Reperfusion

General anesthesia was induced with 0.1 mg/kg BW diazepam (Ziapam 5 mg/kg, Ecuphar GmbH, Greifswald, Germany) and 2.2 mg/kg BW ketamine (Narketan, Vétoquinol GmbH, Ismaning, Germany) after premedication with 0.7 mg/kg BW xylazine (Xylavet 20 mg/ml, CP-Pharma GmbH, Burgdorf, Germany). Anesthesia was maintained with isoflurane (Isofluran CP, CP-Pharma GmbH) in 100% oxygen, and continuous rate infusions with lactated Ringer's solution (Ringer-Laktat EcobagClick, B. Braun Melsungen AG, Melsungen, Germany) and dobutamine (Dobutamin-ratiopharm 250 mg, Ratiopharm GmbH, Ulm, Germany) were given to effect, to maintain the mean arterial blood pressure between 60 and 80 mmHg. A routine pre-umbilical median laparotomy was performed in dorsal recumbency following aseptic preparation. Segmental small intestinal ischemia was induced in 1.5 m jejunum by occlusion of the mesenteric vessels with umbilical tape. The ligature was tightened under monitoring of intestinal microperfusion with microlightguide spectophotometry and laser Doppler flowmetry (O₂C, LEA Medizintechnik GmbH, Giessen, Germany), and the ligature was tied when the blood flow was reduced by 90% of the pre-ischemic measurement. The ischemia was maintained for 90 min. In group C, the ligature was released without manipulation of the vessels and reperfusion was initiated without delay. In group IPoC, postconditioning was implemented after release of ischemia by clamping the mesenteric vessels for three cycles of 30 s, alternated with 30 s of reperfusion. This was followed by 120 min of reperfusion in both groups. Subsequently, the horses were euthanized with 90 mg/kg BW pentobarbital intravenously (Release 50 mg/mL, WDT eG, Garbsen, Germany) without regaining consciousness.

Full thickness intestinal samples were taken at the end of the pre-ischemia period (pre-ischemia sample, P), at the end of ischemia (ischemia sample, I), and at the end of reperfusion (reperfusion sample, R). At reperfusion point, an additional sample was taken just proximal to the post-ischemic intestinal segment (proximal sample, PR). One segment of each sample was fixed in a 4% formaldehyde solution for 24–36 h and subsequently embedded in paraffin.
Immunohistochemical staining was performed for HIF1α and HIF2α. In short, the slides were deparaffinized and subsequently the antigen retrieval was done using citrate buffer with a pH of 6.0 at 95°C for 20 min, followed by blocking for unspecific binding with 20 % goat serum. The slides were incubated overnight with 1:500 polyclonal rabbit antibody against HIF1α (HIF-1 alpha Antibody NB100-134 1.0 mg/ml, Novus Biologicals LLC, Centennial, USA) or 1:100 monoclonal mouse antibody against HIF2α (Anti-Hypoxia Inducible Factor 2 α Antibody clone 190b, Sigma Aldrich, Darmstadt, Germany). Subsequently, the slides were incubated with secondary antibody (1:200 goat-anti-rabbit or 1:200 goat-anti-mouse, respectively), followed by incubation with the ABC reagent (Vectastain ABC, Biozol diagnostics Vertrieb GmbH, Eching, Germany). As negative isotype control, the control slides were incubated with rabbit IgG (IgG from rabbit serum I5006, Sigma Aldrich) for HIF1α and with mouse IgG1 (Clone MOPC-21, BioLegend, San Diego, USA) for HIF2α instead of the primary antibody. Equine kidney tissue and equine squamous cell carcinoma tissue was used as a positive control. The slides were incubated with 3,3′-diaminobenzidine and counterstained with modified hematoxylin (Delafield Hemalaun).
All slides were scanned to a digital format using a microscopic scanner at 20× magnification (Axio Scan.Z1, Carl Zeiss GmbH, Oberkochen, Germany), and subsequently evaluated using the accompanying software (Zen Blue 3.0, Carl Zeiss GmbH). In addition to the descriptive evaluation, a semi-quantitative score was developed for comparison between the groups and time-points. The enterocytes in the crypts and the villi were separately graded for staining intensity of both the cytoplasm and the nucleus with the following score for immunoreactivity: grade 0–no staining; 1–weak staining (staining hardly visible); 2–mild staining (light brown); 3–moderate staining (medium brown); 4–intense staining (dark brown; Supplementary file 1). To quantify the difference between the cytoplasmic and nuclear staining within one slide, the nucleus/cytoplasm ratio was calculated. The same score was used for the myocytes of the tunica muscularis. Microscopic photographs were used as color reference, and the evaluation was performed at fixed color settings by one observer, who was blinded for the identity of the slides. Because many sections showed a varying amount of focal cytoplasmic staining close to the nucleus, a separate score was added to quantify the proportion of cells with this perinuclear staining: grade 0–<1%; 1–1 to 25%; 2–26 to 50%; 3–51 to 75%; 4–76 to 100%.