- Email: [email protected]
Mechanism of calcification in atherosclerosis
Mechanism of Calcification in Atherosclerosis Linda L. Demer, Karol E. Watson, and Kristina Bostriim
Calcification is commonly associated with atherosclerosis, and it has impotiant clinical implications, especially in coronary arteries. The mineral has been identified as the same mineral as in bone, hydroxyapatite, and several features of its development suggest a mechanism similar to osteogenesis and not merely passive precipitation. The artery wall has been shown to contain several bone-related proteins, including those for osteopontin, osteonectin, and osteocalcin, as well as proteoglycan core proteins homologous with bone biglycan. Our laboratory recently demonstrated that a potent osteogenic differentiation factor, bone morphogenetic protein 2a, is expressed in calcified human atherosclerotic lesions, suggesting that arterial calcification may be initiated by an osteogenic differentiation. In addition, a cell capable of calcium mineral formation in vitro has been isolated from bovine and human aorta and identified by immunostaining as having a surface marker characteristic of microvascular pericytes. These findings suggest the possibility that plaque calcification develops when a signal from atherosclerotic plaque or a factor associated with atherosclerosis induces expression of bone morphogenetic protein, leading to osteogenic differentiation of pluripotential, pericytelike cells located in the arterial intima, which then produce bonelike matrix and hydroxyapatite mineral. These findings also raise questions as to whether osteogenic-promoting factors used to prevent osteoporosis may also increase risk of arterial calcification. (Trends Cardiovasc Med 1994;4:45-49)
l
Artery Wall Calcification
Calcification is a common feature of atherosclerotic lesions. In some cases, the mineral has the histologic appearance of fully formed, mineralized bone tissue, including trabeculations, suggestive of remodeling. Occasionally, the trabeculations
include cellular elements
resembling marrow. Such observations led 19th-century pathologists to describe artery wall calcification as “ossification” (Virchow 1863). Linda L. Demer, Karol E. Watson, and Kristina Bostrom are at the Division of Cardiology, Department of Medicine, UCLA School of Medicine, Los Angeles, CA 90024-1679, USA. TCM Vol. 4, No. I, 1994
The mechanism of atherosclerotic plaque calcification remains unknown. It is a form of ‘dystrophic calcification,” which refers to soft tissue calcification of unknown etiology in normocalcemic individuals, as opposed to metastatic calcification, which occurs in the setting of hypercalcemia. Plaque calcification is widely considered to be a passive, degenerative, and inevitable, end-stage process of aging, suggesting little value to its study, yet these same words were used to describe atherosclerosis only a few decades ago. It was even suggested that calcification represented the primary process and that “atheromatous plaques were probably a secondary phenomenon . . . resulting from the calcification”
01994,
Elsevier
Science
Inc.,
1050-I
73&J/94/$7.00
(Lansing et al. 1948). Although calcification is now considered the secondary phenomenon, current studies suggest that it will soon be recognized as an active, regulated, multifactorial, and potentially preventable process, just as research has shown for atherosclerosis. Calcification is most commonly located at the base of atherosclerotic plaque, but it frequently occurs at other sites throughout the plaque as well as in the media and adventitia. Total calcium content increases IO- to 80-fold in coronary arteries when atherosclerotic lesions develop (Fleckenstein et al. 1990). Early investigators used light microscopy to demonstrate that calcium deposits were associated with elastic tissue. More recently, Kruth and colleagues used scanning electron microscopy to demonstrate that the calcium deposits were associated with small crystals of unesterified cholesterol (Hirsch et al. 1993). New imaging methods have shown that coronary calcification is more common than previously believed and is not limited to a minority of older patients with end-stage disease. The general impression that it is uncommon may come from its infrequent appearance in textbook examples of sections of atherosclerotic lesions, which may simply reflect selection bias in that calcified specimens are poorly suited for histologic sectioning. By intravascular ultrasound, >80% of significant coronary lesions are calcified (Honye et al. 1992). Furthermore, Detrano and colleagues found by ultrafast computed tomographic (CT) scanning that 90% of patients with coronary artery disease had coronary calcification (Agatston et al. 1990). This method has shown that calcification occurs so commonly in atherosclerosis that its greatest value is as a negative test: the absence of calcium deposits by ultrafast CT is highly predictive of absence of obstructive atherosclerotic disease (>75% narrowing) (Simons et al. 1992). Presence of calcification in coronary atherosclerosis has been associated with a less favorable prognosis. Early studies of the link between calcification and myocardial infarction are limited, because only large deposits of calcium could be detected by fluoroscopy. Greater sensitivity is achieved with intravascular ultrasound or ultrafast CT, but even these methods probably only recognize
45
macroscopic mineralization. Autopsystudies have shown that calcification increases the risk of myocardial infarction independently of age (Beadenkopf et al. 1964). Others have shown increased likelihood of ischemia in peripheral vascular lesions that are calcified (Niskanen et al. 1990). The clinical importance of coronary calcification may derive from its contribution to plaque rupture (Sharma et al. 1993). According to biomechanical analyses, plaque rupture occurs at sites of highest stress concentration. In general, stress concentration is greatly increased at interfaces between rigid and compliant structures such as between calcium mineral and normal arterial wall or fibrous plaque. Other possible mechanisms are that it interferes with vasomotion and, possibly, compensatory enlargement (Glagov et al. 1987). Loss of normal vasomotion is probably due to loss of distensibility (Demer 1991). Reduced distensibility may also interfere with the process of compensatory enlargement (Glagov et al. 1987), in which the artery wall dilates to maintain lumen size as plaque increasingly fills space.
l
Bone-Related Factors in Atherosclerotic Calcification
In addressing the question of whether artery wall calcification occurs by a mechanism similar to osteogenesis, our laboratory recently demonstrated that calcified human atherosclerotic lesions contain cells that express bone morphogenetic protein 2a (BMP-2a), a potent bone differentiation factor (Bostrom et al. 1993). BMP activity was discovered by Urist 20 years ago (Urist 1965, Urist and Strates 1971), and it was characterized biochemically and molecularly (Wozney et al. 1988) in the past few years. BMP-2a, one of eight members of the BMP family, is capable of inducing bone tissue formation from other mesenchymal tissue such as skeletal muscle, both in vitro and in vivo (Khouri et al. 1991). The presence of this bone differentiation factor in atherosclerotic lesions suggests that artery wall calcification may begin with osteogenic differentiation of artery wall cells. Several features of atherosclerotic plaque are consistent with the concept of an osteogenic origin of calcium deposits. X-ray crystallography and electron mi-
46
croprobe analysis of mineral contained within atherosclerotic lesions have identified it as mature and immature hydroxyapatite, the same calcium phosphate mineral found in bone, rather than simply amorphous calcium phosphate (S&mid et al. 1980). By transmission electron microscopy, Anderson (1983) and colleagues have identified structures in atherosclerotic plaque that resemble matrix vesicles involved in bone formation. Bone matrix vesicles are proteolipid spheres -200 nm in diameter that contain alkaline phosphatase activity. They are believed to be derived from osteoblasts and to serve as the initial nidus for hydroxyapatite crystal formation. Osteocalcin, also known as bone gla protein because of its content of carboxyglutamic acid residues, as well as another gla-containing protein have been identified within mineralized atherosclerotic plaques (Levy et al. 1979, Gijsbers et al. 1990). Osteocalcin is believed to be synthesized exclusively by osteoblasts, yet elevated serum levels of osteocalcin, believed to indicate bone remodeling, are found in patients with atherosclerosis and no bone disease (Koyama et al. 1991). Osteopontin, an acidic phosphoprotein present in bone, has been found to be associated with artery wall cells in two separate cDNA-subtraction library studies. By this method, Giachelli and colleagues ( 199 1) made the unexpected observation that osteopontin is expressed by neointimal cells in rat carotid arteries after balloon injury. It was then found to be expressed as a cell-cycle-specific event in cultured smooth muscle cells (Gadeau et al. 1993). Preliminary evidence suggests that osteopontin is also present in human atherosclerotic lesions (Ingram et al. 1993). Although osteopontin is not strictly specific to bone, its presence in the artery may contribute to an osteogenic milieu. Acidic phosphoproteins such as osteopontin are believed to have an important role in hydroxyapatite crystal formation. According to one proposed model, the acidic portions of osteopontin attract local calcium ions that are then drawn into close proximity with the protein’s phosphate groups when its tertiary structure is altered by binding to collagen. This rapid association of ions may trigger calcium phosphate crystal formation. Osteopontin also contains Arg-Gly-Asp amino acid sequences by 01994. Elsevier Science Inc., 1050-1738/94/$7.00
which osteoclasts, and possibly osteoblasts, adhere to bone matrix. Osteonectin, a bone glycoprotein also known as SPARC (secreted protein, acidic and rich in cysteine), is also present in the artery wall. It binds tightly to hydroxyapatite and collagen. One proposed function of osteonectin is the release of cellular adhesion to bone matrix. Vascular smooth muscle cells also produce a proteochondroitin core protein that shares sequence homology with the biglycan core protein from human bone (Marcum and Thompson 1991). Other proteins, including the collagen subtypes that predominate in bone matrix, are found both in cultured vascular cells and in atherosclerotic lesions.
l
An Artery Wall Cell with Osteoblast and Pericyte Features
There has been little previous information about the cell in the artery wall responsible for producing calcium mineral deposits. We recently isolated and cloned a cell from the aortic media that produces hydroxyapatite mineral in vitro (Bostrom et al. 1993). The deposits form within cellular nodules after l-2 weeks in culture. Nearly identical nodules are produced by osteoblastlike cells in culture, and similar nodules occur in vivo as the starting point for embryologic development of bone (Rosen and Thies 1992). Similar aggregates with calcification are produced by embryonic limb bud cells and marrow stromal cells. Formation of nodules by vascular smooth muscle cells in culture had been described previously by Bjorkerud (1987), but calcification had not been recognized. In the past, nodule formation in smooth muscle cell cultures has been considered a sign of senescence, and cultures containing them are discarded. By trypsin digestion, we were able to isolate nodules from the surrounding monolayer. When the nodules were replated, cells grew out as from an explant, and these new cultures formed nodules at lo- to IOO-fold greater density than the primary cultures or cloned smooth muscle cells on the same substrate. By dilutional cloning, we isolated the predominant cell present in nodules. Cloned cultures also formed nodules producing calcium mineral deposits at high density, suggesting that the cells cloned from the nodules are responsible for vascular calcification. TCM Vol. 4, No. I, 1994
By immunofluorescent staining, nodtdederived cells were similar to endothelial cells in expression of f%actin, but different in their lack of von Willebrand factor. They were similar to smooth muscle cells in expression of a-actin, but distinct in having strong expression of fl-actin and a surface ganglioside recognized by monoclonal antibody 3G5 (Nayak et al. 1992), a marker for vascular pericytes (Figure 1). To assess whether these unusual cellular features were not simply phenotypic modulation due to culture conditions, effects of nodule formation, or manipulation in cloning, we examined bovine and human thoracic aortic tissue by immunohistochemistry using 3G5. Positive staining cells were found within the aortic intima as scattered single cells and as clusters. Positive cells were also found, but less often, in the media and adventitia. These results suggested that pericytelike cells are present in adult arteries. The relation of these pericytelike cells in the aortic wall to microvascular pericytes is not clear. Pericytes are defined anatomically by their location in the capillaries, within the basement membrane of endothelial cells. Ultrastructurally, they resemble fibroblasts, but they produce basement membrane and have more extensive branching processes. In many respects, their ultrastructural characteristics are most similar to those of embryonic mesenchymal cells (Rhodin 1968), and it has been suggested that they are derived from dedifferentiated smooth muscle cells. One possibility, considering the anatomic continuum between microvascular pericytes, arteriolar intermediate smooth muscle cells, and aortic intima, is that intimal pericytelike cells represent an embryonic remnant. Interestingly, retinal microvascular pericytes also form nodules and hydroxyapatite mineral in vitro (Schor et al. 1990). Our finding that a subpopulation of aortic wall cells is capable of osteoblastlike hydroxyapatite production is consistent with functional evidence that pericytes retain pluripotentiality. In the periosteum surrounding bone, pericytes were shown to become osteoblasts in response to bone injury (Brighton et al. 1992). Other cells potentially arising from pericytes include fibroblasts, smooth muscle cells, chondroblasts, and preadipocytes (Meyrick et al. 198 1, DiazFlores et al. 1990).
TCM Vol. 4, No. 1, 1994
b Figure 1. Immunofluorescence images showing positivity for pericyte marker in calcifying vascular cells. Cells were cloned from calcified nodules in bovine aortic smooth muscle cell cultures. (a) Positive surface labeling with monoclonal antibody 3G5, a cell surface marker specific for pericytes in the vasculature. (b) Negative control image of cells from the same clone labeled with secondary antibody alone.
0 Conclusions In summary,
calcification
is commonly
associated with atherosclerosis, and it has important clinical implications, especially features
in coronary arteries. Several of its development suggest a
mechanism similar to osteogenesis rather than mere passive precipitation, includ-
01994,
Elsevier Science Inc., 1050-1738/94/$7.00
ing expression of the potent osteogenic differentiation factor, BMP-2a, in calcified human atherosclerotic lesions. In addition, a cell capable of calcium mineral formation in vitro has been isolated from bovine and human aorta and identified by immunostaining as having a surface marker characteristic of microvascular pericytes. These findings sug-
47
Bjorkexud S: 1987. Agglomeration to nodules modulates human arterial smooth muscle cells to distinct postinjury phenotype via foam cell transition. Am J Path01 127:485498. Bostrom K, Watson KE, Horn S, Wortham C, Herman IM, Demer LL: 1993. Bone morphogenetic protein expression in human atherosclerotic lesions. J Clin Invest 9 1: 18001809. Brighton CT, Lorich DG, Kupcha R, Reilly TM, Jones AR, Woodbury RA: 1992. The pericyte as a possible osteoblast progenitor cell. Clin Orthop 275:287-299. Demer LL: 1991. Effect of calcification on in vivo mechanical response of rabbit arteries to balloon dilation. Circulation 83:20832093.
Figure 2. Possible origin of bone mineral-forming cells in artery wall from pericytelike pluripotential cells. Microvascular pericytes are known to differentiate into smooth muscle cells in response to PDGF and into osteoblasts in response to TCF-8. It is proposed that they form a variety of mature mesenchymal tissues in response to different stresses such as injury or mechanical stress. gest the possibility that plaque calcification develops when a signal from atherosclerotic plaque or a factor associated with atherosclerosis induces expression of bone morphogenetic protein, leading to osteogenic differentiation of pluripotential, pericytelike cells located in the arterial intima, which then produce bone matrix and hydroxyapatite crystals (Figure 2).
l
Future Directions
Future research may focus on what factor in atherosclerotic lesions stimulates bone morphogenetic protein and osteoblastic differentiation of pericytelike cells. For example, transforming growth factor b is present in atherosclerotic lesions, and it is capable of stimulating osteogenesis (Noda and Camilliere 1989). Identification of the promoter for the BMP-2a gene will be of great interest, and the role of modulating factors such as activin needs to be assessed. Further characterization of the proteins in the mineral matrix and the cells responsible for their synthesis is needed. Such studies may also provide clues to the general mechanism of ectopic soft tissue calcification occurring in a variety of chronic inflammatory conditions. A pressing issue for cardiovascular medicine is to determine what doses of
4s
dietary vitamin D and calcium supplements may put patients at increased risk of atherosclerotic calcification as they do at high doses in animals (Demer 1991, Ito et al. 1990, Take0 et al. 1989). This question is particularly important in treatment of postmenopausal women with coronary artery disease who may be receiving multiple calcification-regulating agents including vitamin-D and calcium supplements for prevention of osteoporosis. The possible effects of these agents, whether positive or negative, on artery wall calcification are not yet known.
l
Acknowledgment
This work was supported in part by Public Health Service grants HL-30568, HL-43379, HL-35570, and HL-3 1249. References Agatston AS, Janowitz WR, Hildner FJ, Zusmer NR, Viamonte M, Detrano R: 1990. Quantification of coronary artery calcium using ultrafast computed tomography. J Am Co11Cardiol 15:827-832. Anderson HC: 1983. Calcific diseases: a concept. Arch Path01 Lab Med 107:341-348. Beadenkopf WG, Daoud AS, Love BM: 1964. Calcification in the coronary arteries and its relationship to arteriosclerosis and myocardial infarction. Am J Roentgen01 92:865871.
01994, Elsevier Science Inc., lOSO-1738/94/$7.00
Diaz-mores L, Gutierrez R, Varela H, Rancel N, Valladares F: 1o%l . Microvascular pericytes: a review of their morphological and functional characteristics. Histol Histopathol6:269-286. Fleckenstein A, Fleckenstein-Griin G, Frey M, Thimm F: 1990. Experimental antiarteriosclerotic effects of calcium antagonists. J Clin Pharmacol30: 15 l-l 54. Gadeau A-P, Campan M, Millet D, Candresse T, Desgranges C: 1993. Osteopontin overexpression is associated with arterial smooth muscle cell proliferation in vitro. Arterioscler Thtomb 13:120-125. Giachelli C, Bae N, Lombardi D, Majesky M, Schwartz S: 1991. Molecular cloning and characterization of 2B7, a rat mRNA which distinguishes smooth muscle cell phenotypes in vitro and is identical to osteopontin (secreted phosphoprotein I, 2aR). Biochem Biophys Res Commun 177:867-873. Gijsbers BLMG, van Haarlem JM, Soute BAM, Ebbetink RHM, Vermeer C: 1990. Characterization of a Gla-containing protein from calcified human atherosclerotic plaques. Arteriosclerosis 10:991-995. Glagov S, Weisenberg E, Zarins CK, Stankunavicius R, Kolettis GJ: 1987. Compensatory enlargement of human atherosclerotic coronary arteries. N Engl J Med 3 16: 137 l-l 375. Hirsch D, Azoury R, Sarig S, Kruth HS: 1993. Colocalization of cholesterol and hydroxyapatite in human atherosclerotic lesions. Calcif Tissue Int 52:94-98. Honye J, Mahon DJ, Jain A, et al.: 1992. Morphological effects of coronary balloon angioplasty in vivo assessed by intravascular ultrasound imaging. Circulation 85: 10 121025. Ingram RT, Fitzpatrick LA, Edwards WD, Frye RL, Fisher LW, Schwartz RS: 1993. Calcification in human coronary atherosclerosis is specifically associated with osteopontin, a bone matrix protein [abst]. J Am Co11Cardiol21(Suppl):363A.
TCM Vol.4,No. 1,1994
Ito
M, Cho HS, Kummerow FA: 1990. Effects of a dietary magnesium deficiency and excess vitamin D3 on swine coronary arteries. J Am Co11Nutr 9:155-163.
Khouri RK, Koudsi B, Reddi H: 1991. Tissue transformation into bone in vivo: a potential practical application. JAMA 266: 19531955. Koyama N, Ohara K, Yokota H, et al.: 1991. A one-step sandwich enzyme immunoassay for gamma-carboxylated osteocalcin using monoclonal antibodies. J Immunol Methods 139:17-23. Lansing AI, Blumenthal HT, Gray SH: 1948. Aging and the calcification of the human coronary artery. J Gerontology 3:87-97. Levy RJ, Lian JB, Gallop PM: 1979. Atherocalcin, a gamma-carboxyglutamic acidcontaining protein from atherosclerotic plaque. Biochem Biophys Res Commun 91:4149. Marcum JA, Thompson MA: 1991. The aminoterminal region of a proteochondroitin core protein, secreted by aortic smooth muscle cells, shares sequence homology with the pre-propeptide region of the biglycan core protein from human bone. Biochem Biophys Res Commun 175:706-712. Mevrick B. Fuiiwara K. Reid L: 1981. Smooth muscle myosin in precursor and mature smooth muscle cells in normal pulmonary
arteries and the effect of hypoxia. Exp Lung Res 2:303-313.
retinal microvasculature undergo calcification in vitro. J Cell Sci 97:449-461.
Nayak RC, Attawia MA, Cahill CJ, King GL, Ohashi H, Moromisato R: 1992. Expression of a monoclonal antibody (3G5) defined ganglioside antigen in the renal cortex. Kidney Int 41:1638-1645.
Sharma SK, Israel DH, Kamean JL, Bodian CA, Ambrose JA: 1993. Clinical, angiographic, and procedural determinants of major and minor coronary dissection during angioplasty. Am Heart J 126:39-47.
Niskanen LK, Suhonen M, Siitonen 0, Lehtinen JM, Uusitupa MIJ: 1990. Aortic and lower limb artery calcification in type 2 (non-insulin-dependent) diabetic patients and non-diabetic control subjects: a five year follow-up study. Atherosclerosis 84:6171.
Simons DB, Schwartz RS, Edwards WD, Sheedy PF, Breen JF, Rumberger JA: 1992. Noninvasive definition of anatomic coronary artery disease by ultrafast computed tomographic scanning: a quantitative pathologic comparison study. J Am Coll Cardio120:1118-1126.
Noda M, Camilliere JJ: 1989. In vivo stimulation of bone formation by transfotmation growth factor-beta. Endocrinology 124:29912994.
Takeo S, Anan M, Fujioka K, et al.: 1989. Functional changes of aorta with massive accumulation of calcium. Atherosclerosis 77:175-181.
Rhodin JAG: 1968. Ultrastructure of mammalian venous capillaries, venules, and small collecting veins. J Ultrastruct Res 25:452-500.
Urist MR: 1965. Bone: formation duction. Science 150:893-899.
Rosen V, Thies RS: 1992. The BMP proteins in bone formation and repair. Trends Genet 8:97-102.
Virchow R: 1863. Cellular Pathology: As Based upon Physiological and Pathological Histology [trans Frank Chance, 1971; an unabridged and unaltered republication of the English translation originally published by Dover]. New York, Dover, pp 404-408.
Schmid K, McSharry WO, Pameijer CH, Binette JP: 1980. Chemical and physicochemical studies on the mineral deposits of the human atherosclerotic aorta. Atherosclerosis 37: 199-2 10. Schor AM, Allen TD, Canfield AE, Sloan P, Schor SL: 1990. Pericytes derived from the
by autoin-
Urist MR. Strates BS: 1971. Bone morphogenetic protein. J Dent Res 50:1392-1406.
Wozney JM, Rosen V, Celeste AI, et al.: 1988. Novel regulators of bone formation: molecular clones and activities. Science 242: 1528 TCM 1534.
Why wait for it to circulate? Read your own copy of
If you share TCM with others at work, you know what goes around doesn’t always come around. And if it does, articles are often dated or sometimes clipped. Why miss a single issue, or wait weeks to read one? By starting your personal subscription, for just $79 you’ll be guaranteed to read each and every issue. TRENDS IN CARDIOVASCULAR MEDICINE’s short reviews on ‘hot” topics can make a difference in your research and practice. Commissioned researchers and clinicians report on advances in cardiology and its related areas. Use the business reply cards in this issue to order your personal subscription today-or call Elsevier Customer Service at
(212)633-3950 or FAX (212)633-3880 TCM Vol. 4,No.1,1994
01994, Elsevier Science Inc., 1050-1738/94/$7.00
49
Calcification is commonly associated with atherosclerosis, and it has impotiant clinical implications, especially in coronary arteries. The mineral has been identified as the same mineral as in bone, hydroxyapatite, and several features of its development suggest a mechanism similar to osteogenesis and not merely passive precipitation. The artery wall has been shown to contain several bone-related proteins, including those for osteopontin, osteonectin, and osteocalcin, as well as proteoglycan core proteins homologous with bone biglycan. Our laboratory recently demonstrated that a potent osteogenic differentiation factor, bone morphogenetic protein 2a, is expressed in calcified human atherosclerotic lesions, suggesting that arterial calcification may be initiated by an osteogenic differentiation. In addition, a cell capable of calcium mineral formation in vitro has been isolated from bovine and human aorta and identified by immunostaining as having a surface marker characteristic of microvascular pericytes. These findings suggest the possibility that plaque calcification develops when a signal from atherosclerotic plaque or a factor associated with atherosclerosis induces expression of bone morphogenetic protein, leading to osteogenic differentiation of pluripotential, pericytelike cells located in the arterial intima, which then produce bonelike matrix and hydroxyapatite mineral. These findings also raise questions as to whether osteogenic-promoting factors used to prevent osteoporosis may also increase risk of arterial calcification. (Trends Cardiovasc Med 1994;4:45-49)
l
Artery Wall Calcification
Calcification is a common feature of atherosclerotic lesions. In some cases, the mineral has the histologic appearance of fully formed, mineralized bone tissue, including trabeculations, suggestive of remodeling. Occasionally, the trabeculations
include cellular elements
resembling marrow. Such observations led 19th-century pathologists to describe artery wall calcification as “ossification” (Virchow 1863). Linda L. Demer, Karol E. Watson, and Kristina Bostrom are at the Division of Cardiology, Department of Medicine, UCLA School of Medicine, Los Angeles, CA 90024-1679, USA. TCM Vol. 4, No. I, 1994
The mechanism of atherosclerotic plaque calcification remains unknown. It is a form of ‘dystrophic calcification,” which refers to soft tissue calcification of unknown etiology in normocalcemic individuals, as opposed to metastatic calcification, which occurs in the setting of hypercalcemia. Plaque calcification is widely considered to be a passive, degenerative, and inevitable, end-stage process of aging, suggesting little value to its study, yet these same words were used to describe atherosclerosis only a few decades ago. It was even suggested that calcification represented the primary process and that “atheromatous plaques were probably a secondary phenomenon . . . resulting from the calcification”
01994,
Elsevier
Science
Inc.,
1050-I
73&J/94/$7.00
(Lansing et al. 1948). Although calcification is now considered the secondary phenomenon, current studies suggest that it will soon be recognized as an active, regulated, multifactorial, and potentially preventable process, just as research has shown for atherosclerosis. Calcification is most commonly located at the base of atherosclerotic plaque, but it frequently occurs at other sites throughout the plaque as well as in the media and adventitia. Total calcium content increases IO- to 80-fold in coronary arteries when atherosclerotic lesions develop (Fleckenstein et al. 1990). Early investigators used light microscopy to demonstrate that calcium deposits were associated with elastic tissue. More recently, Kruth and colleagues used scanning electron microscopy to demonstrate that the calcium deposits were associated with small crystals of unesterified cholesterol (Hirsch et al. 1993). New imaging methods have shown that coronary calcification is more common than previously believed and is not limited to a minority of older patients with end-stage disease. The general impression that it is uncommon may come from its infrequent appearance in textbook examples of sections of atherosclerotic lesions, which may simply reflect selection bias in that calcified specimens are poorly suited for histologic sectioning. By intravascular ultrasound, >80% of significant coronary lesions are calcified (Honye et al. 1992). Furthermore, Detrano and colleagues found by ultrafast computed tomographic (CT) scanning that 90% of patients with coronary artery disease had coronary calcification (Agatston et al. 1990). This method has shown that calcification occurs so commonly in atherosclerosis that its greatest value is as a negative test: the absence of calcium deposits by ultrafast CT is highly predictive of absence of obstructive atherosclerotic disease (>75% narrowing) (Simons et al. 1992). Presence of calcification in coronary atherosclerosis has been associated with a less favorable prognosis. Early studies of the link between calcification and myocardial infarction are limited, because only large deposits of calcium could be detected by fluoroscopy. Greater sensitivity is achieved with intravascular ultrasound or ultrafast CT, but even these methods probably only recognize
45
macroscopic mineralization. Autopsystudies have shown that calcification increases the risk of myocardial infarction independently of age (Beadenkopf et al. 1964). Others have shown increased likelihood of ischemia in peripheral vascular lesions that are calcified (Niskanen et al. 1990). The clinical importance of coronary calcification may derive from its contribution to plaque rupture (Sharma et al. 1993). According to biomechanical analyses, plaque rupture occurs at sites of highest stress concentration. In general, stress concentration is greatly increased at interfaces between rigid and compliant structures such as between calcium mineral and normal arterial wall or fibrous plaque. Other possible mechanisms are that it interferes with vasomotion and, possibly, compensatory enlargement (Glagov et al. 1987). Loss of normal vasomotion is probably due to loss of distensibility (Demer 1991). Reduced distensibility may also interfere with the process of compensatory enlargement (Glagov et al. 1987), in which the artery wall dilates to maintain lumen size as plaque increasingly fills space.
l
Bone-Related Factors in Atherosclerotic Calcification
In addressing the question of whether artery wall calcification occurs by a mechanism similar to osteogenesis, our laboratory recently demonstrated that calcified human atherosclerotic lesions contain cells that express bone morphogenetic protein 2a (BMP-2a), a potent bone differentiation factor (Bostrom et al. 1993). BMP activity was discovered by Urist 20 years ago (Urist 1965, Urist and Strates 1971), and it was characterized biochemically and molecularly (Wozney et al. 1988) in the past few years. BMP-2a, one of eight members of the BMP family, is capable of inducing bone tissue formation from other mesenchymal tissue such as skeletal muscle, both in vitro and in vivo (Khouri et al. 1991). The presence of this bone differentiation factor in atherosclerotic lesions suggests that artery wall calcification may begin with osteogenic differentiation of artery wall cells. Several features of atherosclerotic plaque are consistent with the concept of an osteogenic origin of calcium deposits. X-ray crystallography and electron mi-
46
croprobe analysis of mineral contained within atherosclerotic lesions have identified it as mature and immature hydroxyapatite, the same calcium phosphate mineral found in bone, rather than simply amorphous calcium phosphate (S&mid et al. 1980). By transmission electron microscopy, Anderson (1983) and colleagues have identified structures in atherosclerotic plaque that resemble matrix vesicles involved in bone formation. Bone matrix vesicles are proteolipid spheres -200 nm in diameter that contain alkaline phosphatase activity. They are believed to be derived from osteoblasts and to serve as the initial nidus for hydroxyapatite crystal formation. Osteocalcin, also known as bone gla protein because of its content of carboxyglutamic acid residues, as well as another gla-containing protein have been identified within mineralized atherosclerotic plaques (Levy et al. 1979, Gijsbers et al. 1990). Osteocalcin is believed to be synthesized exclusively by osteoblasts, yet elevated serum levels of osteocalcin, believed to indicate bone remodeling, are found in patients with atherosclerosis and no bone disease (Koyama et al. 1991). Osteopontin, an acidic phosphoprotein present in bone, has been found to be associated with artery wall cells in two separate cDNA-subtraction library studies. By this method, Giachelli and colleagues ( 199 1) made the unexpected observation that osteopontin is expressed by neointimal cells in rat carotid arteries after balloon injury. It was then found to be expressed as a cell-cycle-specific event in cultured smooth muscle cells (Gadeau et al. 1993). Preliminary evidence suggests that osteopontin is also present in human atherosclerotic lesions (Ingram et al. 1993). Although osteopontin is not strictly specific to bone, its presence in the artery may contribute to an osteogenic milieu. Acidic phosphoproteins such as osteopontin are believed to have an important role in hydroxyapatite crystal formation. According to one proposed model, the acidic portions of osteopontin attract local calcium ions that are then drawn into close proximity with the protein’s phosphate groups when its tertiary structure is altered by binding to collagen. This rapid association of ions may trigger calcium phosphate crystal formation. Osteopontin also contains Arg-Gly-Asp amino acid sequences by 01994. Elsevier Science Inc., 1050-1738/94/$7.00
which osteoclasts, and possibly osteoblasts, adhere to bone matrix. Osteonectin, a bone glycoprotein also known as SPARC (secreted protein, acidic and rich in cysteine), is also present in the artery wall. It binds tightly to hydroxyapatite and collagen. One proposed function of osteonectin is the release of cellular adhesion to bone matrix. Vascular smooth muscle cells also produce a proteochondroitin core protein that shares sequence homology with the biglycan core protein from human bone (Marcum and Thompson 1991). Other proteins, including the collagen subtypes that predominate in bone matrix, are found both in cultured vascular cells and in atherosclerotic lesions.
l
An Artery Wall Cell with Osteoblast and Pericyte Features
There has been little previous information about the cell in the artery wall responsible for producing calcium mineral deposits. We recently isolated and cloned a cell from the aortic media that produces hydroxyapatite mineral in vitro (Bostrom et al. 1993). The deposits form within cellular nodules after l-2 weeks in culture. Nearly identical nodules are produced by osteoblastlike cells in culture, and similar nodules occur in vivo as the starting point for embryologic development of bone (Rosen and Thies 1992). Similar aggregates with calcification are produced by embryonic limb bud cells and marrow stromal cells. Formation of nodules by vascular smooth muscle cells in culture had been described previously by Bjorkerud (1987), but calcification had not been recognized. In the past, nodule formation in smooth muscle cell cultures has been considered a sign of senescence, and cultures containing them are discarded. By trypsin digestion, we were able to isolate nodules from the surrounding monolayer. When the nodules were replated, cells grew out as from an explant, and these new cultures formed nodules at lo- to IOO-fold greater density than the primary cultures or cloned smooth muscle cells on the same substrate. By dilutional cloning, we isolated the predominant cell present in nodules. Cloned cultures also formed nodules producing calcium mineral deposits at high density, suggesting that the cells cloned from the nodules are responsible for vascular calcification. TCM Vol. 4, No. I, 1994
By immunofluorescent staining, nodtdederived cells were similar to endothelial cells in expression of f%actin, but different in their lack of von Willebrand factor. They were similar to smooth muscle cells in expression of a-actin, but distinct in having strong expression of fl-actin and a surface ganglioside recognized by monoclonal antibody 3G5 (Nayak et al. 1992), a marker for vascular pericytes (Figure 1). To assess whether these unusual cellular features were not simply phenotypic modulation due to culture conditions, effects of nodule formation, or manipulation in cloning, we examined bovine and human thoracic aortic tissue by immunohistochemistry using 3G5. Positive staining cells were found within the aortic intima as scattered single cells and as clusters. Positive cells were also found, but less often, in the media and adventitia. These results suggested that pericytelike cells are present in adult arteries. The relation of these pericytelike cells in the aortic wall to microvascular pericytes is not clear. Pericytes are defined anatomically by their location in the capillaries, within the basement membrane of endothelial cells. Ultrastructurally, they resemble fibroblasts, but they produce basement membrane and have more extensive branching processes. In many respects, their ultrastructural characteristics are most similar to those of embryonic mesenchymal cells (Rhodin 1968), and it has been suggested that they are derived from dedifferentiated smooth muscle cells. One possibility, considering the anatomic continuum between microvascular pericytes, arteriolar intermediate smooth muscle cells, and aortic intima, is that intimal pericytelike cells represent an embryonic remnant. Interestingly, retinal microvascular pericytes also form nodules and hydroxyapatite mineral in vitro (Schor et al. 1990). Our finding that a subpopulation of aortic wall cells is capable of osteoblastlike hydroxyapatite production is consistent with functional evidence that pericytes retain pluripotentiality. In the periosteum surrounding bone, pericytes were shown to become osteoblasts in response to bone injury (Brighton et al. 1992). Other cells potentially arising from pericytes include fibroblasts, smooth muscle cells, chondroblasts, and preadipocytes (Meyrick et al. 198 1, DiazFlores et al. 1990).
TCM Vol. 4, No. 1, 1994
b Figure 1. Immunofluorescence images showing positivity for pericyte marker in calcifying vascular cells. Cells were cloned from calcified nodules in bovine aortic smooth muscle cell cultures. (a) Positive surface labeling with monoclonal antibody 3G5, a cell surface marker specific for pericytes in the vasculature. (b) Negative control image of cells from the same clone labeled with secondary antibody alone.
0 Conclusions In summary,
calcification
is commonly
associated with atherosclerosis, and it has important clinical implications, especially features
in coronary arteries. Several of its development suggest a
mechanism similar to osteogenesis rather than mere passive precipitation, includ-
01994,
Elsevier Science Inc., 1050-1738/94/$7.00
ing expression of the potent osteogenic differentiation factor, BMP-2a, in calcified human atherosclerotic lesions. In addition, a cell capable of calcium mineral formation in vitro has been isolated from bovine and human aorta and identified by immunostaining as having a surface marker characteristic of microvascular pericytes. These findings sug-
47
Bjorkexud S: 1987. Agglomeration to nodules modulates human arterial smooth muscle cells to distinct postinjury phenotype via foam cell transition. Am J Path01 127:485498. Bostrom K, Watson KE, Horn S, Wortham C, Herman IM, Demer LL: 1993. Bone morphogenetic protein expression in human atherosclerotic lesions. J Clin Invest 9 1: 18001809. Brighton CT, Lorich DG, Kupcha R, Reilly TM, Jones AR, Woodbury RA: 1992. The pericyte as a possible osteoblast progenitor cell. Clin Orthop 275:287-299. Demer LL: 1991. Effect of calcification on in vivo mechanical response of rabbit arteries to balloon dilation. Circulation 83:20832093.
Figure 2. Possible origin of bone mineral-forming cells in artery wall from pericytelike pluripotential cells. Microvascular pericytes are known to differentiate into smooth muscle cells in response to PDGF and into osteoblasts in response to TCF-8. It is proposed that they form a variety of mature mesenchymal tissues in response to different stresses such as injury or mechanical stress. gest the possibility that plaque calcification develops when a signal from atherosclerotic plaque or a factor associated with atherosclerosis induces expression of bone morphogenetic protein, leading to osteogenic differentiation of pluripotential, pericytelike cells located in the arterial intima, which then produce bone matrix and hydroxyapatite crystals (Figure 2).
l
Future Directions
Future research may focus on what factor in atherosclerotic lesions stimulates bone morphogenetic protein and osteoblastic differentiation of pericytelike cells. For example, transforming growth factor b is present in atherosclerotic lesions, and it is capable of stimulating osteogenesis (Noda and Camilliere 1989). Identification of the promoter for the BMP-2a gene will be of great interest, and the role of modulating factors such as activin needs to be assessed. Further characterization of the proteins in the mineral matrix and the cells responsible for their synthesis is needed. Such studies may also provide clues to the general mechanism of ectopic soft tissue calcification occurring in a variety of chronic inflammatory conditions. A pressing issue for cardiovascular medicine is to determine what doses of
4s
dietary vitamin D and calcium supplements may put patients at increased risk of atherosclerotic calcification as they do at high doses in animals (Demer 1991, Ito et al. 1990, Take0 et al. 1989). This question is particularly important in treatment of postmenopausal women with coronary artery disease who may be receiving multiple calcification-regulating agents including vitamin-D and calcium supplements for prevention of osteoporosis. The possible effects of these agents, whether positive or negative, on artery wall calcification are not yet known.
l
Acknowledgment
This work was supported in part by Public Health Service grants HL-30568, HL-43379, HL-35570, and HL-3 1249. References Agatston AS, Janowitz WR, Hildner FJ, Zusmer NR, Viamonte M, Detrano R: 1990. Quantification of coronary artery calcium using ultrafast computed tomography. J Am Co11Cardiol 15:827-832. Anderson HC: 1983. Calcific diseases: a concept. Arch Path01 Lab Med 107:341-348. Beadenkopf WG, Daoud AS, Love BM: 1964. Calcification in the coronary arteries and its relationship to arteriosclerosis and myocardial infarction. Am J Roentgen01 92:865871.
01994, Elsevier Science Inc., lOSO-1738/94/$7.00
Diaz-mores L, Gutierrez R, Varela H, Rancel N, Valladares F: 1o%l . Microvascular pericytes: a review of their morphological and functional characteristics. Histol Histopathol6:269-286. Fleckenstein A, Fleckenstein-Griin G, Frey M, Thimm F: 1990. Experimental antiarteriosclerotic effects of calcium antagonists. J Clin Pharmacol30: 15 l-l 54. Gadeau A-P, Campan M, Millet D, Candresse T, Desgranges C: 1993. Osteopontin overexpression is associated with arterial smooth muscle cell proliferation in vitro. Arterioscler Thtomb 13:120-125. Giachelli C, Bae N, Lombardi D, Majesky M, Schwartz S: 1991. Molecular cloning and characterization of 2B7, a rat mRNA which distinguishes smooth muscle cell phenotypes in vitro and is identical to osteopontin (secreted phosphoprotein I, 2aR). Biochem Biophys Res Commun 177:867-873. Gijsbers BLMG, van Haarlem JM, Soute BAM, Ebbetink RHM, Vermeer C: 1990. Characterization of a Gla-containing protein from calcified human atherosclerotic plaques. Arteriosclerosis 10:991-995. Glagov S, Weisenberg E, Zarins CK, Stankunavicius R, Kolettis GJ: 1987. Compensatory enlargement of human atherosclerotic coronary arteries. N Engl J Med 3 16: 137 l-l 375. Hirsch D, Azoury R, Sarig S, Kruth HS: 1993. Colocalization of cholesterol and hydroxyapatite in human atherosclerotic lesions. Calcif Tissue Int 52:94-98. Honye J, Mahon DJ, Jain A, et al.: 1992. Morphological effects of coronary balloon angioplasty in vivo assessed by intravascular ultrasound imaging. Circulation 85: 10 121025. Ingram RT, Fitzpatrick LA, Edwards WD, Frye RL, Fisher LW, Schwartz RS: 1993. Calcification in human coronary atherosclerosis is specifically associated with osteopontin, a bone matrix protein [abst]. J Am Co11Cardiol21(Suppl):363A.
TCM Vol.4,No. 1,1994
Ito
M, Cho HS, Kummerow FA: 1990. Effects of a dietary magnesium deficiency and excess vitamin D3 on swine coronary arteries. J Am Co11Nutr 9:155-163.
Khouri RK, Koudsi B, Reddi H: 1991. Tissue transformation into bone in vivo: a potential practical application. JAMA 266: 19531955. Koyama N, Ohara K, Yokota H, et al.: 1991. A one-step sandwich enzyme immunoassay for gamma-carboxylated osteocalcin using monoclonal antibodies. J Immunol Methods 139:17-23. Lansing AI, Blumenthal HT, Gray SH: 1948. Aging and the calcification of the human coronary artery. J Gerontology 3:87-97. Levy RJ, Lian JB, Gallop PM: 1979. Atherocalcin, a gamma-carboxyglutamic acidcontaining protein from atherosclerotic plaque. Biochem Biophys Res Commun 91:4149. Marcum JA, Thompson MA: 1991. The aminoterminal region of a proteochondroitin core protein, secreted by aortic smooth muscle cells, shares sequence homology with the pre-propeptide region of the biglycan core protein from human bone. Biochem Biophys Res Commun 175:706-712. Mevrick B. Fuiiwara K. Reid L: 1981. Smooth muscle myosin in precursor and mature smooth muscle cells in normal pulmonary
arteries and the effect of hypoxia. Exp Lung Res 2:303-313.
retinal microvasculature undergo calcification in vitro. J Cell Sci 97:449-461.
Nayak RC, Attawia MA, Cahill CJ, King GL, Ohashi H, Moromisato R: 1992. Expression of a monoclonal antibody (3G5) defined ganglioside antigen in the renal cortex. Kidney Int 41:1638-1645.
Sharma SK, Israel DH, Kamean JL, Bodian CA, Ambrose JA: 1993. Clinical, angiographic, and procedural determinants of major and minor coronary dissection during angioplasty. Am Heart J 126:39-47.
Niskanen LK, Suhonen M, Siitonen 0, Lehtinen JM, Uusitupa MIJ: 1990. Aortic and lower limb artery calcification in type 2 (non-insulin-dependent) diabetic patients and non-diabetic control subjects: a five year follow-up study. Atherosclerosis 84:6171.
Simons DB, Schwartz RS, Edwards WD, Sheedy PF, Breen JF, Rumberger JA: 1992. Noninvasive definition of anatomic coronary artery disease by ultrafast computed tomographic scanning: a quantitative pathologic comparison study. J Am Coll Cardio120:1118-1126.
Noda M, Camilliere JJ: 1989. In vivo stimulation of bone formation by transfotmation growth factor-beta. Endocrinology 124:29912994.
Takeo S, Anan M, Fujioka K, et al.: 1989. Functional changes of aorta with massive accumulation of calcium. Atherosclerosis 77:175-181.
Rhodin JAG: 1968. Ultrastructure of mammalian venous capillaries, venules, and small collecting veins. J Ultrastruct Res 25:452-500.
Urist MR: 1965. Bone: formation duction. Science 150:893-899.
Rosen V, Thies RS: 1992. The BMP proteins in bone formation and repair. Trends Genet 8:97-102.
Virchow R: 1863. Cellular Pathology: As Based upon Physiological and Pathological Histology [trans Frank Chance, 1971; an unabridged and unaltered republication of the English translation originally published by Dover]. New York, Dover, pp 404-408.
Schmid K, McSharry WO, Pameijer CH, Binette JP: 1980. Chemical and physicochemical studies on the mineral deposits of the human atherosclerotic aorta. Atherosclerosis 37: 199-2 10. Schor AM, Allen TD, Canfield AE, Sloan P, Schor SL: 1990. Pericytes derived from the
by autoin-
Urist MR. Strates BS: 1971. Bone morphogenetic protein. J Dent Res 50:1392-1406.
Wozney JM, Rosen V, Celeste AI, et al.: 1988. Novel regulators of bone formation: molecular clones and activities. Science 242: 1528 TCM 1534.
Why wait for it to circulate? Read your own copy of
If you share TCM with others at work, you know what goes around doesn’t always come around. And if it does, articles are often dated or sometimes clipped. Why miss a single issue, or wait weeks to read one? By starting your personal subscription, for just $79 you’ll be guaranteed to read each and every issue. TRENDS IN CARDIOVASCULAR MEDICINE’s short reviews on ‘hot” topics can make a difference in your research and practice. Commissioned researchers and clinicians report on advances in cardiology and its related areas. Use the business reply cards in this issue to order your personal subscription today-or call Elsevier Customer Service at
(212)633-3950 or FAX (212)633-3880 TCM Vol. 4,No.1,1994
01994, Elsevier Science Inc., 1050-1738/94/$7.00
49