Introduction

Ten percent of all patients with cancer world-wide receive cisplatin as part of their solid tumor chemotherapy regimen, but unfortunately, it is particularly toxic to certain tissues in the body in which the drug accumulates, including the kidney and the inner ear.1–7 Cisplatin concentrates in the kidney during glomerular filtration and tubular secretion, which results in a fivefold increase in cisplatin concentration compared with serum levels, causing severe damage to kidney proximal tubular cells.8–10 Cisplatin buildup in the inner ear causes hair cell death and damage to the supporting cells, neuron fibers, and stria vascularis cells.7,11,12 Despite its potent antitumor effect, the clinical use of cisplatin is dose-limited by nephrotoxicity, which is commonly manifested as AKI or chronic renal dysfunction.13–17 The renal function impairment can be progressive and eventually lead to CKD.18 Cisplatin can also cause renal tubular dysfunction, salt wasting, hypomagnesemia, and anemia.19 These side effects are observed in roughly one third of patients, despite intensive prophylactic measures, such as hydration, diuresis, magnesium supplementation, and amifostine treatment.1,4,16 Moreover, the incidence and severity of renal failure increases with subsequent courses and can eventually become irreversible.8,20 As a result, discontinuing therapy with cisplatin is generally indicated in those who develop the evidence of progressive renal impairment.21 More than half of the patients also suffer from irreversible hearing loss after cisplatin treatment.22,23 To date, no drugs have been approved by the US Food and Drug Administration (FDA) for protection from cisplatin-induced kidney damage and hearing loss.24–26

In this study, we reasoned that drugs that protect from ototoxicity may also alleviate renal toxicity. Cisplatin has been shown in both tissue types to cause DNA damage, mitochondrial injury, production of reactive oxygen species, triggering of inflammatory responses, and cell death signaling pathways, including pyroptosis and necroptosis.27,28 The kidney and inner ear cells share similar transport proteins, such as ATP6V1B1 and ATP6V0A4, which can contribute to relatable pharmacologic activity of drugs in the two organs.24–26,29,30 Herbal Chinese medicine points to similarities between drugs that work in the kidney and ear tissues.31–33 Interestingly, congenital abnormalities, such as Alport syndrome and branchio-oto-renal syndrome, share renal and hearing defects.18,31

Protein kinases are particularly promising therapeutic targets as they regulate critical cellular functions, and many kinase inhibitors have been approved by the FDA for effective cancer therapy.34,35 Recently, we reported that two protein kinase inhibitors and anticancer drugs, a clinical phase 2 drug CDK2 kinase inhibitor AZD5438,22,36–39 and an FDA-approved drug dabrafenib40–42 protect the postmitotic inner ear cells against cisplatin-induced ototoxicity in mice by oral delivery. Both drugs were identified as top hits in inner ear cell line–based high-throughput screens for protection from cisplatin-induced cell death and were shown to not interfere with cisplatin killing efficacy in lung and neuroblastoma cancer cell lines.22,24 Here, we evaluated the two drugs activity against cisplatin-induced kidney injury in vitro, in human kidney 2 (HK-2) cells, and in vivo in adult mice. The anticancer activity of the drugs with cisplatin was confirmed killing activity of the drugs with or without cisplatin was quantified using a neuroblastoma cell line and two testicular cancer cell lines, tumor types in which cisplatin is the standard of care. We studied the mechanism of action of AZD5438 in the kidney by analyzing CDK2-null mice and by complementing wild type (WT) and knockout (KO) CDK2 mice with the drug. Dabrafenib was shown to inhibit, in vivo, the phosphorylation of the downstream target ERK1/2. Our experimental results identify the two drugs as promising therapeutic candidates for the prevention of both cisplatin-induced kidney injury and hearing loss.

Methods Animals

Friend Virus B/NIH Jackson (FVB/NJ) mice (Strain 001800) were purchased from The Jackson Laboratory and bred in Creighton University animal facility. The germline CDK2 KO mice were derived, as described in Teitz et al., in 2018 by crossing CDK2-floxed/floxed mice with EIIA-Cre on C57BL/6/129 mixed backgrounds, from which offspring possessing the CDK2 deletion and lacking Cre were intercrossed to obtain homozygous CDK2 germline KO mice.22 Heterozygous mixed background CDK2 KO mice were bred in Creighton University animal facility for five generations to the FVB/NJ background. All animal experiments were approved by the Institutional Animal Care and Use Committee of Creighton University.

Cell Lines

Human immortalized proximal tubular HK-2 cells (CRL-2190) were purchased from the American Type Culture Collection (ATCC) and were maintained as per ATCC specifications in keratinocyte serum-free medium (Invitrogen, Gibco, Catalog Number 17005-042) and supplemented with additives 0.05 mg/ml bovine pituitary extract and 5 ng/ml EGF human recombinant EGF, to promote growth, purchased from Invitrogen (Gibco). The cell line was grown at 37°C and 5% CO2 and passaged using 0.05% (w/v) trypsin—0.53 mM EDTA 1× (25300-054, Gibco). Testicular cell lines (JKT-1 and NCCIT) and neuroblastoma cell line (Kelly) were grown in Roswell Park Memorial Institute 1640 (A1049101) medium with the ATCC modification (Gibco) with a final concentration of 10% FBS and 200 µl of 0.04 mg/ml ampicillin.

Drugs

AZD5438 (HY-10012) and dabrafenib (HY-14660A) were obtained from MedChem Express.

CellTiter-Glo Viability Assays

Human immortalized proximal tubular HK-2 cells were plated in 96-well plates in six replicates and incubated overnight at 37°C in 5% CO2. After 24 hours, 9600 cells were pretreated with AZD5438 or dabrafenib, and 1-hour later, cells were treated with 5 µM cisplatin plus drugs AZD5438 or dabrafenib for 48 hours. CellTiter-Glo Luminescent Assay (Promega, G7571) was used to determine the number of viable cells in culture on the basis of the quantitation of the ATP present, an indicator of metabolically active cells. For controls, media alone, cisplatin alone, and drug alone treated cells were tested. Cell viability was calculated as percent survival compared with media alone treated cells.

For tumor cell lines viability experiments, two testicular cell lines (JKT-1 and NCCIT) and a neuroblastoma cell line (Kelly) were used. In total, 9600 cells per well were plated in six replicates in 96-well plates and allowed to attach overnight at 37°C in 5% CO2. The following day, the tumor cell lines were pretreated with 3 µM of dabrafenib or AZD5438 alone for 1 hour. The cells were then treated with 10 µM cisplatin and incubated for 48 hours. Cell viability was then measured using the CellTiter-Glo (Promega) assay. Medium alone, cisplatin alone, and drug alone were used as controls, and the percent viability was calculated as the viability compared with the medium alone treated cells. Drug plus cisplatin-treated cells were compared with cisplatin alone treated cells to determine whether dabrafenib or AZD5438 interfered with cisplatin's tumor-killing ability.

Mouse Model for Cisplatin-Induced AKI

Ten mg cisplatin (479306; Sigma-Aldrich) powder was dissolved in 10 ml sterile saline (0.9% NaCl) at 37°C for 40–60 minutes. Cisplatin at doses (15, 20, 25, and 30 mg/kg) was administered by intraperitoneal (IP) injection to 8–10-week-old FVB/NJ mice, females, and males in equal ratio. To decrease the dehydration during cisplatin treatment, mice were injected subcutaneously with sterile saline 1 ml 1 day before cisplatin injection and twice daily throughout the experimental period. The cisplatin-treated mice were placed on a heating pad, and fresh water-mushed food was given daily. Body weights were monitored daily. Blood samples were collected by cardiac puncture on day 3 or day 21 after cisplatin treatment for quantifying kidney injury markers in serum. Kidney tissues were harvested for histopathology and Western blot analysis.

Administration of Drugs by Oral Gavage

AZD5438 and dabrafenib mesylate were dissolved in a solution containing 10% DMSO (BP231-100, Fisher), 5% Tween 80 (P1754, Sigma-Aldrich), 40% PEG-E-300 (91462, Sigma-Aldrich), and 45% saline (0.9% NaCl), as described previously.24 The mice were pretreated with AZD5438 (35 mg/kg) or dabrafenib (12 mg/kg) by oral gavage 45 minutes before injecting with cisplatin. The mice were given AZD5438 by oral gavage twice daily for three consecutive days. Dabrafenib was administered three times daily, 6 hours apart, for three consecutive days. Mice were sacrificed after 72 hours (day 3). Blood was collected by cardiac puncture. Kidneys were harvested and transferred immediately to 10% neutral buffered formalin for histological studies, and tissue was stored at −80°C for Western blotting. Different doses of AZD5438 (30, 35, and 40 mg/kg) were tested for its protective effects. In addition, 12 mg/kg body weight dose of dabrafenib was used.

For 14-day mouse survival studies, cisplatin (22 mg/kg) was IP-injected and AZD5438 and dabrafenib was given orally twice daily for three consecutive days. For longer-term studies of 21 days, mice were injected with 30 mg/kg of cisplatin and given 12 mg/kg of dabrafenib twice daily for three consecutive days by oral gavage.

Kidney Injury Markers BUN, Neutrophil Gelatinase-Associated Lipocalin, and Creatinine to Determine Renal Function

BUN serum levels were analyzed with BUN colorimetric detection kit (EIABUN, Thermo Fisher Scientific). Neutrophil gelatinase-associated lipocalin (NGAL) serum levels were analyzed by ELISA using Invitrogen Lipocalin-2 mouse ELISA KIT (EMLCN2). Creatinine serum levels were analyzed by Colorimetric Assay Kit (Cayman Chemical, 700460). Control proteins provided in the kits were used to graph a dose-response curve.

Kidney Histologic Examination

Mouse kidney tissues were fixed in 10% neutral buffered formalin solution for 24 hours, embedded in paraffin, sectioned (3 µm), and stained with hematoxylin and eosin (H&E) or periodic acid–Schiff (PAS). Three sections from each mouse were observed under a microscope (Nikon Eclipse Ci) for histologic examination. Semiquantitative pathological scoring system was used, as described in references.43,44 The grading system used scores 0, 1, 2, and 3, which indicate the percentage of damage in each section and were examined by an experienced pathologist in a double-blinded manner. Grade 0—no visible injury and normal kidney morphology; Grade 1—mild tubular dilation, presence of casts, condensed nuclei, and partial loss of brush borders in the membranes in <1/3 tubules; Grade 2—clear dilation of many tubules, loss of brush border membranes, loss of nuclei, and presence of casts in <2/3 of tubules; and Grade 3—severe dilation of most tubules and complete loss of brush border membranes and loss of nuclei in >2/3 of tubules.

Terminal Deoxynucleotidyl Transferase dUTP Nick End Labeling Assay

Cell death in the mouse kidney sections was performed by terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay kit (Abcam, ab66110) according to manufacturer instruction. In brief, harvested mouse kidneys were fixed in 10% formaldehyde for 24 hours and then embedded in paraffin. Three-micron thick sections were then sliced, deparaffinized, rehydrated using ethanol gradient (100%, 95%, 85%, 70%, 50%), and transferred to 0.85% NaCl solution for 5 minutes and PBS for 5 minutes. After washing with PBS, 20 µg/ml proteinase K is added and fixed with 4% formaldehyde. The sections were labeled with 5-Bromo-2´-Deoxyuridine 5´-Triphosphate- for 60 minutes at room temperature and treated with anti–BrdU-Red antibody for half an hour. After coating with an antifading solution (Fluoromount-G, Thermo Fisher Scientific), the sections were covered with a glass cover slip and sealed. The images are scanned using Olympus BX61VS; Olympus America, Center Valley, Pennsylvania, fluorescent microscope, and the stained area is quantified using National Institutes of Health software ImageJ.45 Each experimental group had 3–5 mice. Three kidney sections were analyzed from each mouse, and five images from each kidney section (covering more than 80% of the kidney total area) were quantified for mean fluorescence intensity (MFI) using ImageJ software. The nuclei number was analyzed from each image using “Analyze Particle” mode. The ratio of MFI to the number of nuclei was used for each image for statistical analysis.

Proliferating Cell Nuclear Antigen and pERK Immunohistochemistry

Tissue sections were analyzed for the expression of proliferating cell nuclear antigen (PCNA) and pERK1/2 by immunostaining using a standard protocol.46 Antigen retrieval was performed using heat-induced epitope retrieval buffer at 95°C for 20 minutes. The sections were blocked in PBS containing 0.25% Triton-X and 10% FBS for 30 minutes at room temperature. Primary antibodies, anti-PCNA (MA5-11358) or anti-pERK1/2 (9101), both purchased from Cell Signaling Technology, were used at a dilution of 1:500 in 10% FBS in PBS solution for overnight incubation. After washing with PBS, the slides were incubated for 1 hour with respective secondary antibodies anti-mouse Alexa Fluor 647 (A31571) or anti-rabbit Alexa Fluor 488 (A11034), purchased from Invitrogen, at dilution of 1:500. Kidney sections were then counterstained with the nuclei marker, 4′,6-diamidino-2-phenylindole (DAPI) for 10 minutes, and covered with glass coverslip using antifading solution and sealed. Slides were imaged using a fluorescent microscope (Olympus BX51; Olympus America, Center Valley, PA). Each experimental group had six mice, and three kidney sections were analyzed from each experimental mouse. Five square areas from each kidney section, covering more than 80% of the total kidney area, were imaged and used for quantification of MFI using ImageJ software. The nuclei number was analyzed from each image using “Analyze Particle” mode. The ratio of MFI to the number of nuclei was used for each image for statistical analysis.

Western Blotting

HK-2 cells and kidney tissue protein lysates were prepared in cell lysis buffer (9803, Cell Signaling Technology), containing protease (complete ULTRA Tablet 05892791001) and phosphatase (Phos STOP 04 906 837 001) inhibitors (Roche). The lysates were centrifuged for 10 minutes at 15,000 g at 4°C, and the supernatants were collected. Protein concentrations in supernatants were determined with the bicinchoninic acid assay protein kit (23235, Thermo Fisher Scientific). Forty micrograms of total cell lysates for HK-2 cells or 60 micrograms of kidney lysates were loaded on 10% Tris-Glycine-polyacrylamide gel electrophoresis gels. The following antibodies used for immunoblotting were anti-B-Rapidly Accelerated Fibrosarcoma (BRAF) (14814), anti-pBRAF (Ser445, 2696), anti-ERK1/2 (4695), anti-pERK1/2 (Thr202/Tyr204, 9101), anti-MEK1/2 (9122), anti-pMEK1/2 (Ser217/221, 41G9), anti-RIP3 (95702), anti-RIP1 (3493), pMLKL (3733), gasdermin E (GSDME) Cleaved (38821), GSDME (AB215191) purchased from Abcam, CASPASE-3 Cleaved (14220) were obtained from Cell Signaling Technology, anti–β-actin (C4; SC-47778), was purchased from Santa Cruz Biotechnology, and anti-α Tubulin (T9026, Millipore sigma) and total mixed lineage kinase domain-like (MLKL) (MAB C604) was purchased from EMD Millipore. The antibodies were used at dilutions ranging from 1:500 to 1:1000. Anti-mouse (P0447) and anti-rabbit (P0448) secondary antibodies were purchased from Dako Laboratories and diluted 1:5000. Images were quantified, and band intensities were recorded as a ratio to loading control. Blot intensities were quantified using National Institutes of Health Image J software. As positive controls in Western blots, Phospho-MLKL Positive Control, product PC-PMLKL, FabGennix International Inc, and Caspase-3 Control, Cell Extracts Cell Signaling Technology 9663, were used.

Statistical Analysis

The results are expressed as mean±SEM using Graph Pad Software, *P < 0.05, **P < 0.01, ***P < 0.001, compared using one-way or two-way ANOVA with the Bonferroni post hoc test to analyze data and the unpaired t test if only two conditions were compared.

Results AZD5438 and Dabrafenib Protected Human Kidney HK-2 Proximal Tubular Cells from Cisplatin-Induced Cell Death

To determine IC50 for cell death with cisplatin in the human HK-2 proximal tubular cell line, cells were treated with increasing concentrations of cisplatin for 48 hours, and cell viability was determined by CellTiter-Glo assay. Treatment with cisplatin showed a dose-dependent increase in cell death with an IC50 of 5 µM (Figure 1A), consistent with previous results published for this cell line.47 Dose-response of AZD5438 or dabrafenib alone showed no cytotoxicity (Figure 1, B and C). HK-2 cells were then pretreated with different concentrations of AZD5438 or dabrafenib for 1 hour and cotreated with 5 µM cisplatin and either drug for 48 hours. The results showed that AZD5438 and dabrafenib pretreatment protected HK-2 cells from cell death with an IC50 of 0.026 and 0.77 µM, respectively (Figure 1, B and C).

Figure 1:

AZD5438 and dabrafenib protect the human proximal tubule cell line HK-2 from cisplatin-induced cell death while not interfering with cisplatin's tumor-killing efficacy. (A) HK-2 cells were treated with increasing concentrations of cisplatin for 48 hours. CellTiter-Glo assay was performed, and dose-dependent increase in cell death of HK-2 cells was observed with an IC50 of 5 µM. (B) HK-2 cells were pretreated with different doses of AZD5438, starting 1 hour before 5 µM cisplatin treatment for 48 hours, and IC50 of 0.026 µM was determined for the protection of cell viability. (C) HK-2 cells were pretreated with different doses of dabrafenib, starting 1 hour before 5 µM cisplatin treatment for 48 hours, and IC50 of 0.77 µM was determined for the protection of cell viability. (D) Three human cancer cell lines, neuroblastoma Kelly, testicular seminoma JKT-1, and pluripotent embryonal carcinoma cell line NCCIT, were studied. The tumor cell lines were initially pretreated with 3 µM of dabrafenib, AZD5438, or both drugs for 1 hour and then treated with 10 µM cisplatin and 3 µM AZD5438, dabrafenib, or both drugs and incubated for 48 hours. Cell viability was measured at 48 hours using the CellTiter-Glo (Promega) assay.

AZD5438 and Dabrafenib Do Not Interfere with or Enhance Cisplatin Tumor-Killing Efficacy in Tumor Cells

It is essential to test that AZD5438 and dabrafenib cotreatments with cisplatin for reducing AKI damage would not interfere with cisplatin killing efficacy of the treated tumors. Previously in our hearing-focused studies, we have shown that AZD5438 and dabrafenib do not interfere with tumor-killing efficacy of cisplatin in three neuroblastoma and three lung cancer cell lines.24 Here, we broadened our tests to include an additional neuroblastoma genetic cell line Kelly and two testicular cancer cell lines, seminoma JKT-1, and pluripotent embryonal carcinoma cell line, NCCIT. Neuroblastoma and testicular tumors are routinely treated with cisplatin as part of standard medical care.48,49 To evaluate interference with or enhancement of cell survival with drug treatments, cells were treated with the drugs alone or in combination with cisplatin for 48 hours (Figure 1D). Cell viability was measured by CellTiter-Glo assay in which 100% cell viability was the medium alone treated cells. The results indicated no interference with cisplatin killing activity in the three tumor cell lines tested (Figure 1D).

Oral Administration of AZD5438 or Dabrafenib Protected from Acute Cisplatin-Induced Renal Injury in Mice, Measured by Reduction in the Levels of Injury and Histopathology Biomarkers on Day 3

To generate an AKI in vivo model, we used an established protocol in FVB/NJ strain mice.43 Renal injury was determined on day 3 after cisplatin treatment by the serum levels of BUN and creatinine, markers for kidney damage,50–52 and renal NGAL, an early kidney injury marker.43,53

To test the cisplatin-induced nephrotoxicity in our mouse model and to evaluate how it mimics the human ailment, escalating doses of cisplatin (15, 20, 25, 30 mg/kg body weight) were given by IP injection to 8–10-weeks-old FVB/NJ mice, and serum levels of BUN, NGAL, and histopathology markers were measured 72 hours after cisplatin treatment (Supplemental Figure 1).

The histologic scores of the cisplatin-treated kidneys were determined in a double-blind manner, by an experienced pathologist, using a grading system, as described previously.43,44 H&E staining and PAS staining of kidney sections of mice treated with 25 or 30 mg/kg cisplatin showed extensive tissue damage characterized by the presence of cellular casts, loss of brush border, presence of dying cells, and inflammation 72 hours after cisplatin treatment, while lower doses of cisplatin exhibited less damage (Supplemental Figure 1, C and D). The recorded histologic grades in mice treated with escalating cisplatin doses correlated well with the measured BUN and NGAL levels (Supplemental Figure 1, A and B).

Using the mouse AKI model, 35 mg/kg dose of AZD5438 twice daily (Figure 2A) and 12 mg/kg dabrafenib three times daily (Figure 2D) were tested against 25 mg/kg cisplatin for their nephroprotective effects. The doses/regimen of the AZD5438 and dabrafenib drugs were chosen based on the lowest effective in vivo doses determined in our previous studies to be nontoxic to the inner ear cells and to confer protection against cisplatin-induced hearing loss in mice.24 The drug doses were confirmed in this study to be nontoxic to kidney cells by themselves or with cisplatin cotreatment in mice (Figure 2 and Supplemental Figure 2).

Figure 2:

AZD5438 and dabrafenib protect against cisplatin-induced kidney injury in adult FVB/NJ mice. (A) Experimental protocol of cisplatin IP injection, saline administration, and AZD5438 oral delivery. (B) Quantification of kidney function markers and overall kidney morphology. BUN, creatinine, and NGAL levels were measured for kidney function, and histology scores were determined for overall kidney function. (C) Renal tissues were stained with H&E and PAS stain. Kidney histologic scores were obtained by a pathologist blinded to the experimental conditions using a semiquantitative pathologic scoring system described in Methods. (D) Experimental protocol of cisplatin injection, saline administration, and dabrafenib oral delivery. (E) Quantification of BUN, creatinine, and NGAL levels and kidney histology scores. (F) Renal tissues were stained with H&E and PAS stain and scored by a pathologist blinded to the experimental conditions In all bar graphs, dots represent the number of mice (n=6–17). Values expressed are mean±SEM, *P < 0.05, ***P < 0.001 using one-way ANOVA with Bonferroni post hoc test. Scale bar: 50 µm. H&E, hematoxylin and eosin; IP, intraperitoneal; NGAL, neutrophil gelatinase-associated lipocalin; PAS, periodic acid–Schiff.

In the mice, serum kidney injury markers were highly elevated on day 3 with cisplatin treatment, as determined by colorimetric detection of BUN and creatinine, and ELISA measurements of NGAL. Average BUN values were 108±6 mg/dl, creatinine values were 4.3±0.2 mg/ml, and NGAL values were 96±6 ng/ml, consistent with a previous publication using this mouse model.43 In comparison, mice given oral gavage of carrier alone had BUN values of 34±5 mg/dl, creatinine values of 0.2±0.03 and NGAL of 7±2 ng/ml (Figure 2). Importantly, AZD5438 cotreatment resulted in 2.1-fold decrease in BUN values to 51±7 mg/dl, 4.3-fold decrease to 1.0±0.1 mg/dl creatinine, and 1.6-fold decrease in NGAL to 59±8 ng/ml (Figure 2B). For dabrafenib cotreatment, BUN decreased 1.8-fold to 59±8 mg/dl, creatinine decreased 2.1-fold to 2.0±0.2 mg/dl, and NGAL decreased 1.4-fold to 65.3±4.0 ng/ml (Figure 2E). Scoring of H&E and PAS-stained kidney sections indicated reduced values from 3±0 grade kidney damage in cisplatin alone treated mice to 0.5±0.2 with AZD5438 cotreatment (Figure 2, B and C) and 0.8±0.2 with dabrafenib cotreatment (Figure 2, E and F).

AZD5438 Inhibited Cell Proliferation Measured by PCNA Levels and Cell Death Measured by TUNEL in Cisplatin Cotreated Kidneys on Day 3

Previous reports of cisplatin treatment in mice measured the elevated levels of PCNA occurring in postmitotic kidney cells undergoing AKI.43,54 Therefore, we evaluated PCNA levels after AZD5438 cotreatment knowing the drug is a CDK2 kinase inhibitor that blocks proliferation and cell cycle progression that occurs with PCNA upregulation.49 PCNA immunofluorescent staining of kidney sections of cisplatin cotreated mice with AZD5438 indeed showed a 2.2-fold decreased expression of PCNA protein on day 3 compared with cisplatin alone treated mice (Figure 3, A and B).

Figure 3:

PCNA levels and cell death are downregulated with AZD5438, while dabrafenib prevents pERK upregulation and cell death in mouse kidneys treated with cisplatin. Representative immunofluorescent images of (A) PCNA and (C) TUNEL staining in kidney sections of experimental mice treated with carrier, cisplatin, or cisplatin and AZD5438. Quantification of PCNA signals is shown in (B), and TUNEL positive cells in (D) DAPI staining were used to localize nuclei in kidney sections. Intensity of PCNA expression and TUNEL positive cells was determined using ImageJ software. MFI per nuclei was quantified for each specimen. Three sections were quantified for each mouse. The graph represents MFI per cell. Representative fluorescent images of (E) pERK1/2 and (F) graph showing quantification of pERK1/2. (G) TUNEL staining in kidney sections of experimental mice. (H) Graph shows the quantification of TUNEL-positive cells expression. DAPI staining was used to localize nuclei. The intensity of TUNEL-positive cells and pERK staining was acquired, and the MFI per nuclei was quantified from each specimen. In all bar graphs, dots represent the number of mice (n=6). Values expressed are mean±SEM *P < 0.05, **P < 0.01, ***P < 0.001 using one-way ANOVA with the Bonferroni post hoc test. Scale bar: 50 µm. MFI, mean fluorescence intensity; PCNA, proliferating cell nuclear antigen; TUNEL, terminal deoxynucleotidyl transferase dUTP nick end labeling.

To determine whether the mechanism of protection of AZD5438 against cisplatin-induced AKI is through the reduction of cell death in vivo, an immunofluorescence TUNEL assay to detect double-strand DNA breaks generated during cell death was performed on kidney sections of the different mouse experimental groups. As previously reported,55 cisplatin-treated mouse kidneys showed a high presence of DNA breaks per cell (0.06±0.005 MFI/cell), in both proximal convoluted tubule and glomeruli, compared with the carrier alone mouse group with 0.016±0.003 MFI/cell. The renal tubular cells and glomeruli of the AZD5438-treated mice had a 3.0-fold lower number of double-strand DNA breaks (0.02±0.004 MFI/cell) compared with the cisplatin-treated mouse group and close to the average number of DNA breaks of the carrier only mouse group (Figure 3, C and D). Overall, both PCNA protein levels and cell death were significantly reduced in AZD5438 cotreated mice.

Dabrafenib Inhibited Cisplatin-Induced Extracellular Signal-regulated Kinase 1/2 (ERK1/2) Phosphorylation and Cell Death Measured by TUNEL Assay in Kidneys

Mice cotreated with dabrafenib and cisplatin had a reduction in pERK levels compared with cisplatin alone, showing 2.6-fold reduction of pERK with dabrafenib cotreatment compared with cisplatin alone treated mice (Figure 3, E and F). TUNEL assay performed on mouse kidney sections showed increased cell death in cisplatin alone treated mice (average 0.07±0.007 MFI/cell) on day 3 of treatments and significant protection from cell death (average 0.03±0.008 MFI/cell) with dabrafenib cotreatment (Figure 3, G and H). In summary, dabrafenib cotreatment decreased extracellular signal (en)regulated kinase phosphorylation of the kidney tissues at 72 hours after treatment and decreased total cell death at 72 hours in vivo.

Understanding the cellular mechanisms through which cisplatin induces nephrotoxic effects is important for identifying compounds for treatment. For dabrafenib, a specific BRAF inhibitor, we tested first whether the canonical mitogen-activated protein kinase cellular pathway (schematically illustrated in Figure 4A) is upregulated with cisplatin treatment by following pBRAF (S445), pMEK (S217/221), and pERK (T202/Y204) protein expression, the activated forms of the kinases, in kidney HK-2 cells treated with cisplatin. Five micromolar cisplatin treatment of HK-2 (IC50 for death of HK-2 cells by cisplatin, presented in Figure 1A) was found to increase pMEK and pERK expression after 5 hours of cisplatin treatment by Western blotting (Figure 4B). For testing in vivo, mice were IP-injected with cisplatin and pERK1/2 protein was quantified in total kidney protein lysates at time points of 0, 1, 2, 5, and 8 hours after cisplatin treatment and compared with cisplatin and dabrafenib cotreatment (Figure 4, C and D). An additional experiment tested pERK protein levels at 72 hours postcisplatin treatment in individual mice, with or without oral treatment of dabrafenib, 12 mg/kg three times a day for 3 days (Figure 4, E and F). The levels of pERK protein after cisplatin treatment were found to be 6.3-fold upregulated at the 5 hour and 14.5-fold at 72 hour time points compared with carrier alone mice. At both time points, the average pERK protein levels were reduced 3.1-fold and 2.7-fold, respectively, with dabrafenib cotreatment (Figure 4, D and F, and Supplemental Figures 3–5, 8, and 9). The Western blot results at 72 hours were consistent with the immunofluorescence data (Figure 3, E and F), showing 2.6-fold reduction of pERK with dabrafenib cotreatment compared with cisplatin alone treated mice.

Figure 4: Dabrafenib inhibits phosphorylation of ERK1/2 in cotreated mice. (A) Putative pathway for MAPK activated by cisplatin in kidney cells. (B) HK-2 cells were treated with cisplatin, and the expression of pBRAF, pMEK, and pERK was quantified by Western blot. The ratio of phosphorylated protein bands to β-actin bands was marked. (C) Representative Western blot images for the levels of pERK1/2 and ERK1/2 in kidney lysates from mice treated with cisplatin (25 mg/kg) or cisplatin and dabrafenib (12 mg/kg) at—0, 1, 2, 5, and 8 hours post-treatments. (D) Quantification of three independent Western blot experiments (the other two blots are shown in Supplemental Figure 3). The ratio of pERK1/2 levels to α-tubulin was plotted. (E) Western blot images for levels of pERK1/2 and ERK1/2 in kidney lysates from mice treated with carrier, cisplatin (25 mg/kg), or cisplatin and dabrafenib (12 mg/kg) at 72 hours. Dabrafenib (12 mg/kg body weight) was given by oral gavage, three times daily, 4 hours apart. (F) Quantification of Western blot. The ratio of pERK1/2 levels to β-actin was plotted. In all bar graphs, dots represent the number of mice (n=3). Values expressed are mean±SEM, **PPpost hoc test. MAPK, mitogen-activated protein kinase.Deletion of CDK2 Attenuates Cisplatin-Induced AKI and Complementation of the CDK2 KO Mouse with AZD5438 Administration Does Not Confer Additional Protection

To confirm that the mechanism through which AZD5438 offers protection is inhibition of CDK2 and no other off-targets, a germline CDK2 KO mouse model in which the CDK2 gene is deleted in all tissues was used.22 Importantly, the CDK2 KO mouse was backcrossed for five generations to the FVB/NJ background to compare with the genetic background, the AKI model was tested. CDK2 KO mice were viable and had normal kidney structure. CDK2 WT and KO mice littermates were subjected to cisplatin IP injection (25 mg/kg), and on day 3, their blood and tissue samples were analyzed. As shown in Figure 5F, BUN levels in CDK2 WT (93.2±2.5 mg/dl) mice were significantly higher compared to CDK2 KO (67.1±3.7 mg/dl) mice. Histology results showed that the kidneys of CDK2 KO mice exhibited less tissue damage with a histology grade of 1.4±0.8 compared to grade 2.8±0.4 for WT mice (Figure 5, A–C). In addition, CDK2 WT mice treated with cisplatin on day 3 had higher levels of cell death by TUNEL assay (0.060±0.008 MFI/cell), compared with the CDK2 KO mice undergoing the same treatment (0.027±0.009 MFI/cell) (Figure 5, D and E). Significantly, the administration of AZD5438 to the WT and KO CDK2 mice resulted in the similar level of kidney protection, and no additional kidney protection was achieved when AZD5438 was given to the KO CDK2 mice (Figure 5F). Thus, CDK2 is a key target in cisplatin-induced kidney damage, and the protective effect of AZD5438 is primarily through the inhibition of this cellular target.

Figure 5:

CDK2 KO mice are resistant to cisplatin-induced nephrotoxicity compared with their WT littermates. CDK2 WT and KO mice (FVB/NJ background) were injected with cisplatin 25 mg/kg body weight and analyzed 72 hours postcisplatin treatment. Significant difference in (A–C) histology, (D and E) TUNEL MFI per cell, and (F) BUN levels were observed. In all bar graphs, dots represent the number of biologically independent samples (n=6–10 mice). Values expressed are mean±SEM, *P < 0.05, **P < 0.01, compared using the t test (C and E) and one-way ANOVA with the Bonferroni post hoc test (F). Scale bar: 50 µm. KO, knockout.

Dabrafenib Inhibited Kidney Injury on Day 21 Postcisplatin Treatment

Dabrafenib 3-day treatment protected hearing loss in adult FVB/NJ mice after one dose of 30 mg/kg cisplatin that induces permanent hearing loss.24 Functional hearing was measured at the 21-day postcisplatin treatment by Auditory Brainstem Response threshold shifts.24 In this study, we tested whether dabrafenib can also confer kidney protection at the 21-day postcisplatin treatment (Figure 6A). Tissue injury was still extensive in cisplatin alone injected mice on day 21 (Figure 6, B–D), and dabrafenib cotreatment resolved the tissue damage, as demonstrated by scoring H&E and PAS-stained kidney sections (Figure 6, B–D). Cisplatin alone treated mice showed enhanced levels of pERK1/2 protein and TUNEL levels per cell on day 21 after treatment compared with carrier alone treated mice (Figure 6, E–H). By contrast, dabrafenib cotreated mice showed 2.1-fold reduction of pERK protein levels and 1.5-fold reduction of TUNEL signals per cell (Figure 6,