PJ34, a poly-ADP-ribose polymerase inhibitor, modulates renal injury after thoracic aortic ischemia/reperfusion

David H. Stone, MD, Hassan Al-Badawi, MD, Mark F. Conrad, MD, Michael C. Stoner, MD, Fateh Entabi, MD, Richard P. Cambria, MD, and Michael T. Watkins, MD, Boston, Mass

Background. These experiments sought to evaluate the effects of PJ34, a poly-ADP-ribose polymerase inhibitor, on molecular indices of renal injury, mitochondrial function, tissue thrombosis, and fibrinolysis after thoracic aortic ischemia/reperfusion (TAR).

Methods. Forty-three 129S1/SvImj mice were subjected to 11 minutes of TAR followed by 48 hours of reperfusion. Experimental groups included untreated normal saline (NS) controls (UC), (n = 15, 0.5 mL NS ip) or PJ34 (PJ) (n = 17, PJ34 10 mg/kg ip, 1 hour before and after TAR). Sham (SH) mice (n = 11) underwent median sternotomy (heparin, NS ip) without TAR. Forty-eight hours after TAR or sham operation, kidney mitochondrial activity (using 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium [MTT]), D-dimer, and thrombin-antithrombin III (TAT) complex levels were measured. Levels of messenger RNA for neutrophil gelatinase-associated lipocalin (NGAL), a marker for renal injury, were also measured by reverse transcriptase-polymerase chain reaction.

Results. PJ34 improves renal mitochondrial activity after 48 hours of TAR, compared with untreated control animals (UC, 87.6 ± 2.2%; PJ, 151.4 ± 9.5%; P < .001). PJ34 did not alter the increase in renal D-dimer levels by 48 hours reperfusion (UC, 1.37 ± 0.09 U; PJ, 1.1 ± 0.14 U; SH, 0.82 ± 0.06 U; P < .05). TAR did not alter renal levels of TAT expression among groups (UC, 0.103 ± 0.034; PJ, 0.067 ± 0.008; SH, 0.106 ± 0.027; P = .619). The incidence of significantly increased NGAL among UC mice was 1415 ± 823.6 (n = 12), compared with 29.6 ± 20.8 (n = 10) in the PJ34-treated group (P < .014). Conclusions. PJ34 preserves renal mitochondrial activity and decreases steady-state levels of NGAL after TAR. TAR did increase markers of fibrinolysis in renal tissue but their increase did not correlate with renal injury or PJ34 treatment. These studies indicate that PJ34 confers protection against TAR and suggest that PARP may represent a novel target for reducing perioperative renal injury. (Surgery 2005;138:368-74.) THORACOABDOMINAL ANEURYSMS continue to pose a therapeutic challenge to the vascular surgeon, anesthesiologist, and critical care intensivist. Despite advances in endovascular stent–grafting techniques for treating aneurysmal disease, con- ventional surgical repair remains the mainstay of therapy. As such, morbidity and mortality associ- ated with complex thoracoabdominal aortic sur- gery remains high. Operative mortality is generally quoted in the 10% range.1-3 Associated perioper- ative renal dysfunction has been determined to be an independent predictor of operative mortality increasing it 3-fold.4 Currently the incidence of acute renal failure (ARF) after thoracoabdominal aneurysms is reported to range from 4% to 40%, depending on the series.1,5-10 In addition, ARF remains a vexing clinical problem associated with a 20% to 70% mortality rate.11 Perioperative ARF after repair of thoracoabdominal aneurysms is believed to stem from atheroembolism and renal ischemia/reperfusion injury sustained during proximal aortic clamping.12,13 Aortic cross-clamp duration also is causally related to increased aci- dosis, reperfusion injury, and release of inflamma- tory cytokines.6,14 Poor outcomes associated with thoracoabdomi- nal aortic reconstruction have prompted surgeons to incorporate pharmacologic and intraoperative strategies into clinical treatment protocols to min- imize the incidence of postoperative paraplegia and renal dysfunction. Some intraoperative strate- gies include atrial-femoral bypass, cerebral spinal fluid drainage, localized or systemic hypothermia, and renal artery cold perfusion. Pharmacologic strategies include perioperative management with pressors, papaverine, and naloxone administra- tion.15 While such adjuncts offer potential, consid- erable debate remains regarding their efficacy because the aforementioned complications persist in patients undergoing complex aortic surgery. Recently, our laboratory has documented that inhibition of poly-ADP-ribose polymerase (PARP), using PJ34, can effectively modulate spinal cord ischemia incurred during thoracic aortic ische- mia/reperfusion (TAR). DNA strand breaks sus- tained during cellular stress stimulate PARP activity in cell nuclei and mitochondria.16 Upon activa- tion, PARP transfers ADP-ribose from nicotina- mide adenine dinucleotide (NAD) to nuclear proteins, leading to the formation of nicotina- mide. Subsequently, this moiety can be recycled to NAD in an energy-consuming process, thereby depleting the cell of its ATP.17 Devoid of its ATP stores, the cell is then susceptible to dysfunction and subsequent necrosis. PARP activation also has been implicated in inflammatory processes and their related transcription factors, most notably nuclear factor-jB (NF-jB).18 PARP activation ap- pears to play a role in numerous models of injury including myocardial infarction, cerebral ische- mia/reperfusion injury, and renal ischemia/reper- fusion injury.19-21 In addition, PARP activation may contribute to diffuse thrombotic dysfunction,22 leading to local platelet deposition and thrombus formation, which eventually lead to multiple end- organ dysfunction. These experiments were de- signed to determine whether PJ34, an ultra potent inhibitor of PARP, could modulate renal ischemia/ reperfusion injury, as measured by markers of mitochondrial activity, parenchymal renal injury, and tissue indices of thrombosis/fibrinolysis. METHODS Experimental procedure. Thoracic aortic ische- mia/reperfusion was created in male, 8-week-old 129S1/SvImJ mice (Jackson Laboratory, Bar Har- bor, Me) weighing 30 to 35 g. All experimental procedures complied with the guidelines of the Massachusetts General Hospital animal committee and the ‘‘Principles of Laboratory Animal Care’’ (Guide for the Care and Use of Laboratory Animals, National Institutes of Health publication 86-23, 1985) Anesthesia was induced by intraperitoneal injection of Nembutal (75 mg/kg). Animals were allowed to breath room air throughout the exper- iment. Five units of heparin (170 IU/Kg) was injected subcutaneously 5 minutes before the procedure. Body temperature was maintained at 37.5°C ± 0.5°C with the use of a heated platform. A cervical mediastinotomy was performed, as de- scribed by Lang-Lazdunski et al.23 Laser Doppler imaging was used intraoperatively to qualitatively confirm the induction of hindlimb hypoperfusion. Normothermic aortic ischemia was then sustained for 11 minutes. This ischemic time interval was chosen because it reproducibly induces IR-related spinal cord injury. Longer ischemic periods have been associated with increased mortality before removal of the cross clamp. Shorter cross-clamp intervals did not induce consistent neurologic injury. Microvascular clamps were routinely dis- carded after 10 surgeries to avoid the possibility of clamp fatigue and ongoing perfusion. Mice were divided into untreated control (UC, n = 15, 0.5 mL normal saline ip), PJ34 (PJ, n = 17, PJ34 10 mg/kg ip) and Sham (SH, n = 11) animals. Sham mice underwent median sternotomy but did not un- dergo TAR. All groups were administered identical total volumes of fluid in the perioperative period. Mice were allowed to recover from anesthesia after the 11 minutes of ischemia and were then re- turned to their cages. Animals were euthanized after 48 hours of reperfusion by an intraperitoneal injection of Nembutal (100 mg/kg). Laser Doppler imaging. Lower body hypoperfu- sion was monitored to confirm TAR with the use of laser Doppler imaging. The laser Doppler instru- ment was positioned on a mounted rack sus- pended 20 cm above the murine hindlimb and tail. The mice were positioned on a heated (37°C) operating platform. The laser beam (780 nm) generates a computerized, color-coded image by reflecting from moving red blood cells in nutritional capillaries, arterioles, and venules. An- imal hindlimbs and tails were scanned with the laser Doppler Imager (model LDI; Moor Instru- ments Inc, Wilmington, Del) before, during, and after aortic clamping to document baseline per- fusion, ischemia, and finally reperfusion. Any mouse found to have ongoing perfusion during aortic cross clamping was excluded from further analysis. Mitochondrial activity. Mitochondrial activity was assessed by the reduction of a tetrazolium salt to water-insoluble colored formazan crystals by electron carriers and oxidative enzymes in the mitochondria of viable tissue. Kidneys were har- vested after 48 hours of reperfusion and dissected into equal halves to increase exposed surface area and thereby enhance tissue uptake of the tetrazo- lium salt. Each piece was then placed in a tube containing 3 mL of phosphate-buffered saline (pH 7.4) mixed with 300 lL of 1 mg/mL of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium (MTT) (Sigma, St. Louis, Mo). Tissue samples were incubated for 3 hours at 37°C in the dark with the use of a rotating mixer. Samples were then removed, washed with distilled water, and blotted dry. The water-insoluble formazan salt was then extracted in 3 mL of 2-propanol for 6 hours at 37°C in the rotating mixer. After extraction, 200 lL of each sample was transferred onto a micro- plate, where the absorbance was measured at 570 nm. The tissue was then dried at 90°C for 24 hours. The viability index is expressed as the optical density (OD570) relative to dry tissue weight in the sham, untreated, and PJ34-treated tissues, respectively. The mitochondrial activity index for UC and PJ-treated mice were compared as percent sham. Renal protein analysis. Protein samples were extracted from the kidneys of sham, untreated control, and PJ34-treated mice, respectively, and prepared for subsequent D-dimer analysis. Excised kidneys were flash-frozen in liquid nitrogen and stored at —80°C until analysis. Kidney samples were then homogenized in RIPA buffer supplemented with a protease inhibitor cocktail (Sigma-Aldrich, St. Louis, Mo) and sonicated for 20 seconds. Samples were then centrifuged at 16,000g for 10 minutes. The supernatants were aliquotted into Eppendorf tubes and stored at –80°C until analysis. Supernatants were analyzed for D-dimer levels and thrombin-antithrombin III (TAT) complexes via enzyme-linked immunosorbent assays (ELISAs). ELISA kits (D-dimer; R&D Systems, Minneapolis, Minn; TAT complex; Dade Behring, Deerfield, Ill) using the sandwich enzyme immunoassay method were used. ELISA plates were read on the Spectromax plate reader (Molecular Devices, Sun- nyvale, Calif) Results were extrapolated off the standard curve and normalized to the total protein concentration, which was determined with a BCA Protein Assay Reagent Kit (Pierce Biotechnology, Rockford, Ill). RNA isolation. Total RNA was isolated from kidney tissue harvested from sham, untreated control, and PJ34-treated mice. Tissue was homog- enized in trizol reagent (Invitrogen, Carlsbad, Calif) followed by chloroform phase separation and 2-propanol precipitation. Total RNA samples were then purified with the use of Rneasy minikit (Qiagen, Valencia, Calif). Total RNA concentra- tion was then determined by measuring the ab- sorption at 260 nm. RNA purity and integrity were assessed by the ratio between absorbance values measured at 260 nm and 280 nm, respectively. Real-time polymerase chain reaction. Transcrip- tional differences of neutrophil gelatinase-associ- ated lipocalin (NGAL) were measured by real-time polymerase chain reaction (PCR). Equal amounts of total RNA were reverse transcribed with the use of the SuperScript First-Strand Synthesis System (Invitrogen). Quantitative reverse transcriptase (RT)-PCR was performed on 2.0 lL complemen- tary DNA aliquots with the use of a QuantiTect SYBR Green PCR buffer (Qiagen) and iCycler iQ system (Bio-Rad Laboratories, Hercules, Calif). Primer pairs designed for the targeted template were then added to the reaction: NGAL forward primer: 5#-CAC-CAC-GGA-CTA-CAA- CCA-GTT-CGC-3# NGAL reverse primer: 5#-CAA-CAC-TCA-CCA-CCC- ATT-CAG-3# b-actin forward primer: 5 b-actin#-CAG-GTC-ATC- ACT-ATT-GGC-AACG-3# b-actin reverse primer: 5#-CAC-AGA-GTA-CTT-GCG- CTC-AGGA-3# PCR fragments containing the target sequence were used as external standards. The RT-PCR reaction was done under the following conditions: 94°C for 30 seconds, 55°C for 30 seconds, and 72°C for 30 seconds for a total of 40 cycles. A melting curve was generated at the end of the reaction to confirm PCR product specificity. A standard curve was then generated on the basis of the threshold cycle (CT) against the log of the standard concen- tration by using a dilution series of the standards. The concentration of the PCR target was calcu- lated by comparing the CT of the unknowns to the standard curve. NGAL mRNA levels were normal- ized to b-actin mRNA from the same tissue ali- quots. Sham mRNA was extracted from kidney tissue of mice subjected to normothermic medias- tinotomy but not to TAR. RNA levels were normal- ized to b-Actin RNA measured from identical tissue aliquots. Statistical analysis. Statistical analysis was performed with Instat (GraphPad, San Diego, Calif). Renal tissue viability (MTT), renal D-dimer, and TAT III complex levels were compared among sham, untreated (C), and PJ34-treated animals by analysis of variance and unpaired t test. Renal NGAL mRNA expression in untreated mice was compared with expression levels in PJ34-treated mice by the Fisher exact test. Fig 1. Renal mitochondrial activity assay as measured by MTT. Values depict mitochondrial function 48 hours after TAR for untreated control mice and PJ34-treated mice. Values are means ± SD. MTT, 3-(4,5-Dimethylthia- zol-2-yl)-2,5-diphenyl tetrazolium. Fig 2. Renal D-dimer levels depicted 48 hours after TAR. All values are represented as nanogram per milligram. Values reflect means ± SD. Both untreated control and PJ34-treated mice reveal increased D-dimer levels, com- pared with sham animals. RESULTS Renal mitochondrial activity (MTT). Mitochon- drial activity was measured 48 hours after TAR (Fig 1) Renal mitochondrial activity as measured by MTT was diminished in all mice receiving normal saline rather than PJ34 before and after TAR, whereas PJ34-treated mice exhibited a sub- stantial increase in renal mitochondrial activity (UC, 87.6 ± 2.2%; PJ, 151.4 ± 9.5%; n = 15; P < .001). All values are depicted as per cent sham. This data suggests that PJ34 preserves mitochon- drial activity after 48 hours of reperfusion in response to ischemic stress. Renal D-dimer levels. Renal parenchymal D- dimer levels revealed a statistically significant in- crease in D-dimers after TAR in both untreated control mice and PJ34-treated animals, compared with sham (UC, 1.37 ± 0.09 ng/mg; PJ, 1.1 ± 0.14 ng/mg; SH, 0.82 ± 0.06 ng/mg; n = 19; P < .05 UC, PJ vs SH, P = 0.189 UC vs PJ; Fig 2). These findings suggest that fibrinolysis is stimulated in renal tissue after TAR; however, PJ34 did not alter the D-dimer levels. Renal TAT complex levels. Renal TAT complex levels were measured in murine kidneys 48 hours after TAR. Comparison of sham, untreated, and PJ34-treated mice revealed no statistically signifi- cant difference in TAT complex levels (UC, 0.103 ± 0.034 ng/mg; PJ, 0.067 ± 0.008 ng/mg; SH, 0.106 ± 0.027 ng/mg; n = 23; P = .619; Fig 3). These find- ings further support the concept that the observed PJ34-conferred improved renal viability after TAR is likely induced independently from a thrombotic or fibrinolytic process. The exceedingly low levels of measured TAT complex suggest that little to no thrombus formation exist in the kidney after TAR. Renal NGAL mRNA levels. NGAL was measured as a marker for renal injury sustained during TAR- associated renal ischemia and was normalized to tissue actin in UC, PJ and SH mice. There is a marked difference in mean NGAL RNA expression (UC, 1415 ± 823.69, (n = 12); PJ, 29.6 ± 20.8; n = 10; SH, 10.5 ± 2.37, n = 7). The observed variation in this data indicates that every animal that undergoes TAR does not sustain a measurable degree of parenchymal renal injury. To evaluate these findings from a clinical perspective, we set a normal range for NGAL expression within 5 times sham expression (52.5 ± 11.85). Subsequent anal- ysis of the data revealed increased NGAL transcrip- tion levels among 66% (P < .04, Fisher exact test) of untreated control mice. Conversely, among PJ34-treated animals, 80% of this cohort revealed relatively normal NGAL expression. Subsequent analysis of the absolute levels of NGAL expression between UC mice and PJ34-treated mice revealed a substantial increase in the untreated cohort (Fig 4). This comparison reached statistical sig- nificance (P < .014). Fig 3. Renal TAT complex levels depicted 48 hours after TAR. All values are represented as nanogram per milli- gram. Values reflect means ± SD. There is a minimal trend toward less TAT complex in PJ34-treated animals although this did not reach statistical significance. TAT III complex, Thrombin-antithrombin III complex. DISCUSSION This study is the first to provide evidence that the administration of PJ34, a potent specific inhib- itor of PARP, improves renal mitochondrial func- tion and prevents expression of specific markers of renal injury after TAR. Tissue ischemia/reperfu- sion injury is known to result in altered mitochon- drial function. Structural changes, including organelle swelling, space formation between cris- tae, and lipid and denatured protein deposition, have been associated with both ischemia and reperfusion. Alterations in mitochondrial function decrease the ability of reperfused tissue to produce ATP and, thus, compromise the ability of tissues to provide specific functions such as muscle contrac- tion (myocardium, skeletal muscle) and solute reabsorption/excretion (renal tubules). While we have not measured tissue levels of ATP in the kidney subjected to TAR, extensive analysis of myocardium subjected to ischemia/reperfusion confirms a protective effect related to PARP inhi- bition, which is associated with preserved cellular energy levels.24 The finding of nearly doubled mitochondrial activity in PJ34-treated mice (Fig 1) after TAR offers convincing evidence that PARP inhibition provides substantial preservation of an important renal function. To further analyze renal tissue subjected to TAR, we assayed the steady state levels of NGAL, a 25 kD protein member of the lipocalin family and a specific marker of tubular injury. NGAL upregula- tion has been described in the kidney in response to the local release of inflammatory cytokines from neutrophils trapped in the microcirculation in the immediate post--ischemia/reperfusion time pe- riod.25 NGAL expression is markedly induced dur- ing inflammatory or related physiologic processes including models of renal ischemia/reperfusion.25 NGAL induction has been described as predicated on inflammatory cell interaction with affected ep- ithelial cells. An analysis of the pattern of NGAL expression in UC, PJ, and SH mice showed evidence that 66% of untreated mice had substantially in- creased levels of NGAL, compared with sham and PJ34-treated mice. In contrast, 80% of the mice treated with PJ34 showed no evidence of significant increased NGAL expression. In addition, NGAL transcription has been documented to commence at early time intervals after renal ischemia/reperfu- sion injury, making NGAL ideally suited to measure as a biomarker for ischemia-related renal injury, compared with N-acetyl-b-D-glucosaminidase and creatinine. While the animals in these experiments displayed a component of biological variability in their individual NGAL response to TAR, there is a consistent and statistically significant increase in NGAL expression among the UC cohort. The etiology of this observed biological disparity among UC mice is not fully clear and may serve as a useful subject for future investigation. Although routinely used to clinically monitor postoperative renal func- tion, studies in documenting consistent changes in murine serum creatinine at early time intervals have been disappointing.25 Our initial analysis thus confirmed a role of the PARP enzyme in the metabolic and inflammatory responses of renal tissue to TAR. However, the cellular response to ischemia/reperfusion injury is multifaceted and also known to involve prothrombotic and fibrino- lytic pathways. In the setting of ischemia/reperfu- sion injury in heart and lymphatic tissue, PARP activity has been correlated with fibrin deposition in microcirculatory beds.22,26,27 Therefore, markers of tissue fibrinolysis (D-dimers) and thrombosis (TAT complex) were then measured in UC, PJ, and SH mice after TAR. Fig 4. Renal NGAL mRNA levels 48 hours after TAR. There is marked increased NGAL transcription in un- treated control mice, reflecting renal injury. All values were normalized to actin mRNA levels. NGAL, Neutro- phil gelatinase-associated lipocalin. After 48 hours of reperfusion after TAR, D- dimer levels were statistically elevated in both UC and PJ34-treated mice, compared with sham. This finding indicates that TAR results in an increase in fibrinolytic activity in both UC and PJ animals. However, PJ34 administration did not ameliorate D-dimer levels, suggesting that PARP does not modulate indices of renal fibrinolytic activity 48 hours after TAR. The complementary biological consequence of fibrinolysis is usually thrombosis, prompting our analysis of markers of tissue throm- bosis in these 3 treatment groups after TAR. Absolute levels of TAT complex were profoundly low in all groups, suggesting that renal injury sustained during TAR was likely independent of thrombotic processes. While fibrin deposition and hemostatic dysregulation after ischemia/reperfu- sion injury have been described in the kidney, the routine use of heparin in the TAR model may preclude the accurate detection of hemostatic dysregulation. CONCLUSION PJ34, a water-soluble specific inhibitor of PARP activity, preserves renal mitochondrial function and abrogates the expression of NGAL, a marker of renal tissue injury, in a murine model of TAR. While tissue markers of fibrinolysis are increased after TAR, PJ34 administration does not appear to modulate their induction in renal tissue. Tissue thrombosis does not appear to play a major role in renal injury after TAR; however, the use of heparin in these experiments may complicate interpreta- tion of these results. Indeed, the ability of PJ34 to enhance renal mitochondrial function and ame- liorate the induction of NGAL, a marker of renal tissue injury, makes it well suited as a potential therapeutic adjunct in patients undergoing com- plex aortic reconstruction. REFERENCES 1. Svensson LG, Crawford ES, Hess KR, Coselli JS, Safi HJ. Experience with 1509 patients undergoing thoracoabdom- inal aortic operations. J Vasc Surg 1993;17:357-68; discus- sion 368-70. 2. Cambria RP, Davison JK, Zannetti S, L’Italien G, Atamian S. Thoracoabdominal aneurysm repair: perspectives over a decade with the clamp-and-sew technique. Ann Surg 1997; 226:294-303; discussion 303-5. 3. Acher CW, Wynn MM, Hoch JR, Kranner PW. Cardiac function is a risk factor for paralysis in thoracoabdominal aortic replacement. J Vasc Surg 1998;27:821-8; discussion 829-30. 4. Coselli JS, LeMaire SA, Miller CC 3rd, et al. Mortality and paraplegia after thoracoabdominal aortic aneurysm repair: a risk factor analysis. Ann Thorac Surg 2000;69: 409-14. 5. Golden MA, Donaldson MC, Whittemore AD, Mannick JA. Evolving experience with thoracoabdominal aortic aneu- rysm repair at a single institution. J Vasc Surg 1991;13:792-6; discussion 796-7. 6. Kashyap VS, Cambria RP, Davison JK, L’Italien GJ. Renal failure after thoracoabdominal aortic surgery. J Vasc Surg 1997;26:949-55; discussion 955-7. 7. Jacobs MJ, Eijsman L, Meylaerts SA, et al. Reduced renal failure following thoracoabdominal aortic aneurysm repair by selective perfusion. Eur J Cardiothorac Surg 1998;14: 201-5. 8. Schmidt CA, Wood MN, Gan KA, Razzouk AJ. Surgery for thoracoabdominal aortic aneurysms. Am Surg 1990;56: 745-8. 9. Cox GS, O’Hara PJ, Hertzer NR, Piedmonte MR, Krajewski LP, Beven EG. Thoracoabdominal aneurysm repair: a rep- resentative experience. J Vasc Surg 1992;15:780-7; discus- sion 787-8. 10. Hollier LH, Money SR, Naslund TC, et al. Risk of spinal cord dysfunction in patients undergoing thoracoabdominal aortic replacement. Am J Surg 1992;164:210-3; discussion 213-4. 11. Singri N, Ahya SN, Levin ML. Acute renal failure. Jama 2003;289:747-51. 12. Hassoun HT, Miller CC 3rd, Huynh TT, Estrera AL, Smith JJ, Safi HJ. Cold visceral perfusion improves early survival in patients with acute renal failure after thoracoabdominal aortic aneurysm repair. J Vasc Surg 2004;39:506-12. 13. Tang IY, Cox DS, Patel K, et al. Argatroban and renal replacement therapy in patients with heparin-induced thrombocytopenia. Ann Pharmacother 2005;39:231-6. 14. Svensson LG, Coselli JS, Safi HJ, Hess KR, Crawford ES. Appraisal of adjuncts to prevent acute renal failure after surgery on the thoracic or thoracoabdominal aorta. J Vasc Surg 1989;10:230-9. 15. Cambria RP, Davison JK. Regional hypothermia for preven- tion of spinal cord ischemic complications after thoracoab- dominal aortic surgery: experience with epidural cooling. Semin Thorac Cardiovasc Surg 1998;10:61-5. 16. Szabo C. Potential role of the peroxynitrate-poly (ADP- ribose) synthetase pathway in a rat model of severe hemor- rhagic shock. Shock 1998;9:341-4. 17. Goldfarb RD, Marton A, Szabo E, et al. Protective effect of a novel, potent inhibitor of poly (adenosine 5#-diphosphate- ribose) synthetase in a porcine model of severe bacterial sepsis. Crit Care Med 2002;30:974-80. 18. Virag L SC. The therapeutic potential of poly (ADP-Ribose) polymerase inhibitors. Pharmacol Rev 2002;54:375-429. 19. Faro R, Toyoda Y, McCully JD, et al. Myocardial protection by PJ34, a novel potent poly (ADP-ribose) synthetase inhibitor. Ann Thorac Surg 2002;73:575-81. 20. Liaudet L, Szabo A, Soriano FG, Zingarelli B, Szabo C, Salzman AL. Poly (ADP-ribose) synthetase mediates intestinal mucosal barrier dysfunction after mesenteric ischemia. Shock 2000;14:134-41. 21. Abdelkarim GE, Gertz K, Harms C, et al. Protective effects of PJ34, a novel, potent inhibitor of poly (ADP-ribose) polymerase (PARP) in in vitro and in vivo models of stroke. Int J Mol Med 2001;7:255-60.
22. Vallee I, Tait SW, Powell PP. African swine fever virus infection of porcine aortic endothelial cells leads to inhibi- tion of inflammatory responses, activation of the throm- botic state, and apoptosis. J Virol 2001;75:10372-82.
23. Lang-Lazdunski L, Matsushita K, Hirt L, et al. Spinal cord ischemia. Development of a model in the mouse. Stroke 2000;31:208-13.
24. Zingarelli B, Salzman AL, Szabo C. Genetic disruption of poly (ADP-ribose) synthetase inhibits the expression of P-selectin and intercellular adhesion molecule-1 in myocardial ischemia/reperfusion injury. Circ Res 1998;83: 85-94.
25. Mishra J, Ma Q, Prada A, et al. Identification of neutrophil gelatinase-associated lipocalin as a novel early urinary bio- marker for ischemic renal injury. J Am Soc Nephrol 2003; 14:2534-43.
26. Cassada DC, Tribble CG, Long SM, et al. Adenosine A2A analogue ATL-146e reduces systemic tumor necrosing fac- tor-alpha and spinal cord capillary platelet-endothelial cell adhesion molecule-1 expression after spinal cord ischemia. J Vasc Surg 2002;35:994-8.
27. Gomez-Villamandos JC, Hervas J, Moreno C, et al. Subcel- lular changes in the tonsils of pigs infected with acute African swine fever virus. Vet Res 1997;28:179-89.