Interleuldn-l[3 and tumor necrosis factor-c release in normal and diseased human infrarenal aortas William H . Pearce, M D , Iliana Sweis, M D , James S. T. Yao, M D , W a l t e r J. M c C a r t h y , M D , and Alisa E. Koch, M D , Chicago, Ill., and
Cleveland, Ohio
The presence of chronic inflammatory cells in the adventitia and media of abdominal aortic aneurysms and aortic occlusive disease suggest an immunologic response. The purpose of this study is to determine whether normal or diseased infrarenal aortas liberate the inflammatory cytokines tumor necrosis factor-~ (TNF-~) and interleukin-ll3 (IL-113). Twenty-six infrarenal aortic biopsies (5 aortic occlusive disease, 15 abdominal aortic aneurysms, and 6 cadaveric donors) were weighed, minced into small pieces, and incubated in media for 48 hours. Conditioned media was harvested at 48 hours and assayed for IL- 113 or TNF-,, with use of an ELISA assay. Comparison of groups was performed with a one-way analysis of variance. The constitutive IL-113 produced by abdominal aortic aneurysms was significantly different than that in cadaveric donors (908 +- 194 pg/ml [SE] vs 100-+ 30 pg/ml). There was no statistically significant difference between abdominal aortic aneurysms and aortic occlusive disease (908 +- 194 pg/ml vs 604 +- 256 pg/ml) or aortic occlusive disease and cadaveric donor (604 + 256 vs 100 +- 30). In time-course studies for the release of IL-113, abdominal aortic aneurysms demonstrated maximal release at 48 hours. IL-113 release was augmented by lipopolysaccharide in all categories. A dose response curve demonstrated maximal IL-11~ release on stimulation with 5 izg/ml LPS. Constitutive TNF-a production was low, ranging from 13 + 1.5 pg/ml in cadaverie donor, to 20 pg/ml in aortic occlusive disease, and 24 + 11 pg/ml in abdominal aortic aneurysms. There was no augmentation in TNF-~ with lipopolysaccharide. These experiments demonstrate (1) A significant increase in the amount of constitutive IL-113 produced in abdominal aortic aneurysms as compared with cadaveric donor; (2) IL-113 secretion is not different between aortic occlusive disease and abdominal aortic aneurysms, nor aortic occlusive disease and cadaveric donor; (3) TNF-~ production is similar in aU cases. We suggest that the inflammatory infiltrate found in infrarenal aortic disease produces cytokines that may contribute to the pathogenesis of the underlying disease. (J VAse SURG 1992;16:784-9.)
Infammafion may play an important role in the pathogenesis o f human aortic disease. Histologic studies of inflammatory abdominal aortic aneurysms (AAAs) reveal an extensive inflammatory infiltrate with neovascularization and fibrosis) ,2 Even in the more c o m m o n atherosclerotic aneurysms there is a From the Departmentof Surgeryand the FeinbergCardiovascular Research Institute, No~.hwestern University Medical School, Chicago, CaseWestern Reserve,Cleveland(Dr. Sweis),and the Department of Medicine, Northwestern University Medical School, Chicago (Dr. Koch). Supported in part by the AlyceF. Salerno Foundation, Veterans Administration Merit Review Grant, the Gaylord Freeman Fund, and the InternationalSocietyfor CardiovascailarSurgery. Presented at the Fifteenth Annual Meeting of the Midwestern Vascular Surgical Society,Chicago, Ill., Sept. 20-21, 1992. Reprint requests: William H. Pearce,MD, 251 E. Chicago Ave., Suite 626, Chicago, IL 60611. 24/1/37640 784
similar, although less extensive, inflammatory response. 3 By use ofimmunohistology we have shown that these infrarenal aortic aneurysms demonstrate a chronic inflammatory infiltrate with large numbers o f activated monocytes. 4 The lesions are characterized by the lymphoid aggregates o f macrophages and T cells and B cells in the adventitia and media. Furthermore, there appears to be an active participation o f the immune response in the development o f atherosclerosis, s'6 In early atherosclerotic lesions, monocytes invade the arterial wall with subsequent smooth muscle hyperplasia and fibrosis. This inflammatory response is mediated by a complex network o f hormone-like substances called cytokines. These cytokines have both autocrine and paracrine functions within the arterial wall. Interleukin-l[3 (IL-I[3) and rumor necrosis factor-e~ (TNF-~) are two inflamma-
Volume 16 Number 5 November 1992
Interleukin I[3 and tumor necrosisfactor-a release 785
2500
E 2000 C
,2
T
1500
iiiiiiiiii
o
o
1000
o
o
c~. I
50o
No LPS Normal
LPS ~
Occlusive
~
AAA
Fig. 1. IL-113 concentrations in conditioned media with and without lipopolysaccharide stimulation.
tory cytokines that have important roles in regulating smooth muscle cell and endothelial cell ftmction. The purpose o f this study was to determine whether normal or diseased infrarenal aortas liberate the inflammatory cytokines, TNF-a and IL-113. METHODS Twenty-six infrarenal aortic biopsies were obtained from patients undergoing aortic reconstructive surgery for aortic occlusive disease (AOD), five; AAA resections, 15; and cadaveric donors (NL), six. The mean age of NL was 41 -_+ 14 years, AOD 65 _ 5 years, and AAA 70 ___6 years. The specimens were obtained at the time of surgery and placed in media (RPMI-1640 + 10% fetal bovine serum and gentamicin). The samples were minced into small pieces and incubated in the same media for 48 hours. Since lipopolysaccharide has been shown to increase the production of IL-113 and TNF-0t in monocytes, parallel experiments were performed with aortic explants incubated in media containing 5 t~g/ml of lipopolysaccharide. The conditioned media was harvested from both experiments at 48 hours and assayed for IL-113 and TNF-a with use of ELISA assays (IL-113, R and D Systems, Minneapolis, Minn.; TNF-(x, Endogen, Boston, Mass.). After initial experiments revealed elevated levels of IL-II3 secretion, further experiments were performed to elucidate the time course and dose response of IL-II3 secretion in AAAs. A dose response curve was generated by varying the concentration oflipopolysaccharide from 0.005 i~g/ml to 10
I~g/ml (N = 3). The media was harvested at 24 hours and assayed for IL-113. A time-course experiment was performed by assaying cytokine release after 2, 8, 24~ and 48 hours of incubation. In 12 patients sufficient tissue was obtained for histologic evaluation. With use of a previously described scoring system, an inflammatory score (0 to 3) was calculated for each patient. 4 Comparison of the groups was performed by use of a one-way analysis of variance (ANOVA) (Statview, Abacus Concepts, Berkeley, Calif.) with Dunnetts post hoc test. Results are expressed in picograms per milliliter _ SEM. RESULTS
Interleukin-113 secretion from cultured aortic explants was greatest in patients with AAA (980 __+194 pg/ml), which was significantly different fromNL (100 ___ 31 pg/ml);p < 0.05. Therewasno statistically significant difference between AAA and AOD (Fig. 1). The time course studies that were performed in explant cultures revealed maximal IL-113 release at 48 hours (Fig. 2). Since the assays were performed at 24 hours, stimulated levels of IL-113 secretion are lower than the levels reported in 48-hour experiments. Interleukin- 113 production was augmented in all groups with lipopolysaccharide stimulation. A dose response curve was performed in three aortic explants, which demonstrates a plateau at 5 I~g/ml (Fig. 3). When comparing the inflammatory score with the disease category and the amount of IL-113 secretion, there is a progressive increase in
786
~ourn~ o f VASCULAR SURGERY
Pearce et al.
500 E v
T
400
¢,.
._o 300 .m to E 0
o
e~ !
200 f_ 1CO
,-I
~ A
0
I
2
1
8
I
24
64
48
Time (hours)
Fig. 2. Time course study for IL-l[3 release in AAAs. 800
T
700 600 500 .m 400 8 300 0 200 .J
100 I
0 0
0.005
t
0.05
I
i
0.5
5
I
10
15
LPS Concentration (ug/ml)
Fig. 3. A dose response curve fbr IL-lJ3 secretions with lipopolysaccharidestimulation. IL-l[3 secretion and inflammation. For NL the inflammatory score was 0.4; AOD, 1.3; and AAA, 2.4. Tumor necrosis factor-a was found in similar amounts in all categories. The range varied from 13 + 1.5 pg/ml in NL to 24 +_ 11 pg/ml in AAA. Incubation with lipopolysaccharide did not augment TNF-a sccretion. DISCUSSION
Abdominal aortic aneurysms constitutively secrete significantly greater quantifies of IL-1B than normal. Specimens from patients with AOD also release IL-113, however, to a lesser degree. The
amount of IL- 113 released approximately parallels the degree of inflammation seen in histologic studies. Koch et al.4 found a spectrum of inflammatory changes that varied from AOD with mild inflammatory infiltrates in the intima and media to more severe changes in AAAs localized in the media and adventitia. With use of immunohistologic techniques with monoclonal antibodies, activated macrophages were identified scattered throughout the adventitia in lymphoid aggregates associated with CD4-positive T lymphocytes. CD19-positive B lymphocytes were found in the adventitia of both occlusive and aneurysmal aortic disease. Occasional smooth muscle cells expressed class II HLA-DR antigens. Such cells
Volume 16 Number 5 November 1992
Interleukin 113 and ~umor necrosisfactor-a release 787
2000 A i
E m
Q, 1500 C 0
Nofn~l
Occlusive
~AAA
M
Ill
1000 -,-. :v:.: :.:.X.: : :-: : :. • . • • ".v:.v...'., .-... • v . . . v . v , v . v . v : . . . • • .v • ..v:.v...v.v... •
0
¢= o
o
v.....v v, .v..
500
..v.v..., v,v...v.v.-.v
v.-.v
i!i!iiiiiiiiiiii!!i!i!iiiiiiiii.,
I qF~
.a 0
,
::::::::::::::::::::::::::::::::::::::
T 0.4
1.3
2.4
Fig. 4. Inflammatory score, IL-lf3 secretion in relationship to disease category. are able to propagate the immune response by further presenting antigen to T cells. Other immunohistologic studies of aortic aneurysms by Stella et al.1 and Brophy et al. 7 found chronic inflammatory changes with evidence of complement activation and immunoglobulin secretion. The importance of the inflammatory cellular infiltrate found in aortic disease lies in the potential modulation of endothelial and vascular smooth muscle cell function through the cytokine network. Although IL-I[~ and TNF-a have been found in human atheromas, 8,9 this report demonstrates for the first time the presence oflL-l[3 and TNF-a in human aortic disease. Interleukin-1 is a monocyte-derived interleukin with wide ranging effects on vascular cells. Of the two forms of IL-1 (a and J3), macrophages are the principal source of IL-I[3.10 Endothelial cells and vascular smooth muscle cells are also capable of IL-113 synthesis when stimulated with IL-I[3) ° This positive feedback loop may act locally to amplify the inflammatory reaction. In aortic tissue stained with monospecific antibodies, IL- lJ3 is found in association with macrophages, occasional smooth muscle and endothelial cdls. n The important biologic functions of IL-1 include vascular smooth muscle cell proliferation, 12 vasodilation, and alteration in low density lipoprotein metabolismY ,14 Of these fimctions, smooth muscle proliferation is important in early atherogenesis. Interleukin-1 induces smooth muscle cell proliferation and platelet-derived growth factor (PDGF) secretion, whereas interferongamma (IFN-y) inhibits growth. Interleukin-1 also inhibits vascular smooth muscle cell contraction by an endothelium independent mechanism) s17
Whether the inhibition of smooth muscle contraction plays an important role or not in aneurysm formation is unknown. Interleuldn-1 also induces vascular smooth muscle cells to secrete copious amounts of IL-6, which stimulates lymphocyte proliferation and antibody production by B-lymphoeytes)8 The accumulation of immunoglobulin in aneurysmal walls seen in several studies may be the result of increased local levels of IL-6) ,7 Interleukin-l[3 along with TNF-a induce expression of endothelial cell adhesion molecules which further enhance the inflammatory response by recruiting additional inflammatory cells. Both vascular cell adhesion molecule-1 and endothelial leukocyte adhesion molecule-1 are expressed in endothelial cell cultures exposed to IL-1[3 and W N F - 0 t . 19 Finally, IL-I[3 may affect collagen synthesis. In cultured dermal fibroblasts, IL-I[3 induces collagen synthesis, which parallels the increase in type I collagen gene expression in AAAs.2°22 Tumor necrosis factor-a was also detected in AAAs, however, in low levels and without differences between groups. Lipopolysaccharide stimulation did not enhance production. Activated macrophages and vascular smooth muscle cells are known sources of TNF-aY Similar to IL-113, TNF-a induces its own gene expression in vascular smooth muscle cells.24 In low concentrations TNF-ct is both angiogenic and cytotoxicfi Barath et al. 8 suggest that the neovascularization and central necrosis seen in human atheroma is a potential biologic effect of TNF-~ in human atherosclerosis. Neovascularization is common in the adventitia of AAAs and is frequently found in association with atherosclerotic plaques. As with IL-1[3, immunohistologic studies find TNF-ot local-
788
Journal of VASCULAR SURGERY
Pearce et al.
ized to macrophages and smooth muscle cells.l° It is interesting to note that in the present study, normal arteries also released TNF-a. The normal tissue in this study was obtained from cadaveric renal donors with recent trauma. Elevated levels of circulating TNF-e~ are found in response to trauma or exposure to endotoxin and may originate from the vascular wall. Normal aortic rings when exposed to endotoxin in animal studies release TNF-a in quantifies similar to those observed here) 5 In summary, diseased infrarenal aortas release IL-113, which may play an important role in the pathophysiology of these diseases. The structural integrity of the arterial wall is a dynamic balance between synthesis and destruction of the extracellular matrix. The complex network of cytokines may influence this balance by modulating smooth muscle cell proliferation and differentiation. A chronic inflammatory infiltrate is fotmd in the media and adventitia of the diseased aorta with activated macrophages that are the principal source of IL-1 [3. The antigenic stimulation that initiates this immune response is unknown. Degenerating elastin fragments or modified lipoprotein are possible candidates, s,26 However, once initiated, the immune response is controlled by the balance of stimulatory and inhibitory cytokines. Although IL-l[3 is elevated in both AAAs and AOD, a slight imbalance of this cytokine in conjunction with other possible changes in IL-6, TNF-a, and interferon-gamma and inhibitors in the susceptible individual may be sufficient to alter the synthesis and degradation of collagen and elastin leading to aneurysm formation or occlusion. Further studies will be needed to fully characterize the cytokine environment in these two diseases. The role of cytokines and smooth muscle synthetic function is under active investigation in our laboratory and may help explain the changes in the structural protein observed in AAAs. We thank Dipti Parikh and Manisha Shah for their technical contributions to this research.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20. REFERENCES
1. Stella A, Gargiulo M, Pasquinetli G, et al. The cellular component in the parietal infiltrate of inflammatory abdominal aortic aneurysms (IAAA). Eur J Vasc Surg 1991;5:65-70. 2. Leu HJ. Inflammatory abdominal aortic aneurysms: a disease entity? Virchows Archiv [A] 1990;417:427-33. 3. Rizzo RJ, McCarthy WJ, Dixit SN, et al. Collagen types and matrix protein content in human abdominal aortic aneurysms. J VAsc SURG 1989;10:365-73. 4. Koch AE, Haines GK, Rizzo RJ, et al. Human abdominal aortic aneurysms: immunophenotypic analysis suggesting an
21.
22.
inamune-mediated response. Am J Pathol 1990; 137:1199213. Hansson GK, Jonasson L, Seifert PS, Stemme S. Immune mechanisms in atherosclerosis. Arteriosclerosis 1989;9:56778. Roessner A, Herrera A, Honing HJ, et al. Identification of macrophages and smooth muscle cells with monoclonal antibodies in the human atherosclerotic plaque. Virchows Arch [A] 1987;412:169-74. Brophy CM, Reilly JM, Walker-Smith GJ, Tilson MD. The role of inflammation in nonspecific abdominal aortic aneurysm disease. Ann Vasc Surg 1991;5:229-33. 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. Wang A, Doyle M, Mark D. Quantification of mRNA using the polymerase chain reaction. Proc Nail Acad Sci USA 1989;86:9717-21. Moyer CF, Sajuthi D, TuUi H, Williams JK. Synthesis oflL-1 alpha and IL-1 beta by arterial cells in atherosclerosis. Am J Pathol 1991;138:951-60. Pearce WH, Koch AE, Haines GK, Mesh CL, Parikh D. Cellular components and immune response in abdominal aortic aneurysms. Surg Forum 1991;42:328-9. Libby P, Warner SJC, Friedman GB. Interleuldn 1: a mitogen for human vascular smooth muscle cells that induces the release of growth-inhibitory prostanoids. J Clin Invest 1988; 81:487-98. Rasmussen L, Seljefid R. The modulatory effect oflipoprotein on the release of interleukin-1 by human peritoneal macrophages stimulated by WF ]-1, 3-D-polyglucose derivatives. Scand J Immunol 1989;29:477-84. Beutler BA, Cerami A. Recombinant interleukin 1 suppresses fipoprotein lipase activity in 3T3-L1 cells. J Immunol 1985 ;6:3969-71. McKenna TM. Prolonged exposure of rat aorta to low levels of endotoxin in vitro results in impaired contractility. J Clin Invest 1990;86:160-8. Busse K, Mulsch A. Induction of nitric oxide synthase by cytokines in vascular smooth muscle cells. FEBS Letr 1990; 275:87-90. Beasley D, Cohen RA, Levinsky NG. Interleukin 1 inhibits contraction of vascular smooth muscle. J Clin Invest 1989; 83:331-5. Loppnow H, Libby P. Proliferating or Interleukin 1-activated human vascular smooth muscle cells secrete copious interleukin 6. J Clin Invest 1990;85:731-8. Pober JS. What can be learned from the expression of endothelial adhesion moleolles in tissues? Lab Invest 1991; 64:301-5. Elias JA, Freundlich B, Adams S, Rosenbloom J. Regulation of human lung fibroblast collagen production by recombinant interleukin-1, tumor necrosis factor, and intefferon-~. Ann NY Acad Sci 1990;580:233-44. Jaeger E, Rust S, Roessner A, et aL Joint occurrence of collagen mRNA containing cells and macrophages in human atherosclerotic vessels. Atherosclerosis 1991;31:55-68. McGee GS, Baxter BT, Shively VS, et al. Aneurysm or occlusive disease: factors determining the clinical course of atherosclerosis of the infrarenal aorta. Surgery 1991;110: 370-6.
23. Mannel D, Moore R, Magenhagen S. Macrophages as a source
Volume 16 Number 5 November 1992
of mmoricidal activity (nmlor necrosis factor). Infect Immun 1980;30:523-30. 24. Warner SJ, Libby P. Hlmaan vascular smooth muscle cells. Target for and source of minor necrosis factor. J Immunol 1989; 142:100-9. 25. Lebovich SJ, Polverini PJ, Shepard HM, et al. Macrophageinduced angiogenesis is mediated by tumor necrosis factor-c~. Nature 1987;329:630-2.
Interleukin 1 ~ and tumor necrosisfactor-a release 789
26. O'Brian }rp. A concept of diffuse actinic arteritis: the rote of actinic damage to elastin in 'age change' and arterifis of the temporal artery and in polymyalgia rheumatica. Br I Dermatot 1978;98:1-13.
Submitted Oct. 15, 1991; accepted Mar. 3, 1992.
DISCUSSION Dr. Peter Gloviczki (Rochester, Minn.). I congratulate Dr. Pearce on his excellent presentation and the Northwestern group for their continuing interest in the research of this important problem: the etiology of aortic aneurysms. Dr. Pearce clearly demonstrated to us today that an inflammatory process is present in aneurysmal disease and that the production of at least one of the cytokines, IL-113, is elevated in aortic aneurysms. The authors concluded that this cytokine may contribute to the pathogenesis of aortic aneurysms. My first question concerns this conclusion. Are these cytokines really the marker of aneurysmal disease or are they really the marker ofatherosclerosis? I ask this question because I noted that there was no significant difference in the production of cytokines between aneurysms and occlusive disease. I am also interested in knowing if you have any data to support that the inflammatory process is a cause of aneurysm formation and not really a consequence of aneurysmal dilation. My third question concerns your histologic findings. We know that the most important change that we see in the aneurysmal wall is the destruction of elastin in the media of the aneurysm or the aortic wall. Did you find inflammatory infiltrate in the area of elastin destruction? Conversely, did you have a normal elastin staining next to areas o f significant or marked inflammation? The presence o f T N F - a was demonstrated by other investigators to be present and elevated in human atheroma. What is your explanation that you could not confirm an elevated level in your experiments? Finally, how would you explain the role of chronic inflammation in the pathogenesis of aneurysms in those patients who have obvious genetic predisposition? Dr. W i l l i a m Pearce. This issue regarding whether this is a cause and effect or if there is really a separate disease
between atherosclerosis and aneurysm formation may be defined by understanding that it is very hard to draw from a spectrum o f disease any statistical significance between two different sets of aneurysms. I hope not to give you the thought that I think that there is a separate disease or it is the same manifestation o f atherosclerosis. The question regarding the comigration of elastin versus the inflammatory infiltrate- the elastin fibers were in the media far away from the inflammatory process. A possible explanation is that the smooth muscle ceils themselves are the source of elastase and allow the destruction of the elastin. The question you had regarding the production o f these cytokines by normal tissues is intriguing to me. You have noticed that in several cases there were elevated levels of at least I L - t . It has been suggested in trauma patients that there are elevated levels of these cytokines and that the source may be the vascular wall itself. The fourth or fifth question was why was the TNF so low, and that is another intriguing question. We have repeated the study several times with different manufacturer's ELISA kits and we have had elevation on occasion but in general there is no statistical difference. The suggestion is that perhaps there is a different sensitivity in the kit that was initially used in this experiment or possibly that there are inhibitors in the serum that was used to perform this test. The data on the TNF is not as strong as the IL-1 and may be related to an artifact of the test used. Finally, in the issue regarding the genetic questions and whether the inflammatory infiltrate affects these p a t i e n t s the only answer I have is that one rarely finds an aneurysm without the inflanamatory component. I suppose if one were able to find an anemTsm very early in its development without any inflammatory component, one would suggest that the "egg came before the chicken," but at the present time I can not answer that question either.
Cleveland, Ohio
The presence of chronic inflammatory cells in the adventitia and media of abdominal aortic aneurysms and aortic occlusive disease suggest an immunologic response. The purpose of this study is to determine whether normal or diseased infrarenal aortas liberate the inflammatory cytokines tumor necrosis factor-~ (TNF-~) and interleukin-ll3 (IL-113). Twenty-six infrarenal aortic biopsies (5 aortic occlusive disease, 15 abdominal aortic aneurysms, and 6 cadaveric donors) were weighed, minced into small pieces, and incubated in media for 48 hours. Conditioned media was harvested at 48 hours and assayed for IL- 113 or TNF-,, with use of an ELISA assay. Comparison of groups was performed with a one-way analysis of variance. The constitutive IL-113 produced by abdominal aortic aneurysms was significantly different than that in cadaveric donors (908 +- 194 pg/ml [SE] vs 100-+ 30 pg/ml). There was no statistically significant difference between abdominal aortic aneurysms and aortic occlusive disease (908 +- 194 pg/ml vs 604 +- 256 pg/ml) or aortic occlusive disease and cadaveric donor (604 + 256 vs 100 +- 30). In time-course studies for the release of IL-113, abdominal aortic aneurysms demonstrated maximal release at 48 hours. IL-113 release was augmented by lipopolysaccharide in all categories. A dose response curve demonstrated maximal IL-11~ release on stimulation with 5 izg/ml LPS. Constitutive TNF-a production was low, ranging from 13 + 1.5 pg/ml in cadaverie donor, to 20 pg/ml in aortic occlusive disease, and 24 + 11 pg/ml in abdominal aortic aneurysms. There was no augmentation in TNF-~ with lipopolysaccharide. These experiments demonstrate (1) A significant increase in the amount of constitutive IL-113 produced in abdominal aortic aneurysms as compared with cadaveric donor; (2) IL-113 secretion is not different between aortic occlusive disease and abdominal aortic aneurysms, nor aortic occlusive disease and cadaveric donor; (3) TNF-~ production is similar in aU cases. We suggest that the inflammatory infiltrate found in infrarenal aortic disease produces cytokines that may contribute to the pathogenesis of the underlying disease. (J VAse SURG 1992;16:784-9.)
Infammafion may play an important role in the pathogenesis o f human aortic disease. Histologic studies of inflammatory abdominal aortic aneurysms (AAAs) reveal an extensive inflammatory infiltrate with neovascularization and fibrosis) ,2 Even in the more c o m m o n atherosclerotic aneurysms there is a From the Departmentof Surgeryand the FeinbergCardiovascular Research Institute, No~.hwestern University Medical School, Chicago, CaseWestern Reserve,Cleveland(Dr. Sweis),and the Department of Medicine, Northwestern University Medical School, Chicago (Dr. Koch). Supported in part by the AlyceF. Salerno Foundation, Veterans Administration Merit Review Grant, the Gaylord Freeman Fund, and the InternationalSocietyfor CardiovascailarSurgery. Presented at the Fifteenth Annual Meeting of the Midwestern Vascular Surgical Society,Chicago, Ill., Sept. 20-21, 1992. Reprint requests: William H. Pearce,MD, 251 E. Chicago Ave., Suite 626, Chicago, IL 60611. 24/1/37640 784
similar, although less extensive, inflammatory response. 3 By use ofimmunohistology we have shown that these infrarenal aortic aneurysms demonstrate a chronic inflammatory infiltrate with large numbers o f activated monocytes. 4 The lesions are characterized by the lymphoid aggregates o f macrophages and T cells and B cells in the adventitia and media. Furthermore, there appears to be an active participation o f the immune response in the development o f atherosclerosis, s'6 In early atherosclerotic lesions, monocytes invade the arterial wall with subsequent smooth muscle hyperplasia and fibrosis. This inflammatory response is mediated by a complex network o f hormone-like substances called cytokines. These cytokines have both autocrine and paracrine functions within the arterial wall. Interleukin-l[3 (IL-I[3) and rumor necrosis factor-e~ (TNF-~) are two inflamma-
Volume 16 Number 5 November 1992
Interleukin I[3 and tumor necrosisfactor-a release 785
2500
E 2000 C
,2
T
1500
iiiiiiiiii
o
o
1000
o
o
c~. I
50o
No LPS Normal
LPS ~
Occlusive
~
AAA
Fig. 1. IL-113 concentrations in conditioned media with and without lipopolysaccharide stimulation.
tory cytokines that have important roles in regulating smooth muscle cell and endothelial cell ftmction. The purpose o f this study was to determine whether normal or diseased infrarenal aortas liberate the inflammatory cytokines, TNF-a and IL-113. METHODS Twenty-six infrarenal aortic biopsies were obtained from patients undergoing aortic reconstructive surgery for aortic occlusive disease (AOD), five; AAA resections, 15; and cadaveric donors (NL), six. The mean age of NL was 41 -_+ 14 years, AOD 65 _ 5 years, and AAA 70 ___6 years. The specimens were obtained at the time of surgery and placed in media (RPMI-1640 + 10% fetal bovine serum and gentamicin). The samples were minced into small pieces and incubated in the same media for 48 hours. Since lipopolysaccharide has been shown to increase the production of IL-113 and TNF-0t in monocytes, parallel experiments were performed with aortic explants incubated in media containing 5 t~g/ml of lipopolysaccharide. The conditioned media was harvested from both experiments at 48 hours and assayed for IL-113 and TNF-a with use of ELISA assays (IL-113, R and D Systems, Minneapolis, Minn.; TNF-(x, Endogen, Boston, Mass.). After initial experiments revealed elevated levels of IL-II3 secretion, further experiments were performed to elucidate the time course and dose response of IL-II3 secretion in AAAs. A dose response curve was generated by varying the concentration oflipopolysaccharide from 0.005 i~g/ml to 10
I~g/ml (N = 3). The media was harvested at 24 hours and assayed for IL-113. A time-course experiment was performed by assaying cytokine release after 2, 8, 24~ and 48 hours of incubation. In 12 patients sufficient tissue was obtained for histologic evaluation. With use of a previously described scoring system, an inflammatory score (0 to 3) was calculated for each patient. 4 Comparison of the groups was performed by use of a one-way analysis of variance (ANOVA) (Statview, Abacus Concepts, Berkeley, Calif.) with Dunnetts post hoc test. Results are expressed in picograms per milliliter _ SEM. RESULTS
Interleukin-113 secretion from cultured aortic explants was greatest in patients with AAA (980 __+194 pg/ml), which was significantly different fromNL (100 ___ 31 pg/ml);p < 0.05. Therewasno statistically significant difference between AAA and AOD (Fig. 1). The time course studies that were performed in explant cultures revealed maximal IL-113 release at 48 hours (Fig. 2). Since the assays were performed at 24 hours, stimulated levels of IL-113 secretion are lower than the levels reported in 48-hour experiments. Interleukin- 113 production was augmented in all groups with lipopolysaccharide stimulation. A dose response curve was performed in three aortic explants, which demonstrates a plateau at 5 I~g/ml (Fig. 3). When comparing the inflammatory score with the disease category and the amount of IL-113 secretion, there is a progressive increase in
786
~ourn~ o f VASCULAR SURGERY
Pearce et al.
500 E v
T
400
¢,.
._o 300 .m to E 0
o
e~ !
200 f_ 1CO
,-I
~ A
0
I
2
1
8
I
24
64
48
Time (hours)
Fig. 2. Time course study for IL-l[3 release in AAAs. 800
T
700 600 500 .m 400 8 300 0 200 .J
100 I
0 0
0.005
t
0.05
I
i
0.5
5
I
10
15
LPS Concentration (ug/ml)
Fig. 3. A dose response curve fbr IL-lJ3 secretions with lipopolysaccharidestimulation. IL-l[3 secretion and inflammation. For NL the inflammatory score was 0.4; AOD, 1.3; and AAA, 2.4. Tumor necrosis factor-a was found in similar amounts in all categories. The range varied from 13 + 1.5 pg/ml in NL to 24 +_ 11 pg/ml in AAA. Incubation with lipopolysaccharide did not augment TNF-a sccretion. DISCUSSION
Abdominal aortic aneurysms constitutively secrete significantly greater quantifies of IL-1B than normal. Specimens from patients with AOD also release IL-113, however, to a lesser degree. The
amount of IL- 113 released approximately parallels the degree of inflammation seen in histologic studies. Koch et al.4 found a spectrum of inflammatory changes that varied from AOD with mild inflammatory infiltrates in the intima and media to more severe changes in AAAs localized in the media and adventitia. With use of immunohistologic techniques with monoclonal antibodies, activated macrophages were identified scattered throughout the adventitia in lymphoid aggregates associated with CD4-positive T lymphocytes. CD19-positive B lymphocytes were found in the adventitia of both occlusive and aneurysmal aortic disease. Occasional smooth muscle cells expressed class II HLA-DR antigens. Such cells
Volume 16 Number 5 November 1992
Interleukin 113 and ~umor necrosisfactor-a release 787
2000 A i
E m
Q, 1500 C 0
Nofn~l
Occlusive
~AAA
M
Ill
1000 -,-. :v:.: :.:.X.: : :-: : :. • . • • ".v:.v...'., .-... • v . . . v . v , v . v . v : . . . • • .v • ..v:.v...v.v... •
0
¢= o
o
v.....v v, .v..
500
..v.v..., v,v...v.v.-.v
v.-.v
i!i!iiiiiiiiiiii!!i!i!iiiiiiiii.,
I qF~
.a 0
,
::::::::::::::::::::::::::::::::::::::
T 0.4
1.3
2.4
Fig. 4. Inflammatory score, IL-lf3 secretion in relationship to disease category. are able to propagate the immune response by further presenting antigen to T cells. Other immunohistologic studies of aortic aneurysms by Stella et al.1 and Brophy et al. 7 found chronic inflammatory changes with evidence of complement activation and immunoglobulin secretion. The importance of the inflammatory cellular infiltrate found in aortic disease lies in the potential modulation of endothelial and vascular smooth muscle cell function through the cytokine network. Although IL-I[~ and TNF-a have been found in human atheromas, 8,9 this report demonstrates for the first time the presence oflL-l[3 and TNF-a in human aortic disease. Interleukin-1 is a monocyte-derived interleukin with wide ranging effects on vascular cells. Of the two forms of IL-1 (a and J3), macrophages are the principal source of IL-I[3.10 Endothelial cells and vascular smooth muscle cells are also capable of IL-113 synthesis when stimulated with IL-I[3) ° This positive feedback loop may act locally to amplify the inflammatory reaction. In aortic tissue stained with monospecific antibodies, IL- lJ3 is found in association with macrophages, occasional smooth muscle and endothelial cdls. n The important biologic functions of IL-1 include vascular smooth muscle cell proliferation, 12 vasodilation, and alteration in low density lipoprotein metabolismY ,14 Of these fimctions, smooth muscle proliferation is important in early atherogenesis. Interleukin-1 induces smooth muscle cell proliferation and platelet-derived growth factor (PDGF) secretion, whereas interferongamma (IFN-y) inhibits growth. Interleukin-1 also inhibits vascular smooth muscle cell contraction by an endothelium independent mechanism) s17
Whether the inhibition of smooth muscle contraction plays an important role or not in aneurysm formation is unknown. Interleuldn-1 also induces vascular smooth muscle cells to secrete copious amounts of IL-6, which stimulates lymphocyte proliferation and antibody production by B-lymphoeytes)8 The accumulation of immunoglobulin in aneurysmal walls seen in several studies may be the result of increased local levels of IL-6) ,7 Interleukin-l[3 along with TNF-a induce expression of endothelial cell adhesion molecules which further enhance the inflammatory response by recruiting additional inflammatory cells. Both vascular cell adhesion molecule-1 and endothelial leukocyte adhesion molecule-1 are expressed in endothelial cell cultures exposed to IL-1[3 and W N F - 0 t . 19 Finally, IL-I[3 may affect collagen synthesis. In cultured dermal fibroblasts, IL-I[3 induces collagen synthesis, which parallels the increase in type I collagen gene expression in AAAs.2°22 Tumor necrosis factor-a was also detected in AAAs, however, in low levels and without differences between groups. Lipopolysaccharide stimulation did not enhance production. Activated macrophages and vascular smooth muscle cells are known sources of TNF-aY Similar to IL-113, TNF-a induces its own gene expression in vascular smooth muscle cells.24 In low concentrations TNF-ct is both angiogenic and cytotoxicfi Barath et al. 8 suggest that the neovascularization and central necrosis seen in human atheroma is a potential biologic effect of TNF-~ in human atherosclerosis. Neovascularization is common in the adventitia of AAAs and is frequently found in association with atherosclerotic plaques. As with IL-1[3, immunohistologic studies find TNF-ot local-
788
Journal of VASCULAR SURGERY
Pearce et al.
ized to macrophages and smooth muscle cells.l° It is interesting to note that in the present study, normal arteries also released TNF-a. The normal tissue in this study was obtained from cadaveric renal donors with recent trauma. Elevated levels of circulating TNF-e~ are found in response to trauma or exposure to endotoxin and may originate from the vascular wall. Normal aortic rings when exposed to endotoxin in animal studies release TNF-a in quantifies similar to those observed here) 5 In summary, diseased infrarenal aortas release IL-113, which may play an important role in the pathophysiology of these diseases. The structural integrity of the arterial wall is a dynamic balance between synthesis and destruction of the extracellular matrix. The complex network of cytokines may influence this balance by modulating smooth muscle cell proliferation and differentiation. A chronic inflammatory infiltrate is fotmd in the media and adventitia of the diseased aorta with activated macrophages that are the principal source of IL-1 [3. The antigenic stimulation that initiates this immune response is unknown. Degenerating elastin fragments or modified lipoprotein are possible candidates, s,26 However, once initiated, the immune response is controlled by the balance of stimulatory and inhibitory cytokines. Although IL-l[3 is elevated in both AAAs and AOD, a slight imbalance of this cytokine in conjunction with other possible changes in IL-6, TNF-a, and interferon-gamma and inhibitors in the susceptible individual may be sufficient to alter the synthesis and degradation of collagen and elastin leading to aneurysm formation or occlusion. Further studies will be needed to fully characterize the cytokine environment in these two diseases. The role of cytokines and smooth muscle synthetic function is under active investigation in our laboratory and may help explain the changes in the structural protein observed in AAAs. We thank Dipti Parikh and Manisha Shah for their technical contributions to this research.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20. REFERENCES
1. Stella A, Gargiulo M, Pasquinetli G, et al. The cellular component in the parietal infiltrate of inflammatory abdominal aortic aneurysms (IAAA). Eur J Vasc Surg 1991;5:65-70. 2. Leu HJ. Inflammatory abdominal aortic aneurysms: a disease entity? Virchows Archiv [A] 1990;417:427-33. 3. Rizzo RJ, McCarthy WJ, Dixit SN, et al. Collagen types and matrix protein content in human abdominal aortic aneurysms. J VAsc SURG 1989;10:365-73. 4. Koch AE, Haines GK, Rizzo RJ, et al. Human abdominal aortic aneurysms: immunophenotypic analysis suggesting an
21.
22.
inamune-mediated response. Am J Pathol 1990; 137:1199213. Hansson GK, Jonasson L, Seifert PS, Stemme S. Immune mechanisms in atherosclerosis. Arteriosclerosis 1989;9:56778. Roessner A, Herrera A, Honing HJ, et al. Identification of macrophages and smooth muscle cells with monoclonal antibodies in the human atherosclerotic plaque. Virchows Arch [A] 1987;412:169-74. Brophy CM, Reilly JM, Walker-Smith GJ, Tilson MD. The role of inflammation in nonspecific abdominal aortic aneurysm disease. Ann Vasc Surg 1991;5:229-33. 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. Wang A, Doyle M, Mark D. Quantification of mRNA using the polymerase chain reaction. Proc Nail Acad Sci USA 1989;86:9717-21. Moyer CF, Sajuthi D, TuUi H, Williams JK. Synthesis oflL-1 alpha and IL-1 beta by arterial cells in atherosclerosis. Am J Pathol 1991;138:951-60. Pearce WH, Koch AE, Haines GK, Mesh CL, Parikh D. Cellular components and immune response in abdominal aortic aneurysms. Surg Forum 1991;42:328-9. Libby P, Warner SJC, Friedman GB. Interleuldn 1: a mitogen for human vascular smooth muscle cells that induces the release of growth-inhibitory prostanoids. J Clin Invest 1988; 81:487-98. Rasmussen L, Seljefid R. The modulatory effect oflipoprotein on the release of interleukin-1 by human peritoneal macrophages stimulated by WF ]-1, 3-D-polyglucose derivatives. Scand J Immunol 1989;29:477-84. Beutler BA, Cerami A. Recombinant interleukin 1 suppresses fipoprotein lipase activity in 3T3-L1 cells. J Immunol 1985 ;6:3969-71. McKenna TM. Prolonged exposure of rat aorta to low levels of endotoxin in vitro results in impaired contractility. J Clin Invest 1990;86:160-8. Busse K, Mulsch A. Induction of nitric oxide synthase by cytokines in vascular smooth muscle cells. FEBS Letr 1990; 275:87-90. Beasley D, Cohen RA, Levinsky NG. Interleukin 1 inhibits contraction of vascular smooth muscle. J Clin Invest 1989; 83:331-5. Loppnow H, Libby P. Proliferating or Interleukin 1-activated human vascular smooth muscle cells secrete copious interleukin 6. J Clin Invest 1990;85:731-8. Pober JS. What can be learned from the expression of endothelial adhesion moleolles in tissues? Lab Invest 1991; 64:301-5. Elias JA, Freundlich B, Adams S, Rosenbloom J. Regulation of human lung fibroblast collagen production by recombinant interleukin-1, tumor necrosis factor, and intefferon-~. Ann NY Acad Sci 1990;580:233-44. Jaeger E, Rust S, Roessner A, et aL Joint occurrence of collagen mRNA containing cells and macrophages in human atherosclerotic vessels. Atherosclerosis 1991;31:55-68. McGee GS, Baxter BT, Shively VS, et al. Aneurysm or occlusive disease: factors determining the clinical course of atherosclerosis of the infrarenal aorta. Surgery 1991;110: 370-6.
23. Mannel D, Moore R, Magenhagen S. Macrophages as a source
Volume 16 Number 5 November 1992
of mmoricidal activity (nmlor necrosis factor). Infect Immun 1980;30:523-30. 24. Warner SJ, Libby P. Hlmaan vascular smooth muscle cells. Target for and source of minor necrosis factor. J Immunol 1989; 142:100-9. 25. Lebovich SJ, Polverini PJ, Shepard HM, et al. Macrophageinduced angiogenesis is mediated by tumor necrosis factor-c~. Nature 1987;329:630-2.
Interleukin 1 ~ and tumor necrosisfactor-a release 789
26. O'Brian }rp. A concept of diffuse actinic arteritis: the rote of actinic damage to elastin in 'age change' and arterifis of the temporal artery and in polymyalgia rheumatica. Br I Dermatot 1978;98:1-13.
Submitted Oct. 15, 1991; accepted Mar. 3, 1992.
DISCUSSION Dr. Peter Gloviczki (Rochester, Minn.). I congratulate Dr. Pearce on his excellent presentation and the Northwestern group for their continuing interest in the research of this important problem: the etiology of aortic aneurysms. Dr. Pearce clearly demonstrated to us today that an inflammatory process is present in aneurysmal disease and that the production of at least one of the cytokines, IL-113, is elevated in aortic aneurysms. The authors concluded that this cytokine may contribute to the pathogenesis of aortic aneurysms. My first question concerns this conclusion. Are these cytokines really the marker of aneurysmal disease or are they really the marker ofatherosclerosis? I ask this question because I noted that there was no significant difference in the production of cytokines between aneurysms and occlusive disease. I am also interested in knowing if you have any data to support that the inflammatory process is a cause of aneurysm formation and not really a consequence of aneurysmal dilation. My third question concerns your histologic findings. We know that the most important change that we see in the aneurysmal wall is the destruction of elastin in the media of the aneurysm or the aortic wall. Did you find inflammatory infiltrate in the area of elastin destruction? Conversely, did you have a normal elastin staining next to areas o f significant or marked inflammation? The presence o f T N F - a was demonstrated by other investigators to be present and elevated in human atheroma. What is your explanation that you could not confirm an elevated level in your experiments? Finally, how would you explain the role of chronic inflammation in the pathogenesis of aneurysms in those patients who have obvious genetic predisposition? Dr. W i l l i a m Pearce. This issue regarding whether this is a cause and effect or if there is really a separate disease
between atherosclerosis and aneurysm formation may be defined by understanding that it is very hard to draw from a spectrum o f disease any statistical significance between two different sets of aneurysms. I hope not to give you the thought that I think that there is a separate disease or it is the same manifestation o f atherosclerosis. The question regarding the comigration of elastin versus the inflammatory infiltrate- the elastin fibers were in the media far away from the inflammatory process. A possible explanation is that the smooth muscle ceils themselves are the source of elastase and allow the destruction of the elastin. The question you had regarding the production o f these cytokines by normal tissues is intriguing to me. You have noticed that in several cases there were elevated levels of at least I L - t . It has been suggested in trauma patients that there are elevated levels of these cytokines and that the source may be the vascular wall itself. The fourth or fifth question was why was the TNF so low, and that is another intriguing question. We have repeated the study several times with different manufacturer's ELISA kits and we have had elevation on occasion but in general there is no statistical difference. The suggestion is that perhaps there is a different sensitivity in the kit that was initially used in this experiment or possibly that there are inhibitors in the serum that was used to perform this test. The data on the TNF is not as strong as the IL-1 and may be related to an artifact of the test used. Finally, in the issue regarding the genetic questions and whether the inflammatory infiltrate affects these p a t i e n t s the only answer I have is that one rarely finds an aneurysm without the inflanamatory component. I suppose if one were able to find an anemTsm very early in its development without any inflammatory component, one would suggest that the "egg came before the chicken," but at the present time I can not answer that question either.
