin-β is involved in the hypoxia response came froms studies by Litovchick in 2002 (37), who showed that the COOH-terminal tail of meprin-β interacts with oteosarcoma-9 (OS-9), an intraction that is significant becasue OS-9 has also interacts with two proteins (HIF-1α and prolyl hydroxylase) involved in the hypoxia response (2). We subsequently demonstrated that OS-9 is a meprin substrate and that MDCK cells transfected with meprin-β cDNA cleaved OS-9 when subjected to hypoxia mimic, CoCl₂. Furthermore, meprin-β expression blocked hypoxia-induced cytosolic accumulation of OS-9 in cells. The present study tested the hypothesis that meprin-β expression/activity impacts downstream targets of the PKA signaling pathway under hypoxic conditions in vitro and in vivo. The data demonstrate that meprin-expressing kidneys have higher baseline cytosolic levels of ERK1/2 than meprin-β-deficient mice. Furthermore, IR-induced injury was associated with significant increases in p-ERK levels for both genotypes (P = 0.05) in cytosolic- and nuclear-enriched fractions. However, the fold change in p-ERK levels were much higher in βKO mice. Also, p-ERK levels were significantly increased in MDCK cells subjected to 2 h of hypoxia followed by 2 h or normoxia. Previous studies by other groups have shown that after IR injury, the ERK1/2 signaling pathway is activated in kidneys (15, 44, 45). It could therefore be argued that the increased levels of p-ERK1/2 confers protection against IR-induced injury as ERK1/2 has been shown to protect against IR injury (15, 44, 45). The higher fold change in p-ERK in βKO kidneys suggests that meprin-β activity downregulates this ERK phosphorylation, thus blocking this protective response. Another study has shown that PKA-C inactivates ERK1/2 through inhibition of RAF-1 (18).

The presrent data demonstrate that meprin-β enhances IR-induced kidney injury in part by modulating mediators of the PKA-Cβ signaling pathway. While meprin-β proteolytically processes both PKA-Cα and PKA-Cβ in vitro, it is the PKA-Cβ isoform that is impacted by meprin-β expression in IR-induced renal injury. The PKA-Cβ gene encodes 10 different splice variants (35). The antibodies used in this study did not distinguish PKA-Cβ variants. However, it has been previously demonstrated that meprin-β cleaves both PKA-Cβ1 and PKA-Cβ2. The PKA-Cβ1 isoforms are ubiquitously expressed while PKA-Cβ2 is restricted to neuronal tissue in humans (28). The PKA signaling pathway has been shown to modulate ECM metabolism in kidney cells (49, 50), and hepatic fibrosis (51), suggesting a role in modulating fibrosis in renal injury. However, the existing data have mostly focused on glomerular cells. Our data show dramatic shifts in levels of PKA-Cβ in proximal tubular cells, and the role of this PKA signaling pathway in tubular injury needs further investigations. The mouse PKA-Cβ2 variant is highly expressed in lymphoid tissue and could thus be involved in inflammation. In the present study, interstitial cells in presumed to be resident immune cells stained positively for PKA-Cβ in WT but not βKO kidney sections. Additional studies are needed to further understand how PKA-Cβ is regulated in kidney injury to determine whether meprin-β modulation of PKA-Cβ signaling is beneficial or harmful in renal injury. βKO mice express homomeric meprin A (α-α), which was shown to cleave PKA-Cβ1 but not PKA-Cβ2. We have recently backcrossed meprin-α KO mice with βKO mice to obtain meprin-αβ double-knockout mice, which will be used for additional experiments to further understand the isoform-specific interactions between meprins and PKA-C and how the interactions impact kidney injury.

GRANTS

This work was supported by National Institutes of Health Grants SC-GM-3102049 and SC1-GM-118271 (to E. M. Ongeri). F. Ahmed was supported by National Institutes of Health Training Grant T32-AI-007273.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

E.M.O. conceived and designed research; F.A., J.-M.M., M.F., I.Y., and E.M.O. performed experiments; F.A. and E.M.O. analyzed data; F.A. and E.M.O. interpreted results of experiments; F.A. and E.M.O. prepared figures; F.A. and E.M.O. drafted manuscript; F.A., S.A., and E.M.O. edited and revised manuscript; F.A., J.-M.M., M.F., S.A., and E.M.O. approved final version of manuscript.