Expression of Macrophage Migration Inhibitory Factor in Different Stages of Human Atherosclerosis
Background— Atherosclerosis is a chronic inflammatory response of the arterial wall to injury. Macrophage migration inhibitory factor (MIF), a cytokine with potent inflammatory functions, was thus considered to be important in atherosclerotic lesion evolution.
Methods and Results— We studied the presence and distribution of MIF immunoreactivity (MIF-IR) and MIF mRNA in internal mammary arteries with a normal histology and arteries with plaques in different stages of human atherosclerosis. To address a potential role for the coactivator Jab1 as a cellular mediator of MIF effects in vascular tissue, we correlated the expression of MIF to that of Jab1 by using immunohistochemistry and coimmunoprecipitation. We further sought to determine a potential functional role for endothelium-derived MIF in early atherogenesis by studying the effects of oxidized LDL on MIF expression in cultured human umbilical vascular endothelial cells. The results showed that MIF-IR and Jab1-IR are found in all cell types present in atherosclerotic lesions, that MIF-IR is upregulated during progression of atherosclerosis, that MIF is produced locally in the arterial wall, and that all MIF+ cells are simultaneously Jab1+. Coimmunoprecipitation experiments demonstrated in vivo complex formation between MIF and Jab1 in plaques. MIF expression in human umbilical vascular endothelial cells and a macrophage line was upregulated after stimulation with oxidized LDL.
Conclusions— MIF is produced abundantly by various cells in all types of human atherosclerotic lesions and thus may play an important role in early plaque development and advanced complicated lesions. MIF-Jab1 complexes could serve critical regulatory functions in atherosclerotic lesion evolution.
Received November 9, 2001; revision received January 28, 2002; accepted January 28, 2002.
Macrophage migration inhibitory factor (MIF) is a pleiotropic cytokine that regulates a broad spectrum of inflammatory reactions1–9⇓⇓⇓⇓⇓⇓⇓⇓ (summarized in References 10 and 11).10,11⇓ MIF production is not restricted to immune cells because in addition to macrophages12 and lymphocytes,1 tissue cells can produce MIF.10 Vascular endothelial cells (ECs) and smooth muscle cells (VSMCs), both important in the process of atherogenesis, were found to produce MIF after inflammatory stimulation.13,14⇓ MIF plays a critical role in the evolution of several inflammatory diseases, such as septic shock, delayed-type hypersensitivity, chronic glomerulonephritis, and rheumatoid arthritis.2,4,7,8,15⇓⇓⇓⇓
The mechanisms governing MIF production and action are only partly unraveled. Inflammatory stimuli, such as tumor necrosis factor or endotoxin, lead to the release of preformed MIF and de novo MIF synthesis.2,12⇓ MIF exhibits catalytic properties,16,17⇓ and the Erk1/2 signaling pathway was shown to be involved in MIF-mediated cellular regulation.9 A receptor for MIF has not yet been identified. The transcriptional coactivator Jab118 was identified as a molecular target of MIF action.6 MIF-Jab1 complex formation was found to be involved in the regulation of the activating protein-1 (AP1) transcription pathway, c-Jun kinase activity, and cell cycle regulation, with MIF counterregulating Jab1 activity.6
The pathogenesis of atherosclerosis is characterized by a chronic inflammatory fibroproliferative response of the arterial wall to injury.19 Because MIF is a mediator of inflammatory processes and human diseases and because it was suggested that MIF expression may be linked to atherogenesis in rabbits, 14 we studied the presence and distribution of MIF immunoreactivity (MIF-IR) and MIF mRNA expression in different stages of human atherosclerosis. To investigate the possibility that Jab1 could be a cellular mediator of the effects of MIF in vascular tissue, we compared the presence of MIF in human atherosclerotic tissue with that of Jab1 by using immunohistochemistry and coimmunoprecipitation. Because macrophage accumulation in the vessel wall requires the adhesion and transmigration of blood monocytes across the endothelium and immobilization of cells at the site of injury, we sought to determine a potential functional role for endothelium-derived MIF in early atherogenesis by studying the effects of oxidized LDL (oxLDL) on MIF expression in human umbilical vascular endothelial cells (HUVECs).
Atherosclerotic lesions were graded as previously described.20
Histologically normal internal mammary arteries (IMAs) (n=10) from patients undergoing aorto-coronary bypass surgery, coronary arteries with adaptive intimal fibrosis (AIF, type I lesions20) that were taken from the explanted hearts from patients undergoing transplantation for idiopathic dilated cardiomyopathy (n=8), and aortic tissue specimens with AIF (type I lesions20) from the thoracic aorta from autopsied individuals (n=2) served as controls. Aortic fatty streak (FS) lesions (n=10) were obtained from the proximal aorta from autopsied individuals, whereas advanced atherosclerotic plaques (CPs) (n=10) were derived from patients undergoing carotid endarterectomy. Postmortem delay was <12 hours and did not affect the quality nor quantity of staining. Tissue was fixed and prepared by standard methods. Serial sections were stained with hematoxylin and eosin and used for immunohistochemistry. For in situ hybridization (ISH), reverse transcription–polymerase chain reaction (RT-PCR), and coimmunoprecipitations, IMA specimens with a normal histol-ogy and carotid endarterectomy specimens with CPs were snap-frozen after excision.
We used anti-human MIF IgG (1:200, AF-289-PB, R&D Systems), anti-human Jab1 IgG (1:200, FL 334, Santa Cruz Biotechnology), and anti-human CD3 (1:100, CD3, DAKO) as antibodies against MIF, Jab1, and T lymphocytes, respectively, and monoclonal antibodies against α-actin (1:600, M0851, DAKO) and macrophages (1:100, CD68 (KP1), DAKO). Immunohistochemistry was performed as described.21 For CD3 staining, sections were pretreated with proteinase K; for MIF, Jab1, CD68, and actin staining, antigens were unmasked by pressure cooking in 10 mmol/L citric acid, pH 6.
In Situ Hybridization
An MIF probe was produced as described22 with primers; sense: 5′-ATG CCG ATG TTC ATC GTA AAC ACC-3′, antisense: 5′-TTA GGC GAA GGT GGA GTT GTT CCA GC-3′, resulting in a 348 bp product. ISH was performed as described,22 with the use of an anti-digoxigenin antibody-conjugate (1:500, Roche). Controls were derived from hybridization of nonlabeled probes to matched sections and labeled probes to sections incubated with RNAse A before hybridization. No hybridization signals were detectable in the controls (not shown). In addition, Northern analysis with the MIF probe revealed a band of the expected size. To identify cell types showing positive ISH signals, double labeling (DL) was performed. After ISH, immunohistochemistry was performed with the ABC method used as described above with anti-actin or anti-CD68 antibodies.
RNA Isolation and RT-PCR
Tissue was homogenized with an Ultraturrax for 20 seconds in 400 μL of Qiagen RTL lysis buffer. RNA was isolated with the use of RNeasy (Qiagen). For RT, the M-MLV first-strand synthesis reagent of Invitrogen was used. MIF PCR was performed with the use of 35 cycles (94°C/50°C/72°C) with primers; sense: 5′-ATG CCG ATG TTC ATC GTA AAC ACC-3′ and antisense: 5′-TTA GGC GAA GGT GGA GTT GTT CCA GC-3′ (product: 348 bp). As internal reference, a GAPDH PCR was performed as described.22 PCR products were analyzed in 1% gels.
Coimmunoprecipitations of native MIF-Jab1 complexes from CPs were performed with the use of a modification of a published procedure.6 Briefly, specimens were resuspended in 600 μL HEPES-NaCl-Triton X-100 buffer and homogenized with an Ultraturrax for 20 seconds. Samples were divided, and equal portions used for coimmunoprecipitation of MIF-Jab1 complexes by anti-Jab1 antibody (FL-334, Santa Cruz, 90 minutes) and control incubations. Precipitation of immune complexes was achieved by Protein A Sepharose (2 mL HNT; 3 hours; washes in HNT). MIF was revealed by Western blotting with anti-MIF antibody.
OxLDL Treatment and HUVEC Experiments
Human LDL was prepared as described previously.23 Oxidation of LDL was determined by the FOXII assay23 and yielded 528 μmol oxLDL per mg protein. The degree of oxidation was also assessed by two other methods: increase of mobility in an agarose gel (≈1.4-fold compared with native LDL) and formation of thiobarbituric acid–reactive substances (4 mmol/L). HUVECs were purchased from Cell Systems and cultured in endothelial basal medium supplemented with hydrocortisone (1 μg/mL), bovine brain extract (3 μg/mL), gentamicin (30 μg/mL), amphotericin B (50 μg/mL), epidermal growth factor (10 μg/mL), and 10% FCS until the third passage. After detachment, cells were grown in 6-cm dishes for 18 hours. HUVECs were stimulated with oxLDL and cellular protein was prepared as described.24 For Western blot analysis, anti-MIF antibody was used at 1:100 and blots were developed by horseradish peroxidase–coupled antibody/ECL (Amersham). Blots were reprobed with anti-actin antibody and quantified by densitometry. RAW 264.7 macrophages were obtained from the American Type Culture Collection and cultured and stimulated with oxLDL as described.3,12,24⇓⇓ Analysis of MIF from the culture supernatants was performed by sandwich ELISA as described.9 The mouse monoclonal IgG1 used in the ELISA was a gift from Prof Richard Bucala (Yale University School of Medicine, New Haven, Conn).
Morphometric Analysis and Statistical Methods
With the use of morphometric software, 5 random nonoverlapping microscopic high-power fields of histologically distinct regions of each specimen were analyzed, scoring all nuclei and cells stained for MIF. A cell was scored positive when the typical cytoplasmic staining was associated with a cell nucleus. Data represent mean±SD. The Mann-Whitney rank sum test was applied for comparison of data from different plaque types with the controls. Probability values <0.05 were considered significant.
MIF Expression in IMAs With Normal Histology
In the specimens, a weak and diffuse cytoplasmic MIF-IR was present in 10 of 10 cases in 5% of the ECs and 10.7±8.6% of the medial VSMCs (α-actin+) (Figure 1A). Also, we demonstrated the presence of MIF mRNA in 3 of 3 specimens by RT-PCR (Figure 3). Jab1-IR was observed in 10 of 10 tissue specimens in 90% of the ECs and medial VSMCs (Figure 2A). Jab1 staining was present in the cytoplasm but was more prominent in the nucleus. Examination of serial sections showed that all MIF+ cells were simultaneously Jab1+.
Noninflammatory Stage of Atherosclerosis: Coronary Arteries With Adaptive Intimal Fibrosis (Type I Lesions)
In the arteries with AIF, MIF-IR was present in 10 of 10 specimens and was localized to ECs (5%), intimal VSMCs (α-actin+, 2.2±3.8%), and medial VSMCs (α-actin+, 5.9±8.0%) (Figure 1B). Jab1-IR was present in most ECs (80%), intimal VSMCs, and medial VSMCs (90%) (Figure 2B). Serial sections showed that cytoplasmic MIF staining never occurred without Jab1 staining.
Inflammatory Stages of Atherosclerosis I: Aortic Fatty Streaks
In the FS, MIF was present in 10 of 10 cases in ECs, in intimal macrophages (CD68+), intimal VSMCs (α-actin+), intimal T lymphocytes (CD3+, 24.0±0% of the intimal cells were MIF+), and medial VSMCs (α-actin+, 20.5±5.5%). Whenever present, 80% of the ECs above the FS showed strong MIF-IR (Figure 1C). Jab1-IR was present in most ECs (80%), intimal VSMCs, and medial VSMCs (90%) (Figure 2C). As confirmed by DL (MIF/Jab1), all MIF+ cells were also Jab1+ (100%) (Figure 2G).
Inflammatory Stages of Atherosclerosis II: Advanced Carotid Atherosclerotic Plaques
In the intima of the CPs, MIF was present in 10 of 10 cases, mainly in plaque inflammatory regions in the vicinity of the acellular lipid core (74.7±21.3% of intimal cells) and to a lesser extent in plaque fibrotic regions and in the media beneath the plaque (9.2±16.2% of medial VSMCs). Luminal ECs and microvascular ECs showed an inconstant MIF-IR (Figure 1D). DL confirmed that macrophages (CD68+), intimal VSMCs, medial VSMCs (actin+), and T lymphocytes (CD3+) were MIF+ (Figure 1, E through G). Jab1-IR was present in macrophages (CD68+), intimal VSMCs, medial VSMCs (actin+), and T lymphocytes (CD3+, 90% of intimal and medial cells were Jab1+) and occurred both in the cytoplasm and nucleus (Figure 2, D through H). DL (MIF/Jab1) revealed that MIF+ cells were also Jab1+ (Figure 2, G and H). ISH demonstrated the expression of MIF mRNA in the intima of CPs (5 of 5). As shown by DL (MIF-ISH/CD68), intimal macrophages showed strong MIF hybridization signals. Furthermore, we demonstrated the presence of MIF mRNA in another 6 specimens by RT-PCR (6 of 6) (Figure 3). Coimmunoprecipitation assays from atherosclerotic plaque tissue provided evidence for in vivo complex formation between MIF and Jab1. MIF was detected by Western analysis after immunoprecipitation by anti-Jab1 antibody of endogenous MIF-Jab1 complexes from lysates of atherosclerotic tissue (Figure 4) in 3 of 4 patients, whereas complexes were not detectable in any of the IMAs (0 of 4).
Statistical Analysis of MIF Immunostaining
In the intima of FS lesions and advanced plaques, significantly more MIF+ cells were observed compared with the analysis of the intima of the AIF tissue (type I lesions,20 Table). Likewise, there were significantly higher numbers of MIF+ cells in the intima of CPs compared with FS lesions. In summary, a significant upregulation of MIF-IR was seen in the intima of atherosclerotic lesions with chronic inflammation as compared with noninflammatory lesions. Upregulation further progressed with the development of advanced complicated atherosclerotic plaques.
Effect of oxLDL on MIF Expression in HUVECs and Macrophages
As shown in Figure 5, MIF protein in HUVECs was upregulated in a time- and concentration-dependent manner when these cells were incubated with oxLDL. Maximal production was obtained after 24 hours and with 10 μg/mL of oxLDL. OxLDL also led to an enhancement of MIF secretion in macrophages, with maximal induction observed at oxLDL concentrations of ≥45 μg/mL (data not shown). Confirming the coimmunoprecipitations from atherosclerotic plaque tissue (Figure 4), MIF-Jab1 complex formation was also detected in HUVECs (data not shown).
We have investigated the expression of the cytokine MIF over different stages of human atherosclerosis. MIF protein is detected in ECs, VSMCs, T lymphocytes, and macrophages of the arterial wall, confirming that MIF differs from other proinflammatory cytokines by exhibiting significant cellular levels at baseline.2,10⇓ In line with recent studies of atherosclerotic lesions in rabbits,14 our evaluation revealed that MIF is present at low levels in ECs and VSMCs of IMAs with a normal histology and, importantly, is upregulated during the progression of atherosclerosis toward inflammatory stages. The immunostaining studies are supported by evidence of local MIF production in the arterial wall by ECs and VSMCs in IMAs and, at least by macrophages, in CPs. The latter observation is further confirmed by oxLDL-mediated induction of MIF, as seen in a macrophage line (this study and Reference 25).25 Thus, because MIF is produced locally in key cells in all stages of lesion evolution, we conclude that it plays an important role during the progression of human atherosclerosis.
Atherosclerosis is a chronic inflammatory disease characterized by accumulation of blood-derived macrophages and T lymphocytes in the arterial wall.19 In the early stages, atherosclerosis is also characterized by formation of oxLDL in the vascular wall. OxLDL affects ECs, leading to changes in adhesion molecule expression, promoting the immigration of macrophages and other inflammatory cells.26,27⇓ Interestingly, our study demonstrates upregulation of MIF after incubation of HUVECs with oxLDL. Furthermore, recent investigations have indicated that MIF upregulates the expression of intercellular adhesion molecule-1 expression by ECs in vitro.14 Thus, oxLDL may promote the immigration of macrophages into the vessel wall and the immobilization of these cells in early stages of lesion evolution, directly or through mediation by endothelium-derived MIF.
Uptake of oxLDL by macrophages and VSMCs eventually leads to foam cell formation.19 In particular, macrophage-derived foam cells produce and secrete proinflammatory cytokines, which then amplify the inflammatory cascade in early stages of plaque evolution and serve to sustain inflammation in advanced plaques. We show that enhanced levels of MIF are released from macrophages on stimulation with oxLDL in vitro and that in vivo, MIF expression is prevalent in regions with chronic inflammation containing VSMC- and macrophage-derived foam cells, indicating that oxLDL also plays an important role in the induction of macrophage MIF in vivo. It is interesting to note that macrophages stimulated with MIF secrete tumor necrosis factor, IL1-β, and IL-8,3–5,12⇓⇓⇓ proinflammatory cytokines abundantly present in atherosclerotic lesions.28,29⇓ In addition, MIF upregulates nitric oxide production in macrophages3 and matrix metalloproteinases in fibroblasts,30 both key events in lesion evolution. Thus, MIF may play a prominent role in the regulation of the inflammatory response in advanced atherosclerotic lesions.
AP1 proteins are transcription factors with a role in gene regulation, cell proliferation, and mediation of inflammatory signals.31 AP1 activity is regulated by protein kinases31 and interaction with the coactivator Jab1.18 Jab1 was recently identified to specifically bind to MIF and to be a molecular target of MIF action.6 MIF-Jab1 interaction occurs intracellularly and has been demonstrated in macrophages and other cells in vitro. We sought for evidence of an interaction between MIF and Jab1 in human arterial tissue and atherosclerosis in vivo. In fact, immunohistochemistry showed that Jab1 expression is abundant in human arterial tissue, with staining activity mainly occurring in the nuclei of ECs, VSMCs, T lymphocytes, and macrophages. We show that all MIF+ cells of the vascular wall are simultaneously Jab1+. Importantly, coimmunoprecipitation provided specific evidence for in vivo complex formation between Jab1 and MIF in plaque tissue of patients. MIF-Jab1 complex formation in atherosclerotic tissue indicates that MIF-Jab1 complexes could be a critical regulatory system in atherosclerotic lesion evolution and inflammatory diseases in general. For example, MIF signaling effects in human atherosclerotic tissue could be integrated into the AP1 pathway by Jab1. Depending on the stage of atherosclerosis, the cells involved, and the surrounding inflammatory stimuli present, MIF could serve to either enhance9,30⇓ or downregulate inflammation.6,32⇓ Also, because AP1 binding sites are present in the promotors of genes regulating cell differentiation and proliferation31 and in genes controlling expression of E-selectin, collagenase-1, and stromelysins, MIF signaling may contribute to the regulation of expression of these atherosclerosis-associated factors.
In summary, MIF, a cytokine with important inflammation-regulatory activities, is produced abundantly by various cells in all types of human atherosclerotic lesions and may be an important regulator in early plaque evolution and advanced complicated lesions.
Dr Bernhagen was supported by DFG grants BE 1977/1-2/3, Dr Burger-Kentischer by an HSP III grant from the University of Stuttgart, and Dr Kleemann by the Netherlands Heart Foundation grant NHS 99.110. We thank A. Einhaus, B. Plessow-Freudenberg, D. Finkelmeier, and S. Hölsch for technical assistance.
- ↵Bacher M, Metz CN, Calandra T, et al. An essential regulatory role for MIF in T-cell activation. Proc Natl Acad Sci U S A. 1996; 93: 7849–7854.
- ↵Bernhagen J, Calandra T, Mitchell RA, et al. MIF is a pituitary-derived cytokine that potentiates lethal endotoxaemia. Nature. 1993; 365: 756–759.
- ↵Bernhagen J, Mitchell RA, Calandra T, et al. Purification, bioactivity, and secondary structure analysis of mouse and human macrophage migration inhibitory factor (MIF). Biochemistry. 1994; 33: 14144–14155.
- ↵Calandra T, Bernhagen J, Metz CN, et al. MIF as a glucocorticoid-induced modulator of cytokine production. Nature. 1995; 377: 68–71.
- ↵Donnelly SC, Haslett C, Reid PT, et al. Regulatory role for macrophage migration inhibitory factor in acute respiratory distress syndrome. Nat Med. 1997; 3: 320–323.
- ↵Kleemann R, Hausser A, Geiger G, et al. Intracellular action of the cytokine MIF to modulate AP-1 activity and the cell cycle through Jab1. Nature. 2000; 408: 211–216.
- ↵Lan HY, Bacher M, Yang N, et al. The pathogenic role of macrophage migration inhibitory factor in immunologically induced kidney disease in the rat. J Exp Med. 1997; 185: 1455–1465.
- ↵Mikulowska A, Metz CN, Bucala R, et al. Macrophage migration inhibitory factor is involved in the pathogenesis of collagen type II-induced arthritis in mice. J Immunol. 1997; 158: 5514–5517.
- ↵Mitchell RA, Metz CN, Peng T, et al. Sustained mitogen-activated protein kinase (MAPK) and cytoplasmic phospholipase A2 activation by macrophage migration inhibitory factor (MIF): regulatory role in cell proliferation and glucocorticoid action. J Biol Chem. 1999; 274: 18100–18106.
- ↵Bernhagen J, Calandra T, Bucala R. Regulation of the immune response by macrophage migration inhibitory factor: biological and structural features. J Mol Med. 1998; 76: 151–161.
- ↵Mitchell RA, Bucala R. Tumor growth-promoting properties of macrophage migration inhibitory factor. Semin Cancer Biol. 2000; 10: 359–366.
- ↵Calandra T, Bernhagen J, Mitchell RA, et al. The macrophage is an important and previously unrecognized source of macrophage migration inhibitory factor. J Exp Med. 1994; 179: 1985–1902.
- ↵Nishihira J, Koyama Y, Mizue Y. Identification of MIF in human vascular endothelial cells and its induction by lipopolysaccharide. Cytokine. 1998; 10: 199–205.
- ↵Lin SG, Yu XY, Chen YX, et al. De novo expression of macrophage migration inhibitory factor in atherogenesis in rabbits. Circ Res. 2000; 87: 1202–1208.
- ↵Bernhagen J, Bacher M, Calandra T, et al. An essential role for macrophage migration inhibitory factor in the tuberculin delayed-type hypersensitivity reaction. J Exp Med. 1996; 183: 277–282.
- ↵Kleemann R, Kapurniotu A, Frank RW, et al. Disulfide analysis reveals a role for macrophage migration inhibitory factor (MIF) as a thiol-protein oxidoreductase. J Mol Biol. 1998; 280: 85–102.
- ↵Rosengren E, Bucala R, Åman P, et al. The immunoregulatory mediator macrophage migration inhibitory factor (MIF) catalyzes a tautomerization reaction. Mol Med. 1996; 2: 143–149.
- ↵Claret F-X, Hibi M, Dhut S, et al. A new group of conserved coactivators that increase the specificity of AP-1 transcription factors. Nature. 1996; 383: 453–457.
- ↵Ross R. The pathogenesis of atherosclerosis: perspective for the 1990s. Nature. 1993; 362: 801–809.
- ↵Stary HC, Chandler AB, Dinsmore RE, et al. A definition of advanced types of atherosclerotic lesions and a histological classification of atherosclerosis. Arterioscler Thromb Vasc Biol. 1995; 15: 1512–1531.
- ↵Ihling C, Goebel HR, Lippoldt A, et al. Endothelin-1-like immunoreactivity in human atherosclerotic coronary tissue: a detailed analysis of the cellular distribution of endothelin-1. J Pathol. 1996; 179: 303–308.
- ↵Burger-Kentischer A, Muller E, Marz J, et al. Hypertonicity-induced accumulation of organic osmolytes in papillary interstitial cells. Kidney Int. 1999; 55: 1417–1425.
- ↵Chavakis E, Dernbach E, Hermann C, et al. Oxidized LDL inhibits vascular endothelial growth factor-induced endothelial cell migration by an inhibitory effect on the Akt/endothelial nitric oxide synthase pathway. Circulation. 2001; 103: 2102–2107.
- ↵Dimmeler S, Assmus B, Hermann C, et al. Fluid shear stress stimulates phosphorylation of Akt in human endothelial cells: involvement in suppression of apoptosis. Circ Res. 1998; 83: 334–341.
- ↵Atsumi T, Nishihira J, Makita Z, et al. Enhancement of oxidised low-density lipoprotein uptake by macrophages in response to macrophage migration inhibitory factor. Cytokine. 2000; 12: 1553–1556.
- ↵Cybulsky MI, Gimbrone MA Jr. Endothelial expression of a mononuclear leukocyte adhesion molecule during atherogenesis. Science. 1991; 251: 788–791.
- ↵Reape TJ, Groot PH. Chemokines and atherosclerosis. Atherosclerosis. 1999; 147: 213–225.
- ↵Libby P, Ordovas JM, Birinyi LK, et al. Inducible interleukin-1 gene expression in human vascular smooth muscle cells. J Clin Invest. 1986; 78: 1432–1438.
- ↵Barath P, Fishbein MC, Cao J, et al. Detection and localization of tumor necrosis factor in human atheroma. Am J Cardiol. 1990; 65: 297–302.
- ↵Onodera S, Kaneda K, Mizue Y, et al. Macrophage migration inhibitory factor up-regulates expression of matrix metalloproteinases in synovial fibroblasts of rheumatoid arthritis. J Biol Chem. 2000; 275: 444–450.
- ↵Karin M, Liu Z, Zandi E. AP-1 function and regulation. Curr Opin Cell Biol. 1997; 9: 240–246.
- ↵Bucala R. A most interesting factor. Nature. 2000; 408: 167–168.