Increased Tissue Microarray Matrix Metalloproteinase Expression Favors Proteolysis in Thoracic Aortic Aneurysms and Dissections

Transcription

1 Increased Tissue Microarray Matrix Metalloproteinase Expression Favors Proteolysis in Thoracic Aortic Aneurysms and Dissections George J. Koullias, MD, PhD, Pars Ravichandran, MD, Dimitris P. Korkolis, MD, PhD, David L. Rimm, MD, and John A. Elefteriades, MD Section of Cardiothoracic Surgery, Department of Surgical Pathology, and Yale Microarray Facility, Yale New Haven Hospital, New Haven, Connecticut Background. Little information is available regarding the role of matrix metalloproteinases (MMPs) in thoracic aortic aneurysms and dissections. We applied tissue microarray analysis to determine MMP profiles in a large group of surgically resected thoracic aneurysms and dissections. Methods. Specimens from 47 patients undergoing a variety of surgical procedures for thoracic aneurysm (n 30) and dissection (n 17) were included. Expression of MMP-1, MMP-2, and MMP-9 and tissue inhibitors of metalloproteinases (TIMP-1 and TIMP-2) were assessed by aortic tissue microarray immunostaining. Matrix metalloproteinase and TIMP expression in aortic tissue was compared with seven control aortic specimens, free of any vascular disease. Results. Expression of MMP-1 and MMP-9 was significantly increased in aneurysm and aortic dissection patients compared with control specimens (p < 0.05). Expression of TIMP-2 was significantly increased in the entire patient group, compared with control specimens (p < 0.05). Aortic dissection patients had higher MMP-2 and MMP-9 expression than aortic aneurysm patients in areas of disease. Compared with control patients, the MMP-9 to TIMP-1 ratio (a relative index of proteolytic state) was increased in both the aortic aneurysm and dissection groups (p < 0.05). Conclusions. The increased MMP expression in aortic aneurysms and dissections indicates a metamorphosis in the aneurysm wall toward increased proteolysis compared with the normal aorta. Furthermore, we find even higher MMP-2 and MMP-9 presence in aortic dissection. In both aneurysms and dissections, this transformation to a proteolytic state likely plays an important pathophysiologic role in the development and progression of the aortic disease. The recognition of this pathophysiologic mechanism raises the potential for drug therapy to interrupt the cascade of events. (Ann Thorac Surg 2004;78: ) 2004 by The Society of Thoracic Surgeons Matrix metalloproteinases (MMPs) are a group of proteases that play an important role in the protein synthetic lytic equilibrium of connective tissue [1, 2]. They are synthesized by numerous cell types that are routinely found inside, or in the vicinity of, the aortic wall. Matrix metalloproteinases act on diverse substrates, including elastin, fibrillin, and collagen. Matrix metalloproteinases 1, 2, and 9 are the most important and widespread MMPs in abdominal aneurysmal tissue [3, 4]. Matrix metalloproteinase tissue activity is regulated by the presence of naturally present tissue inhibitors of metalloproteinases (TIMPs) [5, 6]. Matrix metalloproteinase mediated breakdown of large matrix molecules, such as elastin and collagen of various types, results in protein fragmentation, with destruction of the normal lamellar architecture in the abdominal aortic wall [7, 8]. Accepted for publication May 6, Address reprint requests to Dr Elefteriades, Section of Cardiothoracic Surgery, Yale University School of Medicine, New Haven, CT 06510; john.elefteriades@yale.edu. This destructive rearrangement of the elastic aortic wall components is preceded by a period of inflammatory response and followed by a longer phase of gradual aortic dilatation in abdominal aortic aneurysm models [9, 10]. For the majority of thoracic aortic aneurysms (TAAs) and dissections (TADs), the process that leads to gradual deterioration of the mechanical properties of the aortic wall [11], diameter increase, and subsequent rupture is of unclear pathologic origin. Furthermore, only two studies have examined the presence of MMPs and their tissue inhibitors in a limited number of TAAs and TADs unrelated to Marfan syndrome [12, 13]. In view of the above, we sought to determine the possible differences in MMP and TIMP immunohistochemical expression among aortic samples from patients with various types of thoracic aortic disease operated on at our institution. We specifically sought to compare proteolytic states of aneurysms with normal aortas, as well as to determine the possible differences in proteolytic and antiproteolytic expression among patients with and without aortic dissection by The Society of Thoracic Surgeons /04/$30.00 Published by Elsevier Inc doi: /j.athoracsur

2 Ann Thorac Surg KOULLIAS ET AL 2004;78: METALLOPROTEINASES IN THORACIC AORTIC DISEASE 2107 Table 1. Patient Characteristics Characteristic Number 47 Mean age (y) 67.4 Females/males 15/32 Pathology Ascending aorta 28 Descending aorta 19 Aortic valve involvement 17 Coronary artery disease 4 Aortic arch involvement 18 Dissection 17 Type A 16 Acute 4 Chronic 8 Intramural hematoma subtype 4 Type B 1 Material and Methods Patient Group Our study group consisted of 47 patients operated on for TAA or TAD at Yale New Haven Hospital between 2001 and Thirty of these patients had a TAA, and 17 had a TAD. Four of the 17 patients in the TAD group presented with intramural hematoma. These individuals were selected from the overall population of thoracic aortic surgical patients because pathologic specimens adequate for analysis remained available. The control group consisted of autopsy aortic samples from 7 young cadavers (mean age, 34.5 years; 3 females, 4 males) with no evidence of any aortic, coronary, or vascular disease. Demographic data, as well as the range of aortic pathologic disease, are summarized in Table 1. Full cardiopulmonary bypass, with or without deep hypothermic circulatory arrest, or left atrial to femoral artery bypass were used, depending on the type and location of the aneurysm or dissection. Construction of Aortic Tissue Microarray and Monoclonal Antibodies Tissue microarray is a method developed in 1998 by Kononen and colleagues [14] and has become a standard method of histopathologic analysis today [15, 16]. This method consists of harvesting three small (0.6 mm in diameter) disks of tissue, using a coring device, from three key points of standard histologic sections of each patient. All tissue disks from all patients and controls are placed in a single arrayed master block. In this fashion, three histologic segments per patient, from multiple patients, can be analyzed in one slide simultaneously [16]. Consequently, such a slide contains the entire patient and control group in triplicate. In our investigation, an aortic tissue microarray was constructed with specimens of 47 patients, using current techniques at the Yale Tissue Microarray Facility [17]. For the patients with TAA, three representative tissue disks were taken from n the belly of the aneurysm. In the TAD group, these three tissue disks were collected from the dissected part of the operative specimen, either the external or the internal component. In the four cases of intramural hematoma, the tissue disks were collected from a location involved with the intramural process. The MMP and TIMP monoclonal antibodies used form a nonreversible complex with their respective antigens, which allows their identification by chromogen agents. The primary MMP antibodies (anti MMP-1 OPA , anti MMP-2 OPA , and anti MMP-9 OPA , ABR Inc, Golden, CO) and TIMP antibodies (anti TIMP-1 OMA , ABR Inc; and anti TIMP-2 AB 19078, Chemicon International Inc, Temecula, CA) were incubated in concentrations of 1:20 to 1:10 for 1 hour in a humidified chamber at 37 C. Histology and Immunohistochemistry Tissue microarray slides were stained with hematoxylin and eosin. In every specimen, the presence and severity of elastic fiber disruption, medial necrosis, and infiltration by inflammatory cells were assessed. For immunohistochemistry, as a first step, aortic tissue microarray slides containing the entire patient and control groups in triplicate were deparaffinized through graded xylol and alcohol solutions. Endogenous peroxidase was blocked with 3% H 2 O 2 in a Tris-buffered saline solution, and nonspecific binding was blocked by incubation with normal goat serum for 45 minutes. All microarray slides were incubated afterward with the primary antibodies in a dilution of 1:20 for the MMPs and 1:10 for the TIMPs for 45 minutes in a humidified chamber at 37 C. Incubation of the secondary antibody was repeated (Dako EnVision, peroxidase mouse; Dako, Denmark) for 45 minutes. At the end, staining was developed with 3,3=-diaminobenzidine and hematoxylin. Immunohistochemical Grading System A standard semiquantitative grading system (0 to 4) was applied blindly by two of the investigators to evaluate the intensity of immunohistochemical staining of MMP-1, MMP-2, and MMP-9 as well as TIMP-1 and TIMP-2. This was done for all three copies of the entire patient and control groups on each slide. The average of these two scores (in triplicate) was used for calculations. Grading was as follows: 0, no presence; 1, mild presence in less than 15% of cells; 2, moderate presence in 15% to 50% of cells; 3, high presence in 50% to 75% of cells; and 4, intense presence in greater than 75% of cells. In this manuscript, we use the term expression to represent the relative level of enzyme presence detected on the basis of the immunohistochemical assessments detailed above. This represents a semiquantitative assessment. Statistical Evaluation Statistics were performed by a staff biostatistician using two commercially available biostatistical software programs: (1) Primer of Biostatistics by Stanton A. Glantz, PhD (Mc Graw-Hill Health Professions Division, 1997)

3 2108 KOULLIAS ET AL Ann Thorac Surg METALLOPROTEINASES IN THORACIC AORTIC DISEASE 2004;78: Fig 1. Graph depicting matrix metalloproteinase (MMP) and tissue inhibitor of metalloproteinase (TIMP) expression scores in aneurysms and controls. (NS not significant.) and (2) STATA (STATA Corp, Stata Statistical Software, release 8.0, College Station, TX, 2003). Data were presented either as mean or mean standard error of the mean. Two-sample Student s t test was used to compare the means for every antigen in different patient groups. The Mann-Whitney rank sum test was also used because our data were ordinal. Statistical significance was set as p less than 0.05 to reject the null hypothesis. The Bonferroni correction was not necessary. Results Patient Groups Matrix metalloproteinases were expressed at inflammatory cells adjacent to areas of fibrolamellar fragmentation and in medial smooth muscles, but generally spared areas of cystic medial necrosis. These latter areas showed intense presence of TIMPs. Figure 1 presents data on expression of MMPs and their tissue inhibitors, TIMPs. When the entire aneurysm patient group was compared to the control group, the patient group had significantly higher MMP-1, MMP-9, and TIMP-2 expression (p 0.05 for all three antigens). The MMP-9/TIMP-1 ratio, used in the literature [13, 18] as a relative measure of the lytic synthetic equilibrium, was calculated by dividing the two immunohistochemical scores. We observed that in normal aortas the ratio was Fig 2. Intense matrix metalloproteinase 1 presence in smooth muscle cells (thin arrows) and inflammatory cells (thick arrows) in a thoracic aortic aneurysm case ( 125). less than 1 (ie, higher relative expression of TIMP-1 than MMP-9; exact ratio, 0.557), indicating a relative shift toward a synthetic state. In contrast, the patient group demonstrated a ratio of 1.49, implying a shift toward wall proteolysis (p 0.05; Table 2). Thoracic Aortic Aneurysm Patients Sections of TAAs in the 30 patients of this group showed extensive elastic network fragmentation with destruction of the fibrolamellar units and variable presence of cystic medial necrosis in the majority of patients. Matrix metalloproteinases and TIMPs were localized in smooth muscle cells, elastin fibers, and inflammatory cells (Fig 2). Compared with normal aortas, TAA patients showed increased expression of MMP-1 and MMP-9 and an increased MMP-9 to TIMP-1 ratio (p 0.01 for all; Table 2). Thoracic Aortic Dissection Patients Matrix metalloproteinases and TIMPs were detected in both dissecting layers of the aortic wall. Intense MMP-9 presence was observed in elastin fiber networks and smooth muscle cells in these patients (Fig 3). In the four cases of intramural hematoma, MMPs and TIMPs were detected in the same cellular and matrix components, as Table 2. Immunohistochemical Score Patient Group MMP-1 MMP-2 MMP-9 TIMP-1 MMP-9/TIMP-1 Ratio TIMP-2 Control score (n 7) TAA patients (n 30) TAD patients (n 17) TAA vs controls p 0.01 NS p 0.01 NS p 0.05 NS TAD vs controls p 0.01 NS p 0.01 NS p 0.05 NS MMP matrix metalloproteinase; NS not significant; TAA thoracic aortic aneurysm (nondissecting); TAD thoracic aortic dissection; TIMP tissue inhibitor of metalloproteinase.

4 Ann Thorac Surg KOULLIAS ET AL 2004;78: METALLOPROTEINASES IN THORACIC AORTIC DISEASE 2109 Fig 3. Intense matrix metalloproteinase 9 presence in a smooth muscle cell network (arrows) in a thoracic aortic dissection patient ( 250). well as in inflammatory cells, near areas of wall disruption. Compared with normal aortas, TAD patients showed increased expression of MMP-1 and MMP-9 and increased MMP-9 to TIMP-1 ratio (p 0.01 for both; Table 2). Aneurysm Versus Dissection Patients When TAA and TAD patients were compared with regard to MMP and TIMP expression, TAD patients showed significantly higher MMP-2 and MMP-9 levels (p 0.05; Fig 4). Comment Matrix metalloproteinases are synthesized by a variety of cells present inside the aortic wall under normal and disease states [2, 3, 13]. These lytic enzymes coexist with their naturally occurring inhibitors, TIMPs. The balance Fig 4. Graphic presentation of matrix metalloproteinase (MMP) and tissue inhibitor of metalloproteinase (TIMP) expression in aneurysms and dissections. (NS not significant.) of these enzymes within the vascular wall regulates the degradation and synthesis of the extracellular matrix in health and disease states [19]. The role of MMP excess in the pathogenesis of abdominal aortic aneurysms has been studied to a greater extent than that in thoracic aortic disease states. Current concepts of aneurysm development in the abdominal aorta postulate a variety of contributing factors collectively termed blood flow induced vascular remodeling [4, 6, 20]. In the blood flow induced vascular remodeling model, mechanical stretching and decreased nitric oxide production, as well as increased MMP-2 and MMP-9 secretion and activation, are thought critical for initiation of the destructive cellular and matrix wall changes that accompany transformation of the vessel to an aneurysm. Continued excessive MMP production and activation result in increased degradation of aortic wall elastin and collagen, with alteration of the mechanical properties of the aortic wall, leading to aneurysmal dilation [21]. Our study extends this type of analysis to the thoracic aorta. In the present study, we have shown increases in MMP-1 and MMP-9 in patients with thoracic aortic disease compared with control specimens. We have shown also that the ratio of MMP-9 to TIMP-1 is increased in thoracic aortic disease as well denoting a shift toward an increased proteolytic state. These findings hold for our whole patient group (TAAs and TADs) as well as for each subgroup (TAAs versus controls and TADs versus controls). We have shown also that patients with TAD manifest even greater proteolytic preponderance than patients with TAA alone. Our work is generally consonant with smaller prior studies of MMP homeostasis in the thoracic aorta [12, 13, 22]. We extend this type of analysis to relatively large numbers of patients. Also, we provide comparison between MMP expression in TADs compared with TAAs without dissection. The tissue microarray technology used in our study represents the current state of the art in antigen detection [15 17]. This method enables specific targeting of the tissue to be examined (examining precisely the abnormal tissue and not the adjacent normal tissue), reduces antibody quantities required, maximizes tissue resources, and manifests reproducibility and uniformity of results. By permitting simultaneous analysis in a single slide of many patients, using three small-core biopsies per patient from blocks originating from the areas of disease, this method minimizes the possibility of methodologic and interpretation errors [15, 16]. A weakness of the present study has to do with the lower age of the control patients. We cannot from this study alone rule out an effect of age alone on MMP expression; however, when our study is interpreted in the context of prior work, it appears likely that the effects we have documented are indeed part of the pathophysiologic cascade of aneurysm development. The enzymatic imbalances in TAAs and TADs detected in this study deserve further quantitative determinations. Furthermore, the role of increased MMP presence in these

5 2110 KOULLIAS ET AL Ann Thorac Surg METALLOPROTEINASES IN THORACIC AORTIC DISEASE 2004;78: lesions, whether a cause or an epiphenomenon, needs to be elucidated. The present study determines, in a large sample of patients, that during the formation of TAAs and TADs, MMP-1 and MMP-9 are overexpressed locally within the aortic wall in areas of disease. We were intrigued to find that affected aortic walls of patients with aortic dissection show even more severely increased MMP-2 and MMP-9 expression compared with nondissecting aortic aneurysms. Inasmuch as these two enzymes lyse collagen and elastin, it is conceivable that their increased local expression may play an important role in setting the stage for aortic dissection. This investigation also suggests the intriguing possibility that substances acting as matrix protease inhibitors [21, 23] may be applied clinically in patients to produce a slowing or halting of the evolution of thoracic aneurysm disease. This study offers hope that, through this means, patients might be protected from the otherwise disastrous consequences of uninhibited TAA formation and dissection. References 1. Parks WC. A confederacy of proteinases. J. Clin Invest 2002;110: Vu TH, Werb Z. Matrix metalloproteinases: effectors of development and normal physiology. Genes Dev 2000;14: Newman KM, Ogata Y, Malon AM, et al. Identification of matrix metalloproteinases-3 and -9 in abdominal aneurysm. Arterioscler Thromb 1994;14: Longo GM, Xiong W, Greiner TC, Zhao Y, Fiotti N, Baxter TB. Matrix metalloproteinases 2 and 9 work in concert to produce aortic aneurysms. J Clin Invest 2002;110: Stetler-Stevenson WG, Hewitt R, Corcoran M. Matrix metalloproteinases and tumor invasion: from correlation and causality to clinic. Cancer Biol 1996;7: Dollery C, McEwan J, Henney A. Matrix metalloproteinases and cardiovascular disease. Circ Res 1995;77: Freestone T. Inflammation and matrix metalloproteinases in the enlarging abdominal aortic aneurysm. Arterioscler Thromb Vasc Biol 1995;15: Ricci MA. Anti-CD 18 monoclonal antibody slows experimental aortic aneurysm expansion. J Vasc Surg 1996;23: Gertz SD, Kurgan S, Eisenberg D. Aneurysm of the rabbit common carotid artery induced by periarterial application of calcium chloride in vivo. J. Clin. Invest 1988;81: Anidjar S. Elastase-induced experimental aneurysms in rats. Circulation 1990;82: Koullias GJ, Modak RK, Korkolis DP, Barash P, Elefteriades JA. Mechanical and elastic properties of the normal and aneurysmal ascending aorta by intraoperative epiaortic echocardiography. J Am Coll Cardiol 2003;41(Suppl II):513a. 12. Lesauskaite V, Tanganelli P, Sassi C, et al. Smooth muscle cells of the media in the dilatative pathology of ascending thoracic aorta: morphology, immunoreactivity for osteopontin, matrix metalloproteinases, and their inhibitors. Hum Pathol 2001;32: Tamarina NA, McMillan WD, Shively VP. Expression of matrix metalloproteinases and their inhibitors in aneurysms and normal aorta. Surgery 1997;122: Kononen J, Bubendorf L, Kallioniemi A, et al. Tissue microarrays for high throughput molecular profiling of tumor specimens. Nature Med 1998;4: Camp RL, Charette LA, Rimm DL. Validation of tissue microarray technology in breast carcinoma. Lab Invest 2000; 80: Rimm DL, Camp RL, Charette LA, Olsen DA, Provost E. Amplification of tissue by construction of tissue microarrays. Exp Molec Pathol 2001;70: Rimm DL, Camp RL, Charette LA, Costa J, Olsen DA, Reiss M. Tissue microarray: a new technology for amplification of tissue resources. Cancer J 2001;7: Senzaki H, Paolocci N, Gluzband YA, et al. -Blockade prevents sustained metalloproteinase activation and diastolic stiffening induced by angiotensin II combined with evolving cardiac dysfunction. Circ Res 2000;86: Victor FJ. New insights into the world of matrix metalloproteinases. Circulation 2002;105: Lehoux ST, Tronc FR, Tedgui A. Mechanisms of blood flow-induced vascular enlargement. Biorheology 2002;39: Petrinec D. Doxycycline inhibition of aneurysmal degeneration in an elastase-induced rat model of abdominal aortic aneurysm: preservation of aortic elastin associated with suppressed production of 92kD gelatinase. J Vasc Surg 1996;23: Segura AM, Rafael LBS, Horiba K, et al. Immunohistochemistry of matrix metalloproteinases and their inhibitors in thoracic aortic aneurysms and aortic valves of patients with Marfan syndrome. Circulation 1998;98(Suppl 3):II Zempo N, Koyama N, Kenargy RD, Lea HJ, Cloves AW. Regulation of vascular smooth muscle cell migration and proliferation in vitro and injured rat arteries by a synthetic matrix metalloproteinase inhibitor. Arterioscler Thromb Vasc Biol 1996;16: INVITED COMMENTARY Thoracic aortic aneurysms (TAAs) and dissections represent significant clinical challenges to the cardiothoracic surgeon and are associated with a high incidence of perioperative morbidity. Thus, identifying the cellular and molecular mechanisms that drive the formation and progression of these processes holds great clinical relevance. To date, most of our knowledge of the pathobiology has not been a product of direct study, but rather an extension of studies regarding abdominal aortic aneurysms. Hence, the study undertaken by Koullias and colleagues in this issue of The Annals is significant in that studies were performed in aortic tissue samples collected from patients with TAAs and dissections. A generalized feature of aneurysm development is that significant remodeling of the vascular wall occurs and is characterized by degeneration of the aortic medial layer and degradation of extracellular matrix (ECM). Important ECM proteins that provide support and distensibility to the aortic wall include elastin, collagens and glycoproteins. A family of proteolytic enzymes responsible for ECM degradation are the matrix metalloproteinases (MMPs). A cause-effect relationship has been demonstrated between increased MMP activity to that of ECM degradation and pathologic tissue remodeling. However, there are more than 25 known MMP types, and those directly causative to vascular remodeling, particularly to 2004 by The Society of Thoracic Surgeons /04/$30.00 Published by Elsevier Inc doi: /j.athoracsur