mitochondrial dysfunction in reperfused kidneys after warm ischaemia

(1)

Donation of kidneys after brain death van Dullemen, Leon

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.

Document Version

Publisher's PDF, also known as Version of record

Publication date:

2017

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):

van Dullemen, L. (2017). Donation of kidneys after brain death: Protective proteins, profiles, and treatment strategies. Rijksuniversiteit Groningen.

Copyright

Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.

Download date: 20-07-2021

(2)

512316-L-bw-Dullemen 512316-L-bw-Dullemen 512316-L-bw-Dullemen 512316-L-bw-Dullemen Processed on: 9-8-2017 Processed on: 9-8-2017 Processed on: 9-8-2017

Processed on: 9-8-2017 PDF page: 155PDF page: 155PDF page: 155PDF page: 155

Lipid catabolism provides an alternative energy source and compensates for

mitochondrial dysfunction in reperfused kidneys after warm ischaemia

Honglei Huang Leon F.A. van Dullemen Mohammed Z. Akhtar Maria-Letizia Lo Faro Zhanru Yu

Alessandro Valli Anthony Dona

Marie-Laëtitia Thézénas Philip D. Charles Roman Fischer Maria Kaisar Henri G.D. Leuvenink Rutger J. Ploeg Benedikt M. Kessler

7

(3)

Processed on: 9-8-2017 PDF page: 156PDF page: 156PDF page: 156PDF page: 156 156

ABBREVIATIONS

ACADSB: Short/branched Chain Acyl-CoA Dehydrogenase AKI: Acute Kidney Injury

AMP: Adenosinemonophosphate APO: Apolipoprotein

BDH1: 3-Hydroxybutyrate Dehydrogenase Type 1 BVR: Biliverdin Reductase A

CD36: Cluster of Differentiation 36 / Fatty Acid Translocase

CD44: Cluster of Differentiation 44 / Homing Cell Adhesion Molecule (HCAM) CPT1A: Carpitine Palmitoyltransferase 1A

DBD: Donation After Brain Death DCD: Donation After Circulatory Death

ECHDC3: Enoyl-CoA Hydratase Domain Containing 3 EGFR: Epidermal Growth Factor Receptor

EIF2: Eukaryotic Initiation Factor 2 FA: Fatty Acids

FABP4: Fatty Acid Binding Protein 4 FFA: Free Fatty Acids

GST: Glutathione S-Transferase HK1: Hexokinase 1

HO-1: Heme Oxygenase 1

HSP70: Heat Shock Protein 70 (HSPA1A) ICAM-1: Intercellular Adhesion Molecule 1 IRI: Ischaemia Reperfusion Injury

LXR/RXR: Liver X- and Retinoid X Receptor NAD: Nicotinamideadeninedinucleotide

NADH: Reduced Nicotinamideadeninedinucleotide PPAR: Peroxisome Proliferator-Activated Receptor RCR: Respiratory Control Ratio

ROS: Reactive Oxygen Species

TCA: Tricarboxylic Acid Cycle / Citric Acid Cycle

(4)

7

157

ABSTRACT

There is an increase in kidneys transplanted from deceased donors after circulatory death (DCD). After withdrawal of support and at circulatory arrest the DCD kidneys are subjected to a period of ischaemia, which is associated with a high rate of delayed graft function after transplantation, for which the underlining molecular mechanism is not well defined. In this study an integrated proteomic and metabolomic approach was used to investigate the effects of Ischaemia Reperfusion Injury (IRI) on protein expression and metabolite levels. Rat kidneys were subjected to 45min of warm ischaemia followed by 4h and 24h reperfusion, with contralateral and healthy kidneys serving as controls.

After 4h and 24h reperfusion, tissue proteomics revealed elevated proteins belonging to the acute phase response, coagulation and complement pathway, and fatty acid (FA) signalling.

Metabolic changes were already evident after 4h reperfusion and showed increased level of lipids and FAs, whilst mitochondrial function was impaired after 24h reperfusion.

The enhanced FA consumption and metabolic switch at 4h post IRI could be a compensatory mechanism for the developing energy deficit and reduced mitochondrial function. Novel strategies to target these early metabolic changes could reduce IRI and improve outcomes of high-risk donor kidneys in kidney transplantation.

(5)

INTRODUCTION

Due to a persistent donor kidney shortage, the transplant community is increasingly using organs from deceased donors, such as donation after brain death (DBD) or donation after circulatory death (DCD). In the UK, annual kidney donation from DCD donors has increased four times from 128 in 2005 to 510 in 2015 according to NHS Blood and Transplant (Annual Report 2015), and now accounts for 27% of all kidney donations (1). Kidney transplantation from DCD and DBD donors can achieve similar graft and patient survival rates, however, DCD derived kidney grafts have a higher risk of developing primary non (and never) or delayed graft function (2). Both types of donors suffer from significant ischaemic cerebral injury, leading to a profound systemic inflammatory response. In the DBD donor herniation of the brain stem is followed by a catecholamine release and frequently haemodynamic instability to be corrected in the Intensive Care Unit. In the DCD donor this autonomic storm does not happen, but instead after withdrawal of (ventilatory) support as further medical treatment has been found to be futile, blood pressure will drop over a period of time causing hypoperfusion and hypoxia until circulatory arrest occurs. This period, in combination with the required time of no-touch until procurement and preservation of organs is allowed to start, includes a harmful time of warm ischaemia (WIT). Therefore, one possible explanation for the increased non function rate of DCD kidneys is the reduced blood flow and diminished oxygen and nutrient supply to the organs during this warm ischaemia, leading to ATP depletion and acidosis as a result of an anaerobic metabolism with lactate overproduction (3,4). ATP is the main energy source to maintain cellular physiology, of which about 95% is provided by oxidative metabolism in the kidney.

For this reason, mitochondria play a pivotal role in maintaining the energy balance (5), and when ATP-dependent ion transport systems are impaired this results in calcium accumulation, cell swelling, and apoptotic or necrotic cell death. Consistent with this finding, clinical studies have shown that prolonged WITs contributes to inferior graft survival (6). Upon reperfusion, restored levels of oxygen stimulate mitochondrial oxidative phosphorylation to produce ATP with the concurrence of harmful reactive oxygen species (ROS) (7). ROS can then induce the release of inflammatory mediators and increase local leukocyte infiltration to ischaemic- injured sites aggravating tissue injury. This process of ischaemia and reperfusion injury (IRI) is an inevitable consequence of micro-vascular disruption during donor management and organ transplantation (8-10). Many studies have examined the role of IRI in kidney transplantation (7,11). However, a more comprehensive study combining proteomic and metabolomic alterations at the initial stages of IRI has not yet been carried out in a more systematic fashion.

To mimic DCD in the context of kidney transplantation, we have applied a non-lethal unilateral long IRI model in the rat (12).

We designed an unbiased systematic proteo-metabolomic study and combined it with mitochondrial function analysis of kidneys exposed to IRI to investigate its effects on the molecular level.

(6)

7

159

MATERIAL AND METHODS IRI animal model

A rat model of Ischaemia reperfusion injury was approved by the Animal Welfare Ethics Review Board, and carried out in accordance with Home Office guidance on Operation of Animals (PPL 30/2750). Adult male Fisher rats (F344) weighing 250 to 320g (Harlan, UK) were anaesthetized with isoflurane and placed on an auto-thermoregulatory heating pad. Continuous monitoring of oxygen saturation and heart rate was performed throughout the surgery. A midline laparotomy was performed with dissection of the right and left renal pedicles. After administration of heparin (10 units/ml) the left renal artery was clamped for 45min inducing warm ischaemia (IRI). Ischaemia was confirmed by visual inspection. The contralateral right kidney served as an endogenous control and remained untouched. During the warm ischaemic period the laparotomy incision was kept approximated and the rat maintained anaesthetized. After 45min the clamp was released and reperfusion of the left kidney was observed under visual inspection. The laparotomy incision was then closed and local anaesthetic was administered to the wound. Rodents were placed in a pre-warmed cage with free access to food and water.

No rats were excluded from this study due to technical failure or death.

To evaluate short- and longer-term effects, after 4h and 24h reperfusion, respectively, rats were killed and both the left (IRI-4h; IRI-24h) and right (C-4h; C-24h) kidneys were retrieved for further analysis. As in this unilaterally applied IRI model, molecular injury afflicted to the left kidney may induce systemic effects, modulating the contralateral right kidney, we also included kidneys from healthy control (HC) rats as a baseline to discriminate between local injury versus systemic effects. Procurement of all kidneys included cannulation of the aorta and flushing with cold saline. All kidney samples were either stored in formalin or snap frozen in liquid nitrogen to be kept at -80^oC until use.

Hence, the following groups of kidneys were included in this experiment: IRI-4h (n=7) and C-4h (n=7); IRI-24h (n=5) and C-24h (n=5); HC (n=4).

Histology and apoptosis staining

Dissected kidneys were fixed in 10% formalin, paraffin embedded, and sectioned (5μm).

Periodic acid Schiff (PAS) staining was processed as using standard procedures. Apoptosis analysis was performed by detecting endonucleolytic cleavage of chromatin (ApopTag, Merck Millipore, UK).

Protein extraction for proteomic analysis

The cortex tissue was dissected (20-30 mg sections) from 4h-IRI (n=5), 4h-C (n=5), 24h-IRI (n=5), 24h-C (n=5), and HC (n=4) kidney samples (Figure 1) and placed in beads-beater tubes containing lysis buffer (8M urea, 50mM Tris-HCL pH=8.5, 5mM DTT, 1% SDS, protease inhibitor) to make it 20mg/ml. Tissues were homogenized for 4 times at 6,500 Hz for 40s in a beads- beater (Stretton, UK). The samples were centrifuged at 10,000g for 5min at 4°C to remove insoluble tissue debris. The protein concentration in the homogenates was determined by BCA assay (Thermofisher, UK) and 100 µg of total proteins were added to a 30kDa filter

(7)

(Merck Millipore, UK). Proteins were reduced by 20mM DTT (Sigma, UK) at 37°C for 1h, and then alkylated with 100mM iodoacetamide (IAA, Sigma, UK) for 45min in the dark, at room temperature. Samples were centrifuged for 20min at 14,000 g to remove DTT and IAA and followed by buffer exchange with 8 M urea twice and 50mM ammonia bicarbonate for 3 times.

One hundred microliters of trypsin was added at a trypsin/protein ratio of 1:50 for digestion at 37°C overnight. Digested peptides were collected by upside down spin and membrane filters washed twice with 0.5M NaCl and water respectively. The peptides were purified by a SepPak C18 cartridge (Waters, UK), dried by Speed Vac centrifugation, and resuspended in buffer A (2%

acetonitrile 0.1% formic acid) for LC-MS/MS analysis (Figure 1).

Protein identification and quantitation by mass spectrometry

LC-MS/MS analysis was carried out by Nano-ultra performance liquid chromatography tandem mass spectrometry analysis using a 75 µm-inner diameter x 25 cm C18 nanoAcquity UPLC column (1.7-µm particle size, Waters, UK). Peptides were separated with a 120min gradient of 3–40%

solvent B (solvent A: 99.9% H₂O, 0.1% formic acid; solvent B: 99.9% ACN, 0.1% formic acid) at 250nl/min and injected into a Q Exactive High Field (HF) Hybrid Quadrupole-Orbitrap Mass Spectrometer (Thermo Scientific) acquiring data in electron spray ionisation (ESI) positive mode. The MS survey was set with a resolution of 60,000 FWHM with a recording window between 300 and 2,000 m/z. A maximum of 20 MS/MS scans were triggered in data-dependent acquisition (DDA) mode.

Analysis of mass spectrometry data

Raw LC-MS/MS data were uploaded to LC-Progenesis IQ software (Waters, UK), and total ion chromatograms (TICs) including retention time and m/z were used for automatic sample alignment. Merged files containing all the MS/MS spectra from all samples were submitted to a Mascot (V2.5) Matrix science search against the Swissprot rat database (UPR_Rattus Norvegicus, 34,165 entries). Proteins were identified with at least one matched peptide, a peptide score above 20, and an estimated false discovery rate (FDR) of 1%, representing confidence of identification. Protein quantities were measured by the sum of the top three unique peptide precursor ion intensities quantities, derived from the same protein after quantity normalization. Statistical analysis was conducted using one-way ANOVA for comparing the independent sample types. Alternatively, the in-house Central Proteome Facility Pipeline (CPFP) software was also used to identify peptides based on MS/MS spectra (13). Three search engines including Mascot, OMSSA, and X!Tandem were used for protein identification and combined with iProphet, with a FDR of 1%. The spectral index count (SINQ) was used for protein quantitation. Significantly (p<0.05) and ≥2-fold changed proteins were clustered using Progenesis software (Nonlinear dynamics, UK) and three biological relevant expression patterns were analysed (Figure 3B): 1. proteins downregulated in 24h-IRI; 2.

proteins upregulated in 4h-IRI and 24h-IRI; 3. proteins upregulated only in 24h-IRI.

(8)

7

161 Ingenuity pathway analysis (QIAGEN, UK) was used to assign all the proteins to canonical pathways. KEGG (Kyoto Encyclopaedia of Genes and Genomes) was used to assign differentially expressed proteins to pathways. PANTHER Classification system was used to perform gene ontology enrichment analysis (over-represented or under-represented). The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD005101 and 10.6019/PXD005101.

Western blot validation of differentially expressed proteins

Western blot analysis was performed on 4h-IRI (n=7), 4h-C (n=7), 24h-IRI (n=5), 24h-C (n=5), and HC (n=4) kidneys after dissecting ten milligram of renal cortex samples and lysis in 300µL of RIPA buffer (150mM NaCl, 1.0% NP-40, 0.5% sodium deoxycholate, 1% SDS, 50mM Tris, pH=8.0) containing a protease- (Roche, USA) and phosphatase inhibitor cocktail (Sigma, UK).

Homogenization was performed on a beads beater at 6,500rpm for 40s. Protein concentration was quantified using BCA Protein (Thermofisher, UK). Samples containing 15µg of protein were denatured at 95°C for 5min in Laemmli buffer and loaded onto 8–12% pre-cast SDS-PAGE gels (Bio-Rad, USA). Proteins separated by SDS-PAGE were transferred to hydrophobic PVDF membranes (Merck Millipore, USA) in transfer buffer (25mM Tris, 192mM glycine and 10%

methanol) overnight. PVDF membranes were blocked for 1h in TBST buffer (25mM Tris, pH=7.5, 0.15 M NaCl, 0.05% Tween 20) containing 5% milk. Membranes were incubated overnight at 4 °C with rabbit anti-rat C4 (25ug/ml), HSP70 (1:1000), GST (1:3000), FABP4 (1mg/ml), BVR (1:5000) (Abcam, UK), and TLR4 (1:250) (Santa Cruz) antibodies. Mouse monoclonal anti β-actin was used as loading control (1:25,000) (Sigma, Germany). Membranes were washed for 30min with 5 changes of wash buffer and then incubated at room temperature for 1h in blocking buffer containing a 1:5000 dilution of Dye-800-conjugated anti-mouse or -rabbit secondary antibody (Li-Cor, USA), and visualization was performed with Odyssey CLx (Li-Cor, USA).

Detected signals were quantified and normalized to the β-actin signal on the same blot using ImageJ (1.47v, USA). Difference in mean was calculated with a Mann-Whitney U test (Prism v5.0, GraphPad Software, USA). Significance was set at a value of p<0.05 and all graphs report results as mean ± standard error of the mean.

Extraction of metabolites from kidney tissue

Fifty micrograms of kidney cortex tissues from 4h-IRI (n=7), 4h-C (n=7), 24h-IRI (n=5), 24h-C (n=5), and HC (n=4) kidneys were placed in beads-beater tubes containing 1.5ml of pre-chilled methanol/water (1:1). The tissues were frozen on dry ice (around 2min) and then loaded onto a beads beater (Stretton, UK). Tissues were homogenised four times at 6,500Hz for 40s and cooled on dry ice for 2min between cycles. Homogenates were centrifuged at 13,000g for 20min at 4°C. Supernatants (aqueous fractions) were transferred to an Eppendorf tube, dried in a Speed Vac and stored at -80°C until use. For organic extraction, 1.6ml of pre-chilled dichloromethane/methanol (3:1) was added to the tissue pellet and chilled on dry ice for 2min. Pellets were further homogenized as previously described, and centrifuged at 13,000g for 20min. Supernatants were collected into glass vials. The organic extraction was dried in a fume cupboard overnight with lids open. The extracted samples were stored at -80°C until analysis (Figure 1).

(9)

Metabolomic analysis using nuclear magnetic resonance spectroscopy (NMR)

Sample preparation and acquisition methods by ^1HNMR were adapted from previously published methods (14,15). In brief, extracted aqueous kidney pellets were resuspended before NMR acquisition in 550µl phosphate buffer (0.2M Na₂HPO₄/0.04 M NaH₂PO₄, pH=7.4 with 0.1%

sodiumazide and 1mM 3-trimethylsilyl-1-[2,2,3,3,-2H4] propionate (TSP) in D₂O), vortexed for 2min, and then transferred to a 5 mm NMR tube for analysis. ^1HNMR spectra were acquired with a Bruker Avance 600MHz nuclear magnetic resonance spectrometer instrument operating at 600.13MHz for 1h at 300K. Tissue extracts were analyzed using water suppressed ^1DNMR spectrum using the NOESYPRESAT pulse sequence (256 transients). Irradiation of the solvent (water) resonance was applied during pre-saturation delay (2.0s) for all spectra and for the water suppressed ^1DNMR spectra also during the mixing time (0.1s). Pulse sequence parameters including the 90° pulse (~ 12µs), pulse frequency offset (~ 2,800Hz), receiver gain (~ 200), and pulse powers were optimised. The spectral width was 20ppm for all spectra. NMR data was processed with an exponential line broadening of 1.0Hz prior to Fourier transformation, which was collected with approximately 32k real data points. For NMR spectral data pre-processing, data sets [-1.0 to 10.0ppm] were imported into MATLAB 7.0 software (MathWorks, USA), where they were automatically phased, baseline corrected and referenced to the TSP peak (0.00 ppm) using scripts written in-house. To reduce analytical variation between samples the residual water signal (4.67 – 4.98ppm) was truncated from the data set. Normalization using probabilistic quotient (median fold change) methods was then performed (16). Assignments of endogenous metabolites from kidney tissue were made by reference to published literature data (17) as well as with use of the in-house and online databases {Lindon, 2005 #660} (18,19).

Following the processing of the NMR data, multivariate statistical analysis was performed using SIMCA-P 13.0. Principle Component Analysis (PCA) and Orthogonal Partial-Least Squares Discriminant analysis (OPLS-DA) using both univariate and mean centred scaling were used to identify specific metabolites pertaining to a particular sample group (20). All OPLS models were run through random permutation testing to assess the validity of the supervised model.

Metabolomic analysis using gas chromatography mass spectrometry (GCxGC-MS)

For chemical derivatisation, extracted organic fractions from kidney cortex were oximated at 30°C for 90min using 20mg/ml of methoxyamine hydrochloride (Sigma) dissolved in pyridine. Subsequently, samples were silylated at 60°C for 60min using N-methyl-N- trimethysilyltrifluroacetamide (MSTFA). Deuterium labelled myristic acid was spiked into tissue (5ml/10mg of wet tissue) to serve as an internal standard for data normalization. Two ml of derivatized metabolites were injected into a GCxGC-qMS instrument (GP2010, Shimadzu) with a split ratio of 1 in 20 in pyridine. The oven temperature was programmed to rise from 80°C to 330°C at a rate of 6°C per min, and the total run time was 35 min. The metabolites were first separated by a non-polar column (30m x 0.25mm i.d), followed by a second polar column (6m x 0.25mm i.d) controlled by a modulator that operated at six second cycles as described (21). In brief, metabolite effluents separated on the first column were accumulated for 6s, then remobilized, focused and injected into the second column. Ions were generated within GC-MS by electron impact (EI) ionization and collected at a mass range of 50m/z to 800m/z using an acquisition rate of 20,000u/sec and 100Hz. Metabolites were identified by

(10)

7

163 matching collected fragment ions to the NIST library using GC solution software (Ver.2.32). The identification confidence score cut off was set to 80, and a panel of collected standards were used to verify their identification (such as palmitic acid, stearic acid). Metabolites quantitation was performed using Chrome square software 2.1.6 (Shimadzu, Japan). Each metabolite quantity was normalized to spiked myristic acid for comparison.

Plasma glucose, lactate, and creatinine measurements

Plasma samples from all animals were collected at 4h and 24h post IRI as well as in HC rats.

Plasma samples were diluted (1:2) in PBS and measured (200µl) in a Roche/Hitachi Modular System (Roche Diagnostics, Netherlands) according to the manufacturer’s instructions. The concentration of glucose, lactate, and creatinine was determined.

Luminescent determination of ATP concentrations in the kidney cortex

ATP was extracted extracted from 4h-IRI (n=7), 4h-C (n=7), 24h-IRI (n=5), 24h-C (n=5), and HC (n=4) kidneys according to protocol (22) and concentrations were measured using ENLITEN ATP assay kits (Promega, UK) and FLUOstar Omega (BMG LABTECH, UK) plate reader.

In brief, 40mg of kidney tissue was homogenised in 1ml ice cold Tris-EDTA (10mM, pH 8.0, 1mM EDTA-saturated phenol (phenol-TE) using an Ultra Turrax homogeniser. Homogenate was transferred to micro-tubes containing 200μl chloroform and 150μl water. After mixing and centrifugation at 10,000xg for 5min at 4°C, the aliquot was diluted 100 times for luciferin- luciferase analysis. The assay was performed by adding 0.1ml of the reagent per well and read on a microplate reader. The assay was performed by adding 90µl of reagent per weel to 10µl of samples and read on a microplate reader. ATP was measured by comparing to a standard curve after correction against blank samples. A Kruskal-Wallis (non-parametric) comparison of multiple groups was performed (Prism v5.0, GraphPad Software, USA).

Mitochondrial complex activity and respiratory control ration measurements

Mitochondria were isolated from fresh 4h-IRI (n=7), 4h-C (n=7), 24h-IRI (n=4), 24h-C (n=4), and HC (n=4) kidney tissue following Frezza’s protocol (23). Mitochondrial complex activity assays were performed on a microplate reader (BMG Labtech) (24). Complex activity was normalized to citrate synthase activity (23). For the oxygen consumption assays, fresh tissues were used from 24h-IRI (n=3), 24h-C (n=3), and HC (n=1) kidneys for mitochondrial isolation. State I respiration was recorded with the addition of 1mg/ml of mitochondrial suspension to the Clark electrode. State II and -III respirations were recorded with the addition of 5mM succinate and 100μM ADP respectively. State IV was measured following ADP consumption from the solution and state V was induced by adding 60nM of carbonyl cyanide p-(tri-fluromethoxy) phenyl-hydrazone (FCCP). The respiratory control ratio (RCR) was calculated by dividing state III by state IV respiration (23). Statistical significance was calculated using Mann-Whitney U test and set at p-value <0.05.

Integrative analysis of proteome and metabolome data

To provide a global view of the proteo-metabolome data, a knowledge-based targeted proteome and metabolome data integration approach was applied making use of the human protein sequence database UniProtKB and the human metabolome database (HMDB)

(11)

(8,25,26). Based on altered metabolome profiles between 4h-IRI, 4h-C, and HC kidneys, all the proteins/enzymes involved in energy metabolism, such as glycolysis, TCA cycle, FA transporter, FA β-oxidation and mitochondrial respiration chain for ATP metabolism were shortlisted. Proteins regulated by metabolites (Ampd2) that were changed in IRI were also included from the proteome data. A total of 125 proteins were selected in the analysis, from which 29 were shortlisted based on their statistical significance and altered abundance (Table S4). From all measured 37 metabolites, we selected the subset (14 proteins) that changed in a significant manner. A non-parametric Mann-Whitney test was applied to probe the significance of differences observed for both proteins and metabolites. We then used R-based clustering of proteins and metabolites to visualise subgroups with different patterns of abundance (Figure 9).

RESULTS

Histological and apoptotic changes post IRI

Tissue injury inflicted by clamping the left kidney artery was confirmed by histology and staining for apoptosis. Kidneys subjected to 45min of warm ischaemia and 24h of reperfusion (24h-IRI) displayed severe injury and tubular necrosis, identified by the reduced number of tubular nuclei (Figure 2A). This trait was not observed in 4h-IRI or the control kidneys (4h-C, 24h-C, HC) (Figure S1). With TUNEL staining we could detect tubular apoptosis as early as 4h post-injury in both cortex and medulla of the kidneys subjected to ischaemia (4h-IRI), eventually resulting in necrosis after 24h (24h-IRI) (Figure 2B). This was consistent with the level of injury observed in the histological sections of the same area stained with PAS.

Proteomic analysis after IRI

Kidney cortex samples of 4h-IRI (n=5), 4h-C (n=5), 24h-IRI (n=5), 24h-C (n=5) and HC (n=4) were analysed by LC-MS/MS. In total, 1,055,427 MS/MS-spectra were acquired, resulting in the identification of 2,798 proteins with an estimated false discovery rate (FDR) of 0.96%.

Four hours post IRI, 55 proteins were found to be differently expressed with ≥ 2-fold change and a p<0.05 between 4h-IRI, 4h-C, and HC (Table S1). After 24h reperfusion 397 proteins were found to be different between 24h-IRI and 24h-C kidneys (Table S2). Combining these proteins yielded 423 different proteins with ≥2-fold and significantly changed expression. Subsequently these proteins were clustered and sorted according to their expression patterns. Using hierarchical clustering analysis we assigned 363 proteins to three groups based on their unique expression patterns (Figure 3A, Table S2). Among these 363 differentially expressed proteins, 140 were decreased in 24h-IRI compared to other groups (Figure 3B, upper panel) and 125 proteins were increased in both 4h-IRI and 24h-IRI compared to 4h-C, 24h-C, and HC (Figure 3B, middle panel). A total of 98 proteins was observed to be only increased in 24h-IRI kidneys compared to other groups (Figure 3B, lower panel).

Ingenuity Pathway Analysis was applied to classify these proteins according to their function.

Figure 3C shows the top 11 canonical pathways assigned for all identified proteins (2798).

(12)

7

165 Assignment of the subset of 363 selected proteins revealed enrichment of unique pathways after IRI, including acute phase response signalling, coagulation and complement pathways, and liver X and retinoid X receptor (LXR/RXR) activation (Figure 3D).

Proteomic changes and validation by Western blot

Proteins differentially expressed as listed in Tables S1 and S2 also included stress proteins.

Heat Shock Protein 70 (HSP70 or HSPA1A) and Heme-oxygenase 1 (HO-1) were significantly upregulated after IRI which was validated using Western blot analysis (Figure 4A, 4B). To validate if our IRI model induced ROS production we measured the expression levels of the antioxidants Glutathione S-transferase (GST) and Biliverdin Reductase A (BVR) in the kidney.

Both GST and BVR protein levels were significantly decreased in 24h-IRI compared to 24h-C and HC.

The complement and coagulation pathway was highly upregulated after IRI. Most of the proteins had not been changed after 4h IRI, but showed significant increases at 24h post IRI (Figure 5A, Table S2). C4 was further validated by Western blot, confirming its increased levels post-IRI (Figure 5B).

LXR/RXR signalling is part of the nuclear receptor family that interacts with peroxisome proliferator-activated receptor (PPAR). Thirteen proteins identified in our proteome data were assigned to this pathway (Figure 6A), including significantly increased levels of upstream regulator fatty acid transporter CD36 and FABP4 in both 4h-IRI and 24h-IRI. The increased expression of FABP4 measured by proteomics was validated by Western blot (Figure 6B).

Other downstream proteins were also changed. The fatty acid (FA) oxidation proteins CPT1A, ACADSB, and ECHDC3 were significantly increased in 4h-IRI but, except for ACADSB, returned to normal levels after 24h (24h-IRI). Proteins involved in lipid transport (APOA1, APOA4, APOE, APOH) were increased in both 4h-IRI and 24h-IRI kidneys, but changes did not reach statistical significance. Proteins involved in ketogenesis were affected. BDH1 was significantly upregulated in 4h-IRI, while BDH2 was significantly downregulated in both 4h-IRI and 24h-IRI.

In addition, we found proteins involved in lipogenesis (ACSL4 and ACSL6) to be significantly downregulated in both 4h-IRI and 24h-IRI as compared to the endogenous controls (4h-C, 24h-C).

Metabolomics changes after IRI

After IRI it can be expected to find metabolomic changes on the short term. Therefore, we performed a metabolomic study to further explore these effects by using GCxGC-qMS and NMR on 4h-IRI (n=7), 4h-C (n=7), and HC (n=4) kidney samples. GCxGC-qMS identified seven significantly changed lipid- and FA metabolites (Table S3): palmitate (Figure 7A-1), stearate (Figure 7A-2), linoleate (Figure 7A-3), 1-monopalmitin (Figure 7A-4), 2-monopalmitin (Figure 7A-5), 2-monostearin (Figure 7A-6), and cholesterol (Figure 7A-7). Metabolic changes were detectable in plasma, showing significantly lower glucose levels in 4h-operated animals (Figure 7B-1) compared to 24h-operated animals. Lactate was also changed and significantly elevated in the plasma of 4h- and 24h-operated animals (Figure 7B-2). Similarly, creatinine in blood was significantly increased both after 4h and 24h post IRI (Figure 7B-3).

(13)

In addition, we identified 28 metabolites from the aqueous fraction using 1HNMR (Table S3).

We noted reduced glucose levels (not statistically significant) in both 4h-IRI and 4h-C kidneys compared to HC (Figure 7C-1). Lactate levels were increased in 4h-IRI compared to 4h-C and HC (Figure 7C-2). UDP-glucose, a glycogen precursor, was significantly reduced in 4h-IRI compared to 4h-C (Figure 7C-3). Two identified TCA cycle intermediates, citrate (Figure 7C-4) and succinate were found to be unchanged between the 4h-IRI and 4h-C groups (Table S3). Furthermore, NAD+ and AMP levels were significantly decreased in 4h-IRI compared to 4h-C and HC (Figure 7C-5, -6). Taurine, an organic compound involved in bile acid conjugation and antioxidation was significantly reduced in 4h-IRI compared to 4h-C (Figure 7C-7). 1HNMR additionally identified seven amino acids (leucine, isoleucine, valine, lysine, alanine, glutamine, and glycine), with only the levels of valine and lysine to be significantly changed between 4h-IRI and 4h-C (Figure 7C-8, -9). Concomitantly, ATP levels were significantly decreased in 24h–IRI kidney tissue as compared to 24h-C and HC (Figure 7D).

Mitochondrial function 24h after IRI

Mitochondria are the main ATP-producing source under aerobic conditions. Because ATP tissue levels were significantly decreased in 24h-IRI kidneys (Figure 7D), we assessed mitochondrial function to determine if that could account for the ATP depletion.

Citrate synthase activity was used for normalisation (Figure 8A). Normalised complex I activity was significantly reduced in 24h-IRI compared to 24h-C or 4h-IRI kidneys (Figure 8B).

Mitochondrial complex II-III and -IV activities were also measured in 4h-IRI, 4h-C, 24h-IRI, 24h-C, and HC but were not significantly changed (Figure S2).

Since mitochondrial complex I activity was impaired at 24h-IRI, we also measured the mitochondrial oxygen consumption and RCR using mitochondria extracted from 24h-IRI (n=3), 24h-C (n=3), and a HC (n=1) (Figure 8C, D). The RCR and oxygen consumption were significantly decreased in 24h-IRI compared to 24h-C.

Proteo-metabolic integrated analysis reveals increased lipid metabolism and β-oxidation post IRI To provide a better understanding of the altered metabolites and proteins in our data, we applied an integrative proteo-metabolomic analysis in which we included metabolites and corresponding enzymes that showed statistical differences between 4h-IRI, 4h-C, or HC.

Fourteen metabolites belonging to the lipid-, glycolysis-, and TCA cycle pathways, and 29 metabolic-related proteins with a p<0.05 in at least one comparison were selected and displayed in a heat map and summarising model (Figure 9, Table S4). Hierarchical cluster analysis revealed the existence of three phenotypes: A) Decreased substrates in 4h-IRI and/or 4h-C compared to HC. This group includes proteins involved in FA biosynthesis (Acsl6, Acsl4), metabolites involved in energy metabolism (NAD and AMP), and UDP-glucose. B) Increased substrates in 4h-IRI and/or 4h-C compared to HC. This group comprises FAs (palmitate, stearate and linoleate), monoglycerides (1-monopalmitin, 2-monopalmitin, and 2-monostearin), cholesterol, lipid transporters (CD36, FABP4, Cpt1a), FA β-oxidation proteins (Acadsb, Echdc3), proteins involved ketone genesis (Bdh2), succinate synthesis (Oxct1), and proteins belonging

(14)

7

167 to mitochondrial complex I (Ndufs1, Ndufa6, Ndufa10, Ndufv1, Ndufb2) and -II (Sdha). C) Increased substrates observed in both 4h-IRI and 4h-C with a trend of accumulation in 4h- C, as compared to HC. Citrate and HK1 only increased significantly in 4h-C while Slc5a1 BDH1, and lipid transport (Apoa1bp) increased in both 4h-IRI and 4h-C kidneys as compared to HC (Figure 9A).

DISCUSSION

To gain better insight in the mechanism of IRI, this study has attempted to evaluate effects of renal IRI at the protein and metabolite level aiming to identify pathways for intervention. In our IRI model in the rat 45min of ischaemia was followed by 4h and 24h reperfusion resulting in ROS-related injury, reflected by increased expression of stress proteins (HSP70, HO-1) (Figure 4), reduced levels of antioxidants (BVR, GST) (Figure 4) and severe tubular injury (Figure 2).

It is known that the coagulation and complement systems become activated post IRI in the deceased kidney donor (27,28). Coagulation plays a critical role in homeostasis but can also further activate the complement system (29,30). Consistent with the literature, we observed a strong upregulation of these pathways including enhanced expression of C3 and C4 and their respective downstream products, C5, C6, C8, and C9 after 24h IRI. Complement induces renal injury post IRI and inhibition can prevent these detrimental effects (31,32). This treatment option is currently being explored in the human transplant setting (33,34).

Most remarkably, proteomics analysis revealed that fatty acid (FA) signalling through LXR and PPAR had changed after IRI (Figure 3D and 6). LXR and PPAR are nuclear receptors involved in regulating nutrient metabolism (35), in particular FA synthesis (LXR) and catabolism (PPAR)(36). Downstream transcription targets of PPARγ signalling are genes involved in lipid metabolism (37), which prompted us to further explore this phenomenon in IRI. Our data reveal an increase in FA levels at 4h post IRI. It is known that free fatty acid (FFA) levels decrease after ischaemia (38), but reperfusion injury is correlated with a persistent elevation of FFAs (37).

The mechanism of elevated FFAs post IRI remains unclear, but could be the result of increased membrane lipid degradation or failure of cells to reutilize or dispose FFAs. Integrative proteo- metabolomic analysis showed increased FA β-oxidation after IRI, which was supported by both increased FAs and FA β-oxidation enzymes. Increased FAs could be derived directly from energy storage-triglyceride (TG) utilization under stress or through increased transport from the systemic circulation by CD36 (37). FA can also be internalized from the extracellular compartment by FA transport proteins (e.g. FABP4) and transported into nucleus, affecting lipid metabolism gene transcription (39). In line with this, we observed increased expression of the fatty acid internalization proteins CD36 and FABP4, which are both upstream proteins of PPAR. Furthermore, we detected a reduction in the levels of FA biosynthesis enzymes Acsl4 and Acsl6, suggesting FA catabolism through β-oxidation to produce ATP. FA β-oxidation occurs in the mitochondrial compartment, which may explain why the FA transport protein Cpt1a was also increased after IRI.

(15)

IRI can induce anaplerotic reactions to compensate for a deficiency of metabolic intermediates (4). The end product of FA β-oxidation, Acetyl-CoA, can either feed the TCA cycle or contribute to ketone genesis. We observed an increase in ketogenesis reflected by an increased 3-hydroxybutyrate dehydrogenase (Bdh1) level. Also, acetoacetate, as a product of ketogenesis, can react with the TCA cycle derived succinyl-CoA and form succinate through the activity of Oxct1. Our results found that both Bdh1 and Oxct1 were increased post IRI, suggesting utilization of ketones and thereby providing an alternative source for succinate. Succinate can also be utilized directly by mitochondrial complex II (Sdha) for electron transport. In the IRI kidney, increased levels of Sdha might suggest an induction of the electron transporters in the mitochondria. In line with this, enzymes belonging to mitochondrial complex I (Ndufa10, Ndufb2, Ndufb6, Ndufv1, Ndufa6, Ndufs1) were found to be elevated in our IRI model.

IRI is an energy consuming process and is associated with impaired glucose metabolism (3), mitochondrial injury (40,41), and ATP depletion (7). In our study, glucose levels were significantly decreased in plasma and tissue at 4h-IRI, although tissue ATP content and mitochondrial function at 4h-IRI were not changed. This may be compensated by an increased expression of the glucose transport protein Slc5a1 and an increased anaerobic activity with elevated lactate levels in blood and tissue. However, ATP levels, mitochondrial complex I activity, and oxygen consumption could not be sustained and were significantly decreased in 24h-IRI kidneys. It is evident that mitochondrial dysfunction resulted in a significant energy shortage as reflected by decreased levels of AMP and NAD+.

Taken together, our integrated -Omics data suggests a metabolomic switch to lipid/FA utilization after 4h, and the increased FA β-oxidation could be an adaptive mechanism to balance energy homeostasis in the short term (4h-IRI). An increased energy demand post IRI along with depletion of glucose stores could explain the necessity of this metabolic switch.

Metabolic adaption after IRI can allow kidney cells to sustain their energy levels (ATP) at an early time point of 4h-IRI. However, this adaptive response is probably not sufficient as shown by a drop in ATP at 24h-IRI. This finding could be explained by ongoing ROS-induced mitochondrial injury, since the anti-oxidant levels are markedly decreased at 24h-IRI and the mitochondria show impaired functionality. In addition, sustaining enhanced lipid metabolism will further impair mitochondrial function by additional ROS generation, a process referred to as lipid toxicity (42-44). After 24h reperfusion, the impaired energy metabolism and decreased ATP levels eventually led to increased creatinine plasma levels, reduced kidney function, potentially resulting in irreversible kidney injury.

A limitation of this study is the sensitivity of the mass spectrometer to detect proteins and metabolites, where low abundant molecules are less likely to be identified. Also, we only compared two reperfusion time-points with endogenous and healthy controls, revealing a limited fraction of the dynamic changes post IRI. Furthermore, we chose to use endogenous controls instead of sham-operated control rats. Using both healthy controls and endogenous controls, it is possible to distinguish between systemic effects after surgery and starvation

(16)

7

169 effects, as well as IRI-related molecular changes. An advantage of using animals as their own control is the possibility to correct for inter-animal variability and thus reduce the number of animals necessary for this study.

In conclusion, this study has described the effects of IRI on kidney tissue using a novel integrated metabolic and proteomic approach after 4h and 24h reperfusion and has revealed a significant switch to lipid metabolism. A clinically relevant consequence could therefore be the optimisation or protection of kidney metabolism, maximising substrate utilisation and ATP production efficiency as this intervention strategy may prevent an energy deficit and improve transplantation outcomes.

ACKNOWLEDGEMENTS

We thank Climent Casals-Pascual, Adam Thorne, and Karl Morten for insightful discussions and critical reading of the manuscript. We are indebted to Petra Ottens for her help with the measurement of metabolites in blood samples. We also thank Marian Bulthuis for her expert help with the PAS staining of the tissue sections. This work was supported by NHS Blood and Transplant Trust Fund TF031 to M.K., COPE EU-FP7 to R.J.P. and a John Fell Fund 133/075 and Welcome Trust grant 105605/Z/14/Z to B.M.K.

DISCLOSURE

The authors have nothing to disclose.

(17)

FIGURES

Figure 1. Experiment design and work flow of the proteome and metabolome study.

A. Ischaemia reperfusion Injury (IRI) animal model in male Fischer F344 rats. Ischaemia was induced for 45min followed by 4h (4h-IRI; n=7) or 24h (24h-IRI; n=5) reperfusion. Contralateral kidneys served as endogenous controls (4h-C, 24h-C). In addition, kidneys removed from rats without IRI served as healthy controls (HC; n=4). B. Proteomic sample preparation, mass spectrometer analysis, and data mining are outlined in the left panel. The middle panel presents the kidney tissue, equipment used for sample preparation, and bioinformatics tools for data analysis. Metabolomic sample preparation, 1HNMR analysis, and data mining are displayed in the right panel.

(18)

7

171 Figure 2. Histology and staining for apoptosis in cortex and medulla sections after IRI.

A. PAS staining for histological changes in 4h-IRI, 24h-IRI, and HC. Intratubular cast formation is evident in 4h-IRI while marked necrosis is observed in 24h-IRI B. Terminal Deoxynucleotidyl Transferase dUTP nick end labelling (TUNEL) staining for apoptosis in 4h-IRI, 24h-IRI, and HC. Break down DNA is shown in red, nuclei stained with DAPI are shown as blue, and co localization is shown as pink. Apoptotic tubular cells were detected in 4h-IRI in both cortex and medulla sections. Massive DNA break down in 24h-IRI indicates extensive cell death.

4h-IRI and 24h-IRI: kidneys subjected to IRI; 4h-C and 24h-C: contralateral controls; HC: healthy controls.

(19)

Figure 3. Hierarchical clustering analysis and canonical pathway analysis of the identified proteins. A. From all 2798 proteins identified, 363 proteins were clustered into three groups (1,2,3) based on the protein expression pattern in 4h-IRI, 24h-IRI, their endogenous and healthy controls. B. The unique expression pattern of three clusters: Cluster 1 shows downregulation of 140 proteins in 24h-IRI vs other experimental groups. Cluster 2 shows 125 upregulated proteins in 4h-IRI and 24h-IRI vs other experimental groups. Cluster 3 shows 98 upregulated proteins in the 24h-IRI only. C. Top 11 canonical pathways of all identified 2798 proteins. The -log(p-value) for pathway activation is displayed on the left Y-axis, the percentage of identified protein members for each pathway is shown on the right Y-axis. D. Canonical pathway analysis of the 363 significantly and differentially expressed proteins. None of these top 11 pathways (D) were among the top 11 pathways in (C), indicating enrichment of pathways relevant to IRI.

(20)

7

173 4h-IRI and 24h-IRI: kidneys subjected to IRI; 4h-C and 24h-C: contralateral controls; HC: healthy controls.

Figure 4. Western blot validation of stress proteins and antioxidants.

A. There was an increased expression of HSP70 in both 4h-IRI and 24h-IRI compared to their endogenous contralateral kidneys (4h-C, 24h-C) and healthy controls (HC). B. Increased expression of HO-1 in 24h-IRI compared to 24h-C and HC. C. Reduced expression of BVR in 24h-IRI compared to 24h-C and HC. D. The expression of GST was significantly reduced in 24h-IRI compared to 24h-C and HC. Protein expression was normalised to β-actin expression.

Data is presented as mean±SEM. *: p-value <0.05.

(21)

Figure 5. Coagulation and complement pathway activation after IRI. A. KEGG pathway mapping shows upregulation of differentially expressed proteins involved in the coagulation and complement pathway after 4h-IRI and 24h-IRI. Most proteins mapped to this pathway were significantly (≥2-fold) upregulated in 24h-IRI, however, the changes in 4h-IRI vs 4h-C were mostly indifferent and not significant. B. There was an increased expression of C4 in both 4h-IRI and 24h-IRI measured by Western blot. Data is presented as mean±SEM. 4h-IRI and 24h-IRI: kidneys subjected to IRI; 4h-C and 24h-C: contralateral controls; HC: healthy controls.

(22)

7

175 Figure 6. Peroxisome proliferator-activated receptor (PPAR)/retinoid X receptor (RXR) pathway activation post IRI.

A. PPAR is a type II nuclear receptor that includes the liver X receptor (LXR) and dimerizes with RXR to form LXR/RXR. LXR/RXR can be activated by fatty acids and its derivatives. CD36 and FABP4 are proteins that transport fatty acids and retinoid acids across the cell- or nuclear membranes. Upon activation, LXR/

RXR acts as a transcription factor and thus plays a role in metabolism and clearance of lipids. In our study, proteins involved in fatty acid oxidation (CPT1A, ACADSB, ECHDC3) were increased in 4h-IRI but restored to normal levels after 24h. FABP4 was elevated in 4h-IRI and 24h-IRI. Lipid transporters (APOA1, APOA4, APOE, APOH) were also increased in both 4h-IRI and 24h-IRI, but did not reach statistical significance.

Ketone body generating protein BDH1 was increased in 4h-IRI, while BDH2 was reduced at 24h-IRI. Proteins involved in fatty acid synthesis (ACSL4, ACSL6) were reduced in both 4h-IRI and 24-IRI. B. Increased expression of FABP4 in 4h-IRI and 24h-IRI compared to 4h-C, 24h-C, and HC measured by Western blot.

Data is presented as mean±SEM. 4h-IRI and 24h-IRI: kidneys subjected to IRI; 4h-C and 24h-C: contralateral controls; HC: healthy controls.

(23)

(24)

7

177 Figure 7. Metabolites quantified by NMR and GCxGC-MS in both tissue and plasma after IRI. A. Metabolites extracted from the lipophilic phase (4h-IRI, 4h-C, and HC) were identified and quantified by GCxGC-MS, including 1. palmitate, 2. stearate, 3. linoleate, 4. 1-monopalmitin, 5. 2-monopalmitin, 6. 2-monostearin, and 7. cholesterol. B. Metabolites measured using conventional clinical measurement from plasma samples from all rats. 1. glucose, 2. Lactate, and 3. creatinine. C. Metabolites extracted from the aqueous phase (4h-IRI, 4h-C, and HC) were identified and quantified by NMR, including 1. glucose, 2. lactate, 3. UDP-glucose, 4. citrate, 5. NAD+, 6. AMP, 7. taurine, 8. Valine, and 9. lysine. D. ATP levels were quantified using a luminescent ATP detection assay kit in kidney tissues. Data is presented as mean±SEM. 4h-IRI and 24h-IRI: kidneys subjected to IRI; 4h-C and 24h-C: contralateral controls; HC: healthy controls. *: p-value<0.05, *: p-value<0.01, ***: p-value<0.001.

(25)

Figure 8. Mitochondrial complex activity and oxygen consumption analysis after IRI.

A. Citrate synthase activity was used for normalization and measured in 4h-IRI and 4h-C (n=7), 24h-IRI and 24h-C (n=4), and HC (n=4). B. Complex I activity was measured in the same subset of samples and normalized to citrate synthase. Significant reduction of normalized complex I activity was observed in 24h-IRI compared to 24h-C and 4h-IRI. C. After adding mitochondria (1) the respiration states were measured by oxygen consumption. State II was recorded by addition of succinate (2), State III-IV by addition of ADP (3), and state V by adding FCCP (5). Kidneys form 24h-IRI showed a significantly impaired oxygen consumption rate. D. Respiratory control ratio (RCR) was measured in 24h-IRI (n=3), 24h-C (N=3), and a HC (n=1) by comparing state III and -IV oxygen consumption levels. There is a significant RCR reduction in 24h-IRI compared to 24h-C.

Data is expressed as mean±SD. *: p-value <0.05, **: p-value <0.01, ***: p-value <0.001.

(26)

7

179 Figure 9. Integrated proteomic and metabolomic analysis 4h post IRI.

A. Metabolites and proteins involved in metabolite production were shortlisted based on their changed expression in 4h-IRI compared to 4h-C and HC. A heat map using both metabolites (Italic font-style) and proteins (normal font-style) indicates a distinct pattern in particular for FA β-oxidation and ketogenesis.

B. The proposed model of energy homeostasis 4h post IRI. Increased FFA metabolism was suggested according to increased FA transporters (CD36, Cpt1a), monoglyceride levels, FA β-oxidation enzymes (Acadsb, Echdc3), and ketogenesis. The metabolites derived from FA β-oxidation (acetyl-CoA) can directly feed into and sustain the TCA cycle. Metabolite generated from the TCA cycle (succinyl-CoA) and products from ketogenesis (acetoacetate) can subsequently form succinate catalyzed by Succinyl- CoA:3-Ketoacid-CoA Transferase (Oxct1). Succinate is an essential substrate that can be fed directly into the electron transport chain (Complex II) and into the TCA cycle, maintaining and providing essential mitochondrial substrates.

(27)

REFERENCES

1. NHSBT. Annual raport on kidney transplantation. 2015; Available from:

http://www.odt.nhs.uk/pdf/organ_specific_

report_kidney_2015.pdf.

2. Rao PS, Ojo A. The alphabet soup of kidney transplantation: SCD, DCD, ECD--fundamentals for the practicing nephrologist. Clin J Am Soc Nephrol. 2009 Nov;4(11):1827–31.

3. Weld KJ, Montiglio C, Bush AC, Dixon PS, Schwertner HA, Hensley DM, et al. Predicting irreparable renal ischemic injury using a real- time marker in the porcine model. J Urol.

2008 Nov;180(5):2218–25.

4. Wei Q, Xiao X, Fogle P, Dong Z. Changes in metabolic profiles during acute kidney injury and recovery following ischemia/

reperfusion. PLoS ONE. 2014;9(9):e106647.

5. Akram A, Schmeidler J, Katsel P, Hof PR, Haroutunian V. Increased expression of RXRα in dementia: an early harbinger for the cholesterol dyshomeostasis? Mol Neurodegener. 2010 Sep 15;5:36.

6. Siedlecki A, Irish W, Brennan DC. Delayed graft function in the kidney transplant. Am J Transplant. 2011 Nov;11(11):2279–96.

7. Devarajan P. Update on mechanisms of ischemic acute kidney injury. J Am Soc Nephrol. 2006 Jun;17(6):1503–20.

8. Kosieradzki M, Rowiński W. Ischemia/

reperfusion injury in kidney transplantation:

mechanisms and prevention. TPS. 2008 Dec;40(10):3279–88.

9. Azuma H, Nadeau K, Takada M, Mackenzie HS, Tilney NL. Cellular and molecular predictors of chronic renal dysfunction after initial ischemia/reperfusion injury of a single kidney. Transplantation. 1997 Jul 27;64(2):190–7.

10. Gueler F, Gwinner W, Schwarz A, Haller H.

Long-term effects of acute ischemia and reperfusion injury. Kidney International.

2004 Aug;66(2):523–7.

11. Eltzschig HK, Eckle T. Ischemia and reperfusion--from mechanism to translation.

Nat Med. 2011 Nov 7;17(11): 1391–401.

12. Tajra LC, Martin X, Margonari J, Blanc- Brunat N, Ishibashi M, Vivier G, et al. In vivo effects of monoclonal antibodies against rat beta(2) integrins on kidney ischemia-reperfusion injury. J Surg Res. 1999 Nov;87(1):32–8.

13. Trudgian DC, Thomas B, McGowan SJ, Kessler BM, Salek M, Acuto O. CPFP: a central proteomics facilities pipeline.

Bioinformatics. 2010 Apr 15;26(8):1131–2.

14. Beckonert O, Keun HC, Ebbels TMD, Bundy J, Holmes E, Lindon JC, et al. Metabolic profiling, metabolomic and metabonomic procedures for NMR spectroscopy of urine, plasma, serum and tissue extracts. Nat Protoc. 2007;2(11):2692–703.

15. Dona AC, Jiménez B, Schäfer H, Humpfer E, Spraul M, Lewis MR, et al. Precision high- throughput proton NMR spectroscopy of human urine, serum, and plasma for large- scale metabolic phenotyping. Anal Chem.

2014 Oct 7;86(19):9887–94.

16. Craig A, Cloarec O, Holmes E, Nicholson JK, Lindon JC. Scaling and normalization effects in NMR spectroscopic metabonomic data sets. Anal Chem. 2006 Apr 1;78(7):2262–7.

17. Bollard ME, Stanley EG, Lindon JC, Nicholson JK, Holmes E. NMR-based metabonomic approaches for evaluating physiological influences on biofluid composition. NMR Biomed. 2005 May;

18(3):143–62.

18. Ichas F, Mazat JP. From calcium signaling to cell death: two conformations for the mitochondrial permeability transition pore.

Switching from low- to high-conductance state. Biochim Biophys Acta. 1998 Aug 10;1366(1-2):33–50.

19. Wishart DS, Tzur D, Knox C, Eisner R, Guo AC, Young N, et al. HMDB: the Human Metabolome Database. Nucleic Acids Res.

2007 Jan;35(Database issue):D521–6.

(28)

7

181 20. Barton RH, Nicholson JK, Elliott P, Holmes E.

High-throughput 1H NMR-based metabolic analysis of human serum and urine for large- scale epidemiological studies: validation study. Int J Epidemiol. 2008 Apr;37 Suppl 1:i31–40.

21. Beens J, Adahchour M, Vreuls RJ, van Altena K, Brinkman UA. Simple, non-moving modulation interface for comprehensive two-dimensional gas chromatography. J Chromatogr A. 2001 Jun 1;919(1):127–32.

22. Chida J, Yamane K, Takei T, Kido H. An efficient extraction method for quantitation of adenosine triphosphate in mammalian tissues and cells. Anal Chim Acta. 2012 May 21;727:8–12.

23. Frezza C, Cipolat S, Scorrano L. Organelle isolation: functional mitochondria from mouse liver, muscle and cultured fibroblasts.

Nat Protoc. 2007;2(2):287–95.

24. Kiebish MA, Han X, Cheng H, Lunceford A, Clarke CF, Moon H, et al. Lipidomic analysis and electron transport chain activities in C57BL/6J mouse brain mitochondria. J Neurochem. 2008 Jul;106(1):299–312.

25. May P, Christian N, Ebenhöh O, Weckwerth W, Walther D. Integration of proteomic and metabolomic profiling as well as metabolic modeling for the functional analysis of metabolic networks.Methods Mol Biol. 2011;694:341–63.

26. Valli A, Rodriguez M, Moutsianas L, Fischer R, Fedele V, Huang H-L, et al. Hypoxia induces a lipogenic cancer cell phenotype via HIF1α- dependent and -independent pathways.

Oncotarget. 2015 Feb 10;6(4):1920–41.

27. Danobeitia JS, Djamali A, Fernandez LA. The role of complement in the pathogenesis of renal ischemia-reperfusion injury and fibrosis. Fibrogenesis Tissue Repair.

2014;7:16.

28. Damman J, Bloks VW, Daha MR, van der Most P, Sanjabi B, van der Vlies P, et al.

Hypoxia and Complement-and-Coagulation Pathways in the Deceased Organ Donor as the Major Target for Intervention to Improve Renal Allograft Outcome. Transplantation.

2014 Nov 25.

29. Amara U, Rittirsch D, Flierl M, Bruckner U, Klos A, Gebhard F, et al. Interaction between the coagulation and complement system.

Adv Exp Med Biol. 2008;632:71–9.

30. Rittirsch D, Flierl MA, Ward PA. Harmful molecular mechanisms in sepsis. Nat Rev Immunol. 2008 Oct;8(10):776–87.

31. Arumugam TV, Shiels IA, Woodruff TM, Granger DN, Taylor SM. The role of the complement system in ischemia-reperfusion injury. Shock. 2004 May;21(5):401–9.

32. Delpech P-O, Thuillier R, SaintYves T, Danion J, Le Pape S, van Amersfoort ES, et al. Inhibition of complement improves graft outcome in a pig model of kidney autotransplantation. J Transl Med. 2016 Sep 23;14(1):277.

33. Internet CG. SC Jordan (MD): Cedars-Sinai Medical Center (USA). 2014 feb 19 - 2019 Jun. Identifyer NCT02134314, Assessing Safety and Efficacy of Preoperative and Post-Transplant C1 Inhibitor (Berinert) vs.

Placebo in Recipients of a Renal Allograft From Deceased High Risk Donors and Its Impact on DGF and IRI. [cited 2016 Nov 25];

Available from: http://clinicaltrials.gov/ct2/

show/study/nct02134314.

34. ClinicalTrials.gov [Internet]. L Fernandez (MD): University of Wisconsin (USA). 2015 Apr 21 - 2018 Dec. Identifyer NCT02435732, A Phase I, Single Center, Randomized, Double- Blind, Placebo-Controlled Study to Evaluate Tolerability of C1 Inhibitor (CINRYZE) as a Donor Pretreatment Strategy in Brain Dead Donors Who Meet a Kidey Donor Risk Index (KDRI) Above 85%. [cited 2016 Nov 25];

Available from: http://clinicaltrials.gov/ct2/

show/study/nct02435732.

(29)

35. Yue L, Ye F, Gui C, Luo H, Cai J, Shen J, et al. Ligand-binding regulation of LXR/RXR and LXR/PPAR heterodimerizations: SPR technology-based kinetic analysis correlated with molecular dynamics simulation. Protein Sci. 2005 Mar;14(3):812–22.

36. Li AC, Glass CK. PPAR- and LXR- dependent pathways controlling lipid metabolism and the development of atherosclerosis. J Lipid Res. 2004 Dec;

45(12):2161–73.

37. Matthys E, Patel Y, Kreisberg J, Stewart JH, Venkatachalam M. Lipid alterations induced by renal ischemia: pathogenic factor in membrane damage. Kidney International.

1984 Aug;26(2):153–61.

38. Ruidera E, Irazu CE, Rajagopalan PR, Orak JK, Fitts CT, Singh I. Fatty acid metabolism in renal ischemia. Lipids. 1988 Sep;23(9):882–4.

39. Stahl A. A current review of fatty acid transport proteins (SLC27). Pflugers Arch.

2004 Feb;447(5):722–7.

40. Rouslin W, Millard RW. Canine myocardial ischemia: defect in mitochondrial electron transfer complex I. Journal of Molecular and Cellular Cardiology. 1980 Jun;12(6):639–45.

41. Tompkins AJ, Burwell LS, Digerness SB, Zaragoza C, Holman WL, Brookes PS. Mitochondrial dysfunction in cardiac ischemia-reperfusion injury: ROS from complex I, without inhibition. Biochim Biophys Acta. 2006 Feb;1762(2):223–31.

42. Satapati S, Kucejova B, Duarte JAG, Fletcher JA, Reynolds L, Sunny NE, et al.

Mitochondrial metabolism mediates oxidative stress and inflammation in fatty liver. J Clin Invest. 2016 Apr 1;126(4):1605.

43. Singh N, Manhas A, Kaur G, Jagavelu K, Hanif K. Inhibition of fatty acid synthase is protective in pulmonary hypertension. Br J Pharmacol. 2016 Jun;173(12):2030–45.

44. Banach M, Rysz J. Current problems in hypertension and nephrology. Expert Opin Pharmacother. 2010 Nov;11(16):2575–8.

(30)

7

183

SUPPLEMENTARY FIGURES AND TABLES

Figure S1. Histology and staining for apoptosis in cortex and medulla sections of endogenous controls.

A. Periodic Acid-Schiff (PAS) staining for histological changes in contralateral kidneys (4h-C and 24h-C) and healthy controls (HC). Formalin fixed kidney tissue from contralateral kidneys were used for PAS staining.

No histological changes were evident after 4h or 24h reperfusion in the contralateral kidneys. B. Terminal Deoxynucleotidyl Transferase dUTP nick end labelling (TUNEL) staining for apoptosis in contralateral kidneys (4h-C and 24h-C) and healthy controls (HC). Breakdown of DNA is shown in red, while nuclei were stained blue with DAPI. Co localization of GFP and DAPI is shown as pink. No apoptotic tubular cells were detected at 4h-C or 24h-C in both cortex and medulla sections.

(31)

Figure S2. Mitochondrial complex II-III and IV activity analysis after IRI.

A. Complex II-III activity was measured after 4h and 24h post IRI in injured (4h-IRI, 24h-IRI), contralateral (4h-C, 24h-C), and healthy controls (HC). B. Normalized complex II-III activity after normalization to citrate synthase. C. Complex IV activity was measured using the same set of samples as A. D. Normalized complex IV activity after normalization to citrate synthase. No statistical significant changes were detected between the different groups.

Data is expressed as mean±SD.