High-Mobility Group box-1 Release Induced by Renal Warm Ischemia Mediates Progression of Ischemia-Reperfusion Injury in Miniature Swine

Abstract

Objective: The prevention of ischemia-reperfusion injury (IRI) is important for both early and long-term graft function, particularly when using extended-criteria or donation after cardiac death (DCD). High-mobility group box 1 (HMGB1) has recently been reported to mediate inflammation and cell injury after IRI; however, there is limited data to characterize the role of HMGB1 on IRI in large animals. The aim of this study was 1) to develop a large animal model of severe renal IRI that would closely mimic the warm ischemia and cold ischemia commonly experienced in human cases of DCD and 2) to determine the role of HMGB1 in IRI, and to evaluate the cytoprotective effects of anti-HMGB1 antibody (Ab) following IRI. Methods: Warm ischemia was induced for 120 minutes in 11 CLAWN miniature swine by clamping the left renal artery and vein. Then, cold ischemia was induced by flushing the left kidney with cold perfusate and topically cooling with ice-slush for 60 minutes. In Group 1 (n=8), the artery and vein were then unclamped and animals were followed without any additional treatment. In Group 2 (n=3), an injection of 1mg/kg of anti-HMGB1 Ab was administered just before unclamping and reperfusion. Renal function was monitored by serum creatinine (Cre) and histologic evaluation of sequential biopsies obtained on postoperative days (PODs) 2, 14, 28 and 56. Serum cytokine levels such as TNF-α, IL-1β, IL-6 and HMGB1 were measured in order to characterize the inflammatory response to IRI. Results: In Group 1, three of 8 animals died within 36 hours after reperfusion. In the 5 remaining animals, serum Cre levels markedly elevated after reperfusion, peaking at a median of 5 days (peak Cre levels: 12 ± 2 mg/dL) and recovered by day 14. Sequential measurements of serum cytokine levels showed that the concentrations of HMGB1 increased as early as 30 minutes after reperfusion, followed by the elevation of TNF-α and IL-6 peaking at 6 hours. In Group 2, all animals recovered and renal function returned to near-normal by day 9. Peak Cre levels (7 ± 0.2 mg/dL) were markedly lower than group 1. The Cre area-under-the-curve during days 0-14 (AUC0-14d) for animals in group 1 was 74 mg/dL·day, but significantly lower (43 mg/dL·day; reduction of 42% (p< 0.05)) in Group 2. Renal biopsies in Group 1 showed widespread severe necrosis of renal tubules at POD2. In contrast, treatment with anti-HMGB1 Ab dramatically decreased renal injury and was associated with prompt regeneration of tubular epithelium was seen in Group 2. Conclusion: In this study, we successfully developed a CLAWN miniature swine model of severe renal IRI using 120-min of warm and 60-min of cold ischemia. We observed that HMGB1 was released as an early mediator of IRI that in turn activated the later release of TNF-α or IL-6 and that anti-HMGB1 Ab reduced IRI in this model. To our knowledge, this is the first demonstration of the beneficial effects of perioperative administration of anti-HMGB1 Ab in order to reduce renal IRI in a clinically relevant, large animal model.