Smooth Muscle Cell Phenotypic Transition Associated With Calcification
Upregulation of Cbfa1 and Downregulation of Smooth Muscle Lineage Markers
Bovine aortic smooth muscle cell (BASMC) cultures undergo mineralization on addition of the organic phosphate donor, β-glycerophosphate (βGP). Mineralization is characterized by apatite deposition on collagen fibrils and the presence of matrix vesicles, as has been described in calcified vascular lesions in vivo as well as in bone and teeth. In the present study, we used this model to investigate the molecular mechanisms driving vascular calcification. We found that BASMCs lost their lineage markers, SM22α and smooth muscle α-actin, within 10 days of being placed under calcifying conditions. Conversely, the cells gained an osteogenic phenotype as indicated by an increase in expression and DNA-binding activity of the transcription factor, core binding factor α1 (Cbfa1). Moreover, genes containing the Cbfa1 binding site, OSE2, including osteopontin, osteocalcin, and alkaline phosphatase were elevated. The relevance of these in vitro findings to vascular calcification in vivo was further studied in matrix GLA protein null (MGP−/−) mice whose arteries spontaneously calcify. We found that arterial calcification was associated with a similar loss in smooth muscle markers and a gain of osteopontin and Cbfa1 expression. These data demonstrate a novel association of vascular calcification with smooth muscle cell phenotypic transition, in which several osteogenic proteins including osteopontin, osteocalcin, and the bone determining factor Cbfa1 are gained. The findings suggest a positive role for SMCs in promoting vascular calcification.
Mineralization of bones and teeth is an exquisitely regulated, cell-mediated process in which the tissue extracellular matrix is embedded with crystalline calcium phosphate deposits. This process gives rise to hard tissues endowed with the mechanical properties required to withstand their normal physiological functions. In contrast, mineralization of soft tissues occurs under pathological conditions with detrimental consequences, particularly when present in blood vessels and heart valves. Calcification of arterial plaques decreases vessel elasticity, augments plaque brittleness, leads to increased plaque rupture during angioplasty procedures,1,2 and is associated with increased risk of myocardial infarction and death.3,4,5
Despite its clinical significance, the molecular mechanisms regulating vascular calcification are unclear. Historically, ectopic calcification has been considered a passive process involving spontaneous calcium phosphate mineral precipitation in necrotic tissue. However, several lines of evidence have recently emerged supporting the concept that ectopic calcification, like mineralization of bones and teeth, is a cell-regulated process. Similarities between hard and soft tissue mineralization include the presence of an apatitic mineral phase, matrix vesicles,6 and noncollagenous bone proteins including osteopontin,7 bone acidic glycoprotein 75 (BAG 75),8 osteocalcin,9 osteonectin,10 and bone morphogenetic protein type 2.11 Furthermore, studies of mutant mice, including matrix GLA protein null (MGP−/−) mouse,12 klotho mouse,13 carbonic anhydrase II deficient mouse,14 and osteoprotegerin null mouse,15 have identified genes whose loss of function increases susceptibility to vascular calcification. Finally, outright ossification has been occasionally noted in extensively calcified vascular tissues.16 Taken together, these findings have greatly strengthened the theory that vascular calcification, like osteogenesis, is a delicately regulated balance between inducers and inhibitors.
The importance of vascular calcification in human disease has fueled interest in identifying both positive and negative mediators of vascular calcification. In this study, we have investigated the properties of arterial smooth muscle cells under mineralizing conditions. We provide evidence for a dramatic phenotypic transition associated with mineralization of bovine aortic smooth muscle cells (BASMCs) in vitro as well as with arteries of MGP−/− mice in vivo.
Materials and Methods
Primary human osteoblasts (a generous gift from Dr Norman Wolf, University of Washington, Seattle, Wash) were maintained in DMEM containing 4.5 g/L glucose and 15% fetal calf serum (HyClone Laboratories Inc, Logan, Utah). Primary BASMCs were isolated as described previously,17 and cultured in DMEM supplemented with 4.5 g/L glucose, 15% fetal bovine serum, 1 mmol/L sodium pyruvate (Gibco BRL), 100 U/mL penicillin, and 100 mg/mL streptomycin (Gibco BRL). Cells used in the experiments were between passages 3 and 8. Calcification of BASMCs was induced, as described previously,18 by addition of 10 mmol/L β-glycerophosphate (βGP) or 5 mmol/L inorganic phosphate (Pi) to the medium. Extracellular calcium deposition was measured and normalized to cellular protein content.17,18
Protein Lysate Preparation and Immunoblotting
BASMC lysates and nuclear extracts were prepared as described.18,19 Arterial protein extracts were prepared by homogenizing the carotids of MGP wild-type (MGP+/+) and MGP−/− mice in 20 mmol/L Hepes containing 420 mmol/L sodium chloride, 1.5 mmol/L magnesium chloride, 0.2 mmol/L EDTA, 25% glycerol, 0.5 mmol/L dithiothreitol, 0.5 mmol/L phenylmethylsulfonylfluoride, and 10 μg/mL leupeptin, pH 7.9. Protein amounts of cell lysates, nuclear extracts, and arterial extracts were measured by the Micro BCA Assay (Pierce). Equal amounts of the proteins were separated by polyacrylamide gel electrophoresis followed by transfer to a 0.45 μm Nitro-Bind membrane (Micron Separations, Inc). Proteins of interest were then blotted using specific antibody, biotin-streptavidin amplification, and Western blot chemiluminescence detection (NEN). Antibodies used were as follows: mouse monoclonal antibody (mAb) against SM22α (gift from Dr S. Sartore, University of Padua, Italy), OP-199 against osteopontin, HHF35 against SM α-actin (gift from Dr Alan Gown, PhenoPath Labs, Seattle, Wash), mouse mAb against human bone alkaline phosphatase (B4-78, Developmental Studies Hybridoma Bank, University of Iowa), rabbit polyclonal antibody against bovine osteocalcin (LF-32, gift from Dr L.W. Fisher, National Institutes of Health, Bethesda, Md), and antibody against core binding factor α1 (Cbfa1) (generated from peptide MASNSLFSAVTPCQQSFFWDPSTSRR, representing amino acids 1 to 26 of the Cbfa1 open reading; Genosys Biotechnologies, Inc).
Protein Identification by Mass Spectroscopy
Cell lysates from BASMC cultures grown in the presence or absence of βGP were prepared and their protein amounts were quantified as described. Twenty-five micrograms of each lysate was separated by a 10% SDS-polyacrylamide gel. The protein with a molecular weight of approximately 25 kDa was excised from the gel and digested with trypsin. Peptides obtained were separated by a gradient capillary LC system using a C18 column (100 mm ID×4 cm). The capillary LC system eluate was directed to a Finnigan MAT TSQ 7000 triple quadrupole mass spectrometer (Finnigan Corporation) using a microelectrospray ionization source constructed in the laboratory. The spray voltage was +1.3 kV and the heated capillary set at 200°C. Collision-induced dissociation (CID) of selected peptide ions was achieved using 3.5 mTorr of Argon in the second quadrupole. Spectra were scanned over the range 400 to 2000 mass units at 1.6-second intervals. Automated peak recognition and daughter ion scanning were performed using the built-in Instrument Control Language (ICL) as described previously.20 Tandem mass spectra were analyzed using the SEQUEST computer program that allows the correlation of observed CID spectra with theoretical spectra generated from known protein sequences.21 The criteria used for a positive peptide identification were a correlation factor (Xcorr) greater than 2.5, a δ cross-correlation factor greater than 0.1 (indicating a significant difference between the best match reported and the next best match), and a high preliminary score. At least 2 positive peptide identifications are considered necessary to unambiguously identify a protein.22 All spectra were searched against the OWL protein database.23
Electrophoretic Gel Mobility Shift Assay
Double-stranded oligonucleotides containing the Cbfa1 DNA binding site (OSE2) from osteopontin promoter (5′-CGCTCTTTGTGC- AAACCACACAG-3′) were synthesized by Gibco BRL and end-labeled with [γ32P] ATP, as directed by Promega. The [γ32P]-labeled OSE2 probe was purified using a Sephadex G-25 column (Pharmacia Biotech, Inc). To detect Cbfa1 binding activity, 6 μg of nuclear extract was incubated at room temperature for 10 minutes with 5 fmol of the labeled OSE2 probe and 133 μg/mL of poly dIdC.dIdC (Sigma). The mixture was then applied onto a 5% polyacrylamide gel to separate the proteins at 100 V for 4 hours. The gels were then dried and the specific protein shifts were visualized by autoradiography. As parallel controls, the nuclear extracts were preincubated for 20 minutes at room temperature with either 50 fmol unlabeled OSE2 probe, nonspecific probe (Osteopontin E-box: 5′-GGAGCAGGTGGCCGGCCGTGG-3′), anti-Cbfa1antibody, or normal rabbit serum (Vector Laboratories, Inc, Burlingame, Calif) prior to addition of the labeled probes.
Reverse Transcriptase–Polymerase Chain Reaction and DNA Sequencing
cDNA reverse transcribed from the RNA of day 10 βGP-treated BASMCs was amplified by Taq DNA polymerase as described by Komori et al,24 using the following Cbfa1 specific primers: 5′CCGCACGACAACCGCACCAT-3′ and 5′-CGCTCCGGC-CCACAAATCTC-3′. The resulting 284 base pair fragment was cloned into pcDNA 3.1 vector and confirmed as the bovine homologue of mouse Cbfa1 using the Dye Terminator DNA cycle sequencing kit (Perkin Elmer).
Total RNA isolated from BASMC cultures grown in either the presence or absence of βGP was transferred to a nylon membrane and hybridized overnight at 42°C with either 32P-labeled bovine Cbfa1 or 32P-labeled bovine osteopontin cDNA. The specific bands were visualized by autoradiography, and the relative band intensities were measured by NIH Image Program.
Tissue Preparation and Histochemistry
MGP−/− and MGP+/+ mice were generated by mating of C57BL/6J MGP heterozygous mice, and genotypes were determined as previously described.12 At the indicated ages, the mice were given a lethal intraperitoneal injection of nembutol (0.3 mg/g mouse), and the carotids and aortas were excised, fixed for 18 to 24 hours in methyl Carnoys fixative (3:1, methanol to acetic acid), and embedded in paraffin. Five-micrometer sections were used for histological analysis. Alizarin Red S (0.5%, pH 9.0; Sigma) and Von Kossa staining were used to detect calcification. For immunohistochemical analysis, sections were first blocked in PBS containing 10% goat serum, 1% bovine serum albumin (BSA), and avidin blocking reagent (Vector Laboratories, Inc). Polyclonal rabbit anti-osteopontin antibody (LF123; gift from Dr L. Fisher) was used to detect osteopontin, mouse monoclonal anti–α-actin antibody (1A4; Sigma) to detect SM α-actin, and polyclonal rabbit anti-SM22α antibody (gift from Dr Michael Parmacek, University of Pennsylvania, Philadelphia, Pa) to detect SM22α.25 After incubation at room temperature for 1 hour, the sections were incubated with biotinylated anti-rabbit or anti-mouse antibodies (Vector Laboratories, Inc) prior to streptavidin-conjugated peroxidase staining. Sections were counter-stained with methyl green (2% methyl green, 2% pyronin Y; Sigma).
Mineralizing BASMCs Lose Smooth Muscle Lineage Markers In Vitro
We previously characterized the morphology and ultrastructure of calcifying BASMC cultures in vitro.17 BASMC cultures grown in the presence of βGP developed widespread matrix mineralization. Ultrastructural analysis confirmed the presence of hydroxyapatite, calcifying collagen, extracellular matrix vesicles, and nodular calcification, similar to that observed in calcified atherosclerotic plaques and heart valves in vivo. Thus, this in vitro model represents a useful tool to further investigate mechanisms of vascular calcification. As shown in Figure 1A, the accumulation of calcium phosphate mineral on BASMC monolayers grown in the presence of βGP was observed by light microscopy within 10 days of culture. The mineral was deposited diffusely throughout the culture and localized to the extracellular matrix. The time course by which BASMC monolayers deposit calcium phosphate mineral in the presence of βGP was also established (Figure 1B). Within 7 days of culture, βGP-treated BASMCs showed significantly more calcium deposition than vehicle-treated controls, and a 5- to 6-fold increase in mineral deposition was further observed by day 10. The effect of βGP on mineral deposition was previously shown to depend on alkaline phosphatase-mediated generation of inorganic phosphate in the culture.17 Consistent with those findings, identical results were obtained when βGP was replaced with 5 mmol/L inorganic phosphate (data not shown).
A comparison of proteins extracted from day 10 cultures showed differences between BASMCs grown under normal or calcifying conditions. One very dramatic difference was the loss of a protein band migrating at an apparent molecular weight of 25 kDa in the βGP-treated extracts (Figure 2A). This protein band, which was strongly expressed in control extracts, was trypsin-digested and analyzed by ESI-tandem mass spectroscopy, and was found to contain 3 peptides that were identified as tryptic fragments of bovine smooth muscle protein SM22α.26 The 3 peptides identified were KYDEELEER, AAEDYGVTK, and EFTESQLQEGK, corresponding to amino acids 1 to 9, 77 to 85, and 133 to 143 of the SM22α, respectively.26,27 A fourth peptide, LVNSLYPDGSKPVK, was also identified and shown to correspond to the known sequence of the chicken SM22α protein (amino acids 66 to 79) but coincided with a region containing sequence ambiguities in the bovine homologue.26,27,28 On the basis of this data, the 25-kDa Coomassie blue–stained band was identified as the bovine homologue of SM22α, a specific marker of smooth muscle cells. The identity of this protein was confirmed by SM22α immunoblotting. As shown in Figure 2B, SM22α expression was strikingly decreased in calcified BASMC cultures by day 7, and was barely detectable by day 10. This observation encouraged us to investigate whether another smooth muscle lineage marker, SM α-actin, was also lost during the mineralization process. Indeed, there was a parallel loss of SM α-actin in mineralizing BASMC cultures (Figure 2B). The time course of SM α-actin and SM22α loss mimicked calcium phosphate mineral deposition in BASMC cultures (Figures 1B and 2⇓B).
Mineralizing BASMCs Develop an Osteogenic Phenotype In Vitro
Because it has been recently shown that vascular calcification in vivo shares similarities with bone mineralization, we investigated whether bone-associated molecules were present during mineralization of BASMCs in vitro. A temporal analysis of alkaline phosphatase and osteocalcin proteins showed expression by day 10 in calcifying cultures, whereas neither protein was detected at any time point tested in the nontreated controls (Figure 3A). Consistent with reports by Shioi et al,18 osteopontin transcripts were dramatically increased in βGP-treated cultures and were conversely lost in the control cultures by day 10 (Figure 3B). By Western blot analysis, no collagen type II expression (indicative of the chondrocyte lineage) was observed at any time point tested (data not shown).
Because osteocalcin and osteopontin are regulated by the bone determining transcription factor Cbfa1,29 we examined the expression and activity of this factor in mineralizing versus nonmineralizing BASMCs. Using runt homology domain oligonucleotide primers targeted to the conserved DNA-binding domain and RNA extracted from day 10 calcifying BASMCs, a 284 base pair band was amplified by reverse transcriptase–polymerase chain reaction (RT-PCR) (Figure 4A). This band was subcloned into the pcDNA3.1 sequencing vector and, by sequence analysis, was found to be 97% and 93% identical to the runt domain of human and mouse Cbfa1 genes, respectively. Thus, the insert was excised, 32P-labeled, and used as a probe for temporal analysis of Cbfa1 transcripts in calcifying and noncalcifying BASMCs. As shown in Figures 4B and 4C, Cbfa1 transcripts were expressed by day 3 in calcifying cultures with a 5- to 6-fold increase over control cultures by day 10. The presence of this factor was also confirmed at the protein level using a human Cbfa1-specific antibody (gift from Dr S. Hiebert, Mt. Sinai School of Medicine, New York, NY, data not shown).
In addition to Cbfa1 mRNA and protein expression, Cbfa1 DNA binding activity was examined by electrophoretic gel mobility shift assay. Using an OSE2-containing oligonucleotide from the human osteopontin gene promoter, we found a specific increase in Cbfa1 OSE2 DNA binding activity under calcifying conditions. The activity of this factor paralleled the appearance of BASMC culture mineralization with elevated OSE-2 binding activity observed by day 7 and highest at day 10 in βGP-treated cells (Figure 5A; arrow). The specificity of this interaction was demonstrated by the potent ability of unlabeled OSE2 (S) and the inability of a nonspecific upstream osteopontin E-box–containing oligonucleotide (NS) to compete for binding of this factor to labeled OSE2 (Figure 5B; arrow). To definitively identify the factor as Cbfa1, the OSE2-protein complex was super-shifted with a Cbfa1-specific antibody (Figure 5B; high molecular weight smear above Cbfa1-OSE2 band in lane Cb). Antibodies directed against the 2 other Cbfa family members, Cbfa2 and Cbfa3, did not result in a super-shift of the OSE2-binding complex (data not shown).
Characterization of Smooth Muscle and Osteogenic Markers in Vascular Calcification In Vivo
To determine whether the loss of the smooth muscle lineage markers and the gain of osteogenic markers observed in vitro also occurred in vivo, we examined arteries from MGP−/− mice that develop spontaneous arterial calcification within 2 weeks of birth.12 At various ages carotids and aortas were collected and examined for calcium deposition, SM α-actin, SM22α, and osteopontin expression. At one month of age, MGP−/− arteries were extensively calcified as demonstrated by positively stained arterial media with Alizarin Red S and Von Kossa (Figure 6A, panels b and j). This calcification was concomitant with a loss of SM α-actin and SM22α expression and a gain of osteopontin expression (Figure 6A, panels d, f, and h). In addition, osteopontin was found deposited extracellularly and appeared to coat the mineral as well as colocalize to the cells of the calcified medial layer (Figure 6A, panels h and i). Interestingly, BM8 staining of the vessel sections indicated that the osteopontin expressing media cells were not macrophages (data not shown).
Conversely, arteries examined from MGP+/+ littermates had no Alizarin Red S or osteopontin staining in the media, but strong SM α-actin and SM22α staining in normal looking smooth muscle as expected (Figure 6A, panels a, g, c, and e). The loss of SM α-actin and SM22α expression in MGP−/− mice is not believed to be a consequence of cellular deficiency because counterstaining of the tissue sections with methyl green positively identified numerous nuclei in the calcified region (Figure 6A, panels i and j). Furthermore, partially calcified arteries from 2-week-old MGP−/− mice retained SM α-actin– and SM22α-expressing medial cells in the areas that had not yet undergone mineralization, indicating that the MGP−/− mice do not have a deficit in SM marker expression or smooth muscle cell differentiation (Figure 6B). Similar to 1-month MGP−/− mice (Figure 6A, panels b and h), osteopontin was found to colocalize with mineral and the cells of calcified area. (data not shown). Finally, no collagen II staining or areas resembling cartilaginous metaplasia were observed in arteries of either 2-week or 1-month-old MGP−/− mice (data not shown).
The dramatic protein expression changes seen in vivo were further characterized by immunoblotting. Protein lysates were prepared from carotids of 4-week-old MGP−/− or MGP+/+ mice, and analyzed for SM α-actin and osteopontin expression. As shown in Figure 6C, loss of SM α-actin and gain of osteopontin expression mirrored and thereby confirmed the histological findings. Because antibodies to Cbfa1 for use in immunohistochemical studies are not yet available, we examined expression of this protein by Western blot analysis. We found that Cbfa1 protein expression was greatly enhanced in lysates prepared from calcified carotids of MGP−/− mice and not detectable in those of MGP+/+ (Figure 6C). These in vivo observations further support the concept resulting from the in vitro data that arterial calcification is associated with a dramatic alteration in smooth muscle cell phenotype, and this alteration may be involved in positively regulating vascular calcification via expression of osteogenic proteins.
We have identified a smooth muscle cell phenotypic transition associated with vascular calcification in vitro and in vivo. Using a BASMC culture model, we found that smooth muscle lineage markers SM22α and SM α-actin were dramatically reduced on mineralization. Conversely, genes normally restricted to mineralized tissue such as bones and teeth, including alkaline phosphatase, osteocalcin, and osteopontin, were increased in these cells. The most interesting finding supporting a phenotype transition was the observation that mRNA and binding activity of Cbfa1, a transcription factor essential for bone morphogenesis, was elevated on smooth muscle cell calcification. The relevance of this smooth muscle phenotype transition was tested in vivo using MGP−/− mice that spontaneously develop calcification in all major arteries. The calcified arteries of MGP−/− mice showed a substantial loss of SM α-actin and SM22α in the calcified medial layer and an upregulation of osteopontin. Finally, Cbfa1 protein expression was induced in the calcified MGP−/− arteries but undetected in noncalcified wild type arteries. The combination of smooth muscle marker loss, osteogenic molecule gain, and expression and activity of the transcription factor Cbfa1 characterizes a distinct smooth muscle cell phenotype in calcified vasculature.
Our studies are in agreement with those of Shioi et al,18 who described increases in osteopontin transcripts and reduction in SM α-actin protein levels using a βGP-induced BASMC calcification model. In the same model, levels of MGP mRNA were decreased during BASMC calcification and were inversely correlated to calcification.30 An increase in osteopontin, alkaline phosphatase,31 and osteocalcin32 expression was also observed in spontaneously calcifying vascular cells derived from bovine aortic media (CVC) when compared with nonmineralizing BASMCs. In addition, Cbfa1 transcripts were present in CVC but were unchanged with mineralization as measured by RT-PCR analysis;32 however, neither Cbfa1 DNA binding activity nor identification of the specific Cbfa1 splice variant was determined, making comparison to the present study difficult. CVC also expressed SM α-actin prior to mineralization, but the effect of mineral deposition on levels of this smooth muscle marker were not investigated.
In contrast, Proudfoot et al33 described an increase in smooth muscle cell markers and low osteopontin expression associated with mineralization of cultured human aortic vascular smooth muscle cells. In those studies, SM α-actin, SM22α, and calponin mRNA levels as measured by Northern blot analysis were higher in cultures containing calcifying nodules when compared with noncalcifying monolayers. However, the immunocytochemical analysis of SM α-actin shown in that report indicated that cells surrounding the mineralizing nodules, rather than those within the nodules, contained the SM α-actin protein. In fact, the mineralizing nodules themselves were devoid of SM α-actin, similar to our findings. Thus, an alternative interpretation of the data is that cells within the relatively rare mineralizing nodules have dramatically lost their α-actin expression, whereas cells in the surrounding nonmineralized monolayer maintained expression, thus accounting for the Northern blot results observed in that report.
The plasticity of the smooth muscle cell phenotype is well documented.34 Although most cell types undergo terminal differentiation, vascular smooth muscle cells do not. This smooth muscle cell trait contributes to the ability of vascular tissue to adapt to various environmental stimuli including pressure and injury. Two characteristic phenotypes of vascular smooth muscle cells include a fibroblastic, synthetic phenotype and a mature, contractile phenotype. In response to vascular injury and/or disease, smooth muscle cells undergo a transition from the contractile phenotype to the synthetic phenotype. The mineralizing SMC phenotype observed in the present study shares some features with the synthetic phenotype of SMCs, including loss of smooth muscle lineage markers and upregulation of osteopontin.35 However, it also differs in a number of important ways, including lack of enhanced proliferation of SMCs11,17,18,31 and elevation of Cbfa1 levels, a transcription factor that is predominantly restricted to bone and cartilage.36 Taken together, these data argue that we have identified a novel SMC phenotype associated with mineralization.
The finding of Cbfa1 expression in calcifying smooth muscle cells is of particular interest. To date, Cbfa1 functional analyses have been primarily limited to the studies of skeletal ontogeny. Developmental studies have shown that Cbfa1 is expressed in the common precursor to osteoblasts and chondroblasts, the osteochondroprogenitor.29,37 The requirement for Cbfa1 in osteoblast differentiation was substantiated by the finding that mice containing a homozygous mutation in the Cbfa1 gene are completely devoid of bone and mature osteoblasts.36,38 More recent studies also indicate that Cbfa1 is involved in chondrogenesis. In addition to osteoblast defects, Cbfa1-deficient mice have reduced numbers of hypertrophic chondrocytes.37 Moreover, continuous expression of Cbfa1 in nonhypertrophic chondrocytes induces differentiation to hypertrophic chondrocytes in vitro and partially rescues Cbfa1 null mice.38 Thus, Cbfa1 appears to be important in the development of both cartilage and bone.
These findings are relevant to our localization of Cbfa1 to the calcified arteries of MGP−/− mice. We observed arterial mineralization concomitant with increases in Cbfa1 and osteopontin protein levels by Western blotting. However, no morphological changes resembling bone or cartilage were observed at the time points studied (2 and 4 weeks), despite extensive mineralization of the arterial medias. In addition, there was no collagen type II present as determined by immunostaining at these time points. Similarly, no type II collagen was expressed in mineralizing BASMCs in vitro. In contrast, calcified arterial medial layers of older (6 to 8 weeks old) MGP−/− mice were found to acquire chondrocytic features including a chondrocyte-like cell morphology, type II collagen fibrils, and proteoglycans reminiscent of cartilage extracellular matrix.12 These findings suggest that the early SMC phenotypic transition observed in the MGP−/− mouse may allow for the development of cartilaginous structures over time. On the other hand, no evidence of bone formation has thus far been noted in the MGP−/− mouse by ourselves or others.12 The absence of bone formation may be because of the fact that MGP−/− mice die suddenly of arterial rupture between 6 and 8 weeks of age, thereby precluding development of this tissue from cartilage as is seen in endochondral bone formation.
Taken together, our findings suggest that, in response to mineralization, SMCs of the arterial wall undergo phenotypic transition to Cbfa1-positive, osteochondroprogenitor-like cells, similar to those identified in early mesenchymal condensations that give rise to the chondrocytes and osteoblasts of the skeleton. These findings may explain the presence of bone- and chondrocyte-related factors, as well as the occasional presence of ossified tissue, in calcified vascular tissues.
This work was supported by NIH Grant HL40079-6A2 (to C.M. Giachelli), NIH Training Grant GM07037 (to S.A. Steitz), NIH Grant AR45548 (to G. Karsenty), NIH Grant AR43655 (to G. Karsenty), a grant from Eli Lilly and Company (to G. Karsenty), and the National Science Foundation Science and Technology Center for Molecular Biotechnology BIR 9214821 AM04 (to R. Aebersold). Dr Giachelli is an Established Investigator of the American Heart Association.
Original received March 20, 2000; resubmission received September 27, 2001; revised resubmission received October 19, 2001; accepted October 19, 2001.
Fitzpatrick PJ, Ports TA, Yock PG. Contribution of localized calcium deposits to dissection after angioplasty: an observational study using intravascular ultrasound. Circulation. 1992; 86: 64–70.
Rumberger JA, Schwartz RS, Simons DB, Sheedy PF, Edwards WD, Fitzpatrick LA. Relation of coronary calcium determined by electron beam computed tomography and lumen narrowing determined by autopsy. Am J Cardiol. 1994; 73: 1169–1173.
Beadenkopf WG, Daoud AS, Love BM. Calcification in the coronary arteries and its relationship to arteriosclerosis and myocardial infarction. Am J Roentgenol. 1964; 92: 865–871.
Loecker TH, Schwartz RS, Cotta CW, Hickman JR. Fluoroscopic coronary artery calcification and associated coronary disease in asymptomatic young men. J Am Coll Cardiol. 1992; 19: 1167–1172.
Puentes G, Detrano R, Tang W, Wong N, French W, Narahara K, Brundage B, Baksheshi H. Estimation of coronary calcium mass using electron beam computed topography: a promising approach for predicting coronary events? Circulation. 1995; 92 (suppl 1): I-313.Abstract.
Kim KM. Calcification of matrix vesicles in human aortic valve and aortic media. Fed Proc. 1976; 35: 156–162.
Giachelli CM. Ectopic calcification: gathering hard facts about soft tissue mineralization. Am J Pathol. 1999; 154: 671–675.
Gura TA, Wright KL, Veis A, Webb CL. Identification of specific calcium-binding noncollagenous proteins associated with glutaraldehyde-preserved bovine pericardium in the rat subdermal model. J Biomed Mater Res. 1997; 35: 483–495.
Levy RJ, Gundberg C, Scheinman R. The identification of vitamin K-dependent bone protein osteocalcin as one of the γ-carboxyglutamic acid containing proteins present in calcified atherosclerotic plaque and mineralized heart valves. Atherosclerosis. 1983; 46: 49–56.
Hirota S, Imakita M, Kohri K, Ito A, Morii E, Adachi S, Kim H-M, Kitamura Y, Yutani C, Nomura S. Expression of osteopontin messenger RNA by macrophages in atherosclerotic plaques: a possible association with calcification. Am J Pathol. 1993; 143: 1003–1008.
Bostrom K, Watson KE, Horn S, Wortham C, Herman IM, Demer LL. Bone morphogenetic protein expression in human atherosclerotic lesions. J Clin Invest. 1993; 91: 1800–1809.
Luo G, Ducy P, McKee MD, Pinero GJ, Loyer E, Behringer RR, Karsenty G. Spontaneous calcification of arteries and cartilage in mice lacking matrix GLA protein. Nature. 1997; 386: 78–81.
Kuroo M, Matsumura Y, Aizawa H, Kawaguchi H, Suga T, Utsugi T, Ohyama Y, Kurabayashi M, Kaname T, Kume E, Iwasaki H, Iida A, ShirakiIida T, Nishikawa S, Hagai R, Abeshima Y. Mutations of the klotho gene leads to a syndrome resembling aging. Nature. 1997; 390: 45–51.
Spicer SS, Lewis SE, Tashian RE, Schulte BA. Mice carrying a CAR-2 null allele lack carbonic anhydrase II immunohistochemically and show vascular calcification. Am J Pathol. 1989; 134: 947–954.
Bucay N, Sarosi I, Dunstan CR, Morony S, Tarpley J, Capparelli C, Scully S, Tan HL, Xu W, Lacey DL, Boyle WJ, Simonet WS. Osteoprotegerin-deficient mice develop early onset osteoporosis and arterial calcification. Genes Dev. 1998; 12: 1260–1268.
Bostrom K, Watson K, Stanford WP, Demer LL. Atherosclerotic calcification: relation to developmental osteogenesis. Am J Cardiol. 1995; 75: 88B–91B.
Wada T, McKee MD, Steitz S, Giachelli CM. Calcification of vascular smooth muscle cell cultures: inhibition by osteopontin. Circ Res. 1998; 84: 166–178.
Shioi A, Nishizawa Y, Jono S, Koyama H, Hosoi M, Morii H. β-Glycerophosphate accelerates calcification in cultured bovine vascular smooth muscle cells. Arterioscler Thromb Vasc Biol. 1995; 15: 2003–2009.
Lee KA, Bindereif A, Green MR. A small scale procedure for preparation of nuclear extracts that support efficient transcription and pre-mRNA splicing. Gene Anal Tech. 1988; 5: 22–31.
Ducret A, Van Oostveen I, Eng JK, Yates III JR, Aebersold R. High throughput protein characterization by automated reverse-phase chromatography/electrospray tandem mass spectrometry. Protein Sci. 1998; 7: 706–719.
Eng J, McCormack AL, Yates III JR. An approach to correlate tandem mass spectral data of peptides with amino acid sequences in a protein database. J Am Soc Mass Spectrom. 1994; 5: 976–989.
Haynes PA, Fripp N, Aebersold R. Identification of gel separated proteins by liquid chromatography-electrospray tandem mass spectrometry: comparison of methods and their limitations. Electrophoresis. 1998; 19: 939–945.
Bleasby AJ, Akrigg D, Attwood TK. OWL: a nonredundant composite protein sequence database. Nucl Acids Res. 1994; 22: 3574–3577.
Komori T, Yagi H, Nomura A, Yamaguchi A, Sasaki K, Deguchi K, Shimizu Y, Bronson RT, Gao Y-H, Inada M, Sato M, Okamoto R, Kitamura Y, Yoshiki S, Kishimoto T. Targeted disruption of Cbfa1 results in a complete lack of bone formation owing to maturational arrest of osteoblasts. Cell. 1997; 89: 755–764.
Zhang JC, Kim S, Helmke BP, Yu WW, Du KL, Lu MM, Strobeck M, Yu Q, Parmacek MS. Analysis of SM22α-deficient mice reveals unanticipated insights into smooth muscle cell differentiation and function. Mol Cell Biol. 2001; 21: 1336–1344.
Lees-Miller JP, Heeley D, Smillie LB. An abundant and novel protein of 22 kDa (SM22) is widely distributed in smooth muscles. Biochem J. 1987; 244: 705–709.
Kobayashi R, Kubota T, Hidaka H. Purification, characterization, and partial sequence analysis of a new 25-kDa actin binding protein from bovine aorta: a SM22 homologue. Biochem Biophys Res Commun. 1994; 198: 1275–1280.
Pearlstone JR, Weber M, Lees-Miller JP, Carpenter MR, Smillie LR. Amino acid sequence of chicken gizzard smooth muscle SM22α. J Biol Chem. 1991; 130: 297–302.
Ducy P, Zhang R, Geoffroy V, Ridall AL, Karsenty G. Osf2/Cbfa1: a transcriptional activator of osteoblast differentiation. Cell. 1997; 89: 747–754.
Mori K, Shioi A, Jono S, Nishizawa Y, Morii H. Expression of matrix Gla protein (MGP) in an in vitro model of vascular calcification. FEBS Lett. 1998; 433: 19–22.
Watson KE, Bostron K, Ravindranath R, Lam T, Norton B, Demer LL. TGF-β and 25-hydrocholesterol stimulate osteoblast-like vascular cells to calcify. J Clin Invest. 1994; 93: 2105–2113.
Tintut Y, Parhami F, Bostrom K, Jackson SM, Demer LL. cAMP stimulates osteoblast-like differentiation of calcifying vascular cells: potential signaling pathway for vascular calcification. J Biol Chem. 1998; 273: 7547–7553.
Proudfoot D, Skeeper JN, Shanahan CM, Weissberg PL. Calcification of human vascular cells in vitro is correlated with high levels of matrix Gla protein and low levels of osteopontin expression. Arterioscler Thromb Vasc Biol. 1998; 18: 379–388.
Katoh Y, Periasamy M. Growth and differentiation of smooth muscle cells during vascular development. Trends Cardiovasc Med. 1996; 6: 100–106.
Giachelli CM, Liaw L, Murry CE, Schwartz SM, Almeida M. Osteopontin expression in cardiovascular diseases. Ann N Y Acad Sci. 1995; 760: 109–26.
Otto F, Thornell AP, Crompton T, Denzel A, Gilmour KC, Rosewell AR, Stamp GWH. Cbfa1, a candidate gene for cleidocranial dysplasia syndrome is essential for osteoblast differentiation and bone formation. Cell. 1997; 89: 765–771.
Kim IS, Otto F, Zabel B, Mundlos S. Regulation of chondrocyte differentiation by Cbfa1. Mech Dev. 1999; 80: 159–170.
Takeda S, Bonnamy JP, Own MJ, Ducy P, Karsenty G. Continuous expression of Cbfa1 in nonhypertrophic chondrocytes uncovers its ability to induce hypertrophic chondrocyte differentiation and partially rescues Cbfa1-deficient mice. Genes Dev. 2001; 15: 467–481.