Targeting brain death-induced injury van Erp, Anne Cornelie

Targeting brain death-induced injury van Erp, Anne Cornelie

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:

2018

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):

van Erp, A. C. (2018). Targeting brain death-induced injury. 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.

Organ-specific metabolic profile during brain death evaluated using hyperpolarized Magnetic

Resonance Imaging:

an initial experience

Anne C. van Erp Haiyun Qi Nichlas R. Jespersen

Marie V. Hjortbak Petra J. Ottens J. Wiersema-Buist Rikke Nørregaard Michael Pedersen Christoffer Laustsen Henri G.D. Leuvenink Bente Jespersen

In preparation

6

CHAPTER

6

ABBREVIATIONS

ALT Alanine Transaminase AST Aspartate Transaminase ATP Adenosinetriphosphate AUC Area Under the Curve BD Brain Death

BSA Bovine Serum Albumin DNP Dynamic Nuclear Polarization FOV Field Of View

ICP Intracranial Pressure IPK Isolated Perfused Kidney IRR Intrarenal Vascular Resistance LDH Lactate Dehydrogenase MAP Mean Arterial Pressure MDA Malondialdehyde

MRI Magnetic Resonance Imaging PDH Pyruvate Dehydrogenase ROS Reactive Oxygen Species Sham Sham-operated

TBARS Tthiobarbituric Acid Reactive Substances TCA Tricarboxylic Acid Cycle

TE Echo Time

TR Repetition Time

UF Ultrafiltrate

6 INTRODUCTION

The global shortage of donor organs results in almost 115.000 people in the United States and 15.000 people in Europe on the organ waiting list

. For most of these patients, organ transplantation would mean a significant improvement of their survival and quality of life

. Most organs worldwide are obtained from brain-dead organ donors

. However, brain death (BD) causes major changes in donor physiology

, leading to impaired graft function, higher rejection rates, and inferior long-term outcomes after transplantation compared to living- donor transplantations

. Therefore, it is important to improve our understanding of the BD-induced injuries. This knowledge would allow us to treat or precondition these organs, in the donor or during organ preservation.

Donor BD affects both systemic as well as organ-specific metabolism. In plasma, metabolite levels shift as evidenced by decreased glucose yet increased lactate and fatty acid levels

. In the liver and kidneys, BD results in tissue injury as well as adenosinetriphosphate (ATP) depletion

. A prior study from our group indicated increased anaerobic breakdown of pyruvate in the kidneys, while the production of glucose from pyruvate decreased

. These changes were likely the result of hypoxic changes in the kidney, as BD caused decreased renal perfusion yet increased oxidative stress

. Meanwhile in the liver, BD resulted in reduced glycogen stores, while glycolysis and oxygen consumption increased

(see Figure 1). Also, several mitochondrial peptides involved in fatty acid oxidation and substrate transport increased in the liver, suggesting that the liver altered its metabolic profile to face an increasing energy demand

. These metabolic changes are potentially significant, as the metabolic status of the liver and kidney prior to transplantation has been correlated to post-transplantation graft viability

. This might be explained by the fact that (changes in) metabolism can influence cell survival, cell growth, and cell death pathways including autophagy and apoptosis

.

Cellular metabolism can now be revealed non-invasively using a novel technique called hyperpolarized magnetic resonance imaging (MRI). Hyperpolarization increases the sensitivity of the MRI system more than 10.000-fold, using techniques such as dynamic nuclear polarization (DNP). DNP first aligns all electron spins by means of a very strong magnetic field (usually around 3.35T) and low temperature (<1.4K), thereby achieving a high polarization level (ie. alignment of electron spins perpendicular to the magnetic field)

. Using microwave irradiation, this energy is then transferred to the carbon-13 spins of a molecule of interest, thereby increasing their polarization up to 50%

. These

C-labeled (enriched) molecules allow for metabolic processes to be observed with compounds such as labelled pyruvate, which are otherwise not visible due to their low

C abundance (~1%

of the total carbon pool) and in vivo concentrations

. Labelling of

C-pyruvate allows

6

for real-time metabolic mapping of pyruvate as well as its metabolites, including lactate, bicarbonate, and alanine (see Fig 1). Injected

C-pyruvate will quickly equilibrate with MRI- inactive

C-pyruvate in the body and enter several metabolic pathways: lactate, formed anaerobically using the enzyme lactate dehydrogenase (LDH) (in the case of low O

levels);

bicarbonate, which is in direct equilibrium with CO

via carbonic anhydrase, as a byproduct of the aerobic breakdown of acetyl-CoA in the tricarboxylic acid (TCA) cycle involving the enzyme pyruvate dehydrogenase (PDH); and alanine, produced using the enzyme alanine transaminase (ALT) in the Cahill (or glucose-alanine) cycle

(see Fig 1). Hyperpolarized

13

C-pyruvate MRI has already showed promise in the diagnosis and monitoring of early renal functional changes

.

Figure 1. Overview of metabolic changes in the liver and kidney following brain death. All changes depicted in green represent an increase in activity or level, whereas all changes marked in red denote decreased activity or levels

. All compounds marked in organ are involved in the metabolization of pyruvate into its metabolites: the pyruvate-alanine (or Cahill) cycle involving enzyme alanine aminotransferase (ALT), the pyruvate-lactate cycle involving lactate dehydrogenase (LDH) as the last step of glycolysis in the absence of oxygen, pyruvate decarboxylation involving the transformation of pyruvate to acetyl-CoA involving pyruvate dehydrogenase (PDH), and finally the formation of bicarbonate (HCO

) from carbon dioxide (CO

) as an end-product of the mitochondrial tricarboxylic acid (TCA) cycle.

Perfusion Oxidative stress Pyruvate

Glucose

Lactate Acetyl-CoA

Oxaloacetate

TCA

PC

LDH

Gluconeogenesis Glycolysis

Fatty acid β oxidation

Acyl-CoA

Acyl-CoA

O₂

O₂ Extracellular

Mitochondrion

P-enolpyruvate PCK-1

PFK-1 Glycogen

ATP

consumption

Alanine ALT

Cahill cycle (gluconeogenic) O₂

CO₂ + H₂O

HCO₃^- + H⁺

6

To prevent poor transplantation outcomes, we would like to predict which organs will function well, and which will function poorly, before the actual transplantation takes place.

However, there currently is a lack of predictive tools or biomarkers that can predict organ quality prior to transplantation. Also, thus far, treatments in the deceased donor have not led to improved organ quality or survival after transplantation

. As we previously observed organ-specific, metabolic changes already in the brain-dead donor, we hypothesized that metabolic assessment could potentially be used to visualize or even predict organ stress and quality. Moreover, we believe that metabolic changes could be of key importance and related to other pathophysiological changes during BD. Therefore, the aim of this study was to investigate the feasibility of hyperpolarized MRI as a way to visualize organ- specific metabolism in a non-invasive, repetitive manner using a brain-dead model in rats.

Furthermore, we assessed whether metabolic assessment during normothermic organ reperfusion could be used to visualize differences between kidneys of brain-dead versus control animals. Using radio-actively labelled glucose, we investigated the effect of BD on glucose oxidation within the TCA cycle in an isolated perfused kidney device. We believe that developing these techniques might allow us not only to visualize metabolic changes, but also to pinpoint exactly which parts of the metabolic pathways are affected. This knowledge could aid in the development of treatment strategies to improve liver and kidney quality prior to transplantation.

MATERIALS AND METHODS

Brain death model

Twenty-two male, Fisher 344 rats were randomised into one of the following groups: a sham- operated (sham) group (n = 8), a BD group (n = 8), and a BD control group (n = 6). The BD and sham group were both exposed to an experimental duration of 4 hrs, whereas the BD control group was terminated immediately after BD induction. The Danish Animal Experimentation Inspectorate approved the experimental protocol. All animals were cared for according to the National Institutes of Health Guide for the Care and Use of Laboratory Animals and the European convention for the Protection of Vertebrate Animals Used for Experimental and Other Scientific Purposes. Animals were kept under the following conditions: ad libitum access to tap water and a standard rodent diet (Altromin, Germany), a 12:12 hr light:dark cycle, a temperature of 2121°C ± 2°C, and humidity of 55% ± 5%.

The BD model previously described by Kolkert et al. was used

. Several adaptations were made to ensure MRI compatibility, as was described in a prior study by our group

. Briefly, animals were anesthetized with 4% sevoflurane in 1 L/min O

and ventilated via a tracheostomy using a MR-compatible Small Animal Ventilator (SA Instruments, USA).

Cannulas were places in the femoral vein and artery for volume replacement and continuous

6

mean arterial pressure (MAP) monitoring, respectively. A frontolateral hole was drilled in the skull to allow insertion of a 4F Fogarthy catheter in the epidural space (Edwards Lifesciences Co, USA). Once the animal was placed in the MR scanner in prone position, BD was gradually induced in all animals from the BD groups through inflation of the balloon cathether using a syringe pump (0.16 ml/min) (Terufusion, Termo Co, Tokyo, Japan). Sham-operated (sham) rats served as controls for the experimental procedure, undergoing an identical procedure except for inflation of the balloon catheter as well as prolonged anesthesia during the entire experiment. BD was confirmed when the MAP rose to a level higher than 80 mmHg, following the drop in blood pressure characteristic for brain-dead animals. Following BD confirmation, the sevoflurane was switched off. This time point marked the beginning of the experimental period (T=0, see Fig 2). In the BD control group, the experiment was ended as soon as the animal received the first pyruvate shot and imaging was completed (described in more detail in the following paragraphs). In sham animals, 30 minutes was considered the equivalent of the BD induction period after which the experimental time was started (T=0, see Fig 2). Sham animals maintained anesthetized for the remaining 4 hrs. After 4 hrs, the experiment was termined as previously described

and the liver, kidneys, plasma, and urine were harvested. The left kidney was used for the subsequent reperfusion model. The remaining tissue, plasma, and urine was stored.

Figure 2. Experimental overview, where T=-2 marks the beginning of surgery, T=-0.5 the beginning of brain death induction, T=0 the beginning of the BD period, and the subsequent hrs the total BD duration. At T=0, T=2, and T=4, hyperpolarized 1-

Cpyruvate MR spectroscopy was performed, each time preceded by an injection of 1-

Cpyruvate. The BD control group was terminated following the first round of MR spectroscopy (shortly after T=0).

MRI assessment of ¹³ C-pyruvate metabolism

All MRI examinations were executed in a 9.4 T pre-clinical MR system (Agilent, Yarnton,

UK) with VnmrJ 4.0A (Agilent, Santa Clara CA), using a dual tuned 13C/1H volume rat coil

(Doty Scientific, Columbia, SC), as previously described

. After a 1H anatomical MR

scan to determine the location of the liver and kidneys, hyperpolarized 1-

Cpyruvate MR

spectroscopy was performed to quantify pyruvate metabolism in the liver and kidneys at

6

three different timepoints, at the start of BD (T=0), after two hours (T=2), and after four hours (T=4, see Fig 2), preceded each time by an injection of hyperpolarized 1-

Cpyruvate.

1-

Cpyruvate was polarized in a SpinLab (GE Healthcare, Brøndby, Denmark) using a method previously described

. Each time, 1.5 ml 1-

Cpyruvate solution was injected into the femoral vein in a period of ten sec, after which scanning was initiated. A multislice

C slice selective free induction decay (FID) sequences with two slices placed on the liver and kidneys, respectively, was used to acquire the spectra from each organ. Parameters for the sequence were: TR = 1 sec, flip angle = 10°, spectral width = 50 kHz, and 2048 complex points, and two 5 mm thick slices covering the liver and kidneys separately. The area under the curve (AUC) of the spectral peaks of [1-

C]pyruvate and each of the metabolites was calculated for the total signal using MATLAB.

Reperfusion model

Following 4 hrs of BD, the left renal artery, vein, and ureter were cannulated. The kidney was then flushed with ice-cold saline solutation, removed from the body and placed on ice.

The kidney was then normothermically perfused for 90 min in a so-called isolated perfused kidney (IPK) device (see Fig 3). The IPK system is a temperature- (37°C) and pressure- controlled device that allows for pulsatile perfusion of the kidney with oxygenated (95%

O

, 95% CO

partial oxygen pressure of at ≥ 60kPA) perfusion medium (perfusate) through the renal artery and vein , as has been described in more detail previously

. “Ultrafiltrate”

(UF) or urine was collected from the ureter cannula. The perfusion medium was prepared by adding 0.08 g/dL creatinine (Sigma-Aldrich, the Netherlands), 0.715 g/dL HEPES (Sigma- Aldrich, the Netherlands), and 5 g/dL Bovine Serum Albumin (BSA) (GE-healthcare, United Kingdom) to 100 mL William’s Medium E GlutaMAX (Life technologies, USA). To allow for glucose oxidization measurements, an additional 5 µl Glucose D-6-

H(N) tracer was added per 100 mL. Finally, the pH was set between 7.35-7.45 at 37°C using NaOH. During the 90 min experimental duration, the flow and oxygenation of the kidney was recorded at T=0, T=15, T=30, T=60, and T=90 min. At those same time points, arterial and venous perfusate and UF samples were collected. After 90 min, the kidney was disconnected and snap-frozen in liquid nitrogen or fixated in paraformaldehyde.

To evaluate renal function, the flow of perfusate through the kidney (“flow”), UF production,

and intrarenal vascular resistance (IRR) were calculated. The UF production was calculated

in mL/min. The IRR is a measure of vascular resistance that has been correlated to impaired

renal function following transplantation

. The IRR was calculated dividing the (constant)

pressure (mmHg) by the flow (mL/min). To determine the glomerular filtration rate (GFR),

6

Figure 3. Experimental setup of Isolated Perfused Kidney (IPK) device.

Glucose oxidation during IPK

Rates of glucose oxidation rates were measured using a tritium labelled glucose isotope (D-6- 3H-glucose) in accordance to a protocol modified from isolated heart perfusions

. A buffer volume of 100 mL (5μL D-6-3H-glucose/100 mL KH buffer) was recirculated. Glucose oxidation was quantified by produced 3H

O separated from labeled glucose by anion exchange chromatography on AG 1-X8 resin columns (Bio-Rad, Hercules, CA, USA). Next, 3H

O was suspended in 10 mL Opti-Phase scintillation solution (Perkin-Elmer, Shelton, CT, USA) and quantified by beta-scintillation on a TriCarb 2900TR liquid scintillation analyzer (Packard, Perkin, IL, USA) in disintegrations per minute (dpm). Results were weighted against the mass of each individual kidney (wet weight) and corrected for flow. Samples taken from before (“arterial”) and after the kidney (“venous”) were sampled at five different time points throughout the protocol (T=0, T=15, T=30, T=60, and T=90 min).

Plasma and urine injury markers

Plasma levels of Aspartate transaminase (AST), alanine aminotransferase (ALT), creatinine,

urea, lactate dehydrogenase (LDH), glucose, and lactate, and urine creatinine were

determined at the clinical chemistry laboratory of the University Medical Centre Groningen

according to standard procedures.

6

Enzyme activity assays and gene expression

The enzyme activities of LDH, PDH, and ALT were determined according to the manufacturer’s instructions (Sigma-Aldrich, USA). For gene expression, RNA was isolated from whole liver and kidney sections using TRIzol (Life Technologies, Gaithersburg, MD) using a method previously described (REF). Primer sets used to amplify Ldha, MCT1, and MCT4 are outlined in Table 1. Gene expression was normalized to the mean mRNA β-actin content and pooled cDNA from brain-dead rats used as an internal reference. Real-Time PCR was done according to methods described previously (REF) and results expressed as 2

(CT - Threshold Cycle).

Table 1. Primer sequences used for Real-Time PCR.

Gene Primers Amplicon size (bp)

Ldha 5’-AATATTACGTGAAATGTAAGATCTGCATATG-3’

5’-TTTTCCTTGGCATGACACTTGAG-3’ 70

MCT-1 5’-CAGTCTGACCTGTGGAGCATGA-3’

5’-TTTCGGATGTCTCGGGTCAC-3’ 52

MCT-4 5’-GAGGTCCAGAGACTGGCAACG-3’

5’-CTCAGCTGGTCCTGTTTGTGAAA-3’ 53

Tissue ATP levels

Hepatic and renal ATP was determined as a measure of the organ energy status. Frozen tissue was cut into slices of 20 µm to obtain around 50 mg of liver and kidney tissue. The tissue was homogenized in SONOP (0.372g EDTA in 130mL H2O and NaOH (pH 10.9) + 370 mL 96% ethanol), sonificated, and centrifuged to remove the precipitate. The protein content in the supernatant was determined (Pierce BCA Protein Assay Kit, Thermo Scientific, USA) and diluted with SONOP to obtain a concentration of 200-300 µg/mL. The supernatant was then mixed with phosphate buffer. ATP content was determined via bioluminescence (Bivro 14020 multilabel counter, PerkinElmer) in 50 µL of this mixture using an ATP Bioluminescence kit CLS II (Mannheim, Germany). Values were corrected for the amount of protein and recorded as µmol/g protein.

Oxidative stress markers

Oxidative damage was determined through estimation of the lipid peroxidation product

malondialdehyde (MDA) in plasma as well as tissue homogenates (each 20 µl). Levels of

MDA were measured fluorescently after binding of MDA to thiobarbituric acid reactive

substances (TBARS), which are formed as a byproduct of lipid peroxidation when reactive

oxygen species react with lipids in plasma membranes. TBARS in plasma and tissue were

determined using methods previously described

. In brief, 20 µL of homogenates or

plasma were mixed with 2% sodium dodecyl sulfate and 5mM butylated hydroxytoluene.

6

Next, this mixture was added to 400 μL 0.1N HCL, 200 μL 0.7% thiobarbituric acid, and 50 μL 10% phosphotungstic acid, and incubated at 97°C for one hour. After adding 800 μL 1-butanol, the samples were centrifuged for 10 min at 960 g. Fluorescence of 200 μL of the supernatant was measured at at 480 nm excitation and 590 nm emission wavelengths (Bivro 14020 multilabel counter, PerkinElmer). Samples were corrected for the amount of protein (Pierce BCA Protein Assay Kit, Thermo Scientific, USA) and expressed as μmol/g protein.

Statistical analysis

For the MRI data, a mixed-effects regression model with a restricted maximum likelihood method with repeated measures over time was used to analyze the impact of treatment (BD or sham) on different metabolites or IPK parameters. Data was checked for a normal distribution of variances. If this assumption was not met, data transformations were applied or extreme values removed from the analysis. The model included fixed effects of time, treatment, the interaction between time and treatment, and a random effect that took between-rat variations into account (STATA).

For the IPK data, a mixed-effects ANOVA was performed to check for significant interaction effects. If the interaction was not significant, we checked for main effects of group (BD or sham) or time. If significant differences were found, Bonferroni’s post tests were done to test for differences between individual groups (IBM SPSS Statistics 23).

For the enzyme activity assay analyses, data was checked a normal distribution. When variances of the dependent variable were equal across groups and there were no outliers, a pooled student T-test was done. If data that was not normally distributed, Mann Whitney tests were done to compare between two groups individually (IBM SPSS Statistics 23). All statistical tests were 2-tailed and p < 0.05 was regarded as significant. Results are presented as mean ± SD (standard deviation).

RESULTS

Brain death parameters

Throughout the experiment, the MAP was measured continuously. BD was confirmed based

on a rise in MAP following the drop in blood pressure that is characteristic for the slow

induction BD model. The induction of BD showed a uniform MAP pattern in all animals that

was in agreement with prior BD studies

, with a mean time to declare BD of 25.1 ± 3.14

min (Fig 4). To maintain a MAP higher than 80mmHg, the use of colloid HAES was required

in three brain-dead and three sham animals (0.38 ± 0.55 vs. 0.44 ± 0.63 mL, p = 0.846), and

the use of noradrenalin was required in two brain-dead animals (0.50 ± 0.87 mL, p = 0.170).

6

Figure 4. Average mean arterial blood pressure profile of sham and brain-dead animals. The induction phase was started at T=0 and showed a characteristic drop in blood pressure in the brain death (BD) group. No significant differences in blood pressure profiles between sham and brain-dead animals were observed during the 4 hr experimental duration after the start of BD.

Plasma injury markers

Several hepatic (AST, ALT, and LDH) and renal injury markers (urea and creatinine) were assessed to validate our model and to evaluate the effect of BD on tissue injury. In plasma of brain-dead animals, AST after 4 hrs (T=4) and creatinine immediately after BD induction (T=0), were significantly higher compared to sham animals (AST: 186.0 ± 90.9 vs. 104.0 ± 56.9 U/L, p = 0.017 and creatinine: 43.3 ± 3.2 vs. 28.4 ± 4.0, p < 0.001, Fig 5). Urine creatinine was significantly lower in brain-dead animals at T=4 compared to sham animals (4.0 ± 0.8 vs. 8.8 ± 1.3 µmol/L, p < 0.001, Fig 5). Plasma LDH and urea were significantly lower in BD animals at T=0 compared to sham animals (LDH: 265.0 ± 29.0 vs. 339.0 ± 150.0 U/L, p = 0.04 and urea: 7.7 ± 0.5 vs. 10.6 ± 1.0 mmol/L, p < 0.001, Fig 5). Finally, AST, ALT, LDH, and urea were significantly higher in brain-dead animals at T=4 (vs. brain-dead animals at T=0, AST:

186 ± 90.9 vs. 73.8 ± 23.0 U/L, p = 0.01; ALT: 144.0 ± 99.3 vs. 43.8 ± 11.3 U/L, p < 0.01, LDH:

580.0 ± 513.0 vs. 265.0 ± 293.0 U/L, p = 0.01; urea: 11.1 ± 1.5 vs. 7.7 ± 0.5 mmol/L, p < 0.001, Fig 5), whereas urine creatinine was significantly lower than brain-dead animals at T=0 (4.0

± 0.8 vs. 8.2 ± 1.5 µmol/L, p < 0.01, Fig 5).

Mean arterial pressure

End induction / start BD period

Normotension: 80 mmHg

Time (min)

mmHg

6

MRI assessment of ¹³ C-pyruvate metabolism

Enriched

C-pyruvate was injected repetitively to evaluate the metabolic profile at T=0, T=2, and T=4 hrs by assessing the relative conversion of pyruvate into lactate, alanine, and bicarbonate. The metabolic profiles of the liver and kidneys are portrayed in Fig 5. In order

Figure 5. Brain death induced hepatic and renal failure. Plasma levels of A. aspartate transaminase (AST), B. alanine transaminase (ALT), C. lactate dehydrogenase (LDH), D. urea, and E. plasma and F. urine creatinine, determined after 4 hrs (in the case of sham and brain-dead animals at T=4) of experimental time or immediately after brain death induction (brain-dead animals at T=0). Results are presented as mean ± SD (*p < 0.05, **p < 0.01, ***p < 0.001).

to determine the relative metabolic shift, each metabolite was normalized to the signal intensity of the sum of the products i.e. lactate, alanine, and bicarbonate. At T=0, brain dead animals showed significantly higher lactate levels, and significantly lower alanine levels in both the liver and kidney compared to sham animals (liver lactate T=0: p < 0.01;

liver alanine T=0: p = 0.03; kidney lactate T=0: p < 0.001; kidney alanine T=0: p = 0.03, Fig 6). At T=4, brain-dead animals showed significantly lower alanine levels compared to sham controls (p < 0.001, Fig 6). In the kidney, bicarbonate levels were significantly lower in brain-dead compared to sham animals at T=0 and T=4 hrs (T=0: p < 0.001, T=4: p = 0.04, Fig 6). Furthermore, the interaction term was also significant, showing a significant interaction term, indicating a significantly different effect of time on alanine levels in each group (p <

0.01, Fig 6). In the liver, lactate levels showed a significant interaction term (p < 0.001, Fig 5).

6

In order to evaluate the balance between alanine production in the Cahill cycle and lactate production, the alanine / lactate ratio was calculated in both organs. At T=0, brain- dead animals had a significantly lower alanine / lactate ratio compared to sham animals, suggesting both organs favored the production of lactate over the production of alanine at

Figure 6. Individual pyruvate metabolites in the liver and kidneys of sham and brain-dead animals.

The signal of each metabolite - A,D. lactate; B,E. alanine or C,F. bicarbonate - is represented as the fraction of the sum of the metabolites (lactate + alanine + bicarbonate). Results are presented as the minimum, first quartile, median, third quartile, for each group at each of the different time points, where T=0 represents the start of the BD period, and T=2 and T=4 two and four hours later, n = 8 per group (*/# p < 0.05, **/## p < 0.01).

the start of BD (liver: p < 0.001; kidney: p < 0.01, Fig 7). Furthermore, a statistically significant interaction was observed in both organs, indicating a significantly different effect of time on alanine / lactate ratio between the groups (liver: p < 0.001, kidney: p = 0.04, Fig 7). To evaluate the relative contribution between the liver and kidney of the total measured signal, we calculated the ratio of the relative signal in the liver divided by the signal in the kidney.

Our results indicate a higher lactate production in the kidney concomitant with a higher

alanine production in the liver (Fig 7). Regarding alanine production, brain-dead animals

favor the production of alanine in the liver over the kidney when compared to sham animals

(p = 0.01, Fig 7).

6 Figure 7. Alanine and lactate ratios in the liver and the kidneys following brain death. Ratio between alanine versus lactate signal in the A. liver and in the B. kidney of sham and brain-dead animals and the ratio between the C. lactate and D. alanine signal in the liver versus the kidney. Results are presented as the minimum, first quartile, median, third quartile, for each group at each of the different time points, where T=0 represents the start of the BD period, and T=2 and T=4 two and four hours later, n

= 8 per group (*/# p < 0.05, ***/### p < 0.001).

Kidney function and glucose oxidation in isolated perfused kidney model

Renal function was assessed in an isolated perfused kidney model following BD by assessing the flow, UF production, IRR, and GFR (creatinine clearance). We found a significant interaction of flow and IRR, indicating a significantly different effect of time on flow and IRR in brain-dead versus sham animals (both p < 0.01, Fig 8). In brain-dead animals, the IRR was significantly lower at T=0 compared to sham animals (IRR: 14.7 ± 7.00 vs. 35.0 ± 21.0 mmHg/

(mL/min), p < 0.001, Fig 8). Finally, glucose oxidation was significantly different between the

two groups, as well as over the course of time, between brain-dead and sham animals (main

effect group: p < 0.01, main effect time: p < 0.001, Fig 8), highlighting the significantly lower

glucose oxidation levels during the reperfusion period in brain-dead animals.

6

Figure 8. Reduced glucose oxidation following brain death in the isolated perfused kidney. A.

Perfusate flow, B. ultrafiltrate production, C.intrarenal vascular resistance, D. glomerular filtration rate (GFR, creatinine clearance), and E. glucose oxidation rates in the kidney of brain-dead and sham animals. Results are presented as mean ± SD, n = 6 per group (*/#/Ɵ p < 0.05, ##/ Ɵ Ɵ p < 0.01, ***

p < 0.001).

Enzyme activities and gene expression

We evaluated the different enzymes involved in pyruvate metabolization to assess whether

the observed changes could be explained by changes in enzyme activities of LDH, PDH, or

ALT or gene expression of LDHA, MCT1 and MCT4. At T=4 hrs, a significant reduction in LDH

enzyme activity was observed in the liver of brain-dead animals (vs. BD (T=0), 0.50 ± 0.15

vs. 0.61 ± 0.27 units/mg, p = 0.04. Fig 9). The activites of the remaining enzymes were not

significantly different between the groups.

6

Figure 9. No difference in LDH, PDH, and ALT enzyme activities following brain death. Enzyme activities of A,B. lactate dehydrogenase (LDH); C,D. pyruvate dehydrogenase (PDH); and E,F. alanine transaminase (ALT) in the liver and kidney of sham and brain-dead animals, determined after 4 hrs (in the case of sham and brain-dead animals at T=4) of experimental time or immediately after brain death induction (brain-dead animals at T=0). Results are presented as mean ± SD (*p < 0.05).

Decreased ATP levels in the liver following brain death

To estimate cellular energy status in the liver and kidney, cellular ATP content was measured.

In the liver of brain-dead animals, ATP levels were reduced at T=0 and T=4 compared to

sham animals (T=0: 29.0 ± 11.5 vs. 47.9 ± 7.53 µmol/g protein, p = 0.01; T=4: 30.8 ± 15.4

µmol/g protein, p = 0.04, Fig 10). ATP levels in the liver were not different between groups

6

(sham: 26.3 ± 10.8 µmol/g protein, T=0: 26.9 ± 11.3 µmol/g protein, T=4: 21.1 ± 8.89 µmol/g protein, Fig 10). These results suggest a decreased bio-energetic efficiency in the liver following BD.

Figure 10. Decreased ATP levels in the liver but not the kidneys following brain death.

A. ATP content in liver and B. kidney tissue determined after 4 hrs (in the case of sham and brain-dead animals at T=4) of experimental time or immediately after brain death induction (brain-dead animals at T=0). Results are presented as mean ± SD, n = 8 per group (* p < 0.05).

Oxidative stress markers in tissue and plasma

Tissue and plasma oxidative stress markers MDA were significantly increased in the plasma of brain-dead animals immediately after BD induction as well as after 4 hrs compared to sham animals (T=4: 3.31 ± 0.38 vs. 2.85 ± 0.41 µM, p = 0.04, T=1: 2.46 ± 0.33 µM, p < 0.01, Fig 10). MDA levels in the kidney of brain-dead animals were significantly different from sham animals (Sham: 0.14 ± 0.03 µmol/g protein, T=0: 0.20 ± 0.10 µmol/g protein, T=4: 0.16

± 0.04 µmol/g protein, Fig 10).

Figure 10. Increased oxidative stress in plasma but not the kidney following brain death.

Malondialdehyde (MDA) in A. plasma and B. the kidney determined after 4 hrs (in the case of sham

and brain-dead animals at T=4) of experimental time or immediately after brain death induction

(brain-dead animals at T=0). Results are presented as mean ± SD, n=8 per group (*p < 0.05, **p < 0.01).

6

DISCUSSION

The primary focus of this pilot study was to investigate the feasibility of two different techniques to visualize metabolic changes in the liver and kidney during BD, and in the kidney following BD in an ex vivo reperfusion model. Our result showed a distinctly different metabolic profile in the liver versus the kidney, with a preference for the production of alanine in the case of the liver, yet lactate in the case of the kidney. Secondly, our data indicate that immediately following BD induction, lactate production increased in both organs, likely reflecting the hypoxic changes as a result of the hemodynamic instability during BD induction. Finally, metabolic assessment during reperfusion of the kidney showed a distinct reduction in glucose oxidation in brain-dead compared to sham animals. Together, these data suggest that hyperpolarization MRI and the measurement of glucose oxidation during renal reperfusion are two promising techniques that can be used to visualize the metabolic profile of the liver and kidney during BD as well as during subsequent renal reperfusion.

Immediately after BD induction in the liver and kidney, we observed a significantly increased

lactate signal, and reduced alanine signal, exemplified by a decrease in the alanine/lactate

ratio. In a similar manner, cardiac ischemia reperfusion injury also resulted in a substantial

increase in

C lactate levels

. In a model of renal ischemia reperfusion injury, 60 min

of ischemia and subsequent 24 hrs of reperfusion resulted in a significant decrease in

metabolite (lactate and alanine) to pyruvate signals as well as significantly lower enzyme

activities, reflecting a decrease in metabolic function

. However, when looking at the

ratios (and thus the preference of the different pathways), IRI resulted in a higher lactate

to alanine ratio as well as increased lactate levels in in the kidney

, similarly to what we

observed during BD. During liver surgery, a reduction in blood flow also resulted in increased

lactate production

. In each of these models, ischemic changes have resulted in increased

lactate production likely reflecting a shift from aerobic to anaerobic glycolysis

. In our

model, the increased lactate production and decreased alanine/lactate ratio were likely the

result of hemodynamic instability, as BD induction is characterized by reduced perfusion

and thus hypoxia of the abdominal organs, including the liver and kidney

. These increased

lactate signals could potentially be of clinical importance, as increased lactate levels during

organ reperfusion were correlated to a higher risk of early allograft dysfunction following

liver transplantation

and delayed graft function following kidney transplantation

. Despite

these studies, there is currently insufficient evidence that lactate or any other specific

biomarker has a strong prognostic value predicting transplantation outcomes

. With this

in mind, it was interesting to observe that the level of glucose oxidation, and not renal

creatinine clearance, was significantly altered in kidneys from brain-dead compared to sham

kidneys. This suggests that (a reduction in) glucose metabolism, as estimated by the amount

of 3H

O liberated through oxidation of D-6-3Hglucose in the TCA cycle, might be a better

6

predictor of organ quality than standard biomarkers such as creatinine clearance. Additional transplantation studies will need to be conducted to assess whether glucose oxidation might have a possible prognostic value.

Using

C pyruvate in the MRI, we were able to visualize a distinctly different metabolic profile in the liver compared to the kidney. Whereas the liver of sham animals consistently showed a higher pyruvate to alanine conversion, the kidney showed an increased pyruvate to lactate conversion. Interestingly, BD only seemed to enhance this difference between the liver and the kidney. Albeit not significantly, we observed a trend towards increased lactate production in the kidneys (Fig 6). These changes could reflect a suppression of gluconeogenesis (and thus increased lactate pools) and an altered redox state as a result of hypoxia in the kidneys

. This is supported by a prior study on the effects of BD in the kidney, where gluconeogenesis-related gene expression was decreased, while anaerobic, glycolytic-gene expression was found to be increased

. Therefore, this higher lactate production in the kidneys might reflect increased anaerobic production of lactate and lower gluconeogenesis in the kidney. This notion is supported by a prior study that showed increased levels of oxidative stress and a decline in renal perfusion as BD progressed

. Conversely, our results indicate that the liver to maintained a steady pyruvate to alanine conversion as well as unaltered enzyme activities. These changes might reflect the ability of the liver to maintain aerobic glycolysis, as increased alanine and ALT activity have been associated with adequate cellular oxygen availability

. This idea is further supported by a prior study that showed increased expression of glycolysis-related gene Pfk-1 as well as higher oxygen consumption in the liver during BD

.

To validate our BD model, we assessed several hepatic and renal injury markers, as well as markers of oxidative stress and tissue energy status. Our result confirm that BD resulted in increased hepatic and renal injury, as was in line with prior animal and human BD studies

. Furthermore, we confirm increased levels of oxidative stress as evidenced by higher MDA levels in plasma following BD. Finally, the cellular energy status was decreased in the liver, immediately after BD as well as four hours later. Even though ATP and MDA levels were not significantly changed in the kidney of brain-dead animals, our data show a trend of decreased energy levels and increased oxidative stress, similar to a prior study from our group

. Altogether, these data confirm that the adjustments made to our model to allow for the administration of labelled pyruvate during BD did not impact the efficacy of our BD model.

This pilot study has several limitations. As the aim of this study was to evaluate the potential

of two metabolic assessment techniques, this study was not powered to statistically support

some of the trends we observed in our data. Furthermore, this study did not evaluate

6

whether these metabolic changes can be correlated to organ quality or can be predictive of transplantation outcome. Finally, we observed no significant alterations in LDH, ALT, or PDH, suggesting that changes in enzyme activities were not responsible for the changes in the metabolic profiles. Therefore, additional analyses on organ perfusion, oxygenation, and several transporter proteins will need to be conducted in a future study to assess whether the observed trends were affected by organ perfusion or limited by the uptake of any of the compounds.

In conclusion, this study shows for the first time that metabolic processes during BD in

the liver and kidney can be visualized non-invasively using hyperpolarized MRI, as well as

afterwards during ex vivo renal reperfusion by assessing glucose oxidation. Our data showed

significant differences in the metabolic profile of both the liver and the kidney during BD

onset. These changes might be in part responsible for the subsequent, albeit more subtle,

metabolic changes as well as other homeostatic perturbances in the organs or afterwards

during reperfusion. This study lays the groundworks for future studies aimed at investigating

and treating BD-induced metabolic changes and may be part of the key to improving

transplantation outcomes.

6 REFERENCES

1. Organ Procurement and Transplantation Network. Available at: https://optn.trans- plant.hrsa.gov/. (Accessed: 6 July 2017) 2. Eurotransplant. Available at: https://www.

eurotransplant.org/cms/. (Accessed: 20 February 2018)

3. Rudge, C., Matesanz, R., Delmonico, F. L.

& Chapman, J. International practices of organ donation. Br J Anaesth 108 Suppl 1, i48–55 (2012).

4. Mascia, L. et al. Ventilatory and hemody- namic management of potential organ do- nors: an observational survey. Critical Care Medicine 34, 321–7– quiz 328 (2006).

5. Fitzgerald, R. D. et al. Intraoperative cate- cholamine release in brain-dead organ do- nors is not suppressed by administration of fentanyl. Eur J Anaesthesiol 20, 952–956 (2003).

6. Adrie, C. et al. Immune status and apopto- sis activation during brain death. Shock 33, 353–362 (2010).

7. Weiss, S. et al. Brain death activates do- nor organs and is associated with a worse I/R injury after liver transplantation. Am J Transplant 7, 1584–1593 (2007).

8. Floerchinger, B., Oberhuber, R. & Tullius, S.

G. Effects of brain death on organ quality and transplant outcome. Transplant Rev (Orlando) 26, 54–59 (2012).

9. Van Erp, A. C. et al. Organ-specific respons- es during brain death: increased aerobic metabolism in the liver and anaerobic me- tabolism with decreased perfusion in the kidneys. Sci Rep 8, 4405 (2018).

10. Roelsgaard, K. et al. Effects of brain death and glucose infusion on hepatic glycogen and blood hormones in the pig. Hepatology 24, 871–875 (1996).

11. Novitzky, D., Cooper, D. K. C., Rosendale, J.

D. & Kauffman, H. M. Hormonal therapy of the brain-dead organ donor: experimental and clinical studies. Transplantation 82, 1396–1401 (2006).

12. Hoeksma, D. et al. Inadequate anti-oxida- tive responses in kidneys of brain-dead rats. Transplantation (2016). doi:10.1097/

TP.0000000000001417

13. Rebolledo, R. A. et al. Anti-Apoptotic Effects of 3,3’,5-Triiodo-L-Thyronine in the Liver of Brain-Dead Rats. PLoS ONE 10, e0138749 (2015).

14. Nijboer, W. N. et al. Effects of brain death on stress and inflammatory response in the human donor kidney. Transplantation Pro- ceedings 37, 367–369 (2005).

15. Westendorp, W. H., Leuvenink, H. G. &

Ploeg, R. J. Brain death induced renal inju- ry. Curr Opin Organ Transplant 16, 151–156 (2011).

16. Bretan, P. N. et al. Pretransplant assess- ment of renal viability by phosphorus-31 magnetic resonance spectroscopy. Clinical experience in 40 recipient patients. Trans- plantation 48, 48–53 (1989).

17. Hené, R. J., van der Grond, J., Boer, W. H., Mali, W. P. & Koomans, H. A. Pre-transplan- tation assessment of renal viability with 31P magnetic resonance spectroscopy. Kidney International 46, 1694–1699 (1994).

18. Wijermars, L. G. M. et al. Defective post- reperfusion metabolic recovery direct- ly associates with incident delayed graft function. Kidney International 90, 181–191 (2016).

19. Nakatani, T., Spolter, L. & Kobayashi, K. Ar-

terial ketone body ratio as a parameter of

hepatic mitochondrial redox state during

and after hemorrhagic shock. World J Surg

6

19, 592–596 (1995).

20. Yamaoka, Y. et al. Evaluation of the liver graft before procurement. Significance of arterial ketone body ratio in brain-dead pa- tients. Transpl. Int. 3, 78–81 (1990).

21. Klionsky, D. J. & Emr, S. D. Autophagy as a regulated pathway of cellular degradation.

Science 290, 1717–1721 (2000).

22. Taylor, R. C., Cullen, S. P. & Martin, S. J.

Apoptosis: controlled demolition at the cellular level. Nat. Rev. Mol. Cell Biol. 9, 231–241 (2008).

23. Van Erp, A. C. et al. The Crosstalk between ROS and Autophagy in the Field of Trans- plantation Medicine. Oxid Med Cell Longev 2017, 1–13 (2017).

24. Golman, K., in ‘t Zandt, R. & Thaning, M. Re- al-time metabolic imaging. Proc. Natl. Acad.

Sci. U.S.A. 103, 11270–11275 (2006).

25. Schroeder, M. A., Clarke, K., Neubauer, S.

& Tyler, D. J. Hyperpolarized magnetic res- onance: a novel technique for the in vivo assessment of cardiovascular disease. Cir- culation 124, 1580–1594 (2011).

26. Morze, von, C. et al. Detection of localized changes in the metabolism of hyperpolar- ized gluconeogenic precursors 13C-lactate and 13C-pyruvate in kidney and liver. Mag- netic Resonance in Medicine 77, 1429–1437 (2016).

27. Hansen, E. S. S., Stewart, N. J., Wild, J. M., Jørgensen, H. S. & Laustsen, C. Hyperpolar- ized 13C,15N2-Urea MRI for assessment of the urea gradient in the porcine kidney.

Magnetic Resonance in Medicine 76, 1895–

1899 (2016).

28. Mariager, C. Ø., Nielsen, P. M., Qi, H., Ring- gaard, S. & Laustsen, C. Hyperpolarized 13C, 15N 2-urea T 2relaxation changes in acute kidney injury. Magnetic Resonance in Medicine 5, 198ra108 (2017).

29. Bhangoo, R. S., Hall, I. E., Reese, P. P. &

Parikh, C. R. Deceased-donor kidney per- fusate and urine biomarkers for kidney allograft outcomes: a systematic review.

Nephrol. Dial. Transplant. 27, 3305–3314 (2012).

30. Moers, C. et al. The Value of Machine Per- fusion Perfusate Biomarkers for Predicting Kidney Transplant Outcome. Transplanta- tion 90, 966–973 (2010).

31. Kolkert, J. L. P. et al. The gradual onset brain death model: a relevant model to study or- gan donation and its consequences on the outcome after transplantation. Lab. Anim.

41, 363–371 (2007).

32. Qi, H. et al. Early diabetic kidney maintains the corticomedullary urea and sodium gra- dient. Physiol Rep 4, e12714 (2016).

33. Bertelsen, L. B. et al. Diabetes induced re- nal urea transport alterations assessed with 3D hyperpolarized13C,15N-Urea. Magnetic Resonance in Medicine 77, 1650–1655 (2017).

34. Nielsen, P. M. et al. In situ lactate dehy- drogenase activity: a novel renal cortical imaging biomarker of tubular injury? Am.

J. Physiol. Renal Physiol. 312, F465–F473 (2017).

35. Mahboub, P. et al. Gradual Rewarming with Gradual Increase in Pressure during Ma- chine Perfusion after Cold Static Preserva- tion Reduces Kidney Ischemia Reperfusion Injury. PLoS ONE 10, e0143859 (2015).

36. Chudek, J. et al. The intrarenal vascular re- sistance parameters measured by duplex Doppler ultrasound shortly after kidney transplantation in patients with immediate, slow, and delayed graft function. Transplan- tation Proceedings 38, 42–45 (2006).

37. Hjortbak, M. V. et al. Influence of diabetes

mellitus duration on the efficacy of isch-

emic preconditioning in a Zucker diabetic

6

fatty rat model. PLoS ONE 13, e0192981 (2018).

38. Rebolledo, R. A. et al. Slow induction of brain death leads to decreased renal func- tion and increased hepatic apoptosis in rats. J Transl Med 14, 141 (2016).

39. Ball, D. R. et al. Metabolic imaging of acute and chronic infarction in the perfused rat heart using hyperpolarised [1-13C]

pyruvate. NMR Biomed 26, 1441–1450 (2013).

40. Theodoraki, K. et al. Transhepatic lactate gradient in relation to liver ischemia/reper- fusion injury during major hepatectomies.

Liver Transpl. 12, 1825–1831 (2006).

41. Hems, D. A. & Brosnan, J. T. Effects of isch- aemia on content of metabolites in rat liver and kidney in vivo. Biochem. J. 120, 105–

111 (1970).

42. Herijgers, P., Leunens, V., Tjandra-Maga, T. B., Mubagwa, K. & Flameng, W. Chang- es in organ perfusion after brain death in the rat and its relation to circulating cate- cholamines. Transplantation 62, 330–335 (1996).

43. Kim, D. G. et al. Clinical significance of lac- tate clearance for the development of early allograft dysfunction and short-term prog- nosis in deceased donor liver transplanta- tion. Clin Transplant 31, e13136 (2017).

44. Lewis, A. J. M. et al. Assessment of Met- formin-Induced Changes in Cardiac and Hepatic Redox State Using Hyperpolar- ized[1-13C]Pyruvate. Diabetes 65, 3544–

3551 (2016).

45. Qi, H. et al. Acute renal metabolic effect of metformin assessed with hyperpolar- ised MRI in rats. Diabetologia 61, 445–454 (2018).

46. Gerich, J. E., Meyer, C., Woerle, H. J. &

Stumvoll, M. Renal gluconeogenesis: its importance in human glucose homeostasis.

Diabetes Care 24, 382–391 (2001).

47. Akhtar, M. Z. et al. Using an Integrated -Omics Approach to Identify Key Cellular Processes That Are Disturbed in the Kid- ney After Brain Death. Am J Transplant 16, 1421–1440 (2016).

48. Laustsen, C. et al. Insufficient insulin ad-

ministration to diabetic rats increases sub-

strate utilization and maintains lactate pro-

duction in the kidney. Physiol Rep 2, e12233

(2014).