The ethics committee of the Norwegian Food Safety Authority approved this study in pigs (approval number: 68/11-3811) and all experiments were performed in concordance with the guidelines from Directive 2010/63/EU of the European Parliament on the protection of animals used for scientific purposes. Housekeeping, anesthesia, euthanasia, and recording of hemodynamic and respiratory parameters were performed in accordance to ARRIVE guidelines as shown in table (Online Resource 1) and as reported previously . Briefly, anesthesia was induced in twenty-one 20 kg pigs by intramuscular ketamine (800 mg), azaperone (80 mg), atropine (1 mg) followed by intravenous (iv) pentobarbital 1–3 mg kg−1 and maintained using iv morphine 1–2 mg kg−1 h−1 and isoflurane 1.0–1.5% in oxygen/air mixture. After sternotomy, a silastic occluding tape was placed around the left anterior descending (LAD) coronary artery distal to the second diagonal branch allowing reversible complete occlusion. Microdialysis catheters (CMA 71, 100 kDa cut-off, 2 cm membrane, 1 µl min−1 flow, M Dialysis, Solna, Sweden) were placed in the LAD dependent area and in a control region supported by the left circumflex artery (Cx).
Ischemia was induced for a total of 40 min by LAD occlusion, except for sham animals. Twenty minutes prior to reperfusion, sixteen animals were randomized to treatment with coversin or saline (NaCl 0.9%, placebo group), n = 8 in each group. Coversin (Akari Therapeutics Plc, London, UK) has a plasma half-life of about 30 h due to stable binding to C5  and was diluted in saline. It was given as a 1 mg kg−1 bolus, and followed by a continuous infusion of 0.036 mg kg−1 h−1 . The control group and the three sham animals received the same amount of saline without coversin. Fifteen minutes before euthanasia, iv magnetic resonance imaging (MRI) contrast agent gadoteric acid (0.4 mM kg−1, Dotarem, Guerbet, Paris, France) was given . Just before euthanasia, LAD was re-occluded and iv Evans Blue (2% in 40 ml phosphate buffered saline, Sigma Aldrich, St. Louis, MO, USA) was given to delineate the area at risk (AAR). Euthanasia was carried out by iv injection of pentobarbital (500 mg), morphine (30 mg), and potassium chloride (50 mmol). After euthanasia, the heart was excised and rinsed in ice-cold saline.
Arterial blood samples were obtained prior to surgery, after stabilization prior to induction of ischemia, at the end of 40 min of ischemia, and every hour throughout the reperfusion period. Samples were taken for blood gas analysis, serum, and EDTA-plasma preparation and were immediately cooled and centrifuged prior to storage at −80 °C. Microdialysis samples and thermal dilution cardiac output were obtained at the same time points. After euthanization, tissue samples were taken from the center of the Evans blue free area (AAR), at the border of the Evans blue free area (border zone) and in the Evans blue stained Cx region (control area) and snap-frozen in approximately 1 ml OCT™ (Sakura Finetek Europe, Zoeterwoude, the Netherlands) prior storage at −80 °C.
Infarct size assessed by magnetic resonance imaging
After tissue sampling, air-filled balloons were placed in the left and right ventricle. MRI analysis was performed using a 3 Tesla scanner (Philips, the Netherlands). T1-weighted images (3D FFE, TR/TE = 5.4/2.3 ms, flip angle 35°, BW = 434 Hz, 125 slices and scan duration = 02:15) with a measured isotropic resolution of 0.8 mm covering the entire heart were acquired using a quadrature head coil. Additionally, T1 measurement sequence was performed (Look Locker sequence: T1w TFE with “shared” inversion pulse, TR/TE = 2.3/4.3 ms, flip angle = 3°, inversion delay = 38.4 ms, phase interval = 65.5 ms, BW = 853 Hz, SENSE factor 2, isotropic resolution of 1 mm), and T1 maps were reconstructed using NordicIce (NordicNeuroLab, Bergen, Norway). The segmentation of the infarcted volumes was done in OsiriX . T1map was used to discriminate infarcted areas with the 3D region-growing tool (threshold of 400). The used threshold lead to inclusion of pericardium and endocardium as well but as the amount is comparable and small in all groups and subjective manual processing would have been necessary, we did not subtract it from the total infarcted volume. Infarction size (ml) was determined in T1 weighted images and compared to the total left ventricular volume.
Infarct size assessed by histological staining
After MRI, the left ventricle was cut in 5 mm thick slices. The non-stained AAR was dissected and immersed in tetrazolium chloride (TTC, 1% in phosphate buffered saline, Sigma Aldrich, St. Louis, MO, USA) at 38°C for 20 min. Slices were placed in 4% formaldehyde solution (Histolab Products AB, Gothenburg, Sweden) on ice for 30 min prior to digital scanning. Infarct size was determined as percentage of AAR as described previously  using Photoshop CS5 (Adobe Systems Software Ltd., Ireland).
Systolic left ventricular function was assessed by echocardiography from a four-chamber view prior to ischemia and at the end of the reperfusion period (GE Vivid 7, Horton, Norway). Peak systolic velocity and systolic displacement of the mitral plane were obtained from pulse Doppler echocardiography and averaged from the septum and the lateral wall (Echopac PC Version 112, GE Vingmed Ultrasound, Horten, Norway).
The snap-frozen tissues were cut into 5 μm thick sections, air-dried for 60 min and fixed with cold acetone for 10 min. They were either processed immediately or stored at −80 °C until further analysis. Then, after hydration, the sections were stained using a two-step indirect immunofluorescence technique. For E-selectin, the following primary and secondary antibodies were used: mouse anti-human E-selectin (Sigma, St. Louis, MO, USA) and goat anti-mouse IgG-Alexa546 (Molecular probes, Carlsbad, CA, USA). The antibodies used for Fibrinogen-like protein 2 (FGL-2) were rabbit anti-FGL2 (Aviva Systems Biology Corp, San Diego, CA, USA) and sheep anti-rabbit IgG-Cy3 (Sigma, St. Louis, MO, USA). A nuclear staining was performed using 4′,6-diamidino-2-phenylindole (DAPI; Sigma, St. Louis, MO, USA). A fluorescence microscope (DMI4000B; Leica, Wetzlar, Germany) was used to analyze the slides and the quantification of fluorescence intensity was performed using Image J software, version 1.50 (https://rsb.info.nih.gov/ij/) on TIFF images. All pictures were taken under the same conditions to allow for correct quantifications and comparison of fluorescence intensities.
In vitro assessment of complement inhibitory effects of coversin and eculizumab
Human and porcine whole blood samples anticoagulated with lepirudin (Celgene, Marburg, Germany) were pre-incubated with coversin or eculizumab (Alexion Pharmaceuticals, CT, USA) in a twofold serial dilution (final concentrations of 1.6, 0.8, 0.4, 0.2 and 0.1 µM) or PBS for the uninhibited control in sterile polypropylene tubes for 5 min at 37 °C. Subsequently, blood specimens were stimulated with zymosan at a final concentration of 50 µg/ml, or PBS for the negative control. After 30 min, the reaction was stopped by adding EDTA (final concentration 10 mM), samples centrifuged (3000g, 15 min, 4 °C). The resulting plasma was stored at −80 °C before analysis of C5b-9. Human and porcine serum samples were pre-incubated with coversin or eculizumab in a twofold serial dilution (final concentrations of 3.2, 1.6, 0.8, 0.4, 0.2 and 0.1 µM) or PBS for the uninhibited control in sterile polypropylene tubes for 5 min (room temperature) before analysis for functional complement activity.
Functional complement activity and C5b-9 (TCC)
Commercially available enzyme immune assay (Complement System Screen Wieslab; Euro Diagnostica, Malmö, Sweden) and murine anti-human C5b-9 antibody (clone aE11, Dako, Glostrup, Denmark) were used according to manufacturer’s instructions to detect functional complement activity and sC5b-9 production in plasma, respectively. Both methods detect the respective human and pig epitopes . In tissue, the membrane form of C5b-9 was visualized in frozen sections from the AAR, border zone and control area. Tissue samples were incubated for 30 min at room temperature using the murine anti-human C5b-9 antibody (clone aE11, Dako, Glostrup, Denmark) diluted 1/25 in Dako antibody diluent (Dako K8006, Glostrup, Denmark), washed in phosphate buffered saline and stained by Ventana ultra View Universal DAB Detection Kit (Ventana Medical Systems, Inc., Tucson, AZ) according to the manufacturer’s instructions. A Nikon Eclipse E1000M microscope was used and photos were obtained with original 40× magnification.
Myocardial metabolism and inflammation
Microdialysis fluid from the AAR and control Cx region and EDTA-plasma was assessed for inflammatory mediators interleukin (IL)-1β, IL-6, IL-8, IL-10, and TNF using a porcine multiplex cytokine assay on a Bio-Plex 100 system (Bio-Rad, Hercules, CA, USA) as previously described . LTB4 from plasma and myocardial tissue was measured using a competitive enzyme immunoassay according to the manufacturer’s instructions (R&D systems, Minnesota, MN, USA).
Markers of cardiac injury
Serum troponin-T levels were determined at the institutional clinical laboratory (Modular E170, Roche Diagnostics, Switzerland). Plasma heart fatty acid binding protein H-FABP levels were measured by ELISA in accordance to manufacture’s instruction (Hycult Biotech, Uden, The Netherlands).
Investigators were blinded to the treatment during the experiments and all analyses.
Two animals died immediately after reperfusion due to ventricular fibrillation (one coversin and one placebo treated animal) and were excluded. Thus, functional complement activity was analyzed in 16 animals subjected to LAD occlusion and three sham-operated animals. Complete inhibition of all three complement pathways by coversin treatment was confirmed in all animals, except for one, which was excluded after statistical confirmation of outlier behavior (Grubbs’ test, p < 0.05). Thus, 15 animals (seven coversin and eight control animals) were used in all further analyses if not stated otherwise.
Two animals (one coversin and one control animal) had significantly smaller AAR determined by Evans Blue staining due to anatomical variations of the LAD and were therefore excluded from MRI analysis. Microdialysis catheters ceased function before 120 min of reperfusion in two coversin and one control animal and statistical comparison was therefore done with five and seven animals, respectively.
If not stated otherwise, values are presented as mean ± standard deviation (SD). Values obtained for coversin treated and control animals were compared at defined time points using Mann–Whitney U test. Two-way ANOVA was used if more than two groups had to be compared. Linear mixed effect model (intervention as fixed effect and subject number as random effect) was used to compare groups throughout the whole study period. Multiple comparisons were post hoc Bonferroni corrected. The Pearson correlation coefficient was calculated to compare infarct sizes determined by TTC and MRI. Statistical analyses were performed using SPSS 22 (IBM, Armonk, NY, USA) and GraphPad Prism 6 (GraphPad Software, La Jolla, CA, USA).