γ‐secretase inhibitor DAPT mitigates cisplatin‐induced acute kidney injury by suppressing Notch1 signaling

Abstract Organ toxicity, including kidney injury, limits the use of cisplatin for the treatment of multiple human cancers. Hence, interventions to alleviate cisplatin‐induced nephropathy are of benefit to cancer patients. Recent studies have demonstrated that pharmacological inhibition of the Notch signaling pathway enhances cisplatin efficacy against several cancer cells. However, whether augmentation of the anti‐cancer effect of cisplatin by Notch inhibition comes at the cost of increased kidney injury is unclear. We show here that treatment of mice with cisplatin resulted in a significant increase in Notch ligand Delta‐like 1 (Dll1) and Notch1 intracellular domain (N1ICD) protein expression levels in the kidneys. N‐[N‐(3,5‐difluorophenacetyl)‐L‐alanyl]‐S‐phenylglycine t‐butyl ester (DAPT), a γ‐secretase inhibitor reversed cisplatin‐induced increase in renal N1ICD expression and plasma or urinary levels of predictive biomarkers of acute kidney injury (AKI). DAPT also mitigated cisplatin‐induced tubular injury and reduction in glomerular filtration rate. Real‐time multiphoton microscopy revealed marked necrosis and peritubular vascular dysfunction in the kidneys of cisplatin‐treated mice which were abrogated by DAPT. Cisplatin‐induced Dll1/Notch1 signaling was recapitulated in a human proximal tubule epithelial cell line (HK‐2). siRNA‐mediated Dll1 knockdown and DAPT attenuated cisplatin‐induced Notch1 cleavage and cytotoxicity in HK‐2 cells. These data suggest that Dll1‐mediated Notch1 signaling contributes to cisplatin‐induced AKI. Hence, the Notch signaling pathway could be a potential therapeutic target to alleviate renal complications associated with cisplatin chemotherapy.

is noteworthy due to its association with poor clinical outcomes including therapeutic failure and cancer recurrence. 2,3,5 Apart from chemoresistance, cisplatin therapy is beset with marked adverse effects that hamper its clinical use. A rapid decline in kidney function (acute kidney injury) is a limiting factor in cisplatin chemotherapy and often results in treatment discontinuation. [6][7][8] Up to 34% of cancer patients treated with cisplatin exhibited various degrees of nephrotoxicity. [9][10][11] Hence, elucidation of the pathogenesis of cisplatin-induced acute kidney injury (AKI) is essential for the development of adjunctive therapies to reduce morbidity and mortality in cancer patients.
The Notch signaling, a highly conserved cellular pathway controls cell fate specification, survival and differentiation. 12,13 The interaction between Notch ligands (Jagged1 and 2; Delta-like1, 3 and 4) and their transmembrane receptors (Notch1-4) on adjacent cells activate proteolytic cleavage of the Notch receptors by γ-secretases thereby releasing the Notch intracellular domain (NICD) into the cytoplasm. 12,13 NICD translocates into the nucleus and forms a complex with transcriptional factors that regulate the expression of target genes. 12,13 Dysregulation of the Notch signaling pathway has been implicated in the pathophysiology of both cancer 12,13 and kidney disease. 14,15 Recent reports suggest that Notch signaling is upregulated in cisplatin-resistant cancer. [16][17][18][19][20] Accordingly, targeting the Notch signaling pathway with pharmacological inhibitors of γsecretase or knockdown of Notch components increased cisplatin efficacy against several cancer cells. 17,[21][22][23][24][25][26][27] Whether enhancement of the anti-cancer effect of cisplatin by Notch inhibition comes at the cost of increased kidney injury is unknown. In the present study, we investigated whether Notch signaling is induced in the kidneys of cisplatin-treated mice. We also tested the hypothesis that pharmacological inhibition of the Notch signaling pathway ameliorates cisplatin-induced AKI.

| Animals
All experimental animal procedures were reviewed and approved by the Animal Care and Use Committee of the University of Tennessee Health Science Center (UTHSC). Male mice (C57BL/6J; 8-10 weeks old; Jackson Laboratories, Bar Harbor, ME, USA) were used in this study.

| In Vivo studies
Cisplatin and DAPT were solubilized in pharmaceutical excipient sulfobutyl ether-β-cyclodextrin (20% Captisol) 28-30 as we have previously described. 31 Mice were randomized into four groups (Figure 1; n = 16/group) and housed in ventilated micro-isolation cages. A group of mice was given a single intraperitoneal (IP) injection of cisplatin (15 mg/kg). Mice in the control groups were treated (IP) with Captisol or DAPT (15 mg/kg) alone. Another group received DAPT 1 hour before cisplatin administration, followed by daily DAPT injection for 4 days. The dose of cisplatin used was chosen based on a previous study that determined the dose-response nephrotoxic effect of cisplatin in mice upon single IP injections. 32 The injection volume was kept at 10 μL/g body weight. On the fifth day, each mouse was weighed and placed on a new 96-well plate inside an empty box for~2 hours. 31,33 Urine samples were collected from the wells and analyzed. Blood was obtained from anesthetized mice via retro-orbital bleeding. Kidneys were collected, weighed and processed after mice had been euthanized with sodium pentobarbital (200 mg/kg; IP) followed by exsanguination.

| Determination of kidney function
To evaluate kidney function, we measured plasma or urinary concentrations of creatinine, urea nitrogen, cystatin C, neutrophil gelatinase-associated lipocalin (NGAL) and albumin. We also determined the glomerular filtration rate (GFR) in the mice. Plasma and urinary creatinine concentrations were evaluated by mass spectrometry (iso- GFR was evaluated using the FIT-GFR Inulin Kit and a one-compartment plasma clearance method (BioPhysics Assay Laboratory; BioPAL, Worcester, MA, USA) according to the manufacturer's instructions and as previously described. [36][37][38] Briefly, mice were injected with GFR-grade inulin (5 mg/kg; IP). Blood samples were then collected at 30, 60 and 90 minutes post injection from the retro-orbital plexus under isoflurane anesthesia. Inulin concentrations in plasma samples were measured using the BioPAL inulin ELISA plate. The data were fit to a one-phase exponential decay equation y = Be −bX , where y is inulin concentration, B is the intercept at time 0, e is the natural logarithm, b is the slope and x is the time.

| Western immunoblotting
Kidney tissue and HK-2 cell samples were homogenized in ice-cold RIPA buffer. Proteins were separated by SDS-polyacrylamide gel (4%-20%) electrophoresis as we have previously described. 41 Immunoreactive proteins were visualized and documented using a gel documentation system (Bio-Rad, Hercules, CA, USA).

| Human proximal tubule epithelial cell line (HK-2)
The use of HK-2 cell line was approved, and experiments were performed in accordance with the guidelines and regulations of the Institutional Biosafety Committee of the UTHSC. The cell line (CRL-2190) was purchased from the American Type Culture Collection (Manassas, VA, USA). The cells were cultured as we have previously described. 31 2.9 | Apoptosis, cytotoxicity and cleaved Notch1 assays Apoptosis of HK-2 cells was examined in real time using the Cell-Player caspase-3/7 reagent, and the IncuCyte ZOOM live content microscopy system (Essen BioScience, Ann Arbor, MI, USA) as we have previously described. 31,42 Cytotoxicity was also determined using the LDH colorimetric assay kit (Life Technologies). LDH release and percent cytotoxicity were quantified according to the manufacturer's instructions. The levels of cleaved Notch1 in HK-2 cell lysates F I G U R E 1 Schematic illustration of experimental groups. A group of mice was given a single IP injection of cisplatin (15 mg/kg). Mice in the control groups were treated (IP) with Captisol (vehicle) or DAPT (15 mg/kg) alone. Another group received DAPT 1 h before cisplatin administration, followed by daily DAPT injection for 4 days. The injection volume was kept at 10 μL/g body weight were evaluated with the PathScan Cleaved (Val1744) Notch1 ELISA kit (Cell Signaling Technology, Danvers, MA, USA).

| siRNA transfection
Complexes consisting of a non-targeting control or a pool of 3 target-specific Dll1 siRNAs (Santa Cruz Biotechnology, Santa Cruz, CA, USA) and TransIT-TKO Transfection Reagent (Mirus Bio, Madison, WI, USA) were prepared in Opti-MEM medium (Life Technologies).
HK-2 cells were transfected with the siRNAs and maintained at 37°C; 5% CO 2 for 72 hours. Western blotting was used to confirm effective knockdown of Dll1.   The mean GFR in captisol-treated control mice was~1.5 mL/ min/g kidney weight which was similar to that in mice treated with DAPT alone ( Figure 3F). Cisplatin reduced GFR in the mice by~51% ( Figure 3F); an effect attenuated by DAPT ( Figure 3F).

| DAPT protects against renal morphological changes induced by cisplatin
The kidney-to-body weight ratio in Captisol-and DAPT-treated mice was similar ( Figure 4A). Cisplatin significantly increased the kidneyto-body weight ratio in the mice; an effect abrogated by DAPT (Figure 4A). There were no noticeable histopathological changes in the group of mice treated with the vehicle (Captisol) and DAPT alone, where tubules and glomeruli were normal in appearance ( Figure 4B).
Cisplatin caused tubular injury exemplified by overt necrosis and vacuolar degeneration which were attenuated by DAPT ( Figure 4B Cisplatin-induced increase in PI fluorescence intensity was reversed by DAPT ( Figure 5A,B). Similarly, fast line-scanning of peritubular capillaries showed distorted blood flow indicated by the variable slope in cisplatin-treated mice compared with the constant slope in Captisol-, DAPT-and DAPT + cisplatin-treated mice ( Figure 5C). SONI ET AL.

| 263
Caspase 3 activity was elevated~2-fold in the whole kidneys of cisplatin-treated mice compared with the mice treated with Captisol and DAPT ( Figure 5D). Cisplatin-induced caspase 3 activity was inhibited by DAPT ( Figure 5D). Also, DAPT ameliorated cisplatin-induced upregulation of pro-apoptotic Bax and downregulation of anti-apoptotic Bcl-2 genes in the kidneys (Figure 6A,B). Together, these findings signify that cisplatin-driven Notch signaling elicits renal cell death and peritubular vascular dysfunction.  The Notch pathway regulates nephrogenesis. 12,43 However, recent studies indicated that increased expression and activity of Notch components are associated with AKI and chronic kidney disease. 14,15 The expression levels of Notch ligands, receptors and transcriptional targets were all increased in rodents subjected to renal ischemia-reperfusion. [44][45][46] The Notch pathway is also induced in animals and human with diabetes, kidney fibrosis and glomerulosclerosis. 14,15 Accordingly, suppression of Notch signaling by inhibiting proteolytic cleavage of Notch receptors alleviated acute and chronic kidney disease. 14,15,44,47,48 Data here signify that the basal protein expression level of N1ICD in mouse kidney is low.
Additionally, we did not detect alterations in basal renal function in mice treated with DAPT alone. These findings corroborate previous reports suggesting that basal Notch signaling is reduced or absent in healthy adult kidneys. 14,15 Cisplatin increased Dll1 and N1ICD expression in mouse
Given that the Notch pathway is regulated by a variety of signaling molecules, including five ligands, our data did not unequivocally determine that Dll1 induction was solely responsible for N1ICD cleavage and renal injury in the mice. We show that cisplatin inversely altered dll1 and Jag1 expression in whole kidneys and cultured HK-2 cells. It is conceivable that cisplatin triggers compensatory expression and function of renal Dll1 and Jag1 as Notch ligands may respond to different cellular signals or stimulate distinct Notchdependent biological processes. 49,50 Jag1 expression was increased in obstructed mouse kidneys and TGF-β1-treated renal cortical epithelial cells. 51 Dll1, but not Jag1, Jag2, Dll3 or Dll4 expression levels were increased in the kidneys of rats subjected to ischemia/ reperfusion. 46 A study has also reported that cadmium chloride stimulated N1ICD cleavage and cell death while reducing Jag1 expression in HK-2 cells. 52 Together, these studies suggest that alterations in Notch ligand expression and activity in the kidney may depend on the type of renal insult.
Notch signaling is induced in hematologic malignancies such as T-cell acute lymphoblastic leukemia and lymphoma, multiple myeloma and acute myelogenous leukemia as well as solid tumors. [53][54][55] On the other hand, Notch signaling exhibited tumor-suppressing activity in hepatocellular carcinoma, chronic myelomonocytic leukemia and head and neck squamous cell carcinoma. [53][54][55] These reports indicate that the Notch signaling pathway elicits oncogenic or tumor suppressive activity, which may be due to variability in cancer celltype, tumor microenvironment, Notch dosage sensitivity and components of the Notch cascades that are activated. [53][54][55] Notch signaling is upregulated in cisplatin-resistant osteosarcoma, head and neck squamous cell carcinoma and lung, gastric and ovarian cancers. [16][17][18][19][20] Suppression of Notch signaling by DAPT increased cisplatin cytotoxic efficacy against colorectal, nasopharyngeal, ovarian and lung cancer cells. 22,[25][26][27] Therefore, DAPT and other inhibitors of the Notch signaling pathway may overcome cisplatin chemoresistance in responsive tumors. 56,57 Mechanisms of the protective effect of DAPT against cisplatininduced nephrotoxicity appears to include inhibition of caspasedependent and -independent cell death as cisplatin-induced necrosis, caspase 3 activity, increase in pro-apoptotic Bax and a compensatory decrease in anti-apoptotic Bcl-2 were all attenuated by DAPT. It is intriguing that DAPT inhibited cisplatin-induced cell death in the kidney but promoted cisplatin-induced cancer cytotoxicity. 22,[25][26][27] Since pathological induction of the Notch signaling pathway can elicit cell growth or death, 12,13 cisplatin may differentially activate Notchdependent proliferation-and cell death-associated genes in cancer and renal cells.
Exploration of combination therapy consisting of γ-secretase inhibitors and cisplatin to circumvent chemoresistance is currently a subject of research interests. 56,57 Given the reversal of cisplatin-induced nephropathy by DAPT, our study suggests that γ-secretase inhibitors may kill the proverbial two birds with one stone by sensitizing responsive cancer cells to cisplatin and at the same time, protecting the kidneys against injury. Although the Notch signaling pathway is a major target of the proteolytic activities of γ-secretases, other F I G U R E 9 Dll1-dependent Notch1 signaling contributes to cisplatin-induced renal tubular cell death. A and B, Western blot images and bar graphs (n = 3) confirming Dll1 knockdown in Dll1 siRNAtransfected HK-2 cells. C, bar graphs summarizing the levels of cleaved Notch1 in scrambled (control; Scrm) siRNA-and Dll1 siRNA-transfected HK-2 cells (n = 6 each). D, kinetic curves (n = 12 each) demonstrating that cisplatin (30 μmol L −1 )induced caspase-3/7 activity in HK-2 is diminished in Dll1 siRNA-transfected HK-2 cells. $ P < 0.05 vs Scrm siRNA; # P < 0.05 vs Scrm siRNA + cisplatin (8-22 h for "D"); *P < 0.05 vs Scrm and Dll1 siRNA + Captisol (8-22 h) transmembrane proteins including the amyloid β precursor protein (AβPP) can be cleaved by the enzymes. 58,59 However, AβPP intracellular domain has been shown to degrade N1ICD, thereby inhibiting Notch1 signaling. 60 Induction of both renal Dll1 and N1ICD by cisplatin and reversal of renal Notch1 cleavage by DAPT indicate that DAPT mitigates cisplatin-induced AKI by suppressing Notch signaling.
In conclusion, our data suggest that Dll1-mediated Notch1 signaling contributes to cisplatin-induced AKI. Pharmacological inhibition of the Notch signaling pathway preserved renal function and morphology in cisplatin-treated mice and viability in cisplatin-treated human proximal tubule cell line. Hence, inhibition of Notch signaling could be a potential therapeutic strategy to alleviate renal complications associated with cisplatin chemotherapy.

ACKNOWLEDGMENTS
We acknowledge a fellowship (to A. Adebiyi) from the O'Brien Center for Advanced Renal Microscopic Analysis (NIH-NIDDK P30DK079312).

CONF LICT OF I NTERESTS
The authors report no conflicts of interest.