当前位置: 首页 > 期刊 > 《动脉硬化血栓血管生物学》 > 2004年第11期 > 正文
编号:11330181
Inhibition of Tumor Necrosis Factor- Reduces Atherosclerosis in Apolipoprotein E Knockout Mice
http://www.100md.com 《动脉硬化血栓血管生物学》
     From the Department of Medicine (L.B., M.N., E.B., J.N., S.J.), University Hospital MAS, Lund University; the Department of Pharmacy (L.H.), The Danish University of Pharmaceutical Sciences; the Department of Cardiology (S.J.), University Hospital MAS, Lund University; and Lund Strategic Research Center for Stem Cell Biology and Cell Therapy (S.J.), Lund Universit.

    ABSTRACT

    Objective— Inflammation plays an important role in atherosclerosis. One of the most potent pro-inflammatory cytokines is tumor necrosis factor- (TNF-), a cytokine identified to have a pathogenic role in chronic inflammatory diseases such as rheumatoid arthritis (RA). The aim of the study was to evaluate the importance of TNF- in atherogenesis.

    Methods and Results— Mice deficient in both apolipoprotein E (apoE) and TNF- were compared regarding their atherosclerotic burden. Mice were fed a Western-style diet (WD) or normal chow. Mice deficient in both apoE and TNF- exhibited a 50% (P=0.035) reduction of relative lesion size after 10 weeks of WD. Bone marrow transplantation of apoEo mice with apoEotnf-o bone marrow resulted in a 83% (P=0.021) reduction after 25 weeks on WD. In apoE knockout mice treated with recombinant soluble TNF receptor I releasing pellets, there was a reduction in relative lesion size after 25 weeks of 75% (P=0.018).

    Conclusions— These findings demonstrate that TNF- is actively involved in the progression of atherosclerosis. Accordingly, TNF- represents a possible target for prevention of atherosclerosis. This may be of particular importance in rheumatoid arthritis because these patients have an increased risk for cardiovascular disease.

    Atherosclerosis is an inflammatory disease. We have shown that deficiency in TNF- reduces atherosclerosis. This implies that TNF antagonizers are interesting candidates for testing in clinical studies of atherosclerosis prevention.

    Key Words: atherosclerosis ? genetically altered mice ? TNF-

    Introduction

    Tumor necrosis factor- (TNF-) participates in the direction of the inflammatory response and influences the balance of blood lipid composition.1,2 TNF- is involved in practically every step of inflammation from initiation to downregulation3 and each characteristic lesion of atherosclerosis represents a different stage of a chronic inflammatory process in the artery.4 By being a potent stimulator of several of the matrix metalloproteinases5,6 and of plasminogen activator inhibitor-1,7 it has been identified as a player in several of the complications of atherosclerosis. TNF- has 2 identified receptors (TNF-RI and TNF-RII), which both exist in transmembranous and soluble forms. The receptors are shared by lymphotoxin- (LT-).8 Although TNF- is mainly synthesized by macrophages, several other cells synthesize it as well. LT-, however, is a cytokine primarily secreted by T cells. In contrast to TNF-, LT- only exist transiently in membrane-bound form.9 The interplay between the ligands and their receptors are only vaguely sketched but there seems to be a regulatory balance in the ratio between them.

    Although inflammation is a natural defense mechanism against toxic compounds such as oxidized proteins and lipids, it presents a major hazard to individuals inflicted by inflammatory diseases such as rheumatoid arthritis (RA) and Crohn’s bowel disease (CBD). Interestingly, patients with RA have increased morbidity and mortality in atherosclerotic complications.10,11 The specific TNF- inhibitor, infliximab, as well as the combined TNF- and LT- inhibitor, etanercept, have been shown to be efficient in the treatment of RA.12 In contrast, CBD patients are helped only by the specific TNF- inhibitor infliximab.13,14 Apparently, there are different roles in different diseases for these ligands, despite the shared receptor population.

    We have studied apoE/tnf- knockout mice compared with apoEo mice with regard to plaque forming ability in response to Western-style diet (WD). The impact of loss of TNF- synthesized by the inflammatory system was studied by bone marrow (BM) transplantation. BM was transplanted from apoEo/tnf-o to lethally irradiated apoEo mice. To study the effect of pharmacological inhibition of TNF-, animals were treated with a subcutaneous pellet containing soluble TNF-RI.

    The apoE knockout mouse has high levels of plasma cholesterol as a result of impaired clearance of cholesterol-enriched lipoproteins, and it develops complex atherosclerotic lesions throughout the arterial tree.15 The TNF-–deficient mice lack splenic primary B-follicles and cannot form organized follicular dendritic cell networks and germinal centers. Their responses to infection and toxins are altered, as well as their hypersensitivity reactions.16 The combined apoE/tnf- knockout is viable and fertile and can be bred in normal animal housing.

    Methods

    Animals and Diet

    C57BL/6J x 129S tnf-o mice (B6,129-Tnftm1Gkl, The Jackson Laboratory, Bar Harbor, Me) were bred to C57BL/6J apoE-deficient mice (C57BL/6JBom-Apoetm1Unc, Bomholtgard Breeding and Research Center, Ry, Denmark). ApoEo/tnf-o, apoEo, tnf-o, and wild-type mice were selected and intercrossed to yield F3 progeny, which served as subjects in the experiments. Food and tap water were administered ad libitum. The local animal care ethical committee approved all animal experiments.

    Tissue Harvesting

    A flat preparation was performed as previously described.17 Briefly, the aorta was dissected, adventitial fat carefully removed, and a longitudinal cut made along the full length of the abdominal and thoracic aorta, which was mounted en face. The heart and aortic arch were also carefully dissected and sectioned from the aortic origin to the branching of the left carotid artery. The novelty of the procedure used is that mouse aortas are mounted on a transparent support where they are accessible for direct microscopical observation. The illumination of the vessel wall allows, in contrast to previous methods, an unequivocal distinction between fatty lesions and adventitial fat.17 The flat preps and the first slide in each series of the sections were stained with Oil Red O (CI 26125; solvent red 27; SIGMA Chemicals).

    Knockout Study

    The mice were weaned at the age of 4 weeks and put on WD (21% cocoa butter, 0.15% cholesterol; Analycen, Sweden). After 10 weeks, the mice were euthanized (apoEo, n=11; apoEo/tnf-o, n=8). A smaller number of animals were kept for 40 weeks on WD before being euthanized.

    Chimeric Soluble Receptor

    A subcutaneously implanted pellet delivered a fusion protein between murine sTNF-RI and human IgG Fc (R&D Systems Europe Ltd). The formula of the receptor delivery pellets was tested for 2 different amounts of recombinant murine sTNF-RI/Fc chimera (R&D Systems Europe Ltd). The pellet composition (ethylene-vinyl copolymer, albumin, and sodium phosphate buffer) was based on a previous publication by Eliaz et al.18 The receptors were mixed into the carrier: 20% albumin, 3.75 or 33.3% albumin, 7.5 mg. Release profiles were obtained by agitating the pellets in vials of phosphate-buffered saline on a shaker. An enzyme-linked immunosorbent assay test recognizing murine sTNF-RI (Quantikine M, R&D Systems Europe Ltd) was performed on the suspensions to test the amount released at different time points: 0, 2, 4, 7, 14, 21, and 28 days.

    Pellet Implantation

    The mice receiving sTNF-RI pellet were male apoEo mice on a C57BL/6J background. The mice were fed normal chow. Each mouse got a pellet implanted subcutaneously in the abdomen at 7 weeks of age. The implantation was performed under fentanyl (0.315 mg/mL)/fluanison (10 mg/mL) (Janssen-Cilag AB) midazolam (5 mg/mL) (Hoffmann-La Roche Ltd) anesthesia. Fentanyl/fluanison was mixed with sterile water 1:1, as was midazolam; the 2 mixtures were combined in equal amounts and administered intraperitoneally, 0.2 mL/20 g bodyweight. The study comprised 3 groups, with each group receiving a specific pellet. Group 1 received placebo, pellet without fusion protein (n=7). Group 2 received 10 μg fusion protein in pellet (n=4). Group 3 received 20 μg fusion protein in pellet (n=5). After 25 weeks, the animals were euthanized and the heart and aorta were perfusion fixed with Histochoise (Amresco), mounted en face and eventually stained with oil red O (CI 26125; solvent red 27).

    BM Transplantation

    BM from male apoEo/tnf-o mice (n=7) was transplanted at age 10 weeks to female apoEo mice (n=7), and as control male apoEo BM (n=8) was transplanted to female (n=8) apoEo mice. The mice were euthanized after 25 weeks on WD. BM cells were isolated by flushing the femurs and tibias from either male apoEo or apoEotnf-o mice with Iscove’s modified Dulbecco’s medium (Invitrogen Ltd) with added 100 U/mL penicillin (Invitrogen Ltd) and 100 μg/mL streptomycin (Invitrogen Ltd), 0.25 μg/mL amphotericin B (Invitrogen Ltd), 20% fetal calf serum and 2 mmol/L L-glutamate (Invitrogen Ltd). The cells were treated with 0.2% NaCl to lyse erythrocytes. Remaining cells were washed in phosphate-buffered saline, centrifuged, and dissolved in the described medium.

    Female apoE-deficient mice were exposed to a single dose of 9 Gy on the day of the transplantation. Irradiated recipients received 0.5x106 BM cells by intravenous tail vein injection. One day before and 1 week after the BM transplantation, the recipient mice were given 2 mg/mL neomycin sulfate (Sigma-Aldrich) in acidified drinking water.

    Polymerase Chain Reaction

    ApoEo/tnf-o, apoEo, tnf-o, and wild-type mice were identified in three-primer polymerase chain reactions (Table I, available online at http://atvb.ahajournals.org). DNA was extracted from tail tips as previously described by Talts et al.19

    Polymerase chain reaction was also used to identify successful BM transplantations. Whole blood was collected at the time of euthanization. Blood cells were pelleted and stored at –80°C until analysis. The pellets were lysed in equal volumes of buffer (10 mmol/L Tris, pH 8, 1.5 mmol/L MgCl2, 1% Tween-20), and 400 mg/mL proteinase K was added for digestion. Proteinase K was inactivated, debris pelleted, and DNA precipitated from the supernatant. A 310-bp band and a 290-bp band were amplified for the X-chromosome and Y-chromosome, respectively.20

    Morphometric Analysis

    Morphometric analysis was performed using computer-aided analysis software. The slides were photographed with an Olympus U-TU1X-2 camera connected to an Olympus CX41 microscope and a computer. The vessel structures (plaques, vessel area) were traced and areas calculated in Imageproplus (Media Cybernetics Inc).

    Determination of Plasma Lipid Profile

    Mice were fasted overnight, and blood was collected under anesthesia (Isoflurane; Abbott Scandinavia AB) by retro-orbital vein puncture. The lipid profile included total cholesterol (TC), triglycerides (TG), and lipoprotein profile as lipid in low-density lipoprotein plus very-low-density lipoprotein, high-density lipoprotein, and free fatty acid. TC and TGs were measured in enzymatic colorimetric assays (Sigma-Aldrich Company Ltd), whereas the lipoproteins were separated on an alkaline-buffered agarose gel and stained with Sudan black (SEBIA). The electrograms were scanned and the bands quantified by computer-aided analysis software Quantity One (Bio-Rad Laboratories AB).

    Immunohistochemistry

    To identify monocyte/macrophage infiltration, the rat–antimurine monocyte/macrophage antibody (MOMA-2 antibody; BMA Biomedicals) was used at the dilution 1:1000. T cells were identified by CD3 staining (monoclonal rat anti-mouse CD3, MAB4841; R&D Systems Inc) at a dilution 1:200; B cells by rat anti-mouse B-cell antigen (MCA-1386; Serotec Ltd) at 1:100, and vascular cell adhesion molecule-1 (VCAM-1) by rat anti-mouse VCAM-1 (BSA-12; R&D Systems Inc) at 1:250. The secondary antibody used in the stainings were mouse adsorbed biotinylated rabbit anti-rat IgG (BA4001; Vector Laboratories Ltd) at 1:70 for MOMA staining and 1:200 (T cells), 1:150 (B cells), and 1:500 (VCAM-1), followed by incubation with avidine–peroxidase complex and diaminobenzene substrate or NovaRED Substrate (Vector Laboratories Ltd). Sections were counterstained with Harris Hematoxylin.

    Statistics

    In normally distributed material, Student 2-tailed t test was used and results reported as mean±standard deviation (SD). When the material was skewed, Mann–Whitney rank sum test was performed and results reported as median followed by the mid-interquartile range withing parentheses: median (25%; 50%). P<0.05 was considered significant. All the calculations were performed with SigmaStat 2.03 (SPSS Inc) and graphs were made in SigmaPlot 2001 (SPSS Inc).

    Results

    Lipid Parameters

    TC and TG values were normally distributed in all groups. No significant differences were detected with Student t test at either time point for TG or TC. Nor did the lipid profiles of high-density lipoprotein, very-low-density lipoprotein, low-density lipoprotein, or free fatty acids differ between groups (Table 1). Although there were no statistical differences in TC values between groups, the levels tended to be higher in the TNF-–depleted mice at 10 weeks (Table 1; 479±273 versus 672±246 mg/dL) and in the pellet-treated mice (Table 1; 445±171 versus 722±197 mg/dL). The mice that underwent BM transplantation mice had, however, lower levels of both TGs and TC compared with apoEo and apoEotnf-o mice and pellet-treated mice. No significant difference in TC or TG was seen between mice transplanted with apoEo compared with mice transplanted with apoEo/tnf-o BM.

    TABLE 1. Lipid Profiles

    Release Profile of Chimeric sTNF-RI/Human IgG:Fc Pellets

    TNF- and LT- share receptors. To signal, a homotrimeric TNF- protein cross-binds at least 2 cell surface receptors. The receptor chimera used in this experiment is dimeric and thus competes strongly with the endogenous monomeric receptors. The murine-soluble TNR-RI is fused to the Fc fragment of human IgG. This construction conveys a longer half-life to the molecule compared with its natural equivalent.21 Pellets with different chimera concentrations were tested for their ability to release active substance to the environment and the formula was approved (Figure I, available online at http://atvb.ahajournals.org).

    Atherosclerotic Burden in Aortas

    At 10 weeks, the apoeEo/tnf-o mice had relative lesion areas of 0.20% (0.05; 0.03) as compared with 0.40% (0.32; 0.67 in the apoEo mice) (Figure 1; Figure II, available online at please see http://atvb.ahajournals.org). These data correlate with data from the smaller substudy in which mice were kept for 40 weeks before being euthanized (apoeEo/tnf-o 6.6% and apoEo 16.45% ; P=0.048) (Figure III, available online at please see http://atvb.ahajournals.org).

    Figure 1. Male mice given WD for 10 weeks. Flat preparations were performed and relative lesion areas calculated. The apoEotnf-o mice had a 50% reduction in relative lesion area compared with apoEo mice (P=0.035).

    Mice, 7 weeks old, were given either a placebo pellet or a pellet with 10 or 20 μg of the chimeric sTNF-RI. The mice were euthanized at age 32 weeks and aortas were harvested. The mice were given ordinary chow throughout the experiment. The mice treated with a 20-μg pellet showed significantly lesser lesion size than mice treated with control pellet. The group that received the 20-μg TNF-RI pellet had a lesion area of 0.80% (0.60; 1.62), whereas the mice given a placebo pellet had a lesion area of 3.20% (1.85; 15.5); P=0.018 (Figure 2). The apoEo mice transplanted with apoeEo/tnf-o BM had 83% lesser relative lesion area than apoEo mice transplanted with apoEo BM (Figure 3) (0.41% versus 2.47% ) (P=0.021). The success of the transplantations were confirmed by polymerase chain reaction (Figure IV, available online at http://atvb.ahajournals.org).

    Figure 2. Male mice were treated with a pellet containing sTNF-RI. After 25 weeks of normal chow, lesion coverage was measured. Mice treated with a 20-μg pellet showed a 75% reduction compared with mice treated with control pellet (P=0.018).

    Figure 3. At 10 weeks of age, female apoEo mice underwent transplantation with BM from apoEo/tnf-o male mice or apoEo male mice. The mice with transplanted apoEo/tnf-o BM had 83% less relative lesion area than those with transplanted apoEo BM after 25 weeks on WD (P=0.021).

    Immunohistochemistry

    Subvalvular plaques were stained with VCAM-1 antibody and MOMA-2 antibody. There were no differences between the apoEo and the apoEo/tnf-o mice with regard to these antigens. The sTNF-RI–treated mice had similar levels of macrophage presence and VCAM staining. Sections of the aortic arch were stained with CD3 and CD19 antibodies. There was no difference in CD19 staining between groups. CD3 staining relative to lesion area is nonsignificantly higher in the apoEo/tnf-o mice than in the apoEo mice. The group that received the highest dose of soluble receptor responded by a significant increase in relative CD3 stained area in the plaques (Table 2; P=0.04).

    TABLE 2. Immunohistochemistry

    Discussion

    Inflammation is a natural defense mechanism against toxic compounds such as oxidized proteins and lipids. It has a well-recognized role in the development of atherosclerotic lesions.4,22 The consequences of inflammation in atherosclerosis are difficult to predict. Although it may be beneficial at early stages, with properties reversing atherogenesis, it may be detrimental to the individual with more aggressive disease progression and plaque rupture at a later stage. Markers of inflammation such as C-reactive protein and TNF- have been associated with an increased risk for cardiovascular events in several clinical studies.2,23,24 TNF- is a key cytokine in the recruitment and in the activation of inflammatory cells. Further, it promotes matrix degradation,5,6 thus facilitating influx of inflammatory cells to the vessel wall.

    This study was performed to test the hypothesis that TNF- has a central role in atherogenesis, which would make it a promising target for future therapies. In fact, TNF intervention has already been proven beneficial in combating other inflammatory disorders such as RA and CBD. This is especially interesting in RA patients because they have increased cardiovascular morbidity and mortality.10,11 Patients affected by RA experience relief from a TNF- antibody (infliximab) or a recombinant TNF receptor (etanercept) alike, whereas in CBD patients only the TNF- antibody inhibits disease development.12–14,25–27 Our apoEo/tnf-o model follows a similar strategy to infliximab—a selective inhibition of TNF-—and our pharmacological inhibition mimics etanercept by being a sTNF-R, which blocks both TNF- and LT-. The inhibitory effect of treatment with recombinant sTNF-RI implies that TNF- plays an important role in disease development. Both experimental settings show a distinct inhibition of the atherosclerotic process. TNF- is a cytokine with a prominent effect on lipoprotein lipase activity.1 No significant differences were seen between TNF-–competent and TNF-–deficient animals in lipid parameters. This may be because of the hypertriglyceridemia of the apoE knockout mice that most likely overrides the expected triglyceridemia from loss of lipoprotein lipase inhibition by physiological levels of TNF-. TC values were nonsignificantly higher in TNF-–depleted groups, which, if anything, means that the loss of TNF- counteracts the pro-atherogenicity of high TC values. This suggests that the retardation of atherosclerosis is not secondary to lipid-dependent mechanisms. The mice that underwent BM transplantation generally had lower TG and TC values than did mice in the other experiments. We have previously noted that mice that underwent BM transplantation tended to have lower TC and TG values than nontransplanted mice (unpublished data). Van Eck et al28 have also reported lowered TGs for up to 10 weeks after transplantation. The effect of TNF- depletion in animals undergoing BM transplantation is, even with the lower TG values, as pronounced as in the mice that did not undergo BM transplantation. This further strengthens the view that TNF- has an atherogenic effect, which is not secondary to lipid levels. The mice undergoing BM transplantation received a WD, and normal chow would have resulted in even lower lipid levels.

    At 10 weeks, atherosclerotic mice that are deficient in TNF- have a lesser lesion area than their TNF-–competent peers do. This trend was also seen in mice kept on WD diet for 40 weeks before being euthanized. The same was true for the apoEo mice with implanted sTNF-RI pellets. Finally, the depletion of TNF- in BM cells gives a similar reduction in atherogenesis. The data from these different experimental setups all show that TNF- depletion or inhibition retard atherosclerotic development. This further strengthens the argument that TNF- inhibition is therapeutic regarding inhibition of atherosclerosis. These results oppose a report from 2001, which states that TNF- has no bearing on the development of atherosclerosis. In that report, Schreyer et al29 studied atherogenesis in female TNF-–deficient mice on a C57BL/6J background, a model which develops atherosclerosis only in the valvular plane and only after an atherogenic diet containing cholate. In those mice, a depletion of TNF- did not retard aortic sinus lesions in contrast to LT-–deficient mice. Because of the differences between models, the findings are not easily comparable. However, Canault et al30 have, in a model similar to the model used by Schreyer, shown an almost complete retardation of aortic sinus lesion development in tnf-o mice on C57BL/6J background and atherogenic diet. Moreover, Canault et al added a mouse model, which, by mutation of the TNF- membrane linker sequence, only expressed membrane-bound TNF-. They established that the mice expressing only membrane-bound TNF- had a lesser atherosclerotic burden than wild-type mice, but more than that of the tnf-0 mice.

    Atherosclerotic propagation is traditionally followed either by studying serial sections of the aortic root31 or by measuring plaque areas in aortas pinned to a black wax background.32 Our procedure differs in that it is a microscopically aided version that facilitates the distinction between fatty lesions and adventitial fat (online Figure II).17

    During the early stages of atherosclerosis, recruitment of inflammatory cells plays a central role. However, in our mice the monocyte/macrophage-staining MOMA-2 antibody and the mononuclear adhesion molecule VCAM-1 did not suggest a difference in the recruitment of monocyte/macrophages between TNF-–deficient and TNF-–competent animals. The observed reduction in lesion size might be explained by an attenuation of TNF-–induced monocyte/macrophage activation and/or differentiation. The sTNF-RI used in our experiments inhibits both TNF- and LT-. We fed these mice normal chow to be able to study the atherosclerotic process with a more diverse set of lesions, fatty lesions as well as intermediate and fibrous, than if fed WD.15 Animals treated with this receptor also showed a slower progression of atherosclerosis compared with the placebo-treated animals. T lymphocytes are the major sources of LT-.33 When the pellet releases sTNF-RI, LT- is inhibited. In response, the atheroma might raise its T-cell and or B-cell populations to restore levels of LT-. This reasoning was strengthened in the animals treated with pellets containing the highest dosage of sTNF-RI, which did exhibit significantly more T-cells relative to lesion size than the control group. Over time, B-cell presence in the plaques decreased in apoEo/tnf-o and apoEo mice, suggesting that local effects of B cells do not explain the difference in plaque progression.

    Although the results from the knockout study are dependent on a loss of TNF-, the pellet substudy relies on a combined inhibition of TNF- and LT-. The BM transplantation study reveals the importance of TNF- derived from the hematopoietic system. In summary, our findings indicate that TNF- depletion and combined LT-/TNF- inhibition do retard atherosclerotic development. With the exception of animals treated with the highest dose of sTNF-RI, the proportion of cell populations is, however, unchanged between different groups, indicating a proportional decrease in plaque composition in mice deficient in TNF- or treated with combined anti-TNF-/anti-LT- pellets. BM transplantation shows that TNF- derived from hematopoietic cells contributes significantly to atherogenesis. These findings urge further studies to unravel mechanisms in more detail. TNF antagonizing therapies may be interesting candidates for clinical studies in patients with aggressive atherosclerosis, especially those with comorbidity in RA.

    Acknowledgments

    This study was supported by grants from Swedish Society of Medicine (grant 2003-442), The Swedish Research Council (grant K2003-71x-13498-04A), The Crafoord Foundation (grant 20020701), The Emil & Vera Cornell Foundation, and Swedish Heart and Lung Foundation (grant 200141781). We thank Marie W. Lindholm, PhD, for valuable scientific advice concerning lipid analysis and Lena Wittgren for assistance with irradiation protocols.

    References

    Beutler B, Cerami A. Tumor necrosis, cachexia, shock, and inflammation: a common mediator. Annu Rev Biochem. 1988; 57: 505–518.

    Jovinge S, Hamsten A, Tornvall P, Proudler A, B?venholm P, Ericson CG. Evidence of a role of tumor necrosis factor alpha in disturbances of triglyceride and. Metabolism. 1998; 47: 113–118.

    Ruuls SR, Sedgewick JD. Unlinking tumor necrosis factor biology from the major histocompatibility complex: lessons from human genetics and animal models. Am J Hum Genet. 1999; 65: 294–301.

    Ross R. Atherosclerosis–an inflammatory disease. N Engl J Med. 1999; 340: 115–126.

    Rajavashisth TB, Xu XP, Jovinge S, Meisel S, Xu XO, Chai NN, Fishbein MC, Kaul S, Cercek B, Sharifi B, Shah PK. Membrane type 1 matrix metalloproteinase expression in human atherosclerotic plaques: evidence for activation by proinflammatory mediators. Circulation. 1999; 99: 3103–3109.

    Galis ZS, Muszynski M, Sukhova GK, Simon-Morrissey E, Libby P. Enhanced expression of vascular matrix metalloproteinases induced in vitro by cytokines and in regions of human atherosclerotic lesions. Ann N Y Acad Sci. 1995; 748: 501–507.

    Schleef RR, Bevilacqua MP, Sawdey M, Gimbrone MA Jr, Loskutoff DJ. Cytokine activation of vascular endothelium. Effects on tissue-type plasminogen activator and type 1 plasminogen activator inhibitor. J Biol Chem. 1988; 263: 5797–5803.

    Bazzoni F, Beutler B. The tumor necrosis factor ligand and receptor families. N Engl J Med. 1996; 334: 1717–1725.

    Browning JL, Dougas I, Ngam-ek A, Bourdon PR, Ehrenfels BN, Miatkowski K, Zafari M, Yampaglia AM, Lawton P, Meier W, et al. Characterization of surface lymphotoxin forms. Use of specific monoclonal antibodies and soluble receptors. J Immunol. 1995; 154: 33–46.

    Mutru O, Laakso M, Isomaki H, Koota K. Cardiovascular mortality in patients with rheumatoid arthritis. Cardiology. 1989; 76: 71–77.

    del Rincon ID, Williams K, Stern MP, Freeman GL, Escalante A. High incidence of cardiovascular events in a rheumatoid arthritis cohort not explained by traditional cardiac risk factors. Arthritis Rheum. 2001; 44: 2737–2745.

    Mikuls TR, Moreland LW. TNF blockade in the treatment of rheumatoid arthritis: infliximab versus etanercept. Expert Opin Pharmacother. 2001; 2: 75–84.

    Sandborn WJ, Hanauer SB, Katz S, Safdi M, Wolf DG, Baerg RD, Tremaine WJ, Johnson T, Diehl NN, Zinsmeister AR. Etanercept for active Crohn’s disease: a randomized, double-blind, placebo-controlled trial. Gastroenterology. 2001; 121: 1088–1094.

    Sandborn WJ. Strategies targeting tumor necrosis factor in Crohn’s disease. Acta Gastroenterol Belg. 2001; 64: 170–172.

    Nakashima Y, Plump AS, Raines EW, Breslow JL, Ross R. ApoE-deficient mice develop lesions of all phases of atherosclerosis throughout the arterial tree. Arterioscler Thromb. 1994; 14: 133–140.

    Pasparakis M, Alexopoulou L, Episkopou V, Kollias G. Immune and inflammatory response in TNF--deficient mice: a critical requirement for TNF in the formation of primary B cell follicles, follicular dendritic cell networks and germinal centetrs, and in the maturation of the humoral immune response. J Exp Med. 1996; 184: 1397–1411.

    Br?nén L, Pettersson L, Lindholm M, Zaina S. A procedure for obtaining whole mount mouse aortas that allows atherosclerotic lesions to be quantified. Histochem J. 2001; 33: 227–229.

    Eliaz R, Wallach D, Kost J. Long-term protection against the effects of tumour necrosis factor by controlled delivery of the soluble p55 TNF receptor. 1996; 8: 482–487.

    Talts JF, Brakebusch C, Fassler R. Integrin gene targeting. Methods Mol Biol. 1999; 129: 153–187.

    Zhang Y, McCormick LL, Desai SR, Wu C, Gilliam AC. Murine sclerodermatous graft-versus-host disease, a model for human scleroderma: cutaneous cytokines, chemokines, and immune cell activation. J Immunol. 2002; 168: 3088–3098.

    Fernandez-Botran R. Soluble cytokine receptors: novel immunotherapeutic agents. Expert Opin Invest Drugs. 2000; 9: 497–514.

    Hansson GK. Immune mechanisms in atherosclerosis. Arterioscler Thromb Vasc Biol. 2001; 21: 1876–1890.

    Ridker PM, Cushman M, Stampfer MJ, Tracy RP, Hennekens CH. Inflammation, aspirin, and the risk of cardiovascular disease in apparently healthy men. N Engl J Med. 1997; 336: 973–979.

    Ridker PM, Glynn RJ, Hennekens CH. C-reactive protein adds to the predictive value of total and HDL cholesterol in determining risk of first myocardial infarction. Circulation. 1998; 97: 2007–2011.

    Taylor PC. Anti-tumor necrosis factor therapies. Curr Opin Rheumatol. 2001; 13: 164–169.

    Murray KM, Dahl SL. Recombinant human tumor necrosis factor receptor (p75) Fc fusion protein (TNFR:Fc) in rheumatoid arthritis. Ann Pharmacother. 1997; 31: 1335–1338.

    LaDuca JR, Gaspari AA. Targeting tumor necrosis factor alpha. New drugs used to modulate inflammatory diseases. Dermatol Clin. 2001; 19: 617–635.

    Van Eck M, Herijgers N, Yates J, Pearce NJ, Hoogerbrugge PM, Groot PH, Van Berkel TJ. Bone marrow transplantation in apolipoprotein E-deficient mice. Effect of ApoE gene dosage on serum lipid concentrations, (beta)VLDL catabolism, and atherosclerosis. Arterioscler Thromb Vasc Biol. 1997; 17: 3117–3126.

    Schreyer SA, Vick CM, LeBoeuf RC. Loss of lymphotoxin-alpha, but not tumor necrosis factor-alpha reduces atherosclerosis in mice. J Biol Chem. 2002; 23: 23.

    Canault M, Peiretti F, Mueller C, Kopp F, Morange P, Rihs S, Portugal H, Juhan-Vague I, Nalbone G. Exclusive expression of transmembrane TNF- in mice reduces the inflammatory response in early lipid lesions of aortic sinus. Atherosclerosis. 2004; 172: 211–218.

    Paigen B, Morrow A, Holmes PA, Mitchell D, Williams RA. Quantitative assessment of atherosclerotic lesions in mice. Atherosclerosis. 1987; 68: 231–240.

    Tangirala RK, Rubin EM, Palinski W. Quantitation of atherosclerosis in murine models: correlation between lesions in the aortic. J Lipid Res. 1995; 36: 2320–2329.

    Ohshima Y, Yang LP, Avice MN, Kurimoto M, Nakajima T, Sergerie M, Demeure CE, Sarfati M, Delespesse G. Naive human CD4+ T cells are a major source of lymphotoxin alpha. J Immunol. 1999; 162: 3790–3794.(Lena Br?nén; Lars Hovgaar)