IgM binding precedes complement activation during skeletal muscle ischemia-reperfusion

282

ASSOCIATION FOR ACADEMIC SURGERY—ABSTRACTS

significant differences in hospital length-of-stay. Glucose levels and HgbA1c levels were not significantly different (Figure). Interestingly, C-peptide levels were similar throughout the first postoperative week. Patient and graft survival were similar. Each group had one patient with acute cellular rejection. Conclusions: Venous drainage via the portal system has similar outcomes to systemic drainage. Since we anticipated enhanced first-pass hepatic clearance in the portal drained group, it was surprising that C-peptide levels were similar between the two groups.

P28. IgM Binding Precedes Complement Activation During Skeletal Muscle Ischemia-Reperfusion. R. K. Chan, M.D., G. Ding, M.D., N. Verna, M.D., S. I. Ibrahim, M.D., S. Oakes, B.S., H. B. Hechtman, M.D., F. D. Moore, Jr., M.D. Brigham and Women’s Hospital. Introduction: Skeletal muscle reperfusion injury is mediated by IgM natural antibodies and by complement activation, as shown by the attenuation of reperfusion injury seen in mice with no natural IgM (J Immunology 2002; 168:3433) and in mice deficient in complement C3 and C4 (J Exp Med 1996; 183:2343). We postulate that tissue, when ischemic, expresses neo-antigens to which pre-formed natural IgM antibodies bind, in turn producing harmful complement activation and reperfusion injury. Methods: C57Bl/6 mice were subjected to 2 hours of tourniquet-induced hind limb ischemia followed by variable periods of reperfusion. Two hours of ischemia and 3 hours of reperfusion produced severe muscle necrosis and edema. Deposition of IgM and C3 in tissue was assessed using immunohistochemistry on both frozen and formalin-fixed tissue samples. Results: IgM binding to the endothelium and muscle bundles of the hind limb began during the ischemic period and continued throughout reperfusion up to 6 h. C3 deposition was not present during ischemia, and in contrast, began to appear at 1h of reperfusion and increased progressively thereafter. Conclusion: These data demonstrate that IgM binding to ischemic tissues precedes the damaging complement activation by a significant period of time. This has important therapeutic implications when considering anti-inflammatory therapy for reperfusion injury. P29. TNF-dependent Upregulation of HCMV Major Immediate Early Promoter (MIEP) During Warm IschemiaReperfusion. T. K. Varghese, M.D., M. Hummel, Ph.D., S. Kim, M.S., S. Yan, M.S., G. Thomas, B.S., Y. Kanwar, M.D., Ph.D., M. Abecassis, M.D. Division of Organ Transplantation, Northwestern University Medical School. Reactivation of latent Human Cytomegalovirus (HCMV) results in morbidity and mortality in immunosuppressed patients. We have previously demonstrated that TNF plays a major role in inducing Immediate Early (IE) gene expression both as a result of soluble TNF and allotransplantation. The major immediate-early promoter (MIEP) of HCMV controls the expression of IE genes. Transgenic mice containing the lacZ gene regulated by the HCMV MIEP are ideal for analysis of factors that regulate promoter activity, and hence the reactivation process. TNF is known to play a role in warm

ischemia-reperfusion injury. We thus sought to determine whether warm ischemia-reperfusion injury could induce upregulation of HCMV MIEP in vivo, and the role of TNF in this process. Methods: 45 minute renal warm ischemia-reperfusion was performed in the left kidney while the right served as control. MIEP-lacZ transgenic mice were divided into 2 groups. Group I (n ⫽ 6): Ischemia ⫹ 24-hr reperfusion; Group II (n ⫽ 3): Sham controls. Group III consisted of transgenic TNF type I & II receptor Knockout (KO) mice (n ⫽ 4). Kidneys were harvested at 24 hrs reperfusion. Quantitative ␤-Galactosidase activity per nanogram of protein in tissues was done in triplicate. Fold-induction was determined by ratio of ischemiareperfused kidney to its control. Histological analysis was performed by H&E, and staining for CD4, CD8 and MAC-1. RT-PCR was done for TNF, IL-1 and IFN-␥. Statistical analysis was done by means of student t-test for independent samples. Results: Warm ischemiareperfusion induces a significant induction of transgene expression in kidneys (14.96 ⫾ 6.51), as compared to sham surgical controls (1.48 ⫾ 0.44) [p ⬍ 0.05]. This was accompanied by an increase in TNF, decrease in IFN-␥, and no change in IL-1. There was a mild interstitial infiltrate of lymphocytes, which were negative for CD4, CD8 and MAC-1. Transgenic TNF KO mice had a significantly decreased level of induction (6.34 ⫾ 3.36) [p ⬍ 0.05]. Conclusions: TNF plays a critical role in this warm ischemia-reperfusion induced HCMV reactivation model. Failure to completely abrogate HCMV MIEP upregulation in KO mice suggests the role of other factors in addition to TNF in the reactivation process.

VASCULAR ORAL POSTER SESSION I P30. Shear Stress Induces Apoptosis of Endothelial Cells. D. Macario, B.A., I. Entersz, G. B. Nackman, M.D., Robert Wood Johnson Medical School. Introduction: Human endothelial cells fail to ultimately cover artificial vascular grafts. Previously published work by others has identified that shear stress can “rescue” endothelial cells from apoptosis, utilizing a model in which apoptosis is first induced by serum starvation. We question the relevance of this model and hypothesize that, in fact, shear stress can induce apoptosis of a stable monolayer of endothelial cells adherent to an artificial biomaterial. Methods: Human Aortic Endothelial Cells (ECs) cultured on Dacron membranes in a defined medium containing serum and growth factors were placed in a parallel plate flow apparatus. ECs were exposed to 0, 1 and 10 dynes/cm2 of shear stress for 6 or 24 hr. Following flow, cells were stained with ethidium bromide, and apoptosis was determined by established changes in nuclear morphology. In addition, apoptosis was also detected by the absence of characteristic mitochondrial aggregates when stained with JC-1 (Molecular Probes). Positive controls for apoptosis were ECs deprived of serum and growth factors. The percentage of apoptotic cells remaining on membranes was determined by image analysis (n ⫽ 5– 6 membranes per group). Results: Shear stress and time of exposure were associated with the induction of EC apoptosis after 6 and 24 hr of flow, by two-way ANOVA, P ⬍ .05. Post-hoc analysis identified that apoptosis increased significantly with increased shear-stress, p ⬍ .05 (0 vs. 10 dynes/cm 2 and 1 vs. 10 dynes/cm 2). Conclusion: Moderate amounts of shear stress induce apoptosis of ECs adherent to Dacron. This may be a critical event antecedent to flow induced detachment from a biomaterial. The prior in vitro models of growth factor starvation followed by shear-stress exposure, which demonstrated decreased apoptosis, may lack physiologic relevance to many in vivo states. The mechanism(s) through which shear stress induces apoptosis of ECs adherent to a biomaterial remains to be elucidated.