Rapamycin protects against gentamicin-induced acute kidney injury via autophagy in mini-pig models

Gentamicin may cause acute kidney injury. The pathogenesis of gentamicin nephrotoxicity is unclear. Autophagy is a highly conserved physiological process involved in removing damaged or aged biological macromolecules and organelles from the cytoplasm. The role of autophagy in the pathogenesis of gentamicin nephrotoxicity is unclear. The miniature pigs are more similar to humans than are those of rodents, and thus they are more suitable as human disease models. Here we established the first gentamicin nephrotoxicity model in miniature pigs, investigated the role of autophagy in gentamicin-induced acute kidney injury, and determined the prevention potential of rapamycin against gentamicin-induced oxidative stress and renal dysfunction. At 0, 1, 3, 5, 7 and 10 days after gentamicin administration, changes in autophagy, oxidative damage, apoptosis and inflammation were assessed in the model group. Compared to the 0-day group, gentamicin administration caused marked nephrotoxicity in the 10-day group. In the kidneys of the 10-day group, the level of autophagy decreased, and oxidative damage and apoptosis were aggravated. After rapamycin intervention, autophagy activity was activated, renal damage in proximal tubules was markedly alleviated, and interstitium infiltration of inflammatory cells was decreased. These results suggest that rapamycin may ameliorate gentamicin-induced nephrotoxicity by enhancing autophagy.

were obtained from the State Key Laboratory for Agrobiotechnology in China Agricultural University. Two animals per cage were housed in a room maintained at a temperature of 18-22 °C and a relative humidity of 30-70% with artificial lighting from 8:00 to 20:00 hours. The experimental protocol was carried out in accordance with the approved guidelines of the Institutional Animal Care and Use Committee at the Chinese PLA General Hospital.

Experimental groups and treatment. Gentamicin (GM) was purchased from Guangzhou
Baiyunshan Tianxin Phamarceutical Co. LTD and rapamycin was purchased from East China Pharmaceutical Limited CO. LTD. In the dose-response study of gentamicin, minipigs received a daily i.m. injection for 10 days of gentamicin sulfate (5, 6.85, 10, 15, 20, 40, 60, 80, 100 and 120 mg/kg) in distilled water. A total of 33 healthy male minipigs were randomly assigned to 11 experimental groups (n = 3 per group). In the 120 mg/kg group, the miniature pigs died 1 day after injection. The 6.85 mg/kg is the clinically relevant doses of gentamicin. For the gentamicin time course study, minipigs received an i.m. injection of 80 mg/kg gentamicin daily for 0, 1, 3, 5, 7 and 10 days. To test the effect of rapamycin in gentamicin-induced acute kidney injury, first group is the control animals, second is the one that was treated with GM (80 mg/kg/i.m./10 days), third is the one that was treated with GM (80 mg/kg/i.m./10 days) and rapamycin (0.3 mg/kg/p.o./10 days).

Serum biochemistry analysis.
At the scheduled termination day, all minipigs were euthanized by pentobarbital, and serum samples were collected by centrifugation at 3,000 rpm for 10 min and stored in the − 80 °C freezer before analyzing. Serum biochemical parameters were analyzed by an autoanalyzer (Cobas8000, Roche, Germany).
Histological and TUNEL assay. The kidneys from pigs were excised, fixed in 4% paraformaldehyde, embedded in paraffin and sectioned at 4 μ m for histological staining with periodic acid-Schiff. Kidney sections were mounted on glass slides and stained with hematoxylin and eosin for histopathologic evaluation. Acute tubular necrosis (ATN) scores were assigned semiquantitatively by light microscopy on a scale of 0-4: 0 = normal histology; 1 − 4 = < 25, < 50, < 75, and > 5%, respectively, of epithelial cells of the proximal convoluted tubules showing necrosis, degeneration, regeneration, tubular dilatation, protein casts, and interstitial lymphocytic infiltration 19 . Apoptosis in renal tissues was identified by a TUNEL assay with In Situ Cell Death Detection Kit (Roche, 11684817910). Some tissue sections were subjected to antigen retrieval by microwaving for 10 min in 10 mM sodium citrate buffer [pH 6.0]. Endogenous peroxidase activity was blocked by incubation with 3% hydrogen peroxide for 30 min. After PBS washing, sections were incubated with 1.5% normal goat serum for 30 min, followed by incubation with Scientific RepoRts | 5:11256 | DOi: 10.1038/srep11256 TUNEL reaction mixture overnight at 4 °C. After three washes with PBS, the samples were incubated with converter-POD at room temperature. After PBS washing, the sections were incubated with DAB followed by examination under the microscope. For estimation of the level of apoptosis, the TUNEL positive renal tubular epithelial cells, as well as unstained cells, were scored in 10 different fields at 400 magnification. The apoptotic index was expressed as the percentage of total cells scored. The results were expressed as the average number of TUNEL-positive cells for each group. All counting procedures were performed blindly.
Immunohistochemistry. The kidneys were fixed in 10% formaldehyde overnight at 4 °C and processed for paraffin-embedding following standard procedures. Sections were cut at 3-μ m thicknesses. For immunohistochemical analysis, some tissue sections were subjected to antigen retrieval by microwaving for 10 or 15 min in 10 mM sodium citrate buffer [pH 6.0]. Endogenous peroxidase activity was blocked by incubation with 3% hydrogen peroxide for 10 min. After PBS washing, sections were incubated with 1.5% normal goat serum for 20 min, followed by incubation with mouse monoclonal anti-8-OHdG antibody (1:50; Santa Cruz Biotechnology Inc., sc-66036) overnight at 4 °C. After three washes with PBS, the samples were incubated with biotin-conjugated goat anti-mouse IgG for 30 min at room temperature. After washing in PBS, the sections were incubated with streptavidin-conjugated peroxidase 30 min at room temperature. After PBS washing, the sections were incubated with DAB followed by examination under the microscope.
Transmission electron microscopy. Kidneys were cut into tissue blocks (1 mm 3 ) and fixed in 2.5% glutaraldehyde in 0.01 mol/L phosphate buffer at 4 °C, followed by 2% osmium tetroxide. They were then dehydrated in a series of graded ethanol solutions. Ethanol was then substituted with propylene oxide and the tissue was embedded in epoxy resin. Ultrathin sections were double-stained with uranyl acetate and lead and examined under a JEM1200EX transmission electron microscope (JOEL) at 80 kV.

Statistical analyses.
Analyses of all data were performed using the SPSS ver. 13.0 (SPSS, Chicago, IL, USA) software. Data are expressed as means ± standard deviation (SD). Comparisons among groups were made using analysis of variance. Values of P < 0.05 were considered to indicate statistical significance.

Gentamicin causes acute kidney injury in minipigs.
Dose-response studies. In a gentamicin dose-response study, the blood urea nitrogen (BUN) and serum creatinine levels of minipigs were significantly higher than control values only at drug doses of 80 and 100 mg/kg (Table 1), which were sufficient to produce AKI. Gentamicin-induced renal lesions included epithelial cell necrosis, degeneration, and regeneration of the proximal convoluted tubules. Dilatation of the tubules, protein casts in the tubular lumina, glomerular vacuolization, and interstitial lymphocytic infiltration, were also observed ( Fig. 1). The renal histopathological scores indicated that the extent of renal injury was significant, compared to controls, after treatment with gentamicin (60, 80, and 100 mg/kg).
Time course studies. In a gentamicin time-course study, BUN and serum creatinine levels were significantly higher than control values only in the 10-day group (Table 2). Simultaneously, epithelial cell necrosis, degeneration, and regeneration of the proximal convoluted tubules were observed, and renal histopathological scoring indicated significant renal injury in that group, compared to controls (Fig. 2).
The role of autophagy in gentamicin-induced nephrotoxicity. Gentamicin inhibits autophagic degradation in minipig kidneys. Autophagic flux refers to all autophagic processes including the formation of autophagosomes, the fusion of autophagosomes with lysosomes, and the subsequent breakdown thereof. P62/SQSTM1 is a ubiquitin-binding protein that interacts with LC3 to mediate the degradation of polyubiquitinated protein aggregates and mitochondria, via the autophagy-lysosome pathway, in mammalian cells. Cellular survival is thus enhanced, because p62/SQSTM1 is sequestered in autophagosomes upon direct interaction with LC3 20 . Therefore, p62/SQSTM1 and polyubiquitinated protein aggregates (poly UB) may act as markers reflecting autophagic degradation activity. Cellular increases in the levels of p62/ SQSTM1 and poly UB reflect a decrease in autophagic degradation activity. We found that the expression levels of p62/SQSTM1 and poly UB were significantly increased in 1-, 3-, 5-, 7-, and 10-day kidneys compared to day-0 kidneys ( Fig. 3), indicating that gentamicin inhibits autophagic degradation in the kidney.
LC3 is the mammalian homolog of yeast Atg8 and is essential for autophagosomal formation. Precursor LC3 can be digested by Atg4 to form LC3-I, which is then activated by Atg7 to form membrane-bound LC3-II. The latter protein is an autophagosomal membrane marker, and LC3-II/I levels are associated with autophagic flow. We first measured the expression levels of LC3 via Western blotting. Expression of both LC3-I and LC3-II was significantly increased in 1-, 3-, 5-, 7-, and 10-day compared to day 0 kidneys (Fig. 3).
Gentamicin inhibits mitochondrial autophagic activity in minipig kidneys. Regulation of autophagy by hypoxia-inducible factor 1α (HIF-1α ) and a transcriptional target thereof, the "BCL-2 19 kDa interacting protein 3" (Bnip3), was first shown in embryonic fibroblasts of HIF-1α knockout mouse cells subjected to hypoxia 21  an atypical "BH3-only" protein linked to mitochondrial dysfunction and cell death. Bnip3 is also a potent inducer of mitochondrial autophagy, which is a protective response 22 . Bnip3 is also induced in an HIF-1α -dependent manner. This HIF-1α -Bnip3 association may contribute to regulation of autophagy  in renal tubular cells 23 . Expression of both Bnip3 and HIF-1α was significantly increased in 1-, 3-, 5-, and 7-day kidneys, but clearly decreased in 10-day kidneys, compared to day-0 kidneys (Fig. 3). PINK1 spans the outer mitochondrial membrane, with the kinase domain facing into the cytoplasm, and is rapidly degraded in healthy mitochondria. PINK1 interacts directly with Parkin in a manner that does not depend on the activity of either PINK1 kinase or Parkin E3-ligase 24,25 . The cytoplasm-facing PINK1 kinase domain can phosphorylate Parkin, causing the latter protein to be translocated to mitochondria. Also, phosphorylated Parkin (p-Parkin) ubiquitinates mitochondrial substrates. PINK1 is a sensor of damaged mitochondria. We found that the expression of PINK1 was significantly increased only in 10-day kidneys (compared with day 0 kidneys; Fig. 3).
Parkin is considered to be a substrate of the mitochondrial E3 esterase and may induce mitochondrial autophagy. One such pathway appears to be activated by Parkin after translocation thereof from the cytosol to (specifically) dysfunctional mitochondria 23 . Parkin is involved in removal of such mitochondria. We found that although the Parkin levels were similar in 0-, 1-, 3-, 5-, 7-, and 10-day kidneys, the p-Parkin levels were significantly higher only in 10-day than day-0 kidneys (Fig. 3).  Autophagy-promoting protein AMBRA1 ("activating molecule in Beclin1-regulated autophagy") binds to depolarized mitochondria in a Parkin-dependent manner. Overexpression of AMBRA1 enhances removal of depolarized mitochondria, but only in the presence of Parkin 26 . We found that expression of AMBRA1 was significantly increased only in 10-day compared to day 0 kidneys (Fig. 3).
Gentamicin induces oxidative damage in mini-pig kidneys. Oxidative stress caused by reactive oxygen species (ROS) can damage both macromolecules including DNA, proteins, and lipids; and mitochondrial structure. To characterize the DNA damage induced by gentamicin in mammalian cells, we used Western blotting to quantify the levels of γ -H2AX, a core histone protein that becomes phosphorylated when DNA is damaged. The level thereof was consistently and significantly increased in 10-day kidneys compared with day-0 kidneys (Fig. 4). Also, we quantified the levels of 8-hydroxy-2′ -deoxyguanosine (8-OHdG), an oxidized byproduct of DNA. As was true of γ -H2AX, the levels of 8-OHdG were significantly elevated after 10 days of gentamicin treatment (Fig. 4).
To further explore the nature of cellular oxidative damage, we measured the levels of carbonylated proteins and 4-hydroxy-2-nonenal (4-HNE), a product of lipid peroxidation caused by oxidative damage. We found that the levels of these materials were the highest in 10-day kidneys, compared with kidneys obtained at other times (Fig. 4).
Gentamicin causes mitochondrial oxidative damage in minipig kidneys. We further explored the effects of gentamicin on mitochondrial structure via transmission electron microscopy. Obvious injury was apparent, including swelling and disintegration of cristae, in 10-day kidneys (Fig. 4).

Activation of autophagy by rapamycin reduces gentamicin-induced AKI in minipig kidneys.
Effect of rapamycin on renal function. In the rapamycin group, the levels of BUN and serum creatinine were significantly lower than those in 10-day kidneys (Table 3). Rapamycin significantly inhibited gentamicin-induced tubular necrosis and cast formation. The histological scores of kidney sections from gentamicin-treated animals which also received rapamycin were significantly lower than those of animals receiving gentamicin alone (Fig. 7).
Effect of rapamycin on autophagy in the kidneys of gentamicin-treated minipigs. The levels of LC3-I, LC3-II, p62, Poly UB, PINK1, p-Parkin, and AMBRA1 were significantly lower in the kidneys of rapamycin-than gentamicin-treated animals. The Parkin levels were, however, similar. Bnip3 and HIF-1α levels in the kidneys of rapamycin were higher than gentamicin-treated animals (Fig. 8).
Effect of rapamycin on gentamicin-induced oxidative damage to minipig kidneys. The levels of 8-OHdG and γ -H2AX were significantly elevated after 10 days of gentamicin treatment (Fig. 9). To further explore the accumulating cellular oxidative damage, we measured the levels of protein carbonyls and 4-HNE, a lipid peroxidation product resulting from oxidative damage. We found that the levels of these materials were clearly lower in the kidneys of rapamycin-than gentamicin-treated kidneys (Fig. 9). Thus, rapamycin inhibited the mitochondrial structural damage, including swelling and disintegration of cristae, caused by gentamicin.

Effect of rapamycin on gentamicin-induced apoptosis and inflammation in minipig kidneys.
The numbers of terminal uridine nick end-labeled (TUNEL)-positive cells, and the levels of cleaved caspase-3, were significantly lower in the kidneys of rapamycin-than gentamicin-treated animals (Fig. 10). Compared to the gentamicin-treated animals, the expression level of nuclear factor (NF)-κ B (p65) was significantly decreased in the rapamycin -treated animals (Fig. 10).

Discussion
The kidney has both filtration and secretion functions, making it the main excretory organ of many drugs. Gentamicin is widely used to treat infectious diseases caused by gram-negative bacteria in animals and humans. However, nephrotoxicity has limited its application. Therefore, an in-depth understanding of gentamicin-induced kidney damage mechanisms is essential. Nephrotoxicity models have been established in mouse and rat, but due to genetic, biochemical and physiological differences between rodents and humans, it is necessary to establish an aminoglycoside kidney damage model using large animals, such as pigs or monkeys, which are more similar to human beings.
Oxidative stress plays a critical role in the pathophysiology of bactericidal antibiotic-induced injury. Bactericidal antibiotics include quinolones, aminoglycosides, and β -lactams 27 . Bactericidal antibiotics (quinolones, aminoglycosides, and β -lactams) caused mitochondrial dysfunction and ROS overproduction in mammalian cells, ultimately leading to the accumulation of oxidative tissue damage. Autophagy, a lysosomal degradation pathway, plays a crucial role in removing protein aggregates as well as damaged organelles to maintain intracellular homeostasis in renal pathophysiology 28,29 . Autophagy is particularly important for maintenance of post-mitotic cells, such as podocytes, which have only a limited capacity for regeneration 29 .
The exact mechanism of nephrotoxicity caused by aminoglycoside antibiotics, such as gentamicin, is not yet clear. In this study, we investigated the role of autophagy, particularly mitochondrial autophagy, in the pathogenesis of gentamicin nephrotoxicity. First, we successfully established a gentamicin-induced nephrotoxicity model in mini-pigs, and found that serum creatinine and blood urea nitrogen were significantly increased in the 80 and 100 mg/kg gentamicin-injected groups. Renal pathological analysis revealed significant renal tubular epithelial cell necrosis in the 80 and 100 mg/kg gentamicin-injected groups. Although the ATN score in the 60 mg/kg gentamicin-injected group was significantly increased compared with the CON group, serum creatinine and blood urea nitrogen were not significantly increased. Additionally, all of the mini-pigs were dead on the third day after 100 mg/kg gentamicin treatment (i.m.). Therefore, in the gentamicin-induced injury study time course, we decided to observe changes in renal function, autophagy, inflammation, apoptosis, and oxidative damage at 0, 1, 3, 5, 7 and 10 days after 80 mg/kg gentamicin treatment (i.m.).  Table 3. Effect of rapamycin on renal function in gentamicin-induced minipig. BUN, blood urea nitrogen; Scr, serum creatinine. CON, control group, GEN, gentamicin, Rapa, rapamycin. Values are presented as means ± SD (n = 6). *p < 0.05, vs. GEN.
We found a compensatory increase in autophagosome formation-related proteins (LC3I and LC3II) in mini-pigs administered gentamicin; however, the increased expression of p62/SQSTM1 and polyubiquitin aggregates indicated that autophagic degradation capacity was decreased in the kidneys of gentamicin-treated mini-pigs. Bnip3 was identified as a potent inducer of autophagy in cells. Bnip3 is an important regulator of mitochondrial turnover via autophagy in the myocardium 30 . HIF-1α -Bnip3 may contribute to the regulation of autophagy in renal tubular cells 31 . In our study, we observed that Bnip3 and HIF-1α were significantly elevated in the early stage of gentamicin treatment, and markedly decreased in the 10-day group. These results indicated that mitophagy was activated in the early stage, while mitophagy function declined in the gentamicin-induced AKI. Animal studies showed that ER stress or oxidative stress induces adaptive autophagy upregulation in the early phase, which helps to restore intracellular homeostasis by disposing of a number of harmful molecules, such as unfolded or misfolded proteins in the ER lumen, cytosolic proteins damaged by ROS, or even dysfunctional ER and mitochondria [32][33][34] . Upon mitochondrial damage, mitochondria-localized PINK1 recruits and phosphorylates the E3-ligase, parkin, to mitochondria. At the same time, parkin can polyubiquitinate voltage-dependent anion channel (VDAC)1, followed by p62 binding to VDAC1 on the mitochondria and ATG8/LC3/ GABARAP on the developing autophagosome, resulting in mitochondrial sequestration and removal by the autophagic machinery 26 . Parkin reportedly surveys mitochondrial quality by translocating to depolarized mitochondria and inducing their selective macroautophagic removal (mitophagy). In addition, p-parkin interacts with AMBRA1, a protein that promotes autophagy in the vertebrate central nervous system 35 . Prolonged mitochondrial depolarization strongly increases the interaction of parkin with AMBRA1. We found that mitophagy markers, PINK1, p-parkin and AMBRA1, were markedly increased in 10-day gentamicin-treated mini-pig kidneys, indicating that mitochondrial damage can significantly enhance mitophagy activity in the late stage of gentamicin-induced AKI in mini-pig kidney.
Mitophagy is the only pathway that removes damaged mitochondria from cells. Therefore, inhibition of mitochondrial autophagy function leads to the accumulation of damaged mitochondria. Because mitochondria are the main intracellular site for ROS production, damaged mitochondria accumulation will lead to more ROS production, and further promote pro-inflammatory cytokine expression and apoptosis.
In humans, it is likely that oxidative stress and related oxidative cellular damage induced by bactericidal antibiotics underlie many of the adverse side effects associated with these antibiotics [36][37][38] . In our experiment, oxidative damage markers, including 4HNE, protein carbonyls, γ -H2AX and 8-OHdG, were significantly elevated in the 10-day group. Meanwhile, the results showed obvious injury to mitochondrial structures, such as swelling and disintegration of cristae in the 10-day gentamicin-treated group. Under normal circumstances, the damaged mitochondria are degraded by autophagy. If the impaired  mitochondria are not removed, they can generate more ROS, which further aggravates oxidative damage, forming a vicious cycle 39,40 . Therefore, mitochondria play a central role in tissue damage caused by gentamicin 40 .
It was shown that gentamicin can induce apoptosis and inflammation in renal tubular cells [41][42][43][44] . In accordance with previous reports, our results showed that the level of cleaved caspase-3 and the number of TUNEL(+ ) cells were significantly increased in gentamicin-induced AKI. Meanwhile, the expression levels of nuclear factor (NF)-κ B (p65) and inflammatory cytokines were significantly elevated in the gentamicin-induced AKI mini-pigs.
Characterization of bactericidal antibiotic effects on mammalian cells points to widespread cellular oxidative damage 27 . According to the results of our experiments, autophagy plays a central role in removing cellular oxidative damage caused by gentamicin. We next identified whether the activation of autophagy by rapamycin protects against gentamicin-induced AKI. Therefore, we explored the possibility that using rapamycin, a pharmacological inhibitor of mTOR, could enhance autophagy to alleviate these deleterious effects. Identifying bactericidal antibiotics as a cause of ROS overproduction and autophagy in mammalian cells provides a basis for developing therapeutic strategies that could help alleviate adverse  side effects associated with antibiotics. For instance, by coadministering rapamycin, we showed that autophagy, particularly mitophagy, and renal function could be recovered; oxidative damage, apoptosis and inflammation induced by bactericidal antibiotics could be abrogated. When appropriate for patient health, coadministering a pharmacological autophagy activator could be a simple treatment strategy aimed at preventing cellular oxidative damage. It will be important to extend our investigation to human subjects to confirm our findings, demonstrate their relevance to humans, and maximize the translational value of our work. Epidemiologic studies could provide valuable insight into the clinical implications of antibiotic-induced oxidative damage and define some of the risks associated with bactericidal antibiotic exposure. In humans, it is likely that oxidative stress and related oxidative cellular damage induced by bactericidal antibiotics underlie many of the adverse side effects associated with these antibiotics [36][37][38] . It will be intriguing to explore these possibilities with appropriate clinical trials, with the goal of developing effective antibacterial therapies with minimal adverse side effects.