Role of hemodynamic forces in the ex vivo arterialization of human saphenous veins

Role of hemodynamic forces in the ex vivo arterialization of human saphenous veins Xavier Berard, MD, PhD, a Sébastien Déglise, MD, b Florian Alonso, PhD, b François Saucy, MD, b Paolo Meda, MD, c Laurence Bordenave, MD, PhD, a Jean-Marc Corpataux, MD, b and Jacques-Antoine Haefliger, PhD, b Bordeaux, France; and Lausanne and Geneva, Switzerland Background: Human saphenous vein grafts are one of the salvage bypass conduits when endovascular procedures are not feasible or fail. Understanding the remodeling process that venous grafts undergo during exposure to arterial conditions is crucial to improve their patency, which is often compromised by intimal hyperplasia. The precise role of hemodynamic forces such as shear stress and arterial pressure in this remodeling is not fully characterized. The aim of this study was to determine the involvement of arterial shear stress and pressure on vein wall remodeling and to unravel the underlying molecular mechanisms. Methods: An ex vivo vein support system was modified for chronic (up to 1 week), pulsatile perfusion of human saphenous veins under controlled conditions that permitted the separate control of arterial shear stress and different arterial pressure (7 mm Hg or 70 mm Hg). Results: Veins perfused for 7 days under high pressure (70 mm Hg) underwent significant development of a neointima compared with veins exposed to low pressure (7 mm Hg). These structural changes were associated with altered expression of several molecular markers. Exposure to an arterial shear stress under low pressure increased the expression of matrix metalloproteinase (MMP)-2 and MMP-9 and tissue inhibitor of metalloproteinase (TIMP)-1 at the transcript, protein, and activity levels. This increase was enhanced by high pressure, which also increased TIMP-2 protein expression despite decreased levels of the cognate transcript. In contrast, the expression of plasminogen activator inhibitor-1 increased with shear stress but was not modified by pressure. Levels of the venous marker Eph-B4 were decreased under arterial shear stress, and levels of the arterial marker Ephrin-B2 were downregulated under high-pressure conditions. Conclusions: This model is a valuable tool to identify the role of hemodynamic forces and to decipher the molecular mechanisms leading to failure of human saphenous vein grafts. Under ex vivo conditions, arterial perfusion is sufficient to activate the remodeling of human veins, a change that is associated with the loss of specific vein markers. Elevation of pressure generates intimal hyperplasia, even though veins do not acquire arterial markers. (J Vasc Surg 2013;57:1371-82.) Clinical Relevance: The pathological remodeling of the venous wall, which leads to stenosis and ultimately graft failure, is the main limiting factor of human saphenous vein graft bypass. This remodeling is due to the hemodynamic adaptation of the vein to the arterial environment and cannot be prevented by conventional therapy. To develop a more targeted therapy, a better understanding of the molecular mechanisms involved in intimal hyperplasia is essential, which requires the development of ex vivo models of chronic perfusion of human veins. The number of patients with peripheral arterial occlusive disease is increasing worldwide in the aging population. 1 Today, endovascular procedures are preferred to bypass grafting as first-line therapy for critical limb ischemia because of lower morbidity and mortality rates. However, when endovascular therapy fails or is not applicable (eg, in the case of extensive occlusive lesions), bypass grafting of autologous saphenous vein remains the rescue therapy of choice. 2 Failure of such a vein graft is mostly related to intimal hyperplasia (IH). 3 This condition represents a blood vessel wall response to an injury. 4 The IH is a chronic alteration of the wall of the grafted blood vessel, which leads to the formation of a thickened fibrocellular layer between the endothelium and the inner elastic lamina, 5 as a response to the hemodynamic conditions imposed on the arterial compartment. This IH results in a narrowing of the lumen of the graft, reducing blood flow in the vascularized territory. Physical forces and structural and biochemical events, including the proliferation of smooth muscle cells (SMCs) 6 From the Department of Vascular Surgery, Pellegrin Hospital, University of Bordeaux, Bordeaux a ; the Department of Thoracic and Vascular Surgery, University Hospital, Laboratory of Experimental Medicine, Lausanne b ; and the Department of Cell Physiology and Metabolism, University of Geneva, School of Medicine, Geneva. c The authors are supported by grants from the Swiss National Science Foundation (31003A-138528/1) and the Novartis and Muschamp Foundations. P.M. is supported by grants from the Swiss National Science Foundation (310030_141162; CR32I3_129987; IZ73Z0_127935), the Juvenile Diabetes Research Foundation (40-2011-11; 5-2012-281), and the European Union (BETAIMAGE 222980; IMIDIA C2008-T7; BETATRAIN 289932). Author conflict of interest: none. Reprint requests: Jacques-Antoine Haefliger, PhD, Department of Thoracic and Vascular Surgery, c/o Department of Physiology, Bugnon 7a, Bureau 03.018, 1005 Lausanne, Switzerland (e-mail: Jacques-Antoine. Haefliger@chuv.ch). The editors and reviewers of this article have no relevant financial relationships to disclose per the JVS policy that requires reviewers to decline review of any manuscript for which they may have a conflict of interest. 0741-5214/$36.00 Copyright Ó 2013 by the Society for Vascular Surgery. http://dx.doi.org/10.1016/j.jvs.2012.09.041 1371

1372 Berard et al May 2013 secondary to the proteolytic activity of the fibrinolytic, 7 metalloproteinase, 8 and inhibitor systems, 9 are involved in IH. 10 These changes cause the stiffening of the vein graft, limiting its ability to adapt to the arterial circulation. However, the involved molecular mechanism has been only marginally elucidated. 11 Some grafts escape this deleterious evolution because of adequate adaptation to the arterial environment. 3 Among the various hemodynamic forces encountered in the arterial system, shear stress seems the most prominent component, 12 whereas the role of pressure is less clear. Most previous studies monitored development of IH under low shear stress conditions, 13 eliminating the involvement of arterial pressure. We used an ex vivo vein support system (EVVSS) that allowed us to study the wall remodeling of segments of human saphenous veins exposed to various hemodynamic conditions that characterize the arterial compartment. 14,15 Using such a system, we previously reported that IH increased with plasminogen activator inhibitor (PAI)-1 activity in veins exposed to arterial conditions. 7 We now provide evidence that arterial shear stress and pressure differentially impinge on the mechanisms involved in this arterialization. METHODS Human saphenous veins. We obtained 17 surplus segments of nonvaricose human saphenous veins from 15 patients (eight men and seven women) with a median age of 69 years (interquartile range, 51 80), who underwent lower limb bypass surgery for critical limb ischemia. A segment >3 cm of the greater saphenous vein was harvested and immediately stored at 4 C in a RPMI-1640 medium supplemented with GlutaMAX and 30% fetal calf serum (Gibco, Grand Island, NY). A segment of at least 2 cm length was perfused for 7 days (arterial shear stress), and another segment (control) was either fixed in formalin (one half) or rapidly frozen in liquid nitrogen (the other half). The ethical committee of the University of Lausanne approved the experiments, which are in accordance with the principles outlined in the Declaration of Helsinki for use of human tissue. Ex vivo perfusion system. The EVVSS 7 consists of a gearing pump (Reglo-Z; Ismatec, Glattbrugg, Switzerland) that induces a pulsatile signal of 60 pulses/min and constant amplitude generating an unidirectional flow of 80 6 15 ml/min independently from the pressure applied in the system. 7,14,15 This cardioid signal was produced by an arbitrary waveform generator controlled by a computer (National Instruments PCI-6024 E Acquisition card; National Instruments, Austin, Tex). The driving software integrated constant acquisition and monitoring of pressures, flow velocity, pulse rate, and signal (Labview; National Instruments). The vein segment was connected to the perfusion pump by peroxide-treated silicone tubing (internal diameter 3.2 mm; Ismatec) and maintained at 37 6 0.1 C inside a perfusion chamber, which was placed inside a cell culture incubator (Model 310; Forma Scientific Inc, Marietta, Ohio). In all experiments, the culture medium of both the chamber (250 ml) and the perfusate (150 ml) was RPMI-1640, supplemented with Glutamax, 30% fetal calf serum (Gibco), 8% 70-kDa dextran (Sigma), and 1% antibiotic-antimycotic solution (penicillin G, 10,000 units/ml; streptomycin sulfate, 10,000 mg/ml; amphotericin B, 25 mg/ml; and gentamicin, 0.5 mg/ml). The ph value was kept constant at 7.40 6 0.01 using aco 2 /ph algorithm based on the Henderson-Hasselbach equation. Vein segments were perfused for 7 days, 16 and the culture media changed every 2 days. In this system, the shear stress (SS) was given by SS ¼ 4 mq/pr 3, where m is the viscosity (dyn$s/cm 2 ); Q, the flow rate (ml/s); and r, the radius of the vein segment (cm). The medium viscosity was 3.73 10 2 dyn$s/cm 2, as measured in a Coulter viscometer (Coulter Electronics, High Wycombe, UK). Given that the radius of the perfused veins was not altered during the perfusion, the shear stress could be considered essentially constant, whatever the pressure applied to the system. This stress was set at a value of 9 to 15 dyn/cm 2, which corresponds to the value expected in the arterial system. Under these conditions, the system allows for an evaluation of the effects of different imposed pressures. Mean pressure (MP) was given by MP ¼ (systolic pressure þ 2 diastolic pressure)/3. In a first series of experiments (eight vein segments), we tested the effects of a systolic/diastolic pressure of 8 6 1/6 6 1 mm Hg (mean pressure ¼ 7 mm Hg) obtained by the perfusion pump that produced the arterial flow. In a second series of experiments (nine vein segments), under the same arterial flow, the level of systolic/diastolic pressure was increased to 90 6 5/65 6 5 mm Hg (mean pressure ¼ 70 mm Hg), using a second pump, to be representative of a normal arterial compartment. This mean pressure, reproducing the normal mean arterial pressure in humans, could be maintained throughout the 7 days of in vitro perfusion. After the 7-day perfusion, the veins were removed from the system, and the distal and proximal 5-mm segments, which served to attach the vessel to the perfusion equipment, were discarded. A central, 5-mm-thick ring was sampled for morphometry. The two remaining vein fragments were reduced into powder for reverse transcriptase polymerase chain reaction (RT-PCR) and Western blot analysis. Morphometry. A 5-mm-thick ring was harvested before (control) and after a 7-day-long perfusion (arterial shear stress) from the central portion of each vein segment. Hematoxylin-eosin and Van Gieson elastin stains were used for histologic and morphometric analysis. For morphometry, digital pictures were taken and measurements performed with software KS 400 (Carl Zeiss, Oberkochen, Germany). For each sample, 24 measurements of intima and media thickness were processed at a magnification of 20 and 1.25. Quantitative RT-PCR. Quantitative RT-PCR was performed on RNA of human saphenous vein treated for 30 minutes in the presence of DNase I (DNA-free kit; Ambion, Cambridge, UK). For reverse transcription, 1 mg of RNA was used (Promega, Madison, Wisc). Human matrix metalloproteinase (MMP)-2 and MMP-9, tissue

Volume 57, Number 5 Berard et al 1373 Table. Human primers for quantitative reverse transcriptase polymerase chain reaction (RT-PCR) Gene Sense primer (5 0-3 0 ) Antisense primer (5 0-3 0 ) PAI-1 GGCTGGTGCTGGTGAATG ATCGGGCGTGGTGAACTC MMP-2 CAGAGCCACCCCTAAAGAGA TGTGAAAGGAGAAGAGCCTGA MMP-9 GCCACTTCCCCTTCATCTTC GTCGTCGGTGTCGTAGTTGG TIMP-1 TGACATCCGGTTCGTCTACA TGCAGTTTTCCAGCAATGAG TIMP-2 AAGCGGTCAGTGAGAAGGAA TCTCAGGCCCTTTGAACATC Ephrin-B2 CTTTGGAGGGCCTGGATAAC CTGTTGCCGTCTGTGCTAGA Eph-B4 CGCAGACCAAAGAGAGTGTG GGGACTACAAACCCCAATGA GAPDH AACTTTGGTATCGTGGAAGG CAGTAGAGGCAGGGATGATGT GADPH, Glyceraldehyde 3-phosphate dehydrogenase; MMP-2/-9, matrix metalloproteinase-2/-9; PAI-1, plasminogen activator inhibitor-1; TIMP-1/-2, tissue inhibitor of metalloproteinase-1/-2. inhibitor of metalloproteinase (TIMP)-1 and TIMP-2, plasminogen activator inhibitor PAI-1, and Ephrin-B2 and Eph-B4 mrna levels were determined by quantitative RT-PCR, using the Fast SYBR Green Master Mix (Applied Biosystems, Carlsbad, Calif) in a ViiA7 Instrument (Applied Biosystems). The primers used to amplify specific cdnas from human saphenous veins are given in the Table and were designed using the free online software Primer3 (http://frodo.wi.mit.edu/primer3/). All experiments included negative controls (amplification of distilled water or RNA samples that had not been subjected to reverse transcription). 17,18 Western blots. Segments of saphenous veins were reduced to powder and homogenized in lysis buffer as published previously. 17,18 Samples of total vessel extracts (25 mg) were resolved by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) (10%) and transferred to a polyvinylidene difluoride membrane (Immobilon-P; Millipore, Billerica, Mass). The membrane was incubated with a primary antibody overnight at 4 C, washed in TBS-T, and incubated for 1 hour with a horseradish peroxidase-linked antibody against mouse (Southern Biotech, Birmingham, Ala), rabbit (Thermo Fisher Scientific, Waltham, Mass), or goat IgGs (Sigma-Aldrich, St. Louis, Mo), whichever was deemed adequate. Immunostaining was revealed by chemoluminescence, using ECL-plus kit (GE Healthcare, Uppsala, Sweden). The following primary antibodies were used: mouse monoclonal against MMP-2 (Chemicon, Temecula, Calif), mouse monoclonal against MMP-9 (Chemicon this antibody recognizes both the pro- and the active forms of MMP-9), mouse monoclonal against TIMP-2 (R&D Systems, Minneapolis, Minn), rabbit polyclonal against PAI-1 (Novus Biologicals, Littleton, Colo), and goat polyclonal against TIMP-1 (R & D Systems). The former four antibodies were used at a dilution 1:1000. The latter three were used at a dilution 1:500. To check the total protein levels, membranes were probed with polyclonal antibodies against alpha actin (BD Transduction, Lexington, Ky). Immunocytochemistry. The above-mentioned primary antibodies were also used at a dilution of 1:100. Binding sites of the different antibodies were visualized using an avidin-biotinylated horseradish-peroxidase complex (Vectastain Elite ABC Kit; Vector Laboratories, Burlingame, Calif). After rinsing in phosphate-buffered saline, the product of the peroxidase reaction was visualized using 0.1% 3,3-diaminobenzidine tetrahydrochloride dehydrate, dissolved in phosphate-buffered saline containing 0.3% H 2 O 2. Antibody binding was revealed using the avidinbiotin peroxidase technique. When needed, sections were counterstained with hemalum to visualize cell nuclei. Zymography. For zymography, 20-mg total protein extracts of veins were incubated in Laemmli buffer devoid of mercaptoethanol and resolved on 10% SDS-PAGE containing 0.1% gelatin. The gels were incubated at 37 Covernight in the developing buffer (50 mm Tris, ph 7.4, 0.2 M NaCl, 5 mm CaCl 2, 0.1% triton X-100) to allow for gelatin degradation. The gels were stained with 0.25% Coomassie blue (50% methanol, 10% acetic acid) and destained (50% methanol, 10% acetic acid). Recombinant human MMP-2 (0.5 ng; Calbiochem, Darmstasdt, Germany) and MMP-9 (0.25 ng; Calbiochem) were used as positive controls. The gels were scanned and intensities were determined with the National Institutes of Health 1.62 image analyzer software. Statistical analysis. All experiments were quantitatively analyzed, and results were shown as mean 6 standard error of the mean. Student t-test and two-way analysis of variance were performed to compare the mean values between groups, using the post hoc Bonferroni test, as provided by the Statistical Package for the Social Sciences (SPSS 17.0; Chicago, Ill). Statistical significance was set at P <.05. RESULTS Intimal hyperplasia is initiated by high pressure. In preliminary experiments (n ¼ 3) dedicated exclusively to morphometry analysis (at both low-pressure and highpressure conditions), no regional variations in the presence of hyperplasia were observed all along the human saphenous vein segment (data not shown). In another series of preliminary experiments (four veins), we observed no IH after 3 days of perfusion. Morphometric analysis showed that compared with the segments of freshly isolated veins (control), the segments of the same veins that were exposed for 7 days to arterial shear stress and high pressure underwent significant IH, despite stable thickness of the media layer (Fig 1). As a result, the intima-to-media ratio was increased (Fig 1). In contrast, no IH was

1374 Berard et al May 2013 Fig 1. Intimal hyperplasia develops within 7 days of ex vivo pulsatile perfusion. Upper panel, Representative histological sections show the normal intima thickness (arrow) in nonperfused veins (control) and veins perfused at low pressure (arterial shear stress [SS] 7 mm Hg). In contrast, intimal hyperplasia (arrow) is evident in veins perfused at high pressure (arterial SS 70 mm Hg). L, Lumen; M, media. Bar represents 50 mm. Lower panels, Measurements of intima thickness show that intimal hyperplasia was induced by perfusion at high pressure (70 mm Hg), despite unchanged thickness of the media layer. Data represent mean 6 standard error of the mean of two series of eight and nine experiments. *P <.05 and ***P <.001 vs the nonperfused segments (controls) of the same veins. observed in veins exposed to the same shear stress but lowpressure conditions (Fig 1). Staining with hematoxylin-eosin showed the proper outline of endothelial cell nuclei along the lumen of all vessels and the abnormal presence of SMC nuclei in the regions of IH (Fig 2). Immunostaining using antibodies against von Willebrand factor and alpha-smooth muscle cells further confirmed the lining of endothelial cells at the luminal surface of all veins and the SMCs in the media, whatever the experimental protocol (Fig 2). Numerous SMCs further colonized the intima at regions of IH (Fig 2). Immunolabeling using antibodies against KI-67, a marker specifically expressed by proliferating cells, revealed the presence of proliferating SMCs in the regions of IH (Fig 2). Each vessel segment was assessed for viability of both intima and media layers before and after perfusion

Volume 57, Number 5 Berard et al 1375 Fig 2. Human saphenous veins after 7 days of ex vivo perfusion show persistence of endothelial cells and altered distribution of smooth muscle cells (SMCs). Upper row, Hematoxylin-eosin staining (HE) reveals the lining of the lumen by nuclei of endothelial cells and the nuclei of SMCs in the media layer. Nuclei of the latter cells were also observed within the intimal hyperplasia regions (right panel). Middle rows, Immunolabeling using antibodies against von Willebrand factor (vwf) or alpha-smooth muscle actin (SMA) demonstrate the presence of endothelial cells at the luminal surface of all veins and of SMCs in the media of all vessels. SMCs were also detected in the intimal hyperplasia regions (right panels). Lower row, Immunolabeling using antibodies against KI-67 (dark blue) revealed the presence of proliferating SMCs mainly in the intimal hyperplasia regions. L, Lumen; M, media. Bars represent 100 mm in figures and 400 mm in the insets.

1376 Berard et al May 2013 Fig 3. The expression of matrix metalloproteinase (MMP)-2 and MMP-9 is increased by arterial shear stress (SS). A, MMP-2 transcripts are upregulated by arterial shear stress, and this upregulation is increased by high pressure (70 mm Hg). B, Western blots showed increased levels of both MMP-2 and MMP-9 after perfusion, under both low and high hemodynamic pressure. C and D, Both pro-mmp-2 and MMP-2 protein were significantly increased by perfusion and high pressure. E and F, MMP-9 transcript and protein levels were also upregulated by perfusion. However, the MMP-9 protein did not appear to be further induced by high pressure. Data represent mean 6 standard error of the mean of two series of eight and nine experiments. GADPH, Glyceraldehyde 3-phosphate dehydrogenase. *P <.05; **P <.01; ***P <.001 vs nonperfused veins and veins exposed to low pressure. and checked for changes in isometric tension induced by incubation in the presence of vasoactive drugs as previously described. 7,14 Expression of MMP-2 and MMP-9 is increased by perfusion. In the presence of low pressure, the expression of MMP-2 and MMP-9 increased at both transcript and protein levels, together with the activity of these metalloproteinases (Fig 3). Both MMP-2 and MMP-9 transcripts increased by the perfusion and further enhanced in the presence of high pressure (Figs 3 and 4). At the protein level, both pro-mmp-2 and MMP-2 forms increased in parallel with the transcript in veins perfused at low pressure (Fig 3). However, although the levels of the MMP-2 protein were increased further under high-pressure conditions, MMP-9 levels were not (Fig 3). Immunostaining further demonstrated that the expression of the two metalloproteinases was mostly increased in the media layer (Fig 4). Quantitative evaluation of the zymographic gels (Fig 4) showed that perfusion increased the lytic effect of MMP-2, but not pro-mmp-2, in the presence of both low and high pressure (Fig 4). MMP-9 was also upregulated after perfusion, whatever the hemodynamic pressure (Fig 4). Markers TIMP-1, TIMP-2, PAI-1, Ephrin-B2, and Eph-B4 are differentially modulated by the hemodynamic pressure. At both transcript and protein level, the expression of TIMP-1 was significantly enhanced by

Volume 57, Number 5 Berard et al 1377 Fig 4. The lytic activity of matrix metalloproteinase (MMP)-2 and MMP-9 is increased by arterial shear stress (SS) and blood pressure. A, Zymographic analysis distinguished the gelatin lysis induced by pro-mmp-2, MMP-2, and MMP-9. B-D, Quantitative analysis showed that the activities of these enzymes increased in veins exposed to arterial shear stress. However, only MMP-2 and MMP-9 were further enhanced by high pressure. Data represent mean 6 standard error of the mean of two series of eight and nine experiments. *P <.05; **P <.01; ***P <.001 vs control veins or veins submitted to low pressure. E, Immunostaining showed that both MMP-2 (top row) and MMP-9 (bottom row) increased mostly within the media layer of the perfused veins. L, Lumen; M, media. Bars represent 100 mm. perfusion of the veins but not modified further by high pressure (Fig 5). In contrast, the expression of the TIMP-2 transcript was downregulated after the 7-day-long perfusion, whatever the hemodynamic pressure (Fig 5). The levels of the cognate protein were not altered under the low-pressure regimen but were significantly increased after

1378 Berard et al May 2013 Fig 5. The inhibitors of tissue inhibitor of metalloproteinase (TIMP)-1 and TIMP-2 are differentially affected by arterial shear stress (SS) and pressure. A and B, Veins exposed to arterial shear stress featured increased expression of TIMP-1 mrna but decreased expression of the TIMP-2 transcripts. C, Consistent with the transcript change, Western blots showed upregulation of the TIMP-1 protein in the presence of both low and high pressure. D, In marked contrast to the mrna changes, the levels of the TIMP-2 protein were markedly increased only in the presence of high pressure. Data represent mean 6 standard error of the mean of two series of eight and nine experiments. *P <.05; **P <.01, ***P <.001 vs control veins. E, Immunostaining confirmed that the abundance of TIMP-1-positive and TIMP-2- positive smooth muscle cells increased under arterial hemodynamic conditions. GADPH, Glyceraldehyde 3-phosphate dehydrogenase; L, lumen; M, media; PAI, plasminogen activator inhibitor. Insets show the boxed areas at higher magnification. Bars represents 100 mm in figures and 400 mm in the insets.

Volume 57, Number 5 Berard et al 1379 Fig 6. The levels of plasminogen activator inhibitor (PAI)-1 increase in human saphenous vein exposed to arterial shear stress (SS). A, PAI-1 transcripts were increased in human saphenous veins submitted to arterial shear stress. B and C, Western blot analysis and immunocytochemistry showed that PAI-1 mostly increased in the media layer of the arterializing veins. GADPH, Glyceraldehyde 3-phosphate dehydrogenase; L, lumen; M, media. Bar represents 50 mm. Data represent mean 6 standard error of the mean of two series of eight and nine experiments. **P <.01; ***P <.001 vs control veins. exposure to high pressure (Fig 5). Immunocytochemistry further demonstrated an increase in TIMP-1 and TIMP-2 expression in the SMCs of the media (Fig 5). The expression of PAI-1 was elevated with the application of flow, without additional effect of high pressure (Fig 6). Exposure to perfusion was sufficient to downregulate Ephrin-B2 and Eph-B4 transcripts, whichever the pressure (Fig 7). The cognate Eph-B4 protein was decreased in all perfused veins (Fig 7). In contrast, the levels of Ephrin-B2 were significantly decreased only under high-pressure conditions (Fig 7). DISCUSSION Despite technological advances in endovascular approaches, bypass graft using saphenous veins still offers valuable clinical results at both midterm and long term 19 and retains a place of choice for vascular surgeons. However, graft failure remains the main limiting factor during the arterialization of the grafted veins, notably because of the development of aberrant IH. In this context, shear stress plays a pivotal role, especially in the regulation of vein caliber. 20 The aim of this study was to evaluate the effect of arterial pressure in the phenomenon of arterial remodeling encountered by a vein placed in arterial conditions because most of the models reproducing IH in human vessels so far have been developed under conditions of low shear stress. 10,11,15 These models partially reproduce the conditions encountered in zones of blood turbulence, such as anastomoses or valves, where the shear stress is altered, and the risk of IH and focal stenosis is great. The EVVSS model we tested cannot reproduce these local hemodynamic conditions. However, it specifically allows testing of the effect of arterial pressure in the presence of high shear stress (ie, the conditions that prevail in the arterial compartment). Using this model, we showed that under rigorously constant conditions of medium, flow rate, and shear stress, IH develops within 1 week when the mean hemodynamic pressure was that observed in the arterial but not the venous compartments

1380 Berard et al May 2013 Fig 7. The expression of Ephrin-B2 and Eph-B4 decreases in the presence of high pressure. A and B, The transcription of Ephrin-B2 and Eph-B4 decreased in veins submitted to arterial shear stress (SS). C and D, The levels of the cognate proteins were also reduced under these conditions. In vessels exposed to low pressure, only Eph-B4 was found significantly reduced. Data represent mean 6 standard error of the mean of two series of eight and nine experiments. GADPH, Glyceraldehyde 3-phosphate dehydrogenase. **P <.01; ***P <.001 vs control veins. of the circulatory system. These conditions, which reproduce the conditions a vein faces when grafted on an arterial bed, differ from those provided by previously published models, which provided different interpretations about the role of hemodynamic pressure 21-24 on the vessel wall. Dobrin et al 21 explored vascular remodeling in venous graft interposed in the femoral position in dogs after 3 months and found an increase in medial thickness with no influence on intimal thickness. Our ex vivo system also allows for compensation for the possible influence that different anatomical locations may have on the arterialization of grafted veins, providing a unique platform to explore the different factors that may contribute to modify the grafted vessels. 3,25 Remodeling requires the integrated effects of the fibrinolytic system, MMPs, and their inhibitors. We have shown that elevation of PAI-1 26 is induced by the arterial shear stress and, in contrast to IH, is not markedly influenced by the hemodynamic pressure. If PAI-1 promotes IH in response to vascular injury, 27-31 this factor is unlikely to be the primary event triggering the pressure-dependent IH remodeling of the vein wall under arterializing conditions. During this remodeling, degradation of the extracellular matrix by enzymes of the MMP family could conceivably facilitate the migration of SMCs from the media to the intima layer. 32 Consistent with this scenario, MMP-2 expression has been reported to increase in saphenous veins exposed to arterializing conditions analogous to the conditions we have tested. 33 We document that the arterial shear stress associated with elevated flow rate of perfusion is sufficient to increase the expression of both MMP-2 and MMP-9, via an upregulation of the transcription of the cognate genes. We further show that the high pressure, which plays a key role in initiating IH, further increases the production of both MMPs, resulting in a marked enhancement of their lytic activity with a more pronounced effect on the MMP-2 active form. This observation is consistent with the report that MMP-2 is the predominant metalloproteinase involved in the early phase of development of IH. 34 Our study confirms in human veins that compared with MMP-2, MMP-9 seems likely to have a lesser effect, given that its expression was not altered by high pressure. However, this difference could be explained by a difference in the temporal profile of MMP expression. TIMPs counterbalance the role played by MMPs in the degradation of many extracellular matrices. 35

Volume 57, Number 5 Berard et al 1381 TIMPs have been shown to be downregulated in veins grafted in various arterial compartments 36 but to increase with MMP-2 and MMP-9 in an ex vivo system of high flow. 36 We now document that arterial shear stress is sufficient to increase the expression of TIMP-1 and that a concomitant increase in hydrodynamic pressure is further required to activate the expression of TIMP-2. After 7 days of perfusion, the levels of TIMP-2 RNA were reduced, whereas the cognate protein was increased. This apparent discrepancy may be accounted for by a different temporal regulation profile of TIMP-2 transcript and protein, consistent with the findings in other models that tested experimental protocols different from the one we investigated. 34,37 These findings suggest that TIMP-2 plays a relevant role, together with MMP-2 and MMP-9, in initiating or sustaining IH, or both. The relationship between the TIMP-2 changes and IH is complex, as exemplified by the finding that IH is attenuated by exposure of vein grafts to the inhibitor. 38 A previous study indicated that adaptation of vein grafts to an arterial environment is also associated with loss of Eph-B4, a venous marker, but not with the gain of Ephrin-B2, an arterial marker. 39 Our ex vivo experiments extend these observations by showing that the combination of pulsatile flow and high pressure was insufficient to induce an extensive remodeling of all layers of the grafted vessel to replace an arterial segment fully. 40 We have also shown that the reduction of Eph-B4 expression that corresponds to the loss of venous identity is associated with the presence of a condition of arterial perfusion stress and that this phenomenon is independent of the hemodynamic pressure. The IH is not associated with a further decrease in Eph-B4 levels under conditions of high shear stress. These findings suggest the possibility that a pharmacological approach may be able to foster the extensive wall remodeling that would be required for veins to replace arterial segments fully. 11 It is recognized that the EVVSS model we used also features limitations. First, the paucity of the human material prevented us from perfusing vein segments under conditions notably low shear stress mimicking those of the venous compartment. Under such conditions, it is possible that vein segments may have developed alterations that were not detected in the freshly sampled, nonperfused controls we used. Second, based on a set of preliminary experiments, we tested the venous arterialization at a single time point (7 days of perfusion). This single time point prevents us from determining the precise time course of the alterations we document here, which our data suggest should start after 3 to 7 days of continuous perfusion. Third, the ex vivo conditions of the model prevent us from determining the concurrent participation of extravascular cell types, notably bone marrow derived cells. 6 Because we could not isolate sufficient amounts of endothelial cells and test venous segments devoid of the intima layer, our data do not definitively establish whether SMCs and/or endothelial cells are mainly responsible for the alterations of the venous wall during IH. A predominant increase of the former cell type is favored by the increased numbers of SMCs in the regions of IH and their altered expression in metalloproteases and their inhibitors. CONCLUSIONS Our data provide evidence that arterial shear stress and hydrodynamic pressure have distinct effects on the remodeling of the wall of human veins during the first week of arterialization, and we report the crucial role of high pressure on the development of IH. The data have begun elucidating some of the molecular participants of this complex process, whose expression is differentially regulated by various hemodynamic parameters. The ex vivo system we have developed is suitable to explore innovative, molecular-targeted strategies for the prevention of IH. We thank Drs Richard Daculsi, Martine Lambelet, and Jean-Christophe Stehle for their excellent technical assistance. AUTHOR CONTRIBUTIONS Conception and design: J-AH, XB, SD, LB, J-MC Analysis and interpretation: J-AH, XB, SD, PM, LB, J-MC Data collection: FA, FS, XB, SD Writing the article: J-AH, XB, SB, PM, J-AH Critical revision of the article: J-AH, SD, XB, J-MC, PM Final approval of the article: J-AH, XB, SD, PM_LB, J-MC, FA, FS Statistical analysis: FA, J-AH, XB Obtained funding: J-MC, J-AH, LB Overall responsibility: J-AH, XB, SD XB and SD contributed equally to this work, and J-MC and J-AH contributed equally to this work. REFERENCES 1. Lloyd-Jones D, Adams RJ, Brown TM, Carnethon M, Dai S, De Simone G, et al. Heart disease and stroke statistics 2010 update: a report from the American Heart Association. Circulation 2010;121: e46-215. 2. Norgren L, Hiatt WR, Dormandy JA, Nehler MR, Harris KA, Fowkes FG. Inter-Society Consensus for the Management of Peripheral Arterial Disease (TASC II). J Vasc Surg 2007;45(Suppl S):S5-67. 3. Owens CD. Adaptive changes in autogenous vein grafts for arterial reconstruction: clinical implications. J Vasc Surg 2010;51:736-46. 4. Davies MG, Hagen PO. Pathophysiology of vein graft failure: a review. Eur J Vasc Endovasc Surg 1995;9:7-18. 5. Wang X, Chai H, Lin PH, Lumsden AB, Yao Q, Chen C. Mouse models of neointimal hyperplasia: techniques and applications. Med Sci Monit 2006;12:RA177-85. 6. Zhang L, Freedman NJ, Brian L, Peppel K. Graft-extrinsic cells predominate in vein graft arterialization. Arterioscler Thromb Vasc Biol 2004;24:470-6. 7. Saucy F, Probst H, Alonso F, Berard X, Deglise S, Dunoyer-Geindre S, et al. Ex vivo pulsatile perfusion of human saphenous veins induces intimal hyperplasia and increased levels of the plasminogen activator inhibitor 1. Eur Surg Res 2010;45:50-9. 8. Sharony R, Pintucci G, Saunders PC, Grossi EA, Baumann FG, Galloway AC, et al. Matrix metalloproteinase expression in vein grafts: role of inflammatory mediators and extracellular signal-regulated kinases-1 and -2. Am J Physiol Heart Circ Physiol 2006;290:H1651-9.

1382 Berard et al May 2013 9. Eefting D, Seghers L, Grimbergen JM, de Vries MR, de Boer HC, Lardenoye JW, et al. A novel urokinase receptor-targeted inhibitor for plasmin and matrix metalloproteinases suppresses vein graft disease. Cardiovasc Res 2010;88:367-75. 10. Mitra AK, Gangahar DM, Agrawal DK. Cellular, molecular and immunological mechanisms in the pathophysiology of vein graft intimal hyperplasia. Immunol Cell Biol 2006;84:115-24. 11. Muto A, Yi T, Harrison KD, Davalos A, Fancher TT, Ziegler KR, et al. Eph-B4 prevents venous adaptive remodeling in the adult arterial environment. J Exp Med 2011;208:561-75. 12. Meyerson SL, Skelly CL, Curi MA, Shakur UM, Vosicky JE, Glagov S, et al. The effects of extremely low shear stress on cellular proliferation and neointimal thickening in the failing bypass graft. J Vasc Surg 2001;34:90-7. 13. Heise M, Kruger U, Ruckert R, Pfitzman R, Neuhaus P, Settmacher U. Correlation of intimal hyperplasia development and shear stress distribution at the distal end-side-anastomosis, in vitro study using particle image velocimetry. Eur J Vasc Endovasc Surg 2003;26:357-66. 14. Paroz A, Probst H, Saucy F, Mazzolai L, Rizzo E, Ris HB, et al. Comparison of morphological and functional alterations of human saphenous veins after seven and fourteen days of ex vivo perfusion. Eur Surg Res 2004;36:274-81. 15. Rey J, Probst H, Mazzolai L, Bosman FT, Pusztaszeri M, Stergiopulos N, et al. Comparative assessment of intimal hyperplasia development after 14 days in two different experimental settings: tissue culture versus ex vivo continuous perfusion of human saphenous vein. J Surg Res 2004;121:42-9. 16. Labadie RF, Antaki JF, Williams JL, Katyal S, Ligush J, Watkins SC, et al. Pulsatile perfusion system for ex vivo investigation of biochemical pathways in intact vascular tissue. Am J Physiol 1996;270(2 Pt 2): H760-8. 17. Alonso F, Krattinger N, Mazzolai L, Simon A, Waeber G, Meda P, et al. An angiotensin II- and NF-kappaB-dependent mechanism increases connexin 43 in murine arteries targeted by renin-dependent hypertension. Cardiovasc Res 2010;87:166-76. 18. Alonso F, Boittin FX, Beny JL, Haefliger JA. Loss of connexin40 is associated with decreased endothelium-dependent relaxations and enos levels in the mouse aorta. Am J Physiol Heart Circ Physiol 2010;299:H1365-73. 19. Siracuse JJ, Giles KA, Pomposelli FB, Hamdan AD, Wyers MC, Chaikof EL, et al. Results for primary bypass versus primary angioplasty/stent for intermittent claudication due to superficial femoral artery occlusive disease. J Vasc Surg 2012;55:1001-7. 20. Skelly CL, Meyerson SL, Curi MA, Loth F, Schwartz LB. The hemodynamics of vein grafts: measurement and meaning. Ann Vasc Surg 2001;15:110-22. 21. Dobrin PB, Littooy FN, Endean ED. Mechanical factors predisposing to intimal hyperplasia and medial thickening in autogenous vein grafts. Surgery 1989;105:393-400. 22. Geary RL, Williams JK, Golden D, Brown DG, Benjamin ME, Adams MR. Time course of cellular proliferation, intimal hyperplasia, and remodeling following angioplasty in monkeys with established atherosclerosis: a nonhuman primate model of restenosis. Arterioscler Thromb Vasc Biol 1996;16:34-43. 23. Hui DY. Intimal hyperplasia in murine models. Curr Drug Targets 2008;9:251-60. 24. Zhang LN, Parkinson JF, Haskell C, Wang YX. Mechanisms of intimal hyperplasia learned from a murine carotid artery ligation model. Curr Vasc Pharmacol 2008;6:37-43. 25. Zilla P, Moodley L, Scherman J, Krynauw H, Kortsmit J, Human P, et al. Remodeling leads to distinctly more intimal hyperplasia in coronary than in infrainguinal vein grafts. J Vasc Surg 2012;55:1734-41. 26. Ha H, Oh EY, Lee HB. The role of plasminogen activator inhibitor 1 in renal and cardiovascular diseases. Nat Rev Nephrol 2009;5:203-11. 27. Hasenstab D, Lea H, Clowes AW. Local plasminogen activator inhibitor type 1 overexpression in rat carotid artery enhances thrombosis and endothelial regeneration while inhibiting intimal thickening. Arterioscler Thromb Vasc Biol 2000;20:853-9. 28. Hasenstab D, Lea H, Hart CE, Lok S, Clowes AW. Tissue factor overexpression in rat arterial neointima models thrombosis and progression of advanced atherosclerosis. Circulation 2000;101:2651-7. 29. DeYoung MB, Tom C, Dichek DA. Plasminogen activator inhibitor type 1 increases neointima formation in balloon-injured rat carotid arteries. Circulation 2001;104:1972-81. 30. Diebold I, Kraicun D, Bonello S, Gorlach A. The PAI-1 paradox in vascular remodeling. Thromb Haemost 2008;100:984-91. 31. Ji Y, Strawn TL, Grunz EA, Stevenson MJ, Lohman AW, Lawrence DA, et al. Multifaceted role of plasminogen activator inhibitor-1 in regulating early remodeling of vein bypass grafts. Arterioscler Thromb Vasc Biol 2011;31:1781-7. 32. Muto A, Model L, Ziegler K, Eghbalieh SD, Dardik A. Mechanisms of vein graft adaptation to the arterial circulation: insights into the neointimal algorithm and management strategies. Circ J 2010;74:1501-12. 33. Dummler S, Eichhorn S, Tesche C, Schreiber U, Voss B, Deutsch MA, et al. Pulsatile ex vivo perfusion of human saphenous vein grafts under controlled pressure conditions increases MMP-2 expression. Biomed Eng Online 2011;10:62. 34. Berceli SA, Jiang Z, Klingman NV, Pfahnl CL, Abouhamze ZS, Frase CD, et al. Differential expression and activity of matrix metalloproteinases during flow-modulated vein graft remodeling. J Vasc Surg 2004;39:1084-90. 35. Raffetto JD, Qiao X, Koledova VV, Khalil RA. Prolonged increases in vein wall tension increase matrix metalloproteinases and decrease constriction in rat vena cava: potential implications in varicose veins. J Vasc Surg 2008;48:447-56. 36. Wong AP, Nili N, Jackson ZS, Qiang B, Leong-Poi H, Jaffe R, et al. Expansive remodeling in venous bypass grafts: novel implications for vein graft disease. Atherosclerosis 2008;196:580-9. 37. Hasenstab D, Forough R, Clowes AW. Plasminogen activator inhibitor type 1 and tissue inhibitor of metalloproteinases-2 increase after arterial injury in rats. Circ Res 1997;80:490-6. 38. Hu Y, Baker AH, Zou Y, Newby AC, Xu Q. Local gene transfer of tissue inhibitor of metalloproteinase-2 influences vein graft remodeling in a mouse model. Arterioscler Thromb Vasc Biol 2001;21:1275-80. 39. Kudo FA, Muto A, Maloney SP, Pimiento JM, Bergaya S, Fitzgerald TN, et al. Venous identity is lost but arterial identity is not gained during vein graft adaptation. Arterioscler Thromb Vasc Biol 2007;27:1562-71. 40. Germain S, Eichmann A. VEGF and ephrin-b2: a bloody duo. Nat Med 2010;16:752-4. Submitted May 30, 2012; accepted Sep 10, 2012.