UW Supplementation with AP39 Improves Liver Viability Following Static Cold Storage

Static cold storage of donor livers at 4°C incompletely arrests metabolism, ultimately leading to decreases in ATP levels, oxidative stress, cell death, and organ failure. Hydrogen Sulfide (H2S) is an endogenously produced gas, previously demonstrated to reduce oxidative stress, reduce ATP depletion, and protect from ischemia and reperfusion injury. H2S is difficult to administer due to its rapid release curve, resulting in cellular death at high concentrations. AP39, a mitochondrially targeted, slow-release H2S donor, has been shown to reduce ischemia-reperfusion injury in hearts and kidneys. Thus, we investigated whether the addition of AP39 during 3-day static cold storage can improve liver graft viability. At the end of storage, livers underwent six hours of acellular normothermic machine perfusion, a model of transplantation. During simulated transplantation, livers stored with AP39 showed reduced resistance, reduced cellular damage (ALT and AST), and reduced apoptosis. Additionally, bile production and glucose, as well as energy charge were improved by the addition of AP39. These results indicate that AP39 supplementation improves liver viability during static cold storage.


INTRODUCTION
Liver transplantation is the only viable treatment option for patients in end-stage liver failure.However, its broad application is limited by the number of available donor organs [1].A signi cant limiting factor to the expansion of the donor pool is the loss of viability occurring during transport/preservation.The duration of ischemic cold storage correlates with early allograft dysfunction (EAD) and reduced longterm survival of the grafts (Giwa, S).As a result, thousands of organs are discarded each year (Haugen, C).This clinical problem suggests that better preservation techniques are needed to improve graft quality and help combat the global donor organ shortage crisis [2].
Hydrogen sul de (H 2 S) is an endogenously produced gaseous molecule through both enzymatic degradation of cysteine via cystathionine γ-lyase (CGL), or non-enzymatic degradation of thiol-containing molecules [3].H 2 S is proangiogenic, reduces mitochondrial stress, and can regulate the eNOS-NO pathway [4,5].H 2 S also has anti-in ammatory and antioxidant properties, and can reversibly inhibit the mitochondrial electron transport chain, thus reducing reactive oxygen species (ROS) formation during reperfusion [6].During ischemia, H 2 S promotes glucose uptake and glycolytic ATP production [7].Mice lacking endogenous H 2 S production showed increased damage and mortality following renal ischemiareperfusion injury, and the introduction of exogenous H 2 S (NaHS) was shown to reverse this effect [8].
Similarly, it has been shown that the introduction of exogenous NaHS in wild-type mice reduces both hepatic and renal ischemia-reperfusion injury [8,9].The addition of NaHS to University of Wisconsin solution (UW) preservation solution during SCS reduced necrosis and apoptosis, improving kidney function after transplantation in rats [10].Despite great success in mitigating the effect of ischemia, NaHS is limited in its application due to the rapid, uncontrollable rate of H 2 S production, resulting in inhibition of mitochondrial electron complexes I and IV and cellular death at high concentrations [11,12].
AP39 is a mitochondrial-targeting, slow-release H 2 S donor synthesized to improve the mito-protective effects of H 2 S via extended release, and sustained, low-dose release for up to 10 days.Additionally, the introduction of a TPP moiety [13,14] targets H 2 S at the mitochondria.In a rat kidney transplant model, SCS with 200 nM AP39, resulted in approximately three times increase in survival at 7 days, and increased creatinine clearance [15].Consistently, the addition of AP39 during porcine kidneys subnormothermic perfusion (21°C) for 4 hours with an O 2 carrier (Hemopure) improved urine output and graft oxygenation [16].In renal epithelial cells, the addition of 400 nM AP39 during SCS reduced ROS production [15].Similarly, in a heterotopic mouse heart transplant model, 200 nM AP39 improved left ventricular ejection fraction and reduced brosis following transplantation 24 hours after SCS [17].
In the liver, hepatocytes, make up 20-25% of overall cellular volume.Mitochondria are the main energy source in hepatocytes and are at the center of many of the signaling pathways that mediate hepatocyte injury during ischemia.Thus hypothesized that AP39 supplementation could improve liver viability during SCS.In this study, we tested the bene ts of AP39 during liver SCS for 3 days.Following storage, liver viability was evaluated using acellular machine perfusion, allowing real-time assessment of perfusion quality and molecular injury (Fig. 1).

Liver Procurement
This study is reported in accordance with ARRIVE guidelines.Female Lewis rats (250-300g, Charles River Laboratories, Boston MA, USA) were socially housed in controlled, standard conditions (12-hour light/day cycle, 12C, 30-70% humidity, pathogen-free HEPA ltered ventilated cages, mixed paper/cellulose bedding).All rats had unfettered access to sterile water and chow, as in accordance with National Research Council Guidelines.All rats were cared for by the Massachusetts General Hospital (MGH) Center for Comparative Medicine (CCM).The experimental protocol was approved by the Institutional Care and Use Committee (IACUC) of MGH (Protocol #2011N000111), and all experiments were performed in accordance with established guidelines.Livers were procured as previously described [18].Brie y, donor rats were anesthetized under 3% iso urane and maintained at 1%.A transverse abdominal incision was made and the ligaments connecting the superior and inferior portions of the liver were dissected.The gastric and splenic branches of the portal, as well as the hepatic artery, were ligated with 6 − 0 silk (Fine Science Tools inc, Foster City CA, USA).The bile duct was then partially dissected and cannulated with PE-10 tubing (Fisher Scienti c).0.1 U/g heparin was injected into the inferior vena cava through a 30G insulin syringe (Westnet, Canton MA, USA). 5 minutes later, the portal vein was cannulated with a 16G cannula (Westnet), and the liver was immediately ushed with 50 mL UW at approximately 10mL/min, either with or without 200 nM AP39.The remaining connective tissue was then dissected, and the liver was freed from the abdomen.The liver was immediately weighed, and subsequently either perfused for 6 hours at 37°C as described below (fresh control, n = 4), or ushed with UW with 200 nM AP39 (MedChemExpress, Monmouth Junction NJ, USA, n = 6) or vehicle (0.13% v/v dichloromethane) and stored on ice in the same respective solution for 3 days.

Machine Perfusion
Livers were perfused on a homemade machine perfusion system as previously described [19].Brie y, a roller pump (Master ex L/S, Vernon Hills IL, USA) circulated perfusate from a 500 mL basin using 16G in and out ow tubing (Master ex).Before reaching the liver, the circuit entered a double-jacketed oxygenator (Radnoti, Covina CA, USA), followed by a bubble trap (Radnoti).The system was heated to 37°C by a circulating water bath (PolyScience, Niles Il, USA).In ow perfusate oxygen concentration was maintained between 500-600 mmHg by a 21% O2, 5% CO2, balance N2 tank (Airgas, Radnor PA, USA).
The liver intravascular pressure was zeroed according to system pressure using a portable pressure monitor (Sciatica, London ON, Canada), continuously monitored throughout the perfusion.The liver was hand-ushed with 50 mL lactated ringers (Baxter, Deer eld Il, USA), and attached to the system at a ow of 5mL/min.After a short (1-2 min) adjustment period, the ow was rapidly raised to 30 mL/min, maintaining a pressure below 11 mmHg.Out ow samples were collected from the suprahepatic IVC, every 30 minutes, and in ow samples were taken from a side port immediately before the arterial cannula perfusing the liver.Samples were analyzed using a Siemens Rapidpoint 500 (Siemens, Munich, Germany).Oxygen consumption was calculated according to the following equation: OUR = (in ow O2out ow O2) * ow rate / initial weight.Resistance was calculated according to the following equation: R = pressure/ ow rate/initial weight.Pressure and ow were recorded every 15 minutes for the rst 2 hours, and every 30 minutes thereafter.At the end of perfusion, two biopsies were taken from the peripheral left lateral lobe; one of which was stored in 1% formalin, while the other was immediately snap-frozen in liquid nitrogen.

ALT and AST Assay
AST and ALT levels were measured using a commercially available colorimetric activity assay (Cayman Chemicals, Ann Arbor MI, USA) according to the manufacturer's instructions and as previously published [20].Out ow perfusate from 1, 3, and 6 hours was incubated with LDH enzyme, and the oxidation of NADH was measured over time according to the absorbance at 340 nm.

Histological Analysis
Histology samples were moved from formalin to 70% ethanol after 24 hours.Sections were stained with hematoxylin and eosin (H&E) and terminal deoxynucleotidyl dUTP nick end labeling (TUNEL) as previously published [21,22].Slides were then imaged at 20X on a Nikon Eclipse E800.On H&E slides, liver sinusoidal endothelial cells (LSEC) sloughing and congestion of the portal vein were analyzed.
TUNEL staining was quanti ed using the Weka trainable segmentation plugin in Fiji [23].Brie y, nuclei were classi ed as apoptotic if they were stained brown, or alive if they were stained purple.A probability map of live and dead cells was produced and particle count was applied to nd the ratio of dead cells to live cells.All tissue processing was performed at the MGH Histology Molecular Pathology Core Facility (Boston, MA, USA).

Metabolite analysis
Metabolites were analyzed as previously described [24,25].Liver samples were crushed in liquid nitrogen, and metabolites were extracted using an established procedure [26].All mass spectrometry experiments were performed on a Triple TOF 6600 system (AB Sciex) hooked with a Shimadzu HPLC LC20AD (Shimadzu America) system.Compounds were separated on an analytical Luna NH2 column 2 x 150 mm, 3 um, 100Å equipped with a 2.0 x 4 mm guard column (Phenomenex) using the following conditions: mobile phase A − 100% 5mM ammonium acetate in water, adjusted to pH 9.9 with ammonium hydroxide; mobile phase B − 100% acetonitrile (ACN).Brie y, injection was performed at 20% A, followed immediately by a linear gradient to 100% A over 20 min, hold at 100%A for 4 min, drop to 20% A over 1 min and hold for 5mins at 20%A.The ow rate was set at 0.2ml/min; column temperature was 25°C; injection volume was 2 µl, and autosampler temperature was 4°C, with a total runtime of 30 min including mobile phase equilibration.The mass spectrometer was set to acquire TOF MS spectrum followed by a dedicated product ion spectrum in high sensitivity mode for all nine metabolites of interest.This work ow is also referred to as MRMHR by the vendor.MS spectrum dwell time was 250 msecs and each product ion spectrum was 100msecs.All mass spectrometer experiments were performed in positive electrospray ionization mode.The instrument was set to autocalibrate after acquisition of 5 samples.Brie y, autocalibration was performed by injecting 1 µl of a solution containing 0.5 µmolar AMP, 0.3 µmolar GSSG & 0.2 µmolar FAD.Ion source parameters were as follows: nebulizer gas (gas 1) was 50 psi, heater gas (gas 2) was 55 psi, source temperature 450°C, ionspray voltage was 5500 V, mass range for each experiment 100-900 m/z.Once the mass spec data are recorded, MultiQuant 3.0.2(AB Sciex) software was used for quantitation by generating chromatographic peak areas.Concentrations of metabolites in unknown samples were determined from standard curves constructed for each metabolite in the MultiQuant software.An eight point standard curve was generated each time prior to running samples using a mixture of known concentrations of the metabolites.All compounds eluted between 13 and 22 min.

Statistical Analysis
Statistical analysis and graphing were performed using Prism 10 version 10.0.3 (GraphPad Software, San Diego CA, USA).All data was analyzed using ordinary one-way ANOVA followed by Tukey's multiple comparison test to compare groups and determine signi cance.Data was reported as means with standard deviation, differences were considered signi cant when p < 0.05.

Addition of AP39 to UW Improves Liver Perfusion Following 3 Days of Static Cold
Storage First, we examined the effect of AP39 during SCS.Livers that were immediately harvested (fresh) or stored on ice for 3 days with (AP39) or without AP39 (SCS) were subsequently evaluated during a 6-hour ex-vivo normothermic perfusion, previously successfully employed to model transplantation [27].After 3 days of SCS oxygen uptake was reduced compared to fresh livers.No difference was observed between livers stored with and without AP39.(Fresh 49.1 uL O2/min*g ± 15.8, SCS 33.9 uL O2/min*g ± 13.1, AP39 34.6 uL O2/min*g ± 11.0, p = 0.9937, Fig. 2a).Importantly, AP39 supplementation reduced vascular resistance (0.019 mmHg*min/L*g ± 0.012, p = 0.0457) compared to SCS (0.027 mmHg*min/L*g ± 0.012), and was comparable to fresh livers (0.012 mmHg*min/L*g ± 0.010, p < 0.0001, Fig. 2b).No difference between the groups was observed in perfusion ow rate or edema at the end of perfusion (Figure S1a-c).Consistently, alanine aminotransferase (ALT, Fig. 2c) and aspartate aminotransferase (AST, Fig. 2d) liver transaminase re ecting cellular injury, were reduced in AP39-treated livers to levels similar to freshly perfused liver.No difference was observed in out ow pH, out ow lactate, or out ow glucose (Figure, S1d-f).All perfusate electrolytes remained within normal range throughout perfusion (Figure S2).

DISCUSSION
In this study, the addition of the slow-releasing, mitochondrial targeting H 2 S donor AP39 to UW storage solution during SCS reduced post-reperfusion injury and improved cellular function in rat livers.While AP39 was shown to improve heart, kidney, and pancreas function following cold storage, this is the rst study investigating its impact during SCS in a liver model [17,30,31].
The current clinical limitation for liver storage prior to transplantation is between 9 and 12 hours, restricted by the persistence of metabolism at 4°C, inexorable consumption of cellular energy stores, and ROS production during reperfusion [32].Surprisingly we observed that oxygen consumption was reduced in livers treated with AP39.Oxygen consumption is a measurement of oxygen extraction during perfusion, shown to correlate with transplant outcomes in rat models [33].Following cold storage, rat livers are known to exhibit reduced oxygen consumption, which is thought to re ect mitochondrial dysfunction [22,34].Similarly, H 2 S transiently inhibits oxygen consumption in the absence of cellular injury and induces a suspended animation state via inhibition of oxidative phosphorylation at complex I and IV [4,[35][36][37].Consistently, with the bene t of transition inhibition of oxidative phosphorylation by H 2 S, AP39 was also associated with a reduction in hepatic transaminase (ALT and AST), a surrogate of hepatocellular injury [38].Additionally, apoptosis was reduced in AP39 treated liver, which might be associated with a reduction in the release of damage-associated molecular patterns following reperfusion [39].
Vascular resistance was improved in livers treated with AP39.Of interest, LSECs are highly susceptible to IR injury [40].In addition, AP39 was shown to promote vasorelaxation through modulation of NOsignaling; indicating that AP39 may directly improve LSEC outcome following reperfusion [41,42].H 2 S has been shown to directly improve LSEC health during sepsis by reducing defenestration [43].LSEC sloughing is a common indicator of LSEC following reperfusion injury [44].Similarly, a reduction in LSEC sloughing was observed on H&E.Consistently, portal congestion is commonly observed when blood ow is impaired [45].Interestingly, a reduction in portal congestion was observed in livers stored with AP39 [46].Compared to fresh livers, SCS resulted in a reduction in bile production.Although not signi cant, this was improved by AP39.Of interest, the relationship between bile production and bile glucose demonstrated 2 distinct clusters between livers stored with and without AP39.Potentially, a larger difference and stronger clustering may have been observed if bile glucose were able to be measured for all livers stored without AP39 (no bile production).Consistent with the protective role of AP39, bile production, and quality were associated with both post-transplant liver health and functionality [47,48].Additionally, reduction in bile glucose and bile volume during normothermic machine perfusion was previously associated with liver viability [49].
While the direct ratio of ATP:ADP and ATP:AMP was similar in livers stored with AP39, tissue energy charge, previously shown by our team to be associated with transplant outcome [29], was improved by AP39.Consistently, H 2 S is shown to improve post-reperfusion energy stores in rat hearts following warm ischemia [28,50].Despite improvements in energy charge, a reduction in NADH:NAD + ratio was observed in livers stored with AP39.NADH has been shown to accumulate during the storage of rat livers as the electron transport chain is compromised, with reduced complex I function [51].Decreased NAD + levels, when combined with improved energy charge, indicate that following reperfusion, mitochondria with AP39 show improved ATP production [52].
Limitations are acknowledged.First the use of an acellular media strictly limits the IR injury caused by immune response such as neutrophil activation and subsequent parenchymal migration, inducing degranulation and protease activity [53].Second, despite similarities in microarchitecture between rodent and human livers, several key differences are present such as a reduction in connective tissue in rodent livers, increased perfusate delivery through the portal vein as compared to human livers, and a different surface area to volume ratio [54].Additionally, the lack of transplantation, and relatively short perfusion period limits our understanding regarding the long-term impact of AP39 supplementation.
Future studies should perform further work into elucidating the mechanism by which AP39 ameliorates ischemia-reperfusion-injury in larger human organs and in the context of transplantation.
Altogether this study demonstrates that AP39 supplementation in UW during SCS reduced hepatocellular injury during preservation, improving graft function during simulated transplantation.These results provide the basis for next-generation organ preservation solutions and should be translatable to human livers.

Figure 1 See
Figure 1

Figure 2 See
Figure 2

Figure 3 See
Figure 3

Figure 4 See
Figure 4