|Home | About | Journals | Submit | Contact Us | Français|
To examine the roles of the membrane attack complex of complement and its sole membrane regulator, CD59, in atherosclerosis.
C6 (C6−/−) deficient and CD59a (Cd59a−/−) knockout mice were separately crossed onto the apolipoprotein E knockout (apoE−/−) background. The double knockout mice were fed high-fat diet in order to study the effects of absence of C6 or CD59a on the progression of atherosclerosis.
C6 deficiency significantly reduced plaque area and disease severity. CD59a had the opposite effect in that deficiency was associated with a significant increase in plaque area, correlating with increased membrane attack complex (MAC) deposition in the plaque and increased smooth muscle cell proliferation in early plaques.
Our results demonstrate that the MAC contributes to the development of atherosclerosis, C6 deficiency being protective and CD59a deficiency exacerbating disease.
Over the last two decades it has become increasingly clear that atherosclerosis is associated with chronic inflammation (Fan and Watanabe, 2003; Nilsson and Hansson, 2008; Ross, 1999; Seifert and Kazatchkine, 1988). While the causes of the initial insult to the artery wall remain unclear, evidence has accumulated implicating the complement system in disease progression (Meuwissen et al., 2006; Niculescu et al., 2004; Vlaicu et al., 1985; Yasojima et al., 2001). Despite these findings, studies on the role of the complement system in atherosclerosis using animal models have so far produced conflicting results (Bhatia et al., 2007; Patel et al., 2001; Persson et al., 2004; Schmiedt et al., 1998). The deletion of C3 in apolipoprotein-E-deficient (apoE−/−) and C1q in low density lipoprotein receptor (ldl-r)-deficient mice caused a worsening of the disease in both cases, implying that C1q which initiates the classical pathway of complement, and C3, the central orchestrator of all complement activation pathways, have anti-atherogenic roles (Bhatia et al., 2007; Persson et al., 2004). In contrast a recent paper by Leung et al. looked at the effect of DAF deficiency in the ldl-r mouse model of atherosclerosis. They found a worsening of disease in the absence of DAF (Leung et al., 2009). C5 deficiency had little or no effect in the apoE-deficient model, leading the authors to suggest that the membrane attack complex (MAC) had no role in this model of atherosclerosis (Patel et al., 2001). Contrasting data were obtained from studies in C6-deficient rabbits which, when fed a high-fat diet, developed less atherosclerosis than their C6 sufficient controls (Schmiedt et al., 1998). MAC formation is blocked in both C5-deficient mice and C6-deficient rabbits, suggesting that there may be species differences in the roles of MAC in atherosclerosis; alternatively, it is possible that C5a, the other product of C5 cleavage, has a hitherto unrecognised protective role in atherosclerosis.
Regulation of the terminal pathway has also recently been examined in mice through the generation of animals deficient in CD59a, the major regulator of MAC formation in the mouse. CD59 functions by blocking the interaction of the C5b-8 complex with C9, thereby preventing the formation of the lytic membrane attack complex (MAC). These laboratories employed different models of the disease, utilising either the ldl-r or the apoE knockout mouse respectively (Wu et al., 2009; Yun et al., 2008; An et al., 2009). Mice have two genes for CD59: Cd59a and Cd59b. Cd59a is widely expressed and the main regulator of membrane attack complex assembly in the mouse (Baalasubramanian et al., 2004), while Cd59b is highly expressed only in testis (Donev et al., 2008). Yun et al. (2008) tested ldl-r−/− mice deficient in Cd59a and showed increased plaque formation, while Wu et al. (2009) examined apoE−/− mice deficient in Cd59a and Cd59b and similarly demonstrated a worsening of disease. The latter paper also showed that treatment of the triple-deficient mice while on high-fat diets with the anti-C5 monoclonal antibody BB5.1 caused a significant lessening of disease severity. This result contradicts the finding that deficiency of C5 in the apoE−/− mouse did not inhibit progression of atherosclerosis (Patel et al., 2001).
This study was initiated to clearly define the role of MAC during the development of atherosclerosis by characterising the effect of C6 deficiency on the progression of atherosclerosis in the apoE−/− mouse model. We have undertaken a direct comparison of the effects of C6 and CD59a deficiency in the same colony of apoE−/− mice. C6-deficient mice are particularly relevant for investigating the pathogenic roles of MAC, since they have previously been used to implicate MAC as a causative agent in a wide range of diseases, including reperfusion injury, glomerular damage, and xenograft hyperacute rejection (Falk et al., 1983; Fondevila et al., 2008; McCurry et al., 1995). ApoE deficient mice lacking C6 were markedly protected from atherosclerosis compared to C6 sufficient controls. In contrast, CD59a deficiency significantly exacerbated atherosclerosis, in agreement with recent published studies and supportive of a key role of MAC in disease progression.
Additional methods are provided in Supplementary material.
Male apoE−/−/C6−/−; apoE−/−/Cd59a−/− mice together with litter-matched controls were fed high-fat diet containing 21% (wt/wt) pork lard and supplemented with 0.15% (wt/wt) cholesterol (Special Diet Services, Witham, UK) for 8 or 12 weeks starting at 8 weeks of age. Animals were housed in a specific pathogen-free environment. All studies and protocols were approved by the institutional Ethics Review Committee and by the United Kingdom Home Office and conformed to the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised 1996).
Animals were anaesthetised and processed as previously described (Rosenfeld et al., 2002). Brachiocephalic arteries were removed with a piece of the aortic arch and the stump of the right subclavian artery still attached to aid orientation during histological processing. These were immediately embedded in optimum cutting temperature (OCT) compound (Raymond A Lamb Limited, Eastbourne, UK) and snap-frozen in liquid nitrogen.
Serial transverse sections, of 7 μm thickness, were cut along the brachiocephalic artery, starting from the proximal end. Sections were stained with Miller's elastin/van Gieson or oil red O (both from Sigma–Aldrich, Poole, UK). Sections were immunostained for the following complement regulators: DAF (13 μg/mL) (rat anti-mouse DAF, 2C6, prepared in-house using standard immunisation procedures) (Spiller et al., 1999); Crry (2.6 μg/mL) (rat anti-mouse Crry mAb 5D5, a generous gift from Dr. M. Holers, Denver, USA) (Li et al., 1993); and Cd59a (10 μg/mL) (rat anti-mouse CD59a mAb, mCD59a.7, a kind gift from Dr. C.L. Harris, Cardiff University). Macrophages, smooth muscle cells and T cells were identified using anti-murine macrophage antibody (0.1 μg/mL) (MOMA-2, Serotec, Oxford, UK), anti-α-smooth muscle actin (diluted 1:400) (clone α-1-A4; Sigma–Aldrich) and hamster anti-CD3 (20 μg/mL) (48-2B; Santa Cruz) respectively.
Immunostaining was also performed for complement components C3 (20 μg/mL) (rat anti-mouse C3 mAb 11H9; Hy-Cult Biotechnology, The Netherlands), and C9/MAC (2 μg/mL) (rabbit anti-rat C9 prepared in-house using standard immunisation procedures).
In each case positive staining was expressed as a percentage fractional area of the lesion as analysed by a computerised image-analysis program (Image ProPlus™ software version 6.3, Media Cybernetics, Carlsbad, CA, USA).
Morphometric analyses were carried out on elastin-stained sections. One section was quantified per mouse at the same position along the brachiocephalic artery, following the established method of Johnson et al. (2005). Morphometry was performed using Image ProPlus™ software as above. The lengths of the internal and external elasticae were recorded. These were used to derive the media area by assuming them to be the circumferences of perfect circles. The plaque area was measured directly and was subtracted from the area enclosed by the internal elastic lamina to derive the true lumen area.
Mouse blood (~1 mL) was collected into tubes without anticoagulant, allowed to clot at room temperature and then incubated on ice for 1 h. Serum was separated by centrifugation and analysed for triglyceride, cholesterol and lipaemia index on an automated analyser (Clinical Biochemistry Laboratories, Cardiff University Hospital).
Oil red O-stained sections were used to determine plaque lipid content. The total stained area in the plaque was expressed as a percentage of the total plaque area to give the fractional lipid content.
Data are expressed as mean ± SEM and significance tested by two-tailed unpaired Student's t-test (GraphPad Prism software version 3.0), with significance assumed at P < 0.05.
To delineate the role of the MAC during atherosclerosis we crossed apoE−/− mice with C6−/− mice generating apoE−/−/C6−/− and litter-matched controls (apoE−/−/C6+/+). Having no C6, these mice were incapable of MAC formation from birth. The mice were fed a high-fat diet for 12 weeks to induce severe disease, in the expectation that this would amplify differences between the test and control groups. Analyses of the brachiocephalic arteries revealed that the mean plaque cross-sectional area in apoE−/−/C6−/− mice was significantly decreased compared to the apoE−/−/C6+/+ controls. Fig. 1A and C shows representative sections from apoE−/−/C6−/− mice while Fig. 1B and D shows representative sections from apoE−/− control mice. Additionally, (C) and (D) shows the area defined as “plaque”, superimposed onto the image. While Fig. 1E shows the pooled data (66.01 ± 23.2 × 103 μm2 versus 179.5 ± 14.1 × 103 μm2; P < 0.001). Table 1 summarises the morphometric analyses, including plaque, lumen, media and vessel area. The mean lumen and media areas were not different between the groups. However the total vessel area was significantly decreased in the apoE−/−/C6−/− mice compared to apoE−/− controls (349.9 ± 39.6 × 103 μm2 versus 469.8 ± 28.2 × 103 μm2; P < 0.05). MAC deposition was abundant in plaques of control mice, both on the endothelium (inset, thick white arrow) and in the necrotic core (inset, large non-cellular area) (Fig. 1G). In contrast, no staining for MAC above background levels was found in plaques from apoE−/−/C6−/− mice (Fig. 1F). Fig. 1H shows the pooled data for MAC deposition (apoE−/−/C6−/−: 12.8 ± 3.5% versus apoE−/−: 43.0 ± 6.5%; P ≤ 0.01). Note that the absence of MAC staining in the C6−/− mice provides verification for the use of the anti-C9 antiserum as a MAC marker because these mice have normal plasma levels of C9 (data not shown).
To examine the role of regulation of MAC formation during atherosclerosis we generated mice deficient in both apoE and Cd59a and subjected them (together with their controls) to a period of high-fat feeding. Analysis of the brachiocephalic arteries revealed that average plaque size in apoE−/−/Cd59a−/− mice was double that seen in gender, age, strain and litter-matched apoE−/− controls on an identical diet. Fig. 2A and B shows representative sections from apoE−/−/Cd59a−/− and apoE−/− mice respectively. The pooled data are shown in Fig. 2C (59.9 ± 13.1 × 103 μm2 versus 28.2 ± 7.9 × 103 μm2; P < 0.05). Table 2 summarises the morphometric analyses, including plaque, lumen, media and vessel area. The mean lumen area, the media area and the total vessel area were not different between the groups. There were no significant differences in body weight, heart:body weight ratios, serum triglyceride or cholesterol levels, the lipaemia index, or plaque lipid content between apoE−/−/Cd59a−/− and apoE−/− mice (supplemental data, Table 1 and Fig. 1).
The extent of terminal complement pathway activation was examined by staining for C9 deposition as a surrogate marker of MAC (verified as described above). Complement activation was also assessed by staining for C3 fragment deposition. MAC staining was absent from unaffected vessel walls (i.e. those with no plaque), but clearly present in early and late stage plaques obtained from both apoE−/−/Cd59a−/− (Fig. 3A) and apoE−/− mice (Fig. 3B). To better visualise and analyse the expected increase in MAC staining in apoE−/−/Cd59a−/− mice, fluorescence intensity detection limits were set at a high sensitivity level, hence the apparently low levels of MAC staining recorded in apoE controls. MAC deposits were more than 10-fold increased in the apoE−/−/Cd59a−/− mice compared to the apoE−/− controls (Fig. 3C; 28.9 ± 9.4% versus 2.7 ± 0.8%; P ≤ 0.05). C3 deposition was detected weakly in unaffected artery walls and strongly in plaques from the brachiocephalic arteries of both apoE−/−/Cd59a−/− and apoE−/− mice (Fig. 3D and E, respectively). There was no significant difference in levels of C3 between the apoE−/−/Cd59a−/− mice and their apoE−/− controls (Fig. 3F). To assess whether increased MAC deposition in apoE−/−/Cd59a−/− mice would cause an increase in the numbers of inflammatory cells within the plaque we stained for macrophages and T cells. There were no significant differences in the proportions of plaque area staining for macrophages or T cells (as revealed by staining with MOMA-2 and CD3, respectively) between the two groups (Supplemental Data Figs. 2 and 3).
The membrane-bound complement regulators Crry and decay accelerating factor (DAF) were expressed in normal vessel walls and plaques in both groups of mice (Fig. 4A, B, D and E). There was no gross difference in the pattern or abundance of expression of either Crry or DAF between the groups. Cd59a expression was absent from apoE−/−/Cd59a−/− mice and present within the brachiocephalic arteries of apoE−/− mice (Fig. 4C and F, respectively).
Because of the known effect that the MAC has on smooth muscle cell proliferation (Benzaquen et al., 1994) we decided to quantify smooth muscle cell content by staining for α-actin in plaques from the brachiocephalic arteries of apoE−/−mice lacking the MAC regulator CD59a. Grouping the plaques into “early” (fatty streaks and fibrous plaques with cross-sectional areas <80 × 103 μm2) and “advanced” (complex plaques with cross-sectional areas ≥80 × 103 μm2) we found a significant increase in the smooth muscle cell content of “early” plaques from apoE−/−/Cd59a−/− mice as compared to apoE−/−mice (% α-actin staining 62.0% ± 7.1% versus 21.4 ± 6.1%; P < 0.01). Fig. 5A and B shows representative pictures of early plaques stained for smooth muscle cell α-actin. By contrast, in “advanced” plaques smooth muscle cell content was significantly reduced in the plaques from apoE−/−/Cd59a−/− mice versus apoE−/− controls (% α-actin staining 15.3 ± 4.8% versus 36.6 ± 6.7%; P < 0.05). Fig. 5C and D shows representative pictures of advanced plaques stained for smooth muscle cell α-actin. Combined data are presented in Fig. 5E. The proportion of plaque staining for α-actin in the apoE−/−/Cd59a−/− mice fell more than three-fold between “early” and “advanced” plaques (62.0% ± 7.1% versus 15.3 ± 4.8%; P < 0.01). Fig. 5F–I shows C9 deposition in smooth muscle cell regions of an atherosclerotic plaque from an apoE deficient mouse.
In this paper we set out to unequivocally define the roles of the MAC in atherosclerotic plaque formation in the apoE−/− mouse model. Importantly, we clearly show, for the first time, that deficiency of C6 is strongly protective against progression of atherosclerosis in apoE−/− mice. This result is consistent with published studies in fat fed C6-deficient rabbits and strongly implicates the MAC in both of these models of atherosclerosis (Schmiedt et al., 1998). Secondly, we tested the effects of deficiency of the major murine regulator of MAC formation, CD59a, and showed that apoE−/−/Cd59a−/− mice developed much larger plaques on fat feeding when compared to closely matched apoE−/− controls; MAC deposition in the plaques was also markedly increased in apoE−/−/Cd59a−/− mice. Of note we have also shown in these mice that the smooth muscle cell content of advanced plaques is significantly less than that seen in apoE−/− controls. This finding suggests a possible mechanism whereby MAC formation could contribute not only to plaque development but also to plaque instability and hence the incidence of vessel rupture and consequent acute infarcts or strokes.
The data that we have presented here on the contrasting effects of C6 and CD59a deficiency on plaque development in apoE−/− mice clearly support the hypothesis that the MAC is an important contributor to atherosclerotic plaque development in the apoE−/− mouse model. In fat fed apoE−/−/C6−/− mice, plaque size was markedly reduced and staining for MAC deposition within the plaques was also significantly reduced; the converse was true in apoE−/−/Cd59a−/− mice with much larger plaques compared to controls and increased MAC deposition. The mean total vessel area in apoE−/−/C6−/− mice was significantly smaller than that seen in apoE−/− controls, while the lumen area remained the same between the two groups. This is due to arterial remodelling which occurs during atherosclerosis and maintains the lumen area in the face of increased obstruction caused by plaque growth. The extent of disease in these mice correlated with the extent of MAC deposition. In contrast to the markedly elevated MAC deposition, C3 staining was seen strongly in plaques from both apoE−/−/Cd59a−/− and apoE−/− controls, an observation which has been previously noted (An et al., 2009). Most of the MAC deposited in the plaques is associated with extracellular debris and lipid, which accumulate in the core of advanced plaques. Activation of complement at this location is probably due to the presence of oxidised and enzymatically modified LDL, known to activate complement through the alternative pathway (Bhakdi, 1998). Of note, the expression of the other broadly expressed murine membrane complement regulators DAF and Crry was not different between the two groups, indicating that deficiency of CD59a was not compensated by increased expression of other regulators. Plaque lipid content and macrophage and T cell infiltration were also similar in test and control groups.
Smooth muscle cell proliferation is a characteristic feature of plaque formation. In small, simple plaques from apoE−/−/Cd59a−/− mice, smooth muscle cell accumulation was markedly increased compared to apoE−/− controls, suggesting accelerated plaque development. In addition we observed an association between C9 deposition and regions of smooth muscle cell within atherosclerotic plaques. The MAC has previously been shown to cause aortic smooth muscle cell proliferation in vitro (Niculescu et al., 1999), which strengthens our impression that this is a MAC triggered event. In contrast, larger and more advanced plaques in apoE−/−/Cd59a−/− mice contained fewer smooth muscle cells than similar sized plaques from apoE−/− controls, likely due to MAC-induced cytolysis. Reduced smooth muscle cell number in these late plaques in apoE−/−/Cd59a−/− mice will render the plaques unstable and vulnerable to rupture. These observations are compatible with those of Wu et al. (2009) who noted that the plaques present in their apoE−/−/Cd59a−/−/Cd59b−/− mice had a more vulnerable phenotype than those of the apoE−/− controls.
MAC deposition is a characteristic feature of human atherosclerotic lesions; the pattern of MAC deposition in murine plaques closely resembles that seen in humans (Rus et al., 1988). It is therefore reasonable to propose that MAC contributes to plaque formation and progression in human atherosclerosis in a similar manner to that which we and others have now demonstrated occurs in mice. Our data are in agreement with Wu et al., who also implicated the terminal pathway in the progression of atherosclerosis. More recently, Leung et al. (2009) have shown that DAF deficiency exacerbates disease in the ldl-r mouse model of atherosclerosis again implying that unregulated complement activation accelerates the progression if atherosclerosis. However, these findings are at odds with a previous study which had showed no effect of terminal pathway disruption (at the level of C5) on the progression of atherosclerosis in fat fed apoE−/− mice (Patel et al., 2001). These contrasting findings highlight the need for further investigation into the role of C5 and C6 in atherosclerosis.
Inhibition of MAC formation is a potential strategy for therapy of atherosclerosis and a realistic one given that terminal pathway inhibitors are already in use in the clinic (Davis, 2008), with more under development (Song et al., 2003).
Funding: This work was supported by the British Heart Foundation [Studentship number: FS/05/087/19466] and the Wellcome Trust [Programme Grant number: 068590 to BPM].
We thank Marieta Ruseva, Rhodri Turner, Dr Simone Meuter, Dr James Neal and Dr Claudia Calder for help and advice with immunohistochemistry, Dr Paul Brennan and Dr Claire Harris for critical reading of the manuscript. We also thank all the staff at the Biomedical Services Unit at Cardiff University for technical support.
Appendix ASupplementary data associated with this article can be found, in the online version, at doi:10.1016/j.molimm.2009.10.035.