From the Peripheral Vascular Surgery Society Valproic acid reversed pathologic endothelial cell gene expression profile associated with ischemia– reperfusion injury in a swine hemorrhagic shock model Marlin Wayne Causey, MD,a Shashikumar Salgar, PhD,b Niten Singh, MD,a Matthew Martin, MD,a and Jonathan D. Stallings, PhD,b Tacoma, Wash Background: Vascular endothelial cells serve as the first line of defense for end organs after ischemia and reperfusion injuries. The full etiology of this dysfunction is poorly understood, and valproic acid (VPA) has proven to be beneficial after traumatic injury. The purpose of this study was to determine the mechanism of action through which VPA exerts its beneficial effects. Methods: Sixteen Yorkshire swine underwent a standardized protocol for an ischemia–reperfusion injury through received (6 ؍ hemorrhage and a supraceliac cross-clamp with ensuing 6-hour resuscitation. The experimental swine (n Aortic .(5 ؍ and injury-control models (n (5 ؍ VPA at cross-clamp application and were compared with sham (n endothelium was harvested, and microarray analysis was performed along with a functional clustering analysis with gene transcript validation using relative quantitative polymerase chain reaction. Results: Clinical comparison of experimental swine matched for sex, weight, and length demonstrated that VPA significantly decreased resuscitative requirements, with improved hemodynamics and physiologic laboratory measure- ments. Six transcript profiles from the VPA treatment were compared with the 1536 gene transcripts (529 up and 1007 down) from sham and injury-control swine. Microarray analysis and a Database for Annotation, Visualization and Integrated Discovery functional pathway analysis approach identified biologic processes associated with pathologic vascular endothelial function, specifically through functional cluster pathways involving apoptosis/cell death and angiogenesis/vascular development, with five specific genes (THBS1, TNFRSF12A, ANGPTL4, RHOB, and RTN4) identified as members of both functional clusters. This study also examined gene expression of transforming growth factor (TGF)-␤ (TGF-␤1, TGF-␤2, and TGF-␤-releasing thrombospondin 1 [THBS1]) and genes expressing vascular endothelial growth factor (VEGF) C, VEGFD, and VEGFR1 and found that these genes were involved in the endothelial functional preservation associated with VPA administration. Conclusions: VPA minimized pathologic endothelial cell function through the TGF-␤ and VEGF functional pathways. This study also implicates that integrated functional modeling and analysis will enable advancements in endothelial dysfunction using a systems biology approach. (J Vasc Surg 2012;55:1096-103.) Clinical Relevance: The functional pathways associated with vascular ischemia–reperfusion are not fully understood. This study found that valproic acid (VPA), administered along with the application of a supraceliac cross-clamp, improved clinical, physiologic, and histologic changes in a swine model of hemorrhage and ischemia–reperfusion: 35% blood volume hemorrhage, 50-minute supraceliac aortic cross-clamp, and a 6-hour resuscitation. A functional analysis of endothelial dysfunction, with relative quantitative polymerase chain reaction validation, demonstrated the initial changes in endothelial cell dysfunction occurred through vascular endothelial growth factor and transforming growth factor-␤ gene members. This study demonstrates that pharmaceutical agents administered with an ischemic injury are beneficial after ischemia and target two common functional endothelial pathways. The molecular pathways altered with vascular ischemia once it had occurred.1-3 During the past 2 decades, how- are not fully understood. Previously, basic science research ever, there has been a transition in researching free radical into endothelial and vascular ischemia–reperfusion injury scavenging to an analysis of pharmacologic agents that centered on free radical scavenging and minimizing injury would minimize autogenous injury through a reduction in From the Departments of Surgerya and Clinical Investigation,b Madigan Competition of interest: none. Healthcare System, Joint Base Lewis-McChord. Additional material for this article may be found online at www.jvascsurg.org. The views expressed are those of the author(s) and do not reflect the official Reprint requests: Marlin Wayne Causey, 9040 Jackson Ave, Tacoma, WA policy of the Department of the Army, the Department of Defense, or the 98431 (e-mail: [email protected]). U.S. Government. The investigators adhered to the policies for protection The editors and reviewers of this article have no relevant financial of human subjects as prescribed in Title 45 of the Code of Federal relationships to disclose per the JVS policy that requires reviewers to Regulations, Part 46. decline review of any manuscript for which they may have a competi- This manuscript was the Norman Rich Award recipient for the 2010 tion of interest. Military Vascular Surgery Society and was presented at the podium at 0741-5214/$36.00 the 2011 Peripheral Vascular Surgery Society meeting, Chicago, Ill, Published by Elsevier Inc. on behalf of the Society for Vascular Surgery. Jun 15, 2011. doi:10.1016/j.jvs.2011.08.060 1096 JOURNAL OF VASCULAR SURGERY Volume 55, Number 4 Causey et al 1097 cellular metabolic demand before an ischemic insult. This is survival factors15 and offers significant HDAC inhibi- important, because maintenance of physiologic homeostasis tion-mediated neuroprotection in models of cerebral after major resuscitative efforts is crucial, and most current ischemia.16 studies have focused on the effect of intravenous crystalloids VPA may mitigate changes in vascular gene expression and use of vasopressors because these are widely used to patterns and preserve overall vascular endothelial function, augment adequate tissue perfusion by maintaining homeo- because several in vitro models have demonstrated that static cardiovascular and hemodynamic parameters.4,5 VPA attenuates expression of both angiogenic and vascular The vascular endothelial cell layer is now widely recog- permeability factors.6 The in vivo profile of VPA in humans nized as a critical component in maintaining physiologic is well characterized. The drug is easily administered and homeostasis and regulating normal as well as pathologic shows good pharmacokinetics, with a clinically useful half- responses to injury. Disruption or injury of the vascular life of 15.9 to 17.3 hours.17 endothelial cell layer initiates a broad cascade of local and We initiated this follow-on study to determine whether systemic processes that involve the coagulation system, the VPA altered endothelial pathway members, such as TGF-␤ inflammatory cascade, and the cellular immune system. and VEGF, in vivo and to determine the clinical, physio- This has been implicated in adverse outcomes among a logic, and hemodynamic benefit in our porcine model of wide variety of patient populations, including acute and hemorrhagic shock with transient ischemia and subsequent chronic vasculopathy, vascular and endovascular surgery, reperfusion. Here, we report changes in endothelial gene and after major traumatic injury. It is likely that the patho- expression and the identification of altered functional clus- genesis of temporary ischemia involves oxidative damage ters as the result of treatment with VPA in six swine that and metabolic disturbances that ultimately lead to dysregu- received this injury compared with the model validation lation of endothelial gene expression.6 Because vascular protocol that our group previously published.6 homeostasis is linked with the acquisition of aberrations in the gene-regulatory programs,7,8 our group has used a microarray approach to examine gene expression changes METHODS in injured aortic endothelial cells after hemorrhagic shock This study was approved by the Madigan Army Medical as part of a model validation.6 Center Institutional Animal Care and Use Committee. Our porcine surgical model of hemorrhagic shock and Hemorrhagic shock and ischemia-reperfusion in- functional clustering analysis demonstrated significant jury model. Sixteen Yorkshire swine were purchased from changes in members of the transforming growth factor-␤ a U.S. Department of Agriculture–approved swine research (TGF-␤) and vascular endothelial growth factor (VEGF) facility and housed in accordance with the Guidelines for the gene families in the endothelium, suggesting that angio- Care and Use of Laboratory Animals.18 genesis, vascular permeability, and apoptotic pathways are Animals were anesthetized, intubated, and vascular ac- critical to endothelial cellular dysfunction as the result of cess was obtained to continuously monitor central and hemorrhagic shock. Major mechanisms involved in the alter- peripheral cardiovascular hemodynamics. A midline celiot- ation and regulation of gene expression include local chroma- omy permitted placement of catheters in the bladder and tin remodeling and alterations in nucleosomal packaging of inferior vena cava and enabled application of a supraceliac DNA, primarily through covalent posttranslational modifica- aortic cross-clamp. Hemorrhage (35% of total circulating tions of nucleosomal histones by acetylation, methylation, blood volume) was induced through the inferior vena cava phosphorylation, and ubiquitination.9 catheter over 30 minutes, followed by 50 minutes of isch- In addition to the previously identified TGF-␤ and VEGF emia through the supraceliac aortic cross-clamp,