Inhibition of Complement Component C3 Reduces Vein Graft Atherosclerosis in Apolipoprotein E3–Leiden Transgenic Mice
Background— Venous bypass grafts may fail because of development of intimal hyperplasia and accelerated atherosclerosis. Inflammation plays a major role in these processes. Complement is an important part of the immune system and participates in the regulation of inflammation. The exact role of complement in the process of accelerated atherosclerosis of vein grafts has not yet been explored, however.
Methods and Results— To assess the role of complement in the development of vein graft atherosclerosis, a mouse model, in which a venous interposition was placed in the common carotid artery, was used. In this model, vein graft thickening appeared within 4 weeks. The expression of complement components was studied with the use of immunohistochemistry on sections of the thickened vein graft. C1q, C3, C9, and the regulatory proteins CD59 and complement receptor-related gene y could be detected in the lesions 4 weeks after surgery. Quantitative mRNA analysis for C1q, C3, CD59, and complement receptor-related gene y revealed expression of these molecules in the thickened vein graft, whereas C9 did not show local mRNA expression. Furthermore, interference with C3 activation with complement receptor-related gene y–Ig was associated with reduced vein graft thickening, reduced C3 and C9 deposition, and reduced inflammation as assessed by analysis of influx of inflammatory cells, such as leukocytes, T cells, and monocytes. In addition, changes in apoptosis and proliferation were observed. When C3 was inhibited by cobra venom factor, a similar reduction in vein graft thickening was observed.
Conclusions— The complement cascade is involved in vein graft thickening and may be a target for therapy in vein graft failure disease.
Received August 1, 2005; de novo received February 14, 2006; revision received September 28, 2006; accepted October 5, 2006.
Bypass graft surgery with venous grafts is one of the most frequently used therapies, in both cardiac and peripheral vascular surgery, to treat atherosclerotic occlusive disease. However, graft patency is often compromised by formation of intimal hyperplasia and accelerated atherosclerosis, resulting in vein graft thickening. Failure rates as high as 60% after 10 years1 have been reported, and reinterventions are often required.
Clinical Perspective p 2838
The process of vein graft thickening is characterized by adhesion and influx of inflammatory cells and migration of vascular smooth muscle cells (SMCs) to the intima of the vein graft.2 Moreover, macrophages in the intima take up oxidized lipoproteins and become foam cells contributing to the development of accelerated atherosclerosis.3–5 Little information is available about the mechanisms underlying these processes, but it is universally assumed that inflammation and consequently the immune system play a pivotal role. 5–7
A major component of the immune system is the complement cascade.8,9 Complement consists of a group of proteins, membrane-bound receptors, and regulatory enzymes. Central in the complement cascade is complement component C3. Cleavage of C3 can be induced via 3 separate pathways: the classic pathway, alternative pathway, and lectin pathway, all activated by specific substrates. Activation via 1 of the 3 pathways results in formation of C3 convertases. The convertases are capable of cleaving C3 into C3a and C3b, starting a cascade that ultimately results in activation of C5 and subsequently in formation of terminal complement component C5b-9 (also called membrane-attack complex). Furthermore, cleavage of C3 eventually leads to formation of potent chemotactic factors, such as C5a.
The role of complement in several inflammatory conditions is well recognized and described. It plays a major role in host defense to microorganisms, hyperacute rejection after organ transplantation, ischemia/reperfusion injury, and several autoimmune diseases. Atherosclerosis, like vein graft disease, is a form of vascular inflammation accompanied by intimal thickening, and several studies have been published suggesting a role for the complement system in the process. This hypothesis is supported by the detection of complement components in human atheroma, and expression of several of these complement components is upregulated in atherosclerotic tissue. Furthermore, animal studies using knockout mice and rabbits deficient in complement have been performed; these, however, show conflicting results.10,11 Data concerning the role of complement activation in vein graft thickening are lacking.
As mentioned above, C3 is the central component in complement activation. Inhibition of C3 activation provides good insight into the role of complement as a whole because formation of biologically active end products (eg, C5b-9, C5a, and C3a) is blocked. Several substances can be used to modify activation of C3. One of the most widely used compounds is cobra venom factor (CVF), derived of venom of the Naja species. It functions as a C3b-like molecule and leads to unregulated C3 activation, resulting in depletion of the complement cascade; it is associated with generation of phlogistic component fragments from C3 and C5, however, indirectly leading to tissue activation. A more elegant approach to block C3 activation is the use of complement receptor-related gene y (Crry)–Ig, a recombinant protein of the mouse membrane complement inhibitor Crry fused to IgG1-hinge.12 Crry-Ig demonstrates decay-accelerating activity for both the classic and alternative pathways of complement as well as cofactor activity for factor I–mediated cleavage of C3b and C4b and thus prevents increased activation of C3.
To study the role of complement activation in vein graft disease, we used a mouse model for vein graft disease in hypercholesterolemic apolipoprotein E3–Leiden mice. This mouse model highly resembles graft morphology in patients regarding leukocyte adhesion and influx, foam cell accumulation, calcification in the vessel wall, and development of a thin fibrous cap.13 Not only was expression of several complement factors, on both RNA and protein levels, in the thickened vein graft shown, but it was also demonstrated that treatment with Crry-Ig, which inhibits C3 activation, resulted in a significant decrease of intimal hyperplasia and accelerated atherosclerosis in murine vein grafts. In the present study, we show that activation of complement cascade plays a pivotal role in the development of vein graft thickening.
All experiments were approved by the Animal Welfare Committee of TNO, Quality of Life, Leiden, The Netherlands. For all experiments, male C57bl6 apolipoprotein E3–Leiden mice bred in our laboratory,14 aged 15 to 20 weeks, were used. During the experiment, animals were fed a high-fat, high-cholesterol diet15 starting 3 weeks before surgery to induce hypercholesterolemia. All mice received water and food ad libitum. Cholesterol levels in serum were determined 1 day before surgery and at the time of euthanasia. Mice were anesthetized by an intraperitoneal injection with a combination of midazolam (5 mg/kg; Roche, Woerden, the Netherlands), medetomidine (0.5 mg/kg; Orion, Espoo, Finland), and fentanyl (0.05 mg/kg; Janssen, Berchem, Belgium).
Vein graft surgery was performed as described previously.16 In summary, caval veins were harvested from genetically identical donor mice to serve as grafts and were preserved in 0.9% NaCl containing 100 U/mL of heparin at 4°C. The right carotid artery was cut in the middle, and a polyethylene cuff was placed at both ends of the artery. The artery was everted around the cuff and ligated with a silk 8.0 suture. Then the graft was sleeved over the 2 cuffs and ligated. Pulsations of the vein graft confirmed successful engraftment. Generally, this procedure is performed in 30 minutes.
Crry-Ig was produced as described.12 Crry-Ig treatment started 1 day before surgery, and animals received 3 mg IP of Crry-Ig every other day during the complete study period. The control group received a monoclonal murine IgG antibody (reactive with human C-reactive protein and with no cross-reactivity in mice) of the same subtype in the same concentration every other day. This approach was described previously to be the proper control for Crry-Ig treatment.17
Cobra Venom Factor
CVF (Quidel Corporation, San Diego, Calif) was dissolved in sterile 0.9% NaCl. Animals received daily injections of 20 IU/kg per day IP of CVF to deplete C3 throughout the whole study period starting 1 day before surgery. Animals in the control group received daily injections with sterile 0.9% NaCl.
Morphometric Assessment of Vein Grafts
Mice were euthanized after either 7 or 28 days after surgery. Vein grafts were in vivo perfused with 4% formaldehyde, harvested, and embedded in paraffin. Serial perpendicular cross sections were made of the specimen. All samples were routinely stained with hematoxylin-phloxine-saffron.
Morphometric analysis of vein grafts, harvested after 28 days, was performed with the use of image analysis software (Qwin, Leica, Wetzlar, Germany). Because only a few layers of cells are in the media of murine veins and no morphological border exists between neointima and media, vein graft thickening, ie, the region between lumen and adventitia, was used to define lesion area. For each mouse, 5 equally spaced cross sections were used to determine vessel wall thickening.
To detect expression of complement factors, immunohistochemistry was performed on paraffin-embedded sections of vein grafts harvested after 28 days. The presence of C1q was assessed with the use of rabbit anti-mouse primary antibodies (Roche Applied Biosciences, Basel, Switzerland), and C3 was detected with a rabbit anti-mouse antibody (developed in our laboratory18). Anti-C9 (rabbit anti-rat, cross-reactive with mouse C919) was a kind gift of Professor B.P. Morgan (Cardiff University, Cardiff, UK). The antibody used to detect the complement regulatory enzyme CD59a (monoclonal rat anti-mouse) was a kind gift of Dr C.L. Harris (Cardiff University), and Crry was detected with a rat anti-mouse anti-Crry antibody (BD Biosciences, Alphen a/d Rijn, the Netherlands).
Complement components were quantified with the use of computer-assisted morphometric analysis (Qwin) and expressed as total immunopositive area as percentage of total vein graft area in cross sections.
In grafts, SMCs were stained with the use of mouse anti-rat anti–smooth muscle α-actin antibodies (cross-reacts with mouse; Roche Applied Biosciences). Collagen was histochemically stained by Sirius red. Leukocytes were detected with the use of anti-CD45 antibodies (Pharmingen, San Diego, Calif). Different subsets of leukocytes were specified. The amount of macrophage-derived foam cells within the thickened vessel wall was visualized by AIA31240 antibody (Accurate Chemical, Westbury, NY).
T cells were detected with the use of anti-CD3 antibodies (Serotec, Raleigh, NC).
Cellular proliferation was quantified by proliferating cell nuclear antigen staining (Calbiochem, San Diego, Calif), whereas apoptotic cells were identified by in situ DNA nick end labeling (Roche Applied Biosciences).
Quantification of all stainings, except those for SMCs and foam cells, occurred by counting the positive number of cells per slide and was expressed as percentage of total number of nuclei (determined by nucleus red staining). SMC, collagen, and foam cell quantification, because of their high occurrence in the vein grafts, was performed in a manner similar to the quantification of complement components.
RNA Isolation, cDNA Synthesis, and Reverse Transcription Polymerase Chain Reaction
RNA was isolated from caval veins and vein grafts harvested at several time points (time=6 and 24 hours after surgery and 3, 7, and 28 days after surgery; n=4 per time point) with the use of RNA Isolation Mini Kits for Fibrous Tissue (Qiagen, Venlo, the Netherlands) according to the manufacturer’s protocol. DNase treatment was included (RNase Free DNase set, Qiagen). RNA (250 ng) was reverse-transcribed with the use of the Ready-To-Go You-Prime-First-Strand Beats (Amersham Biosciences, Uppsala, Sweden).
Reverse-transcribed products were studied with the use of semiquantitative reverse transcription polymerase chain reaction (Robocycler Gradient96, Stratagene, Cedar Creek, Tex), with primers for C1q, C3, C9, CD59, and Crry (for sequences, see Table 1). Samples were amplified for 35 cycles (30 seconds at 94°C, 30 seconds at 56°C, and 90 seconds at 65°C) after an initial denaturation cycle for 2 minutes at 94°C. The last cycle was followed by extension of 4 minutes at 74°C. PCR products were visualized on 1.2% agarose gel containing ethidium bromide.
All data were presented as mean±SEM. Statistical analysis was performed with the use of SPSS 11.5 for Windows. Differences between groups were analyzed with a Student t test. Probability values <0.05 were regarded as statistically significant.
The authors had full access to the data and take full responsibility for its integrity. All authors have read and agree to the manuscript as written.
Presence of Complement Components in Intimal Hyperplasia of Murine Vein Grafts
Twenty-eight days after surgery, mice (n=6) were euthanized, and vein grafts were harvested to detect complement factors and regulatory enzymes by immunohistochemistry (Figure 1). These factors were elicited on the basis that each represents a specific part of the complement cascade (eg, C1q; early classic pathway, C3; central component, C9 part of the membrane attack complex, CD59, and Crry; regulatory molecules).
Massive presence of C1q was seen in thickened vein grafts, mainly deep in the intimal hyperplasia colocalizing with foam cells. Furthermore, inflammatory cells attached to the endothelium and in the adventitia of the veins showed positive staining. C3 was predominantly expressed in the same regions as C1q, ie, in macrophage-derived foam cells, adherent and adventitial inflammatory cells, and the endothelium. In addition, C9 colocalized highly with macrophages in the thickened intima but was not expressed by endothelial cells.
Complement inhibitor CD59 was detected as diffusely distributed in all layers of the vessel wall. The membrane-bound Crry showed a patchy distribution in cells of the media and adventitia. Endothelial cells showed positive staining in approximately one third of cells.
Presence of RNA Coding for Complement Components in Murine Vein Grafts
Local production of complement component was examined by RNA analysis of vein grafts. Sixteen mice underwent vein graft surgery and were euthanized on several time points (time=6 hours, 24 hours, 3 days, 7 days, and 28 days). In addition, normal, not yet interposed, caval veins of donor mice were included. Reverse transcription polymerase chain reaction was performed for C1q, C3, C9, CD59, and Crry (Figure 2).
Minimal expression of C1q mRNA was seen in caval veins. In vein grafts, C1q mRNA was not detectable in the first 24 hours after surgery. C1q mRNA was clearly present 3 days after surgery, however, and expression remained constant for the rest of the study period.
C3 showed an explicit expression in vein grafts in all stages of remodeling. Interestingly, an equal expression was seen in caval veins, showing that local production of C3 also occurs without a remodeling process going on in the vessel. mRNA expression of complement component C9 could not be detected in vein grafts, whereas expression in liver tissue (positive control) appeared clearly positive, indicating that C9 protein is not produced locally in vein grafts.
Crry mRNA was clearly detectable in normal caval veins. Downregulation of Crry mRNA expression was seen in the first days after surgery, whereas after 7 days, an increased expression was seen at the level of normal caval veins.
Complement inhibitor CD59 showed robust expression in normal caval veins, whereas expression in the remodeling vein graft was present but appeared lower.
Effect of Crry-Ig Treatment on Vein Graft Thickening and Complement Expression
The aforementioned studies suggest involvement of the classic pathway on vein graft thickening. Earlier studies demonstrated efficient blockage of complement activation by Crry-Ig. Twelve apolipoprotein E3–Leiden mice were randomly divided into 2 treatment groups. One group received 3 mg IP Crry-Ig every other day; controls received an injection with 3 mg of nonrelevant monoclonal antibody of the same isotype every other day. No significant differences were observed between groups in preoperative body weight, body weight at time of euthanasia, and cholesterol levels before surgery and at time of euthanasia.
Vein grafts were harvested after 28 days, and vein graft thickening was quantified. Crry-Ig–treated mice exhibited ≈50% less vein graft thickening than controls (control, 0.36±0.07 mm2; Crry-Ig, 0.18±0.01 mm2; P=0.028). Luminal area was equal in both groups (control, 0.53±0.04 mm2; Crry-Ig, 0.40±0.06 mm2; P=0.28). Consequently, total vessel wall area was significantly larger in the control group (0.56±0.05 mm2) compared with Crry-Ig–treated vein grafts (0.89±0.07 mm2; P=0.016), indicating reduced outward remodeling in the Crry-Ig–treated group.
C1q quantity in vein grafts was not affected by Crry-Ig and demonstrated no differences compared with control IgG-treated mice (control, 25±4%; Crry-Ig, 25±2%; P=0.48).
A significant decrease of C3 deposition was observed in Crry-Ig–treated mice compared with controls (control, 37±4%; Crry-Ig, 21±3%; P=0.007). Moreover, presence of C9 protein was reduced in Crry-Ig–treated vessels (control, 19±3%; Crry-Ig, 9±2%; P=0.007). Data are summarized in Figure 3 and Table 2.
Effect of Crry-Ig Treatment on Cellular Composition of Remodeling Vein Grafts
To study the effect of blocking complement activation on cellular composition, 24 mice underwent surgery, were treated with either Crry-Ig or a nonrelevant monoclonal antibody (as described above), and were euthanized after 7 or 28 days (n=6 in each group).
Overall, as expected, vein grafts harvested after 28 days contained more cells than the ones harvested after 7 days, and Crry-Ig–treated grafts contained fewer cells than their untreated controls after 28 days (control 7 days, 228±59; Crry-Ig 7 days, 227±32 [P=NS]; control 28 days, 2296±565; Crry-Ig 28 days, 1143±69 [P=0.044]).
SMC content was analyzed in vein grafts harvested after 28 days because in the first week after engraftment all SMCs disappear in this model.13 In the Crry-Ig–treated group, thickened vein grafts had a significantly higher relative SMC content than the control group (control, 33±5%; Crry-Ig, 47±3%; P=0.04). The amount of collagen deposition, however, did not significantly differ between both groups (control, 33±4%; Crry-Ig, 29±2%; P=0.5).
Influx of CD45-positive leukocytes was seen especially after 7 days and was strongly reduced by Crry-Ig treatment (control, 31.3±10.2%; Crry-Ig, 9.6±1.5%; P=0.034). At 28 days, the relative amount of CD45-positive leukocytes was much lower and was not significantly different between groups (control, 1.8±0.4%; Crry-Ig, 1.9±0.9%; P=0.44). In the 7-day vein grafts, monocytes in both groups were predominantly adhering to the vessel wall and invading the vein graft. Crry-Ig treatment resulted in a 42% (although not significant) reduction in adhering monocytes (control, 7.0±1.8%; treated group, 4.1±0.9%; P=0.08). After 28 days, AIA31240-immunostaining revealed mainly macrophage-derived foam cells in the plaque area. A significantly lowered foam cell content was seen in vein grafts of Crry-Ig–treated animals (Crry-Ig, 16±1%; control, 25±5%; P=0.043) after 28 days. In regard to CD3-positive T cells, a significant difference could be seen after 7 days (control, 1.9±0.6%; Crry-Ig, 0.7±0.1%; P=0.05); however, after 28 days, this difference could not be observed (control, 1.8±0.6; Crry-Ig, 1.2±0.3%; P=0.21). Results are summarized in Table 3.
Effect of Crry-Ig Treatment on Cytokinetics of Remodeling Vein Grafts
To obtain insight on the possible mechanisms of the effect of complement inhibition on vein graft remodeling, we analyzed the effects of Crry-Ig treatment on cell proliferation and apoptosis. In regard to proliferation, Crry-Ig treatment resulted in reduced cellular proliferation after 7 days as assessed by proliferating cell nuclear antigen immunohistochemistry (control, 10.6±1.3%; Crry-Ig, 6.7±1.2%; P=0.036). After 28 days, cellular proliferation declined, and no differences were seen between the groups (control, 2.3±0.7%; Crry-Ig, 1.8±0.6%; P=0.3).
Surprisingly, Crry-Ig treatment led to increased numbers of in situ DNA nick end labeling–positive apoptotic cells in vein graft 7 days after surgery (control, 0.7±0.4%; Crry-Ig, 3.0±0.6%; P=0.005). However, after 28 days, the number of apoptotic cells was decreased in both groups to such a level that no conclusion was justified. All data are summarized in Table 3.
Effect of CVF Treatment on Vein Graft Thickening
To demonstrate that the observed effect is not Crry-Ig specific, an alternative treatment affecting C3 activation was tested. Therefore, CVF was administered. Vein grafting was performed in 12 mice, and 6 were treated with CVF (20 IU/kg per day). Mice were euthanized after 28 days. Quantification of vein graft thickening revealed a 63% reduction of intimal hyperplasia in CVF-treated animals (0.15±0.01 mm2) compared with controls (0.41±0.10 mm2; P=0.016). Luminal surface did not differ significantly between groups (CVF, 0.49±0.06 mm2; control, 0.59±0.07 mm2; P=0.47).
The present study argues for a causal role for complement activation in general and C3 activation in particular in the process of venous bypass graft thickening. To our knowledge, the present study is the first that not only demonstrates presence of complement components in vein grafts but also provides evidence that inhibition of C3 activation results in a marked decrease of vein graft thickening in vivo.
Intimal hyperplasia formation and accelerated atherosclerosis in vein grafts is considered to be the result of an inflammatory process.5–7 This process is initiated by mechanical vessel damage and hypoxia during surgery and by altered shear stress, which the vein is subjected to after surgery. Several animal models have been developed to mimic this process. In this study a mouse model is used, in which caval veins of donor mice are placed as interposition in the common carotid artery.16 When performed in hypercholesterolemic mice, vein graft thickening occurs with signs of accelerated atherosclerosis.13 The morphology of the observed lesions strongly resembles that seen in human vein grafts, underscoring the relevance of this model.
Because the complement cascade is an important part of the innate immune system and is also involved in initiation of adaptive immune responses, it may be one of the mediators of inflammatory processes in remodeling vein grafts.
The presence of complement components (C1q, C3, and C9) and regulatory molecules (CD59a and Crry) in murine vein grafts was shown by immunohistochemistry. Furthermore, we analyzed local synthesis of complement components, defined as mRNA expression in vein grafts. Although synthesis of complement components is believed to be localized in the liver, evidence of extrahepatic synthesis of complement components has been published recently. Reverse transcription polymerase chain reaction analysis of vein grafts at various time points after surgery revealed local expression of mRNA coding for C1q, C3, CD59, and Crry. No local production of C9 could be detected, indicating that C9 protein deposited in vein grafts, as seen by immunohistochemistry, is produced elsewhere.
Although C3 mRNA expression was detectable in the vessel wall of caval veins, no upregulation or downregulation of C3 on the mRNA level could be detected in vein grafts. Several explanations for this finding are possible. One explanation may be that, although mRNA C3 is produced locally in vein grafts, this may not tell us much about C3 activity and protein presence. Furthermore, C3 is produced mainly in the liver, as are most of the complement components, and is abundantly present in plasma. This liver-derived C3 may be crucial for mediating the effects on vein graft thickening, whereas the role of locally produced C3 is unclear. In addition, factors involved in regulating local levels of (active) C3 and the effects of C3 downstream molecules on C3 levels in vein graft remodeling are not completely understood.
We were unable to detect activated complement factors or membrane-attack complex because of the lack of specific antibodies against murine-activated complement components, eg, murine iC3b, C5b-9, or C5a.
Because C3 is the central protein in the complement cascade, it was used as a specific target for our interventional experiments. C3 activation was blocked with Crry-Ig, an inhibitor for all activation pathways by the inhibition of C3 convertases. Crry-Ig treatment resulted in a significantly lower C3 and C9 deposition in vein grafts, accompanied by an inhibition of vein graft thickening of ≈50%. Analysis of cytokinetics and cellular composition of vein grafts at 7 and 28 days provided insight into the mechanism by which complement inhibition may affect vein graft remodeling. When C3 was inhibited by Crry-Ig, one of the first phenomena seen was reduced adherence and influx of leukocytes, possibly as a result of absence of (complement activation–derived) chemotactic stimuli. This suggests tempered inflammatory activity in remodeling vein grafts. Reduction in inflammatory activation of the vein graft wall is thought to be the main cause of reduced vein graft thickening. Reduction of leukocyte adhesion and influx is accompanied by reduced numbers of proliferating cells in the remodeling vein graft and induction of apoptosis. This induction of apoptosis is unexpected because reduced inflammation is usually accompanied by reduced apoptosis. No specific cell type or location could be associated with the observed apoptosis.
After 28 days, plaque composition was significantly altered in the Crry-Ig–treated group, favoring SMCs and reduced numbers of foam cells. Increased numbers of SMCs in thickened vein grafts reflect a desirable state because this is believed to be one of the main contributions to plaque stability.20 The observation that vein graft thickening can also be hampered by CVF (another potent C3 activation inhibitor) shows that the phenomena seen after Crry-Ig treatment are not Crry-Ig specific but are caused by C3 inhibition. Inhibition of vein graft thickening after treatment with Crry-Ig also suggests that circulating complement components play a role in the remodeling process.
The demonstrable C1q in vein grafts suggests involvement of the classic pathway. Interestingly, we found reduced C1q mRNA expression in the early phases after surgery, followed by enhanced expression in later stages. Regrettably, very little is known about in vivo C1q mRNA regulation in various cell types. Because C1q in this model is predominantly expressed by monocytes and macrophages, however, and macrophage C1q mRNA expression has been shown to alter during the various stages of macrophage maturation in vitro,21 this reduced C1q mRNA expression in the early stages may be due to a temporary lowered C1q mRNA expression as a consequence of ongoing maturation during the transition of adhering monocytes into tissue macrophages.
Furthermore, we found that in the early phases after surgery, Crry mRNA expression was decreased and therefore diminished complement inhibition potential of the vein graft. This may be one of the causes responsible for complement activation in this model.
However, other questions need to be considered. For instance, what triggers complement activation in venous bypass grafts, and which of the activation routes is accountable for activation? There are several possibilities.
A venous interposition can be seen as an autologous transplant; it undergoes ischemia/reperfusion injury during the procedure. In other models for ischemia/reperfusion, the production of “natural antibodies” by CD5-positive B1 cells is described, leading to deposition of IgM antibodies and thereby triggering the classic pathway.22 This is confirmed by the study of Fitzmaurice and Ratliff,23 who were able to detect immunoglobulins and C3 depositions in human saphenous vein grafts. We show the presence of C1q protein and mRNA in various phases of the remodeling process, suggesting that the classic pathway actually may play a role. In addition to the classic pathway as the major activating pathway of the complement system in ischemia/reperfusion, Stahl et al24 recently described the role of the alternative pathway in (intestinal) ischemia/reperfusion injury, and Thurman et al25 found that the alternative pathway alone was required for renal ischemia/reperfusion injury.
In addition, low-density lipoprotein depositions in the vessel wall may be responsible for complement activation in this model. Torzewski et al26 previously described colocalization of C5b-9 and enzymatically altered low-density lipoprotein in deeper parts of the intima of early atherosclerotic lesions. Furthermore, it was shown that this enzymatically modified low-density lipoprotein is atherogenic and can induce complement activation via the alternative route in vitro.27,28 However, because fat deposition and foam cell formation occur relatively late in the model used in this study and the observed effects on cellular influx were present after 7 days, it is not likely that this route of complement activation plays a role in this model.
Furthermore, the vein graft is susceptible to arterial blood pressure and increased shear stress causing damage to the vein graft wall. In addition, the surgical procedure leads to damage of the vein graft. This results in a mechanical denudation of the endothelium and apoptosis of remaining endothelial cells. It has been described before that apoptotic (endothelial) cells can induce complement activation via both the classic29 and alternative pathways.30
Which of the 3 activation pathways, alone or in combination, is responsible for the activation of complement and by what kind of injury this activation is triggered remain to be determined, and further research is necessary to indicate the role of each of these modalities in vein graft disease.
When data are translated from mouse studies to the human situation, several issues should be taken into account. First, mice do not develop atherosclerosis spontaneously. Therefore, usually genetically altered hypercholesterolemic mice are used to induce atherosclerotic lesions, either spontaneously after prolonged exposure to a hypercholesterolemic environment or in an accelerated way after vascular intervention. Although the apolipoprotein E3–Leiden mice used in this study suffered relatively mild hypercholesterolemia (serum cholesterol 8 to 10 mmol/L), cholesterol levels still exceed serum levels seen in most of the patients. As previous reports have shown, severe hypercholesterolemia might induce increased vascular inflammatory reactions13,31 and consequently atherosclerotic lesion formation. Furthermore, because of anatomic variations between humans and mice, it can be assumed that after engraftment, as a result of the 10-fold increase in blood pressure, more extensive graft distension and subsequent vascular damage and SMC apoptosis are occurring in murine vein grafts (only a few cell layers thick) compared with the human counterpart.
In conclusion, several complement factors (on both protein and mRNA levels) are present in thickened vein grafts, and treatment with Crry-Ig (interfering with C3 activation) results in marked reduction of vein graft thickening. This reduction coincides with reduced numbers of leukocytes in the early stages of vein graft remodeling and increased numbers of SMCs in the later stage. Therefore, we have provided evidence that activation of C3 and thereby the complement cascade is an important early trigger for vein graft thickening. Therapy that interferes with the function of C3 may be an interesting new target to overcome the clinical problem of vein graft disease.
Ronald van der Sluis (Gaubius Laboratory, TNO Quality of Life, Leiden, the Netherlands) and Ngaisah Klar (Department of Renal Diseases, Leiden University Medical Centre, Leiden, the Netherlands) are acknowledged for their excellent technical assistance.
Sources of Funding
This study was supported by the Netherlands Heart Foundation (Molecular Cardiology Program, grant M93.001, to Drs Schepers and Quax).
Bulkley BH, Hutchins GM. Accelerated “atherosclerosis”: a morphologic study of 97 saphenous vein coronary artery bypass grafts. Circulation. 1977; 55: 163–169.
Quigg RJ, Kozono Y, Berthiaume D, Lim A, Salant DJ, Weinfeld A, Griffin P, Kremmel E, Holers VM. Blockade of antibody-induced glomerulonephritis with Crry-Ig, a soluble murine complement inhibitor. J Immunol. 1998; 160: 4553–4560.
Lardenoye JH, de Vries MR, Lowik CW, Xu Q, Dhore CR, Cleutjens JP, van Hinsbergh VW, van Bockel JH, Quax PH. Accelerated atherosclerosis and calcification in vein grafts: a study in APOE*3 Leiden transgenic mice. Circ Res. 2002; 91: 577–584.
van den Maagdenberg AM, Hofker MH, Krimpenfort PJ, de Bruijn I, van Vlijmen B, van der Boom H, Havekes LM, Frants RR. Transgenic mice carrying the apolipoprotein E3-Leiden gene exhibit hyperlipoproteinemia. J Biol Chem. 1993; 268: 10540–10545.
Lardenoye JH, Delsing DJ, de Vries MR, Deckers MM, Princen HM, Havekes LM, van Hinsbergh VW, van Bockel JH, Quax PH. Accelerated atherosclerosis by placement of a perivascular cuff and a cholesterol-rich diet in ApoE*3Leiden transgenic mice. Circ Res. 2000; 87: 248–253.
Park P, Haas M, Cunningham PN, Alexander JJ, Bao L, Guthridge JM, Krauss DM, Holers VM, Quigg RJ. Inhibiting the complement system does not reduce injury in renal ischemia reperfusion. J Am Soc Nephrol. 2001; 12: 1383–1390.
Thurman JM, Ljubanovic D, Edelstein CL, Gilkeson GS, Holers VM. Lack of a functional alternative complement pathway ameliorates ischemic acute renal failure in mice. J Immunol. 2003; 170: 1517–1523.
Torzewski M, Klouche M, Hock J, Messner M, Dorweiler B, Torzewski J, Gabbert HE, Bhakdi S. Immunohistochemical demonstration of enzymatically modified human LDL and its colocalization with the terminal complement complex in the early atherosclerotic lesion. Arterioscler Thromb Vasc Biol. 1998; 18: 369–378.
Bhakdi S, Dorweiler B, Kirchmann R, Torzewski J, Weise E, Tarum-Jensen J, Walev I, Wieland E. On the pathogenesis of atherosclerosis: enzymatic transformation of human low density lipoprotein to an atherogenic moiety. J Exp Med. 1995; 182: 1959–1971.
Seifert PS, Hugo F, Tranum-Jensen J, Zahringer U, Muhly M, Bhakdi S. Isolation and characterization of a complement-activating lipid extracted from human atherosclerotic lesions. J Exp Med. 1990; 172: 547–557.
Tsuji S, Kaji K, Nagasawa S. Activation of the alternative pathway of human complement by apoptotic human umbilical vein endothelial cells. J Biochem (Tokyo). 1994; 116: 794–800.
Aortocoronary and peripheral vein grafts are known to have high failure rates (10% to 40% after 1 year and 50% to 60% after 10 years). As a consequence, relapse of ischemic symptoms may occur, requiring repeat bypass surgery or (more severely) amputation. Late vein graft failure occurs because of the development of intimal hyperplasia and accelerated atherosclerosis, which lead to vein graft thickening and graft failure. Because inflammation seems to play a pivotal role in these processes, the involvement of the complement system in the development of accelerated atherosclerosis was studied, and the results are described in the present article. By unraveling the mechanisms of accelerated atherosclerosis and vein graft thickening in general and specifically the involvement of the complement system, the authors believe that this might lead to potential new targets for therapy to prevent vein graft failure in patients.