Short- and Long-Time Improvement of Graft Function Using Iron-Chelator Supplemented Bretschneider Solution in a Canine Model of Orthotopic Heart Transplantation

An increasing organ demand is facing a constant number of donors. Nevertheless, not all organs are utilized due to a limited time window for heart transplantation (HTX). Therefore, we aimed to evaluate, if iron chelator supplemented Bretschneider solution can protect the graft in a clinically relevant canine model of HTX with prolonged ischemic storage. HTX was performed in foxhounds. The ischemic time was standardized to 4h, 8h, 12h or 16h depending on the experimental group. Left ventricular (LV) and vascular function were measured. Additionally, myocardial high energy phosphate and iron content and in-vitro myocyte force were evaluated. Iron chelator supplementation proved superior at a routine preservation time of 4h as well as for prolonged times of 8h and longer. The supplementation groups recovered quickly compared to their controls. The LV function was preserved and coronary blood ow increased. This was also conrmed by in vitro myocyte force and vasorelaxation experiments. Additionally, the biochemical results showed signicantly higher adenosine-triphosphate content in supplementation groups. The iron chelator LK614 played an important role in this mechanism by reducing the chelatable iron content. This study shows that iron chelator supplemented Bretschneider solution effectively prevents myocardial/endothelial damage during short as well as long term conservation.


Introduction
Heart transplantation (HTX) is the gold standard therapy for end-stage heart failure. Presently the oneyear survival rate of recipients is 90%. 1,2 However, we are confronted with growing numbers of potential recipients and constant donor numbers, leading to an increasing organ de cit. 1 In current literature various options have been suggested to solve this con ict. 3 One approach is the reduction of unused potential donors by extending the maximum conservation time and thereby increasing the pool of suitable recipients.
A predominantly used cardioplegic, conservation solution in HTX is Custodiol (Bretschneider solution, HTK-solution), 4,5 We believe that further development of this well-established solution may improve donor organ conservation. We have already previously described areas for improvement in the Custodiol solution. 6-8 It is common knowledge that ischemia/reperfusion during cardiac surgery and transplantation causes myocardial and endothelial injury. [9][10][11] It has also been shown that, paradoxically, the protective hypothermia used during routine cardiac surgery itself causes potentiated myocardial damage mostly via an increase in chelatable iron concentrations and the resulting formation of reactive oxygen species. 12 Additionally, histidine -one main component used in Custodiol -has demonstrated cytotoxic effects. 13 Taking this into account, a novel amino acid-forti ed and iron chelator-supplemented Custodiol-based solution, Custodiol-N, was developed. (Table 1) In our previous studies, we rst evaluated the effect of iron chelators and N-alpha-acetyl-L-histidine on endothelial function after long-term, cold storage in vitro. We found that endothelial function was preserved by our improvements. 8 Then, we studied Custodiol-N solution's capability in a rodent model of ischemia/reperfusion injury in the heart.
Here we con rmed the superiority of Custodiol-N compared to Custodiol. 7 Finally, Custodiol-N was tested in an animal model of cardiopulmonary bypass. Again, we showed that myocardial, as well as endothelial function, could be effectively preserved by Custodiol-N. 6

Experimental Groups & Study Design
The study was subdivided into three series ( Fig. 1): In the rst series standard cardiac preservation times (4h cold ischemic conservation) were utilized for an initial comparison of Custodiol and Custodiol-N. Baseline hemodynamic measurements, in vivo endothelial function and high energy phosphate content were compared.
In the second series longer conservation periods were evaluated. Hemodynamic measurements and coronary blood ow (CBF) were measured after 8, 12, and 16h of ischemic conservation time. In the 8h and 12h groups we additionally examined histology and cardiomyocyte sensitivity to calcium. Based on the possibility to wean some animals from cardiopulmonary bypass in the 16h group, we examined the in vitro endothelial function of both Custodiol, and Custodiol-N compared to native control hearts after 24h conservation.
In the third series, the effectivity of the iron chelators LK-614 and deferoxamine, key components of Custodiol-N, were tested against Custodiol and native control hearts after 4h and 12h ischemic conservation.

Orthotopic Heart Transplantation
Cardioplegic arrest was induced using either Custodiol or Custodiol-N solution. Non-ischemic hearts served as controls. Donor hearts were preserved with the corresponding organ preservation solution for 4h, 8h, 12h or 16h. Then, orthotopic HTX was performed followed by a 2h reperfusion periode. For measurement of iron content in series 3 ( Fig. 1) heart samples were taken after 10min of reperfusion.

Hemodynamic measurements
A Millar pressure-conductance catheter was used to measure pressure-volume relationships and calculate load independent indices of myocardial contractility. CBF was evaluated by an ultrasonic ow meter placed on the left anterior descending coronary artery of the donor heart.
In series 1 endothelial function was tested in vivo: Coronary endothelium-dependent and endotheliumindependent vasodilatation were assessed after administration of an intracoronary bolus of acetylcholine (10 -7 M) or nitroglycerin (10 -4 M), respectively.

In vitro organ baths functional experiments
Organ bath experiments were performed as previously described. 14 Brie y, after removal of the recipient animals' hearts, the coronary arteries were isolated carefully, transversely cut into 4 mm wide rings, immersed in cold (+ 4 ºC) cardioplegic solution (Custodiol or Custodiol-N) and stored for 24h before evaluation. Coronary rings of the control group were immediately mounted in the organ bath without ischemic storage. Hypochlorite (200 µM, 30 minutes), a highly reactive oxygen species was added to mimic the free radical burst, which usually occurs in vivo during reperfusion. In each vessel ring, endothelium-dependent (acetylcholine) and -independent (sodium nitroprusside) vasorelaxation were investigated.

In vitro force measurements
After mechanical isolation of myocardial tissue samples (original heart of recipient animals served as controls), permeabilization was performed with 0.5 % Triton-X 100 detergent and a single cardiomyocyte was attached to two thin stainless-steel needles, which were connected to a force transducer (SensoNor, Horten, Norway) and to an electromagnetic motor (Aurora Scienti c Inc., Aurora, Canada). Isometric force measurements were performed during repetitive activation-relaxation cycles at 15°C, rst at a sarcomere length (SL) of 1.9µm and then at a SL of 2.3µm. Ca 2+ -contractures were evoked by transferring the myocyte from a Ca 2+ -free relaxing solution to activating solutions of gradually increasing Ca 2+ concentrations. Active isometric and passive forces were normalized for cardiomyocyte cross-sectional area. Isometric force values were normalized for maximal Ca 2+ -activated active force. Then, Ca 2+ -force relations were plotted to determine the Ca 2+ -sensitivity of isometric force production, i.e. pCa 50 .

Histology
Myocardial tissue samples were taken from hearts conserved in Custodiol, Custodiol-N (8h and 12h) as well as from native control hearts, xed in buffered paraformaldehyde solution (4%) and embedded in para n. Then, 5µm-thick sections were stained with hematoxylin/eosin for histopathological evaluation.

Determination of total and chelatable iron concentrations
To avoid a physiologically occurring quick normalization of iron concentations, samples were taken after 10min reperfusion time. A new assay has been established for the determination of chelatable iron, detailed in the following: PhenGreen calibration curve: physiological saline was incubated with Chelex 100 (Bio-Rad) (5g/100ml) overnight to chelate all free iron in the solution. Then PGSK (Life technologies, Darmstadt, Germany) was diluted to 15 different concentrations between 0 and 50µM. Dilutions were spectrophoto uorometrically analyzed at 488nm excitation and 505-530nm.
Sample measurements: 50mg heart tissue was suspended in the pretreated saline at 1mg/10µl weight/volume ratio before homogenization and centrifugation. Supernant was taken and incubated with ascorbic acid and then with 50µM PhenGreen. Afterwards samples were analyzed at 488nm excitation and 530nm emission.

Statistics
All data is expressed as the mean ± standard error of the mean (SEM). Statistical analyses of data were performed using GraphPad Prism 7.02 software (GraphPad Sofware, Inc., CA, USA).

Series 1 & 2:
Individual means between the groups were compared by one-way analysis of variance followed by an unpaired t-test with Bonferroni correction for multiple comparisons. Regarding hemodynamic and CBF measurements the 8, 12 and 16h Custodiol-N groups were compared to baseline results using a paired t-test, as the animals in the Custodiol groups could not be weaned from cardiopulmonary bypass. A value of p < 0.05 was considered statistically signi cant.
Series 3: ANOVA was used to compare the differences between the groups followed by the post-hoc Scheffe's test. A value of p < 0.05 was considered statistically signi cant.

Hemodynamic measurements
Hemodynamic variables did not signi cantly differ among the groups at baseline. After reperfusion, mean arterial pressure (MAP) was signi cantly decreased in Custodiol and Custodiol-N groups compared to baseline without any signi cant difference between the groups. But cardiac output (CO) was signi cantly lower in the Custodiol than in the Custodiol-N group (Table 2). Additionally, systolic cardiac function, in terms of ESPVR and PRSW, was signi cantly better preserved in the Custodiol-N group (Fig. 2). Coronary blood ow While CBF decreased signi cantly after transplantation compared to baseline in the Custodiol group, the Custodiol-N group did not show this difference. Noticeably also in direct comparison of both experimental groups CBF was signi cantly lowered in the Custodiol group (Table 2). Similar results were observed for endothelium dependent vasorelaxarion in response to acetylcholine: Endothelial function was signi cantly reduced compared to baseline in the Custodiol group while no adverse effects were seen in the Custodiol-N group (Fig. 3) Meanwhile, endothelium independent vasorelaxation in response to nitroglycerin showed no signi cant differences (Fig. 3).
High energy phosphates ATP-levels as well as the ECP were signi cantly improved using Custodiol-N compared to the Custodiol group (Table 3).

Hemodynamic measurements
Hemodynamic variables did not signi cantly differ among the groups at baseline.
After HTX, Custodiol groups showed macroscopic signs of severe ischemic injury and no satisfactory contractile function could be detected. Not a single could be weaned from heart-lung machine. Therefore, no hemodynamic measurement is available. However, in the Custodiol-N groups contractility was still preserved after 8h and 12h (Table 4 and Fig. 4). The animals were hemodynamically stable under inotropic support. Even after 16h cold ischemic conservation time 5 out of 9 animals could be weaned from the cardiopulmonary bypass. Two of the remaining four animals underwent massive coronary air embolism and were considered as technical failure. In two cases, the animals could not be weaned from cardiopulmonary bypass, without any evidence of technical failure. Heart rate, LV systolic pressure and CO did not signi cantly differ over the time in any of the Custodiol-N groups. We detected a slight decrease in MAP values in the 8h group, which reached the level of statistical signi cance in the case of 12h and 16h conservation (Table 4). LV contractility in terms of endsystolic-pressure-volume-relationship (ESPVR) and preload-recruitable-stroke-work (PRSW) was preserved (Fig. 4).

Coronary blood ow
After 8h and 12h ischemic conservation with Custodiol-N, CBF was signi cantly increased during reperfusion compared to baseline. However, our results failed to show a signi cance in the 16h group (Table 4).

In vitro vascular function
As a supplementary part of series 2, with 24h cold ischemic preservation time, in vitro endothelial function showed the following results: In the Custodiol group a marked impairment of coronary endothelial function was demonstrated by a reduced maximal relaxation of isolated coronary artery segments to acetylcholine and by a rightward shift of the concentration-response curve when compared with the native control group. Custodiol-N successfully prevented this ischemia/reperfusion induced endothelial impairment. Endothelium independent vasorelaxation was not altered (Fig. 5).

In vitro force measurement
In the Custodiol group, it was di cult to prepare isolated cardiomyocytes for force measurements after 12h ischemic time, as most cells were in a state of ischemic contracture. At 1.9 µm sarcomere length, the calcium-sensitivity of contractile laments of cardiomyocytes was signi cantly lower in the Custodiol group after 12h when compared to native controls. This reduction was reversed in the Custodiol-N group (Fig. 6).

Histology
Histopathological signs of severe ischemic damage (necrosis, myo brillar fragmentation, cellular swelling and tissue edema) were observed in the LV myocardium of transplanted hearts stored for 8h and 12h in Custodiol, which were less pronounced in the Custodiol-N groups (Fig. 7).

Series 3 Determination of chelatable iron concentrations
Chelatable iron concentrations were signi cantly higher in the Custodiol group compared to the Custodiol-N group after 4h as well as 12h of ischemia. Even at 12h Custodiol-N groups did not signi cantly differ from the native control hearts. However, hearts conserved in Custodiol for 12h showed signi cantly higher levels than native control hearts. There was a tendency towards higher iron concentrations after 12h of preservation in comparison to the values after 4h of preservation for both groups, Custodiol as well as Custodiol-N (Fig. 8).

Discussion
Even though, the number of potential HTX recipients has increased by 50% since 2004 15 , the number of donors has remained almost unchanged. Nevertheless, donors are regularly excluded from heart donation because of the limited preservation time to reach a suitable recipient. 1 Improving the utilization rate of grafts is therefore an important goal in transplant surgery.
Ex vivo perfusion systems are one of the strategies which have been developed as solution. Oxygenated blood is used to perfuse the graft and hereby to reduce graft damage. Besides isolated case reports describing a conservation time of 8h [Heartbeat device 16 ] to 10h [OCS by TransMedics 17 ], reliable clinically relevant studies are still missing. As limiting factor activated blood and the further release of metabolic agents from the graft lead to cardiac edema and in ammation as a sign of myocardial impairment. 18 Additionally, in terms of practicability several issues must be considered: Present ex vivo perfusion systems require special equipment and training of surgeons and technicians, which results in considerably higher costs. Therefore, this organ procurement procedure may not be feasible for all hospitals.
Based on these limitations we aimed on an improvement strategy of the well-established Custodiol solution. [4][5][6][19][20][21][22] The primary target of our work was to further reduce myocardial damage during standard conservation times of 4h and secondary to point out an opportunity to safely reach prolonged organ preservation without the need of specialized staff or expensive equipment.
Indeed, the comparison of Custodiol to Custodiol-N in a 4h ischemia/reperfusion setup (series 1, gure 2 and 3) showed that also in routine cardiac surgery with ischemic times of 4h and less Custodiol-N could have an implication, which is indicated by a signi cant improvement of load independent contractility parameters -such as ESPVR and PRSW ( Figure 2). Remarkably, the contractility was recovered almost to baseline levels in this setup.
Also the secondary goal, which was to titrate the absolute borders of Custodiol-N, showed remarkable results: In series 2, long ischemic times of 8h, 12h and 16h were evaluated. All Custodiol groups had macroscopic signs of severe ischemic injury in series 2. None of the animals could be weaned from cardiopulmonary bypass, not even under inotropic support. In contrast, almost all animals from the Custodiol-N group could be weaned from the heart lung machine and were hemodynamically stable under inotropic support.
Improved vascular function, and thereby improved CBF, can contribute to better cardiac performance and better graft outcome.(35) This improvement was already seen as secondary ndings in series 1, and it was con rmed during extended ischemic conservation periods in series 2. Again, to nd the absolute limitations of Custodiol-N we investigated 24h ischemic conservation time and found only a signi cant impairment in the Custodiol group but not in the Custodiol-N group.
Several mechanisms behind these protective effects of Custodiol-N have already been described by our previous studies: Some of the agents used in Custodiol have been replaced by further developed substances. One of these is histidine, which on one hand is an excellent buffer but on the other also contributes to cell toxicity. 7 It was partly replaced by N-α-acetyl-L-histidine, which shares similar capabilities but has a reduced cellular uptake. 6,7,13,20 Furthermore the membrane stabilizing amino acids glycine and alanine have been added to Custodiol-N. 6,7,20,23 Beside these well described mechanisms we were able to explore and prove the key mechanisms of Custodiol-N: It is known that myocardial dysfunction and impairment of force are directly linked to sarcomere function. 24,25 Sarcomere function itself is dependent to its sensitivity to Ca2+, which again can be altered by hyper-or hypo-phosphorilation. This sensitive balance between proteinkinase A, which indirectly increases the phosphorylation of sacromeric proteins, and proteinkinase C, which has contrary effects, is in uenced by ischemia/reperfusion injury. 26 In consequence an in-as well as a de-creased Ca2+ sensitivity indicates a dysregulated phosphorylation of sacromeric proteins and hereby a severe damage to the myocardial tissue. Our in vitro measurements of myocardial sacromeres Ca2+ sensitivity, showed a signi cantly decrease compared to native controls. (Figure 6) Custodiol-N preserved this balanced complex system and thereby myocardial function.
Beside these known mechanistic insides, we now focused on the generation of reactive oxygen species (ROS) by the Fenton reaction as key mechanism. 12,20 The Fenton reaction is boosted by chelatable iron. 7,12 Two different types of iron chelators were introduced, deferoxamine and LK614 [1-(N-hydroxy-Nmethylcarbamide)-3,4-dimethoxybenzol], to address this topic. LK614 is smaller and more lipophilic than deferoxamine, which allows membrane permeability. 7,20,27 Indeed, for the rst time we proved a reduction of chelateable iron content in vivo using a exclusively developed assay. (Figure 8) Together with our previous work, in which we showed a signi cant reduction of myocardial apoptosis, 7 the present results gives us an impression of the link between iron chelators, reduction of iron content and hereby also of apoptosis and nally preservation of myocardial function.
In the past, also other pharmacologic approaches have been made to improve pre-existing cardioplegic solutions for prolonged ischemic storages: Baxter and collegues demonstrated the conservation of rat hearts for 16h using Wisconsin solution supplemented by nitroglycerine. 28 Kevelaitis et al suggested the Na+/H+ exchange inhibitor cariporide as well as the mitochondrial K ATP channel (mitoK ATP ) agonist diazoxide in a rat model as additive to Celsior solution for prolonged storage. 29 Similar results were seen for BMS-180448 -another mitoK ATP agonist -in a rat working heart model. 30 However, none of these approaches could be advanced to a clinical trial. In contrast to that, our ndings were conducted in this clinically relevant canine model and suggested a safety margin between 12h and 16h. These results encouraged us to conduct a presently ongoing clinical trial using.
Summarizing, Custodiol-N has superior protection abilities compared to Custodiol at standard ischemic times. Unlike alternative approaches such as the ex vivo organ perfusion, Custodiol-N can be used very cost e ciently. Additionally, the results are promising for extended ischemic conservation periods. The mechanisms behind these effects are linked to a stabilization of Ca 2+ sensitivity on sarcomere level and on the reduction of the chelateable iron pool. Therefore, our present approach for organ conservation may be an improvement in cardiac surgery. Limitations: The following limitations of this study must be addressed.   In vitro force measurements The calcium-sensitivity index pCa50 of isolated cardiomyocytes at 1.