| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Eur J Cardiothorac Surg 2006;30:451-463
© 2006 Elsevier Science NL
Review |
Department of Cardiac Surgery, Innsbruck Medical University, Innsbruck, Austria
Received 30 September 2005; received in revised form 2 May 2006; accepted 11 June 2006.
* Corresponding author. Address: Anichstrasse 35, 6020 Innsbruck, Austria. Tel.: +43 5125040; fax: +43 51250422528. (Email: Thomas.Schachner{at}uibk.ac.at).
 |
Abstract |
|---|
 Top Abstract 1. Clinical and...2. General considerations on...3. Pig models (saphenous...4. Canine models (jugular...5. Rabbit model (jugular...6. Rat models (inferior...7. Mouse models (vena...8. Other models of...9. ConclusionReferences |
|---|
Abbreviations: ACE = angiotensin converting enzyme • apoE = apolipoprotein E • bFGF = basic fibroblast growth factor • BrdU = 5-bromo-2’-deoxyuridine • CABG = coronary artery bypass graft • cAMP = cyclic adenosine monophosphate • cdc2 kinase = cell division cycle 2 kinase • cGMP = cyclic guanosine monophosphate • CNP = C-type natriuretic peptide • ET = endothelin • ERK-1/2 = extracellular signal-regulated kinase • G-protein = guanine nucleotide binding protein • HVJ = hemagglutinating virus of Japan • ICAM-1 = intercellular adhesion molecule-1 • IL = interleukin • IGF-1 = insulin-like growth factor 1 • ITA = internal thoracic artery • JNK = c-jun N-terminal kinases • LTB4 = leukotriene B4 • MAPK = mitogen-activated protein kinases • MMP = matrix metalloproteinase • mRNA = messenger ribonucleic acid • NOS = nitric oxide synthase • ODN = oligodeoxynucleotide • PCNA = proliferating cell nuclear antigen • PCR = polymerase chain reaction • PDGF = platelet derived growth factor • PGI2 = prostacyclin • PKC = protein kinase C • RCA = right coronary artery • RITA = right internal thoracic artery • Sca-1 = stem cell antigen-1 • Sdi-1 = senescent cell-derived inhibitor 1 (=p21) • SMC = smooth muscle cell • SOD = superoxide dismutase • TEM = transmission electron microscopy • TIMP = tissue inhibitor of MMP • TUNEL = terminal deoxynucleotidyl transferase (TdT)-mediated dUTP nick end labelling • VCAM = vascular cell adhesion molecule • VSMC = vascular smooth muscle cell
Key Words: Vein graft • CABG • Atherosclerosis • Neointima • Neointimal hyperplasia • Bypass
|
1. Clinical and pathophysiological background of vein graft disease |
|---|
|
TopAbstract1. Clinical and...2. General considerations on...3. Pig models (saphenous...4. Canine models (jugular...5. Rabbit model (jugular...6. Rat models (inferior...7. Mouse models (vena...8. Other models of...9. ConclusionReferences |
|---|
Following arterialization vein grafts undergo immediate injury (surgical trauma, ischemia, wall stress) which can cause early graft occlusion/thrombosis in a small percentage of grafted veins. During the next several weeks a typical feature of vein grafts is the development of neointimal hyperplasia, accompanied by early (days) medial SMC loss and consecutive increase of total vascular wall thickness (weeks-months).
Neointimal hyperplasia results from various steps (see Fig. 1 ):
|
|
2. General considerations on methods to study vein graft disease (in vivo models vs in vitro models, large animals vs small animals) |
|---|
|
TopAbstract1. Clinical and...2. General considerations on...3. Pig models (saphenous...4. Canine models (jugular...5. Rabbit model (jugular...6. Rat models (inferior...7. Mouse models (vena...8. Other models of...9. ConclusionReferences |
|---|
In vitro models expand the knowledge about important cell types such as vascular smooth muscle cells in cell culture. The use of organ culture offers the possibility to study human saphenous veins and to study the culture medium reflecting the secretory activity of vein grafts (e.g., the release of growth factors).
Animal models are useful for both studying the pathology of vein graft disease, and to test therapeutic strategies in vivo. Early changes in arterialized veins including thrombosis, inflammatory cell migration, and neointimal hyperplasia can be studied in all of the presented animal models. The process of late vein graft atherosclerosis can partially be simulated by hypercholesterolemic animal models (e.g., cholesterol fed rabbits, ApoE knockout mouse).
This review focuses on animal models of vein graft disease. Neither the field of arterial injury models nor the ex vivo models of vein graft disease are content of this manuscript. The pathological features of vein graft disease and their therapeutic attempts are sorted by the types of animals they were studied in.
The advantage of large animal models is the more human like thick wall of the vein graft compared with the thin wall of rodent's veins. Furthermore, special therapeutic strategies such as external stenting of vein grafts can be best performed in large animals. Functional studies in the organ bath can easily be carried out on veins implanted in the large animal. It is an important fact that animals might have distinct differences in their biochemistry. For example, rat vascular tissues contain ACE only as an angiotensin II-forming enzyme, while vascular tissues of human, monkey, and dog contain chymase in addition to ACE as angiotensin II-forming enzymes. Therefore, testing an ACE inhibitor as vascular antiproliferative agent might be successful in the rat but not in human or dogs due to their chymase-dependent angiotensin II formation [13].
On the other hand, small animal models are cheaper, and proper animal care (including anesthesiological technique) is easier available compared with pigs or dogs. In addition, expensive experimental therapeutic agents such as viral vectors for gene therapy can be applied in a more economic fashion. A unique feature of mouse models is the availability of knock out mice, which offer the advantage of creating distinct pathophysiological conditions.
|
3. Pig models (saphenous vein or jugular vein interposition into carotid artery) |
|---|
|
TopAbstract1. Clinical and...2. General considerations on...3. Pig models (saphenous...4. Canine models (jugular...5. Rabbit model (jugular...6. Rat models (inferior...7. Mouse models (vena...8. Other models of...9. ConclusionReferences |
|---|
|
In the same study with the use of reverse-transcription PCR, higher mRNA levels for PDGF B chain were observed in vein grafts than in ungrafted veins [14].
Dashwood et al. found radioligand binding to endothelin A (ET A) receptors in the neointima and media of vein grafts 4 weeks after implantation. In the same vein grafts ET B receptors were found to a lesser extent in the media but more frequently in the adventitial microvasculature. Interestingly, ET A receptors were abundantly present in ungrafted porcine saphenous veins (in a higher concentration than in vein grafts and adjacent carotid artery segments). This is thought to make the vein susceptive for mitogenic effects of ET 1 (which is increased after mechanical injury and increased wall stress of a vessel) to VSMC after arterial grafting [15].
Early media damage and the role of periadventitial (myo-) fibroblasts were investigated in the following studies:
O’Brien et al. [16] found at 7 to 14 days postoperatively, hypercellular perivascular regions which stained positive for 
-SM actin. At this timepoints one third of periadventitial cells were positive for PCNA.
This group also found a loss of medial SMC at 3–4 days postoperatively. The early SMC loss in the media was due at least in part to apoptosis as demonstrated in another study [17]. At 14 days postoperatively 21% of medial cells were PCNA positive. At 3 months postoperatively all layers of the vein graft were homogenously rich in collagenous tissue (Verhoeff's stain).
Rodriguez et al. [18] contributed a 15% loss of medial SMC 8 h after vein grafting to apoptosis. In the same study medial non-SMC apoptosis also peaked at 8 h postoperatively and was still increased 48 h after grafting whereas for medial SMC the apoptotic rate decreased again 24 h after the arterialization.
Shi et al. performed a two-stage operation where they first marked fibroblasts in the healing wound (after carotid artery exposure) with BrdU and 2 days later did the vein graft procedure. They could demonstrate labeled fibroblasts migrating from periadventitially into the vein graft wall [19].
Neural remodeling in vein grafts was observed by Dashwood et al. who found nerve fibers in the outer media of ungrafted veins, which disappeared after 4 weeks of arterialization. At 4 weeks postoperatively in vein grafts periadventitially many nerves were found which increased in number until 6 months postoperatively. In arterial segments nerves are found primarily periadventitially [20].
The endothelial protective substances PGI2, cAMP, and cGMP (cAMP and cGMP are second messengers for PGI2 and NO) are reduced in veins early after grafting.
Jeremy et al. found an up-regulation of NOS in the endothelium of vein grafts relative to ungrafted saphenous veins. But the generation of the second messengers camp, and cGMP following stimulation with sodium nitroprussid, and prostacyclin PGI2, respectively, was significantly lower in vein grafts 4 weeks postoperatively as compared to native veins and the carotid artery [21].
In another study, Jeremy et al. showed a decreased production of PGI2 (which is an inhibitor of vascular smooth muscle cell proliferation) in vein grafts 4 weeks postoperatively compared to ungrafted veins and the carotid artery. In contrast, this study revealed an increased leukotriene LTB4 synthesis (which may have been derived from infiltrating leukocytes and exerts promitogenic effects on smooth muscle cells) in vein grafts [22].
Matrix metalloproteinases ease SMC migration and proliferative responses to growth factors by proteolytic remodeling of the basement membrane. In 2 days of organ culture, segments of veins 1–4 weeks after bypass grafting produced significantly more (pro-) matrixmetalloproteinases (pro-MMP9, pro-MMP2, active MMP2) compared with ungrafted veins as measured by zymography. Immunohistochemically MMP2 was found in the neointima 1–4 weeks postoperatively. MMP9 was localized in the neointima and media starting from day 2 (and increased until day 28) after bypass grafting [23].
3.2 Therapeutic strategies (ET 1A antagonist, external stenting, brachytherapy, c-myc antisense, iNOS + TIMP-3 gene transfer)
Endothelin 1A receptor antagonism with BSF 302146 that was administered orally 4 weeks after bypass operation significantly reduced the neointimal thickness (Table 1
) [24].
|
Angelini et al. [26] saw that 1.5% of the neointimal cells were foam cells and that 53% of neointimal (especially endothelial) cells were VCAM-1 positive in vein grafts of hypercholesterolemic pigs at 12 weeks postoperatively. In externally (polyester) stented vein grafts they found a reduction of neointimal thickness, as well as a decreased number of foam cells and VCAM-1 amount.
Bartels et al. found no benefit of brachytherapy of veins prior to implantation. There was even a smaller lumen diameter and a higher graft occlusion rate present in veins (at 3 months postoperatively) irradiated with 40 Gray [29].
Treatment of veins with antisense oligonucleotides complementary to the mRNA of c-myc (in saline solution bath before implantation and additional perivascular application after implantation of the vein) showed a significant reduction of neointimal thickness at 3 months postoperatively [30].
Adenovirus-mediated gene transfer of human inducible NOS (incubated for 30 min) in pig jugular vein grafts showed a reduction of neointimal hyperplasia, at 3 weeks postoperatively, that was more prominent in the distal parts of the graft than in the proximal (upstream) parts [31].
Adenovirus mediated gene transfer of TIMP-3 (30 min incubation without distension) into porcine saphenous veins reduced neointimal hyperplasia and was associated with an elevated neointimal and medial cell apoptosis compared with controls [32].
|
4. Canine models (jugular vein into carotid artery interposition, saphenous vein into coronary artery bypass graft, femoral vein into femoral artery, jugular vein into femoral artery interposition) |
|---|
|
TopAbstract1. Clinical and...2. General considerations on...3. Pig models (saphenous...4. Canine models (jugular...5. Rabbit model (jugular...6. Rat models (inferior...7. Mouse models (vena...8. Other models of...9. ConclusionReferences |
|---|
4.1 Pathology (MAPKs, ACE, PGI2)
Mitogen-activated protein kinases (MAPKs) play a key role as the ultimate mediators of extracellular stimuli between the cytoplasm and the nucleus of a cell. Saunders et al. found out that in dog jugular vein grafts ERK-1/2 and p38MAPK were activated at 30 min, 3 h, and 4 days postoperatively compared with contralateral ungrafted jugular veins (Harvest time points: 30 min; 3, 8, and 24 h; and 4, 7, 14, and 28 days). In contrast, JNK was not activated in vein grafts at this study [33].
Yuda et al. [13] demonstrated a 10-fold chymase activity and a doubled ACE activity in arterialized veins compared with control veins at 4 weeks postoperatively.
The vascular biology of a vein graft is flow dependent. Onohara et al. demonstrated a decreased stimulated (with arachidonic acid) PGI2 production in vein grafts with a poor runoff and low shear stress compared with veins grafted into a good runoff area. Furthermore, vein grafts with poor distal runoff showed an increased (
doubled) neointimal hyperplasia compared with control vein grafts [34,35].
4.2 Therapeutic strategies (angiotensin II antagonist, Cobalt-60 radiation, NF kappa B decoy, endothelial cell NOS (ecNOS) gene transfer)
Yuda et al. [13] found out that the angiotensin II antagonist L-158,809 (10 mg/kg per day administered orally 1 week before and 4 weeks after operation) led to a reduced neointimal formation in dog jugular vein grafts. The same group found that the single local incubation of veins with the specific chymase inhibitor Suc-Val-Pro-Phe-(OPh)2 for 20 min inhibited neointimal formation and total angiotensin II-forming activity until 3 months after operation [36].
In contrast to the report of Bartels et al., Ulus et al. [37] demonstrated a reduction of neointimal hyperplasia (at 6 weeks postoperatively) following Cobalt-60 radiation (14 Gy) of dog jugular veins prior to interposition into femoral arteries.
Petrofski et al. demonstrated an effective (although in a patchy pattern) adenovirus mediated beta galactosidase and beta-adrenergic receptor kinase carboxyl terminus gene transfection after exposure of saphenous veins to the virus solution endoluminally for 20 min at only 10 mmHg (Harvest time 7 days postoperatively). By using this approach a traumatic distension of the vein could be avoided [38].
The nuclear transcription factor NF kappa B, upon external stimulation of the cell, modulates the expression of genes for chemotactic factors, cell cycle regulatory proteins, and adhesion molecules. Shintani et al. incubated saphenous veins prior to CABG in dogs for 20 min in a pressurized syringe (at a maximum of 200 mmHg from lumenally and ablumenally) in a NF kappa B decoy ODN solution (40 µmol/l). At 4 weeks after CABG, neointimal hype6rplasia as well as the PCNA amount was reduced in NF kappa B decoy ODN-treated vein grafts compared with controls [39].
Gene transfer of endothelial cell NOS using HVJ-liposomes (10 min incubation at 100 mmHg) in dogs with poor distal runoff limbs led to a significantly decreased neointimal hyperplasia at 4 weeks postoperatively [40].
|
5. Rabbit model (jugular vein into carotid artery interposition) |
|---|
|
TopAbstract1. Clinical and...2. General considerations on...3. Pig models (saphenous...4. Canine models (jugular...5. Rabbit model (jugular...6. Rat models (inferior...7. Mouse models (vena...8. Other models of...9. ConclusionReferences |
|---|
5.1 Pathology (contractile sensitivity, mast cells, hypercholesterolemia, LOX-1, IL-1beta, IL-10, thrombomodulin, G proteins, endothelial injury, reversed vein grafts)
In vitro isometric tension studies showed that rabbit vein grafts (2 weeks postoperatively) exhibited an increased contractile sensitivity to norepinephrine and serotonin, which decreased in reversed vein grafts (reimplanted into venous circulation for 2 weeks). Interestingly both, vein grafts, and reversed vein grafts showed a decreased sensitivity to histamine compared with ungrafted veins. Vein grafts showed no (endothelium-dependent) relaxation to acetylcholine whereas ungrafted veins and reversed vein grafts did relax following acetylcholine exposure [41].
Mast cells are an important source of vasocontractile and chemotactic substances such as histamine, leukotriene C4, and platelet activating factor. Cross et al. found the presence of mast cells in rabbit vein grafts at 4 weeks postoperatively, whereas in ungrafted veins no mast cells were found. Vein grafts showed a contractile response to histamine which was abolished by treatment with the H1 receptorantagonist pyrilamine, but not by treatment with the H2 receptor antagonist cimetidine [42].
Brauner et al. [43] found that 4 weeks postoperatively vein grafts of hypercholesterolemic rabbits (1% cholesterol-supplemented diet started 28 days before the operation and continued postoperatively) exhibited a significantly increased (contractile) sensitivity to serotonin and decreased (vasodilatory) sensitivity to sodium nitroprusside compared with vein grafts of normocholesterinemic rabbits.
In hypercholesterolemic rabbits (serum cholesterol 1898 mg/dl at harvest time) at 4 weeks after operation, neointimal hyperplasia was markedly increased compared with normocholesterolemic rabbits (serum cholesterol 88 mg/dl at harvest time). Moreover, in vein grafts of hypercholesterolemic animals endothelial cells and SMCs contained lipid vacuoles (as determined by TEM), and foam cells that were interspersed between the SMCs [44].
Similar morphologic changes (abundantly lipid vacuoles and lipid-laden macrophages) could be demonstrated by Davies et al. [45] in hypercholesterolemic rabbits.
In contrast, Ge et al. found no significant increase of neointimal thickness in vein grafts of hypercholesterolemic (serum cholesterol 22 mmol/l) rabbits at 12 weeks postoperatively compared with normocholesterolemic animals. Lectin-like oxidized low-density lipoprotein receptor-1 (LOX-1) is a scavenger receptor present on endothelial cells and expression of LOX-1 is increased in the neointima of hypercholesterolemic rabbits compared with normocholesterolemic animals [46].
In low flow vein grafts (90% flow reduction by distal ligation of the internal carotid artery and the first branches of the external carotid artery), neointimal thickness was markedly higher than in high flow vein grafts. The mRNA expression of the proinflammatory IL-1beta was higher in low flow grafts (and peaked at 1 day postoperatively), whereas the expression of the antiinflammatory IL-10 was higher in high flow grafts (and peaked at 14 days postoperatively) [47,48].
Thrombomodulin, by binding thrombin (which thereby loses its fibrinogen cleaving capacity and gains protein C activating ability), is important for the endothelial thromboresistance. Kim et al. [49] could demonstrate a steep decrease of thrombomodulin amounts (below 10% of the amount of ungrafted veins within the first 2 weeks postoperatively) in veins early after arterialization as determined by Western blot.
Davies et al. [50] found a continuous increase of G protein alpha subunits and G protein beta subunits during 4 weeks after vein graft arterialization.
Rabbit jugular vein grafts (4 weeks postoperatively) contracted concentration dependently following exposure to the 1 agonist norepinephrine, which was inhibited by the 1b antagonist chloroethylclonidine but not by the 1a antagonist 5-methylurapidil. The adrenergic contraction of vein grafts showed no response to pertussis toxin (an inhibitor of Gi/Go). Hence, vein grafts appeared to have functional 1B receptors that were associated with pertussis toxin insensitive G proteins [51].
In another study, Davies et al. showed that vein grafts 4 weeks after implantation show no relaxation to acetylcholine (which requires endogenous NO). However, the same grafts did relax to (the non-endothelium-dependent) sodium nitroprusside [52].
Another interesting vasoreactive alteration is the decreased relaxation to dopamine in vein grafts compared with ungrafted veins [53].
An increase of neointimal formation in balloon-injured vein grafts (three passes with a Fogarty catheter) could be demonstrated by Davies et al. At 4 weeks postoperatively the endothelially injured vein grafts showed an increased sensitivity to the vasoconstrictors norepinehrine, serotonin, bradykinin, and angiotesin II. The sensitivity of injured vein grafts to the vasodilator sodium nitroprusside was also increased, but with a decreased maximum relaxation [54].
In a model o6f reversed vein grafts that were reimplanted into venous circulation after 2 weeks of arterialization the following features were found: Within 2 weeks in the veno-venous circulation the neointimal SMCs showed apoptosis like morphology with cellular fragmentation (electron microscopy). Furthermore, the endothelial cells regained a sharp and well-defined border. Reversed vein grafts showed a decrease in neointimal and medial thickness [55].
5.2 Therapeutic strategies (L-arginin, verapamil, tyrphostin AG51, antioxidants, monoterpenes, captopril, angiotensin II receptor antagonism, E2F decoy, tissue factor antibody, thrombomodulin gene transfer, PCNA and cdc2 kinase antisense, PCNA and cdc2 kinase antisense, elafin gene transfer, Sdi-1 gene transfer)
Vein grafts instilled for 15 min with 100 µmol/l L-arginin showed increased NO levels (by increasing the substrate for NOS) and at 4 weeks postoperatively they exhibited a reduced neointimal formation [56]. A similar attenuation of neointimal hyperplasia could be obtained by perivascular application of Spermine/NO as NO donor. In the same study, a decreased amount of IGF-1 was found in NO treated vein grafts compared with controls [57].
In hypercholesterolemic rabbits, L-arginine (2 g/kg, p.o.) treatment showed a 24% reduction of neointimal thickness [58].
Calcium is implicated in a number of events including platelet aggregation, PDGF release, and SMC proliferation. Based on the fact that calcium-channel blockers have been reported to interfere with PDGF induced VSMC proliferation and migration in vitro verapamil was successfully applied in a periadventitial matrix (ethylene-vinyl acetate). In this study, hypercholesterolemia significantly increased vein graft sensitivity to serotonin, which was balanced by the verapamil application. In addition, neointimal hyperplasia was reduced by the verapamil-releasing matrix [43].
In an earlier study, el-Sanadiki et al. observed a significant reduction of neointimal hyperplasia in vein grafts of rabbits treated systemically with verapamil (1.25 mg per day i.v.) at 4 weeks postoperatively compared with controls. However, the vein grafts of verapamil treated animals showed an increased contractile sensitivity to norepinephrine and histamine [59].
Receptors for EGF, PDGF, and bFGF exhibit an intrinsic ligand-stimulated tyrosine kinase activity. Local treatment (ex vivo incubation and perivascular application with pluronic gel) of rabbit vein grafts with tyrphostin AG51, a tyrosine kinase inhibitor, led to an almost 50% reduction of neointimal thickness at 4 weeks postoperatively. Interestingly, tyrphostin-treated vein grafts showed no contractile response to norepinephrine, whereas control vein grafts did contract [60].
Treatment of rabbits with methylaminochroman, a second generation lazaroid with antioxidant properties (10 mg/kg per day p.o.) showed an increased preservation of vasorelaxation following acetylcholin and sodium nitroprusside exposure at 4 weeks postoperatively compared with controls. Morphologically the vein grafts of methylaminochroman-treated animals exhibited a reduced neointimal hyperplasia compared with controls [61].
Malondialdehyde, a marker for lipid peroxidation was more than fivefold increased in vein grafts at 3 days postoperatively compared with ungrafted veins. Antioxidative treatment with polyethylene glycolated SOD (vein immersed in PEG-SOD for 15 min and perivascular application in pluronic gel) reduced both malondialdehyde concentrations in grafts at 3 days postoperatively, and neointimal hyperplasia at 4 weeks postoperatively. In this study, however, the dose-dependent contractile responses to norepinephrine, serotonine, and bradykinin were not different between SOD-treated vein grafts and controls [62].
Monoterpenes exert a cytostatic effect by inhibiting posttranslational isoprenylation of p21ras and other small G proteins. Perillyl alcohol (200 mg/kg per day p.o.) treatment led to a 22% reduction of neointimal thickness at 4 weeks postoperatively compared with controls. Furthermore, the vein grafts of perillyl alcohol-treated animals showed an increased vasoconstrictory response to norepinephrine, serotonine, bradykinin, and histamine and an increased vasodilatory response to acetylcholin compared with controls [63].
Systemic treatment of rabbits with the ACE inhibitor captopril (10 mg/kg per day p.o.) reduced neointimal hyperplasia by 40% at 4 weeks postoperatively. Furthermore, vein grafts of captopril-treated animals showed an increased maximal contractile response to angiotensin I and serotonin exposure compared with controls. However, the maximal contractile response to norepinephrine, histamine, and angiotensin II was not different between the groups [64].
Fulton et al. could demonstrate different responses of rabbit vein grafts following either systemic treatment (10 mg/kg per day p.o., started 5 days before operation and continued until harvest) or local treatment (pluronic gel) with the angiotensin II receptor antagonist L158809. Local treatment of vein grafts with L158809-reduced neointimal thickness by 33%, whereas the medial thickness was not different from controls at 4 weeks postoperatively. Systemic treatment with L158809 led to a 43% decrease of neointimal thickness and to a 44% reduction of medial thickness compared with controls. There was a contractile response to angiotensin of controls and locally treated vein grafts (at 4 weeks postoperatively), whereas there was no contractile reaction following angiotensin stimulation in systemically L158809-treated vein grafts [65].
One way to inhibit the vascular proliferative response is to inhibit cell cycle progression. This has been done by incubating veins prior to implantation (20 min in a non-distendable sheath using 300 mmHg of pressure on the vein from all sides) with a decoy ODN to E2F transcription factor. By using this method E2F decoy ODN-treated grafts showed an attenuated proliferative activity (as expressed by amounts of PCNA and BrdU) and a reduced neointimal thickness both at 6 weeks and at 6 months postoperatively. In the same study by Ehsan et al. [66] in hypercholesterolemic rabbits lipid-rich (Oil red O staining) lesions in the neointima occurred in the long term (6 months), whereas in E2F decoy ODN-treated vein grafts these atherosclerotic lesions were almost absent.
Tissue factor is a transmembrane glycoprotein that binds to and activates factor VII leading to an increased thrombin production. Channon et al. found an increased amount of tissue factor, which was spatially associated with CD18 positive leukocytes that infiltrated the entire vessel wall 1–3 days postoperatively. In contrast, there was no increased tissue factor amount in the neointima of vein grafts at 2 and 4 weeks postoperatively, which showed an intact endothelial lining [67]. The same group, however, noticed no benefit of locally applied antitissue factor antibody (in pluronic gel) regarding the development of neointimal hyperplasia at 4 weeks postoperatively [68].
Kim et al. treated veins by adenovirus mediated thrombomodulin gene transfer (for 1 h under ‘gentle’ distension pressure) and demonstrated an increase of thrombomodulin expression that peaked at 7 days after transfection. However, no reduction of neointimal hyperplasia was observed with this treatment [49].
The development of neointimal hyperplasia is associated with an elevation of G proteins (which dissociate to the G alpha and G beta/gamma subunits and lead to cellular signaling events). Davies et al. [69] demonstrated a reduction of intimal hypeplasia in rabbit vein grafts that were incubated (for 30 min before grafting) with a plasmid which contained cDNA encoding amino acid residues of bovine beta-adrenergic receptor kinase-1 that inhibited the beta/gamma subunit of G proteins.
The same group demonstrated the reduction of neointimal hyperplasia in adenovirus-mediated gene transfer of beta-adrenergic receptor kinase-1 [70].
Another effection site of desensitizing G protein-mediated signaling (and PDGF receptor beta mediated signaling) was chosen by Peppel et al. [71]: Adenovirus mediated G protein-coupled receptor kinase-2 (GRK-2) gene transfer (20 min incubation without pressure) reduced both neointimal thickness and PCNA amount in rabbit vein grafts.
Fulton et al. [72] demonstrated a reduction of neointimal thickness of 26% at 4 weeks postoperatively in rabbit vein grafts treated perivascularly with PCNA antisense ODN (mixed with pluronic gel).
Mann et al. performed a PCNA and cdc2 kinase antisense ODN gene transfer (endoluminal application, 20 min at 100 mmHg) into rabbit jugular veins using HVJ-liposomes. In the transfected vein grafts, PCNA and cdc2 kinase protein levels were reduced compared with controls. Furthermore, an arterial remodeling was found with little neointimal hyperplasia but increased medial thickness. In the same study in control hypercholesterolemic rabbits accumulations of macrophages and foam cells were frequently found (6 weeks postoperatively), which were only occasionally present in PCNA + cdc2 kinase antisense ODN-treated veins [73].
Yamashita et al. performed an adenovirus-mediated antisense bFGF gene transfer (30 min at 120 mmHg from endoluminally), which led to a 40% reduced bFGF protein amount at 1 week postoperatively (and 10% reduction at 3 weeks postoperatively). Furthermore, neointimal hyperplasia was significantly reduced. This study nicely demonstrated the activation of MAPK: ERK1/2, JNK1, and p38 were doubled at 4–7 days postoperatively compared with ungrafted veins, and this increase was still around 1.5-fold at 3 weeks postoperatively [74].
The HVJ-liposome mediated endothelial cell NOS gene transfer (10 min at 70 mmHg from endoluminally) into jugular veins of hypercholesterolemic rabbits showed a 30% reduction of neointimal thickness. The cGMP content in ecNOS-transfected vein grafts was increased compared with (dilated!) control vein grafts but still only about a quarter of the amount of ungrafted veins. The authors addressed this phenomenon to an early SMC loss in the inner media, which resulted in a loss of ecNOS expression. A rarefication of NOS protein in the inner media of vein grafts at 4 days postoperatively supported this hypothesis [75].
Elastases (originating from inflammatory cells or SMCs) act proinflammatory and proliferative to SMCs by activating proforms of cytokines, degrading components of the endothelial basement membrane (facilitating inflammatory cell migration into the vessel wall), and releasing and activating growth factors that are bound to the extracellular matrix. ÓBlenes et al. performed a HVJ-liposome mediated elafin (serine elastase inhibitor) gene transfer (flushed and incubated 10 min without pressure from luminal and adventitial side) into rabbit jugular veins. At 2 days postoperatively elastase activity in vein grafts was fivefold compared with ungrafted veins and could be significantly attenuated in elafin transfected vein grafts. At 2 days postoperatively a significantly decreased amount of inflammatory cells was found in the wall of elafin transfected veingrafts, and at 4 weeks postoperatively the neointimal formation was reduced by 50% compared with controls [76].
Bai et al. [77] transfected rabbit veins with Sdi-1 (which inhibits cell cycle progression by inactivation of cyclin–cyclin-dependent kinase complexes) using HVJ-liposomes (20 min at 100 mmHg) and could achieve a 70% reduction of neointimal thickness at 2 weeks postoperatively compared with control vein grafts.
|
6. Rat models (inferior vena cava into abdominal aorta interposition, superficial epigastric vein into femoral artery interposition, ileolumbar vein into abdominal aorta interposition) |
|---|
|
TopAbstract1. Clinical and...2. General considerations on...3. Pig models (saphenous...4. Canine models (jugular...5. Rabbit model (jugular...6. Rat models (inferior...7. Mouse models (vena...8. Other models of...9. ConclusionReferences |
|---|
6.1 Pathology (PDGF, bFGF, IL-1, TNF alpha, c-jun, c-fos)
Sterpetti et al. showed an association between the increase of neointimal hyperplasia during 12 weeks after vein grafting and the amount of the cytokines PDGF, bFGF, IL-1, and TNF alpha. The regression of neointimal hyperplasia in reimplanted venovenous (after 4 weeks of arterial interposition) rat vein grafts was associated with an increased expression of TGF beta1 and a reduction of PDGF, bFGF, IL-1, and TNF alpha. In another study, this group demonstrated a decreased amount of PDGF and bFGF production in arterial grafts compared with vein grafts [78,79].
Immediate early gene activation is a fundamental response following growth factor-mediated cellular stimulation. Suggs et al. demonstrated a steep increase of the immediate early gene products c-jun and c-fos: There was a fivefold increase of c-fos protein and a 10-fold increase of c-jun protein at 15 min after vein graft arterialization. The c-fos and c-jun levels peaked after 2–6 h and decreased during the following days [80].
6.2 Therapeutic strategies (c-fos and c-jun antisense, heparin)
Suggs et al. have shown that local treatment of the veins with c-fos, and c-jun antisense ODN (in pluronic gel) decreased neointimal hyperplasia. However, an aneurysmal dilatation of 4/20 vein grafts was observed at 2 weeks postoperatively [80].
Heparin is a known inhibitor of SMC proliferation in vitro and inhibits neointimal thickening in injured arteries. However, heparin treatment of rats (800 units/12 h for 3 days s.c.) failed to suppress neointimal hyperplasia in veins at 2 weeks postoperatively [81].
|
7. Mouse models (vena cava into carotid artery interposition, jugular vein patch into abdominal aorta, jugular vein patch into carotid artery) |
|---|
|
TopAbstract1. Clinical and...2. General considerations on...3. Pig models (saphenous...4. Canine models (jugular...5. Rabbit model (jugular...6. Rat models (inferior...7. Mouse models (vena...8. Other models of...9. ConclusionReferences |
|---|
|
In another study, this group found an increase of neointimal hyperplasia and areas of necrosis in the vascular wall in vein grafts of apoE deficient (hypercholesterolemic) mice. In the vein grafts of apoE deficient mice typical atheromas were found featuring infiltrating mononuclear cells, foam cells, cholesterol crystals, and calcified necrotic areas [83].
Mayr et al. found an apoptosis rate (TUNEL) of 13%, 29%, and 21% in mouse vein grafts at 1, 4, and 8 weeks postoperatively, respectively. The apoptosis rate was increased in vein to artery grafts compared with vein to vein grafts and ungrafted veins at 24 h postoperatively indicating that apoptosis was due to biomechanical stress not due to the surgical trauma [84].
In composite vessels (jugular vein patch into abdominal aorta) Sakaguchi et al. could demonstrate an increased neointimal area in vein patches stored for 2 h (in heparinized saline solution) before grafting compared with immediately grafted vein patches. Furthermore, the increased storage time was associated with a decreased cAMP amount in the veins [85].
The delta isoform of PKC was shown to be growth inhibitory and pro-apoptotic in SMCs. PKC delta knockout mice showed increased neointimal hyperplasia which was associated with a decreased apoptosis rate in vein grafts at 8 weeks postoperatively. PKC delta (+/+) veins grafted into PKC delta (–/–) mice showed similar neointimal hyperplasia as PKC delta (+/+) veins grafted into PKC delta (+/+) mice [86].
Increased amounts of the tumor suppressor p53 cause cells either to undergo apoptosis or prolonged cell cycle arrest. Vein grafts of p53 knockout mice exert a doubled neointimal area at 4 weeks postoperatively compared with those of p53 (+/+) mice. In addition, the number of cells (mainly alpha actin positive SMCs) was increased in the neointima of p53 (–/–) mice compared with wild type animals. The increase in the cell number was associated with a decreased apoptotic rate in p53 (–/–) mice [87].
ICAM-1 is a counter receptor that is engaged in leukocyte adhesion and transmigration and it is abundantly present on the endothelial surface of mouse vein grafts during 4 weeks after surgery. Zou et al. [88] found a decreased leukocyte amount and a 40% reduced neointimal thickness in vein grafts of ICAM-1 knockout mice compared with wild-type animals.
In a mouse model of jugular vein patch into carotid artery Shi et al. demonstrated a neointima of 5–8 cell layers with mural fibrin deposits at 7 days postoperatively. At this time point macrophages were abundantly present in the adventitia and in small numbers in the neointima. In the same study the authors found no difference in neointimal hyperplasia between plasminogen knockout mice and wild type animals [89].
An interesting question is the origin of cells in the vein graft showing inflammation and neointimal hyperplasia. Zhang et al. [90] found that 60% of neointimal cells were graft extrinsic deriving from a congenic Lac Z expressing recipient mouse, and 40% of neointimal cells were graft intrinsic (from the wild type donor).
In accordance with this Fogelstrand et al. found an equal amount of neointimal hyperplasia in acellular vein grafts compared with controls at 4 weeks postoperatively. This experiment demonstrated that vein graft extrinsic cells could form a neointima. However, isolated vein grafts (perivascular polyethylene film) showed hardly any graft extrinsic cells but no decrease in neointima. Hence, vein graft intrinsic cells could also form a neointima [91].
Hu et al. [92] used SM-LacZ transgenic mice (which express beta galactosidase in SMCs) and found that 40% of SMCs in vein grafts derived from the recipient (graft extrinsic cells) and 60% derived from the donor (graft intrinsic cells) at 8 weeks postoperatively.
In another study, Hu et al. found that the adventitia of the aortic root harbored abundantly sca-1 positive cells that showed the potential to differentiate into SMCs. Perivascularly applied sca-1 positive progenitor cells were found in neointimal lesions of vein grafts at 2 and 4 weeks postoperatively. This study underlines the potential role of adventitial (progenitor) cells in vein graft disease [93].
Disappearance of ß-gal + staining from donor vessels after vein segments from TIE2-LacZ mice (LacZ gene expression specifically in vascular endothelial cells) were isografted into wild-type animals and the appearance when wild-type veins were grafted into TIE2-LacZ mice showed the recipient origins of endothelial cells of vein grafts. The same study showed that one-third of endothelial cells of vein grafts derived from bone marrow progenitor cells, which were the major source of endothelial repair in mouse vein grafts [94].
7.2 Therapeutic strategies (aspirin, external stenting, suramin, rapamycin, CNP, azathioprin)
Torsney et al. found out that thrombus formation peaked at 1–3 days postoperatively in mouse vein grafts, and decreased thereafter. Locally applied aspirin (in pluronic gel) reduced both thrombus formation and neointimal hyperplasia in vein grafts at 4 and 8 weeks postoperatively [95].
In ApoE*3-Leiden transgenic mice vein grafts showed accelerated atherosclerosis with macrophage and foam cell deposition and excessive lipid accumulation at 4 weeks postoperatively, which was attenuated by external polyethylene stenting of the vein grafts [96].
Suramin is a growth factor (especially PDGF) receptor antagonist. Incubation ex vivo for 20 min and perivascular application of Suramin in pluronic gel led to a more than 50% reduction of neointimal thickness at 4 and 8 weeks postoperatively compared with controls. In the same study, Suramin treatment was associated with a decreased amount of PDGF alpha and PDGF beta receptors in neointimal SMCs compared with control vein grafts [97].
Rapamycin is an immunosuppressant with marked antiproliferative properties due in part to an inhibition of the mammalian target of rapamycin. We could demonstrate a dose-dependent reduction of neointimal thickness with perivascularly applied rapamycin (in pluronic gel) in mouse vein grafts compared with controls [98]. Another study showed that treatment with rapamycin was associated with an increased apoptotic rate in arterialized veins [99].
CNP inhibits VSMC proliferation and shows antiinflammatory and antithrombotic effects. Local application of CNP (in pluronic gel) decreased neointimal hyperplasia in mouse vein grafts. In the same study, the amount of adventitial CD8 positive leukocytes was decreased in CNP-treated vein grafts compared with controls [100].
Hu et al. demonstrated that local application of adenovirus vector resulted in enhanced inflammatory response and neointimal hyperplasia in mouse vein grafts at 4 and 8 weeks postoperatively. Adenovirus-mediated TIMP-2 gene transfer attenuated the virus mediated increase of neointimal area but there was no difference to control vein grafts [101].
Azathioprine, an immunosuppressive and antiinflammatory drug, reduced neointimal thickness of vein grafts by 66% at 2 weeks after arterialization (perivascular application without carrier). In addition, azathioprine-treated vein grafts showed a 30-fold increased apoptotic rate in the vascular wall as compared with controls [102].
|
8. Other models of vein graft disease |
|---|
|
TopAbstract1. Clinical and...2. General considerations on...3. Pig models (saphenous...4. Canine models (jugular...5. Rabbit model (jugular...6. Rat models (inferior...7. Mouse models (vena...8. Other models of...9. ConclusionReferences |
|---|
In a pig model of internal mammary vein to LAD CABG, the steep increase of wall thickness and neointimal hyperplasia of vein grafts (even at 4 days postoperatively) could be demonstrated [104].
Krejca et al. performed radial vein into carotid artery interposition grafts in sheep. By the use of an external dacron mesh both neointimal hyperplasia and the number of proliferating cells (Ki-67 immunohistochemistry) could be reduced compared with control vein grafts [105].
|
9. Conclusion |
|---|
|
TopAbstract1. Clinical and...2. General considerations on...3. Pig models (saphenous...4. Canine models (jugular...5. Rabbit model (jugular...6. Rat models (inferior...7. Mouse models (vena...8. Other models of...9. ConclusionReferences |
|---|
One of the major limitations of experimental models is their homogenous thin wall architecture. In human saphenous vein CABG, the veins show a big variety of wall thickness and preexisting pathologies (e.g., fibrosis, postinflammatory changes, various degrees of varicosis). On the other hand, the standardized experimental settings allow the evaluation of single factors regarding their contribution to vein graft disease. However, animal experiments are prerequisites of going from bench to bedside.
|
Footnotes |
|---|
Presented at the postgraduate course of the joint 19th Annual Meeting of the European Association for Cardio-thoracic Surgery and the 13th Annual Meeting of the European Society of Thoracic Surgeons, Barcelona, Spain, September 25–28, 2005. |
References |
|---|
|
TopAbstract1. Clinical and...2. General considerations on...3. Pig models (saphenous...4. Canine models (jugular...5. Rabbit model (jugular...6. Rat models (inferior...7. Mouse models (vena...8. Other models of...9. ConclusionReferences |
|---|
