Published online before print August 14, 2008, doi: 10.1161/ATVBAHA.108.172387
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© 2008 American Heart Association, Inc.
Cell Biology/Signaling |
Mineral Surface in Calcified Plaque Is Like That of Bone
Further Evidence for Regulated Mineralization
i
From the Departments of Chemistry (M.J.D., T.F., D.G.R., E.R.W.), Medicine (D.P.), and Physiology, Development, & Neuroscience (J.N.S.), University of Cambridge, UK; the Department of Internal Medicine and Cardiology (M.S.), Philipps-University, Marburg, Germany; and the Cardiovascular Division (C.M.S.), Kings College London, UK.
Correspondence to Melinda J. Duer, Department of Chemistry, University of Cambridge, Lensfield Road, Cambridge, CB2 1EW, UK. E-mail mjd13{at}cam.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Objectives— Cell biological studies demonstrate remarkable similarities between mineralization processes in bone and vasculature, but knowledge of the components acting to initiate mineralization in atherosclerosis is limited. The molecular level microenvironment at the organic-inorganic interface holds a record of the mechanisms controlling mineral nucleation. This study was undertaken to compare the poorly understood interface in mineralized plaque with that of bone, which is considerably better characterized.
Methods and Results— Solid state nuclear magnetic resonance (SSNMR) spectroscopy provides powerful tools for studying the organic-inorganic interface in calcium phosphate biominerals. The rotational echo double resonance (REDOR) technique, applied to calcified human plaque, shows that this interface predominantly comprises sugars, most likely glycosaminoglycans (GAGs). In this respect, and in the pattern of secondary effects seen to protein (mainly collagen), calcified plaque strongly resembles bone.
Conclusion— The similarity between biomineral formed under highly controlled (bone) and pathological (plaque) conditions suggests that the control mechanisms are more similar than previously thought, and may be adaptive. It is strong further evidence for regulation of plaque mineralization by osteo/chondrocytic vascular smooth muscle cells.
NMR techniques show that bone and vascular calcifications are very similar with respect to constituents of the mineral/matrix interface, a predominant component being glycosaminoglycans. This implies that regulation of both is very similar, supporting cell biological studies showing pathological calcification of vascular smooth muscle cells to be a regulated process.
Key Words: atherosclerosis • biomineralization • glycosaminoglycans • nuclear magnetic resonance spectroscopy • vascular calcification
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Calcification1 is prevalent in numerous vascular pathologies and is associated with adverse clinical outcomes. Parallels have frequently been drawn between ectopic calcification processes in vessels and normal developmental osteogenesis. In fact vascular calcified material is by many criteria indistinguishable from bone, and its formation involves the participation of many cellular and molecular signaling processes underlying normal osteogenesis.2 These include matrix vesicle release, expression of mineralization-regulating proteins by vascular smooth muscle cells (VSMCs) of the vessel wall, and the association of calcification with lipid and chronic inflammation, a hallmark of atherosclerosis.3 Indeed vascular calcification is recognized as regulated biomineralization instead of merely passive precipitation. However questions are still unanswered concerning the mechanisms leading to the initiation of calcification in the vessel wall particularly as it occurs in different vascular environments (ie, media and intima), which do not mimic that of bone. Moreover, VSMCs can overexpress mineralization inhibitors such as matrix Gla protein (MGP) to effectively block mineralization.4
Aspects of the molecular structure of mineralized tissues5 can be studied by the solid state NMR (SSNMR) technique 13C{31P} rotational echo double resonance (REDOR).6 13C is a stable, nonradioactive, NMR-receptive isotope of carbon occurring naturally at about 1.1% abundance; similarly 31P is a stable NMR-receptive isotope of phosphorus which however is naturally present at 100% abundance. A 13C NMR spectrum records a signal for each chemically distinct carbon site in the sample. In a REDOR experiment a 13C NMR spectrum is recorded with and without a train of radiofrequency pulses applied to the 31P spins before acquisition of the spectrum, forming the so-called nondephased and dephased 13C spectra, respectively. The nondephased 13C spectrum recorded acts as a reference, whereas in the spectrum recorded with the 31P pulse train, signals attributable to 13C spins in close proximity to 31P spins in the material will show a reduction in intensity. Thus, those 13C sites which are spatially close to 31P are immediately identified from the resonance frequency of those signals showing reduction in intensity. By following the reduction in intensity of a given 13C signal as a function of the time for which the 31P pulse train is applied, quantitative determination of the 13C-31P distance is possible. The REDOR effect is only transmitted over interatomic distances of less than 10 Ångstroms (1 nm), so it is a conclusive proof of a true molecular composite material in which individual components are bound in associations mediated by intermolecular interatomic forces (for instance electrostatic and hydrogen bonding) and not merely mixed together. The interface between the organic matrix and calcium phosphate biomineral particles in plaques and bone is particularly amenable, because practically all 13C is in the former and all 31P in the latter. This makes the experiment highly selective for the interface itself. In bone it shows that the molecular constituents of the organic-mineral interface are rich in polysaccharides, most likely glycosaminoglycan (GAG) sugars, rather than proteins.7 Therefore, a comparative analysis of this interface may provide clues to the exact nature of similarities and differences between mineralization processes in the vasculature, and bone, that cannot be revealed by cell biological studies alone.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Calcified Atherosclerotic Plaque
Material was collected with appropriate ethical approval during carotid endarterectomy procedures. The human tissue (n=4) was put into sterile flasks with 15 mL medium (M199 with penicillin and streptomycin) and 1 mL elastase (from a stock solution of 5 mg/mL porcine pancreatic Type IV elastase, Sigma) and 5 mL collagenase (from a stock solution of 9 mg/mL C. hemolyticum Type I collagenase, Sigma) added, shaken for 6 hours in a waterbath at 37°C, centrifuged at 800 r.p.m. for 5 minutes, washed with cold sterile phosphate buffered saline, spun again, and the resultant pellet frozen and stored for analysis. The procedure is necessary as SSNMR is a relatively insensitive technique, so it is highly desirable to remove from the sample as much extraneous material as possible in order to concentrate the sites of interest. To determine whether this pretreatment of the material interfered with the SSNMR measurements, it was compared with untreated tissue that was frozen before grinding. For all SSNMR measurements, thawed material was lightly ground with a pestle and mortar and packed into 4-mm outer diameter zirconia magic angle spinning (MAS) rotors (Bruker).
A bone from an adult horse used for general purpose exercise and euthanized for humanitarian reasons unconnected with this study was used for comparison. Several bone specimens obtained at surgery were found to be identical to those obtained at autopsy, indicating that the difference in harvesting between plaque (surgery) and bone (autopsy) does not introduce noticeable differences. In the cases of both mineralized plaque and bone, the grinding produced particles the order of several tens of micrometers in size; as typical biomineral particles are much smaller than this and the interatomic interactions we are exploring operate over subnanometer length scales, we believe the grinding will not affect our observations of these interactions significantly, if at all. In previous studies on bone, we have explored a variety of ways of preparing the sample material, including slicing/chopping and grating, as well as grinding to different particle sizes and saw no difference in spectra between different preparation methods.
NMR Spectroscopy
This was performed on a Bruker Avance-400 triple channel spectrometer at a polarizing magnetic field strength of 9.4 T (Tesla), with samples packed into 4-mm outer diameter zirconia rotors. These enable the samples to be spun rapidly (at a rotational frequency of 12.5 kHz in these experiments) at an angle of 54.7° to the polarizing magnetic field, the so called "magic angle." This "magic angle spinning" (MAS) is necessary to remove certain magnetic interactions between the NMR active nuclei in the sample which would otherwise broaden the resonances and result in featureless spectra. Where there was insufficient material to fill the rotor, the dead space was packed out with Teflon tape which gives no NMR signal with the data acquisition methods we have used. The following spectral acquisition parameters were used: Resonance frequencies: 1H 400.4 MHz, 13C 100.7 MHz, 31P 162.1 MHz; 1H 
/2 pulse 2.5 µs, mean ramped cross polarization (CP) field strength 70 kHz, CP time 2.5 ms, 1H decoupling field 100 kHz, recycle delay 2 s, MAS rate 12.5 kHz. For the 31P spectra 128 signal accumulations were averaged for satisfactory signal to noise; 13C spectra required averaging of between 20 k and 40 k signal accumulations. This difference reflects the 100% and 1.1% abundance of the 31P and 13C isotopes, respectively. 13C{31P} REDOR was performed with 1H-13C CP, 31P pulses of 8.4 µs, the separation between the center of the 31P pulses pulses being 80 µs and synchronous with the MAS, and a 13C refocusing pulse of 10 µs inserted at the middle of the recoupling 31P pulse train. By convention, the notation "13C{31P}" signifies that it is the signal from 13C magnetic nuclear spins which is observed directly (at 100.7 MHz), whereas the 31P magnetic nuclear spins are caused to interact with them by a suitable combination of nuclear spin energy level excitations applied at 162.1 MHz.
X-Ray Powder Diffraction (XRPD)
This was performed on a Philips X'Pert Pro powder diffractometer equipped with an X'celerator RTMS detector, using Ni-filtered CuK
radiation. Data collection was performed in a 5 to 80° range using samples on a flat plate, with a scanning step size of 0.008°, time per step of 10.8 seconds and scan speed of 0.0985°/s.
Scanning Electron Microscopy (SEM)
Extracted mineral was dusted onto spectroscopically pure double-sided carbon tabs mounted on Cambridge SEM stubs. It was viewed in a FEI Phillips XL30 FEGSEM operated at 5kv.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Figure 1 compares typical 13C SSNMR spectra of calcified plaque, and bone; supplemental Figure I (available online at http://atvb.ahajournals.org) shows data from all samples and clearly shows a high degree of reproducibility. Collagen accounts for most signals.8 Notable exceptions are a broad signal at about 103 ppm, weak but present in all samples, and another at 76 ppm, which tends to be more prominent in mineralized plaque than in bone. These are consistent, respectively, with the anomeric carbons, and some of the other pyranose 6-membered sugar ring carbons, of β-linked polysaccharides like GAGs.9 To highlight the relationship between GAG signals and GAG chemical structure, a chemical representation of the repeat subunit (-glucuronic acid β (1–3) N-acetylgalactosamine-4-sulfate β (1–4)-) of the widely occurring GAG chondroitin-4-sulfate is included. The signal attributable to most of the GAG sugar ring carbon atoms, the signal attributable to the GAG sugar anomeric carbons (marked * in the chemical structural formula and the spectra), and the signal attributable to the GAG acidic carboxylate carbons (marked # in the chemical structural formula and spectra) are identified. The latter also contains signal from protein acidic carboxylate carbons; it presents a prominent shoulder on the high frequency edge of the signal envelope cantred at about 175 ppm, itself attributable to overlap of numerous protein (and GAG) carbonyl signals. Spectra of treated and untreated human carotid endarterectomy material are shown in supplemental Figure II. This clearly shows that collagenase/elastase pretreatment did not affect the GAG composition or structure in the material.
|
Figure 2 shows X-Ray powder diffraction (XRPD) and 31P SSNMR data proving the mineral is essentially hydroxyapatite- and bone-like.10 However the XRPD trace of microcrystalline hydroxyapatite contains numerous sharp reflections consistent with a highly ordered crystalline environment; in consequence of this order the 31P NMR spectrum is also narrow, showing that all phosphorus atoms experience very similar environments. The XRPD trace and 31P spectrum of bone show much broader features consistent with less perfectly crystalline mineral, and greater environmental heterogeneity attributable to molecular disorder and mineral surface effects. The XRPD trace and 31P NMR spectrum of calcified plaque is intermediate between these extremes, probably reflecting greater mineral crystallinity and larger average mineral crystallite size. This is borne out by SEM illustrating the typical heterogeneous micro/nanocrystalline nature of the calcifications.
|
Figure 3 presents typical 13C{31P} REDOR results reflecting the mineral-organic interface composition; more comprehensive data on other samples is presented in supplemental Figure III. The black trace is the reference spectrum, which in this case is largely due to the various components of collagen. As described in the Introduction, this spectrum contains signals from all 13C sites in the sample. The dephased spectrum in red is obtained after the application of radiofrequency pulses to the 31P nuclei in the sample. As already explained, the effect of these pulses is to reduce in intensity any signal due to a 13C nucleus close in space to a 31P; close in space in this context is less than 10 Å. As in bone, the most affected signals, which can therefore be associated with molecules in intimate association with mineral, are the 76 ppm polysaccharide peak, and signals at 175 and 181 ppm which are probably from amide carbonyl, and carboxylate, groups, which are abundant in GAG polysaccharides (as well as in some proteins, of course). The intensity reduction should increase with increasing time for which the radiofrequency pulses are applied to the 31P nuclei; this is illustrated in supplemental Figure IV, and compared with that of bone in supplemental Figure V. Also in common with bone, other signals suffer a smaller amount of intensity loss; some of these are from collagen but others may be from exocyclic 6-carbons (ca. 65 to 70 ppm), aminated or amidated ring carbons (ca. 54 ppm), or N-acetyl methyl groups (ca. 25 ppm) of GAGs and other polysaccharides. Thus, the macromolecules most strongly associated with the mineral phase in calcified plaque are polysaccharides, probably GAGs. We have also examined samples histologically after staining with von Kossa stain, specific for calcified deposits, and alcian blue, which recognizes polyanions; the Inset shows typical data and more appears in the supplemental material. We find that calcification (dark brown) always occurs in alcian blue staining regions which are probably GAG-rich (blue), although not all alcian blue staining regions calcify. Moreover, GAG-rich calcified regions were often found surrounding lipid pools—areas previously described as colocalizing with mineralization.11 This suggests that calcification tends to occur in polyanion-rich regions, where the polyanions are most likely GAGs as these are the predominant polyanions present in these tissues. Coincidence of GAG and calcification at the micron resolution length scale of light microscopy is not synonymous with the subnanometer atomic length scale intermolecular bonding revealed by the NMR, of course, though it is consistent with the hypothesis this article presents.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We have shown that the matrix-mineral atomic interface in calcified atherosclerotic plaque is very similar to that in bone and, surprisingly, is characterized by a predominance of GAGs. Colocalization of GAGs and mineralization has been detected histologically in calcifying tissue,12–14 and importantly there is accumulating evidence that the type of GAG in the matrix may exert tight control over mineralization, particularly during its initiation.15,16 An early event in bone mineralization is release of membrane-bound matrix vesicles which act as the nidus for mineral nucleation by concentrating calcium and phosphate. Matrix vesicles bind GAGs, and this binding is increased in mineralization-competent hypertrophic chondrocyte-derived matrix vesicles, further suggesting that GAGs play a direct role in initiating physiological mineralization.17 However, as mineralization progresses, GAGs may also act to prevent uncontrolled propagation of calcium phosphate crystallization. Further evidence for this comes from human disease where dysregulated expression of GAGs leads to bone defects, whereas a lower incidence of plaque calcification in chronic asthma sufferers, who show elevated levels of circulating GAGs like heparin, has been reported.18
Atherosclerotic lesions contain proteoglycan, intermixed with loosely scattered collagen fibrils;19 as plaques develop secretion and accumulation of GAGs occurs within the extracellular matrix. Also, GAGs bind low-density lipoprotein,20 retaining cholesterol in the plaque and enhancing local proinflammatory signals. Lipid and mineralization are coincident in plaques,4,11 and inflammatory signals at lipid accumulation sites initiate calcification by inducing vascular smooth muscle cell (VSMC) death and vesicle release.21–24 Vesicles contain hydroxyapatitic nanocrystals representing the initial calcification seed. Thereafter, VSMCs that have undergone osteo/chondrocytic conversion are found abutting the microcalcifications.21 Taken together with this study, showing the remarkable similarities between the resultant biomineral found in plaques, and bone, it seems reasonable to postulate that GAG binding to lipid may initiate VSMC injury and death. This leads to an initial calcification nidus, followed by osteogenic phenotypic conversion of VSMCs, although the mineralization process may also trigger GAG production.
Indeed, the similarities between bone and carotid arterial mineralization which we show support the contention that this may be protective, by rapidly dampening unregulated deposition of mineral initiated by injury. Part of this response may be to produce biomineral that is inert, rather than proinflammatory such as in synovial fluid.25 It is plausible that this is a common feature of all vascular calcification but this requires more evidence to establish.
As in bone, GAGs have attracted less attention in plaque than proteins,26 such as matrix Gla protein and osteocalcin,27 decorin,28 osteopontin,29 bone sialoprotein,4 and collagen itself,30 which have all been proposed as important mineralization modulators. Nonetheless, many of these are often heavily glycosylated, and it may be part of their role to target GAGs to specific sites in developing, and mineralizing, matrices. However, it is particularly interesting that these proteins were not more abundant in atherosclerotic mineral and that other vascular specific extracellular matrix proteins such as elastin, often postulated to be a nucleator of hydroxapatite in the vessel wall, were not present in close association with mineral. This further supports the notion that some vessel wall biomineralization may be adaptive, and this has implications for its treatment.
|
Acknowledgments |
|---|
We thank Nikki Figg for histological preparations.
Sources of Funding
This work was supported by UK BBSRC & EPSRC and the British Heart Foundation.
Disclosures
None.
|
Footnotes |
|---|
Original received April 21, 2008; final version accepted August 4, 2008.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1. Abedin M, Tintut Y, Demer LL. Vascular calcification: mechanisms and clinical ramifications. Arterioscler Thromb Vasc Biol. 2004; 24: 1161–1170.
2. Bostrom K, Watson KE, Stanford WP, Demer LL. Atherosclerotic calcification: relation to developmental osteogenesis. Am J Cardiol. 1995; 75: 88B–91B.[CrossRef][Medline] [Order article via Infotrieve]
3. Shanahan CM. Inflammation ushers in calcification: a cycle of damage and protection? Circulation. 2007; 116: 2782–2785.
4. Shanahan CM, Cary NR, Salisbury JR, Proudfoot D, Weissberg PL, Edmonds ME. Medial localization of mineralization-regulating proteins in association with Monckeberg’s sclerosis: evidence for smooth muscle cell-mediated vascular calcification. Circulation. 1999; 100: 2168–2176.
5. Jaeger C, Groom NS, Bowe EA, Horner A, Davies ME, Murray RC, Duer MJ. Investigation of the nature of the protein-mineral interface in bone by solid-state NMR. Chem Mater. 2005; 17: 3059–3061.[CrossRef]
6. Mueller KT. Analytic solutions for the time evolution of dipolar-dephasing NMR signals. J Magn Reson Ser A. 1995; 113: 81–93.[CrossRef]
7. Wise ER, Maltsev S, Davies ME, Duer MJ, Jaeger C, Loveridge N, Murray RC, Reid DG. The organic-mineral interface in bone is predominantly polysaccharide. Chem Mater. 2007; 19: 5055–5057.[CrossRef]
8. Saito H, Yokoi M. A C-13 Nmr-Study on Collagens in the Solid-State - Hydration Dehydration-Induced Conformational Change of Collagen and Detection of Internal Motions. J Biochem. 1992; 111: 376–382.
9. Mucci A, Schenetti L, Volpi N. H-1 and C-13 nuclear magnetic resonance identification and characterization of components of chondroitin sulfates of various origin. Carbohydrate Polymers. 2000; 41: 37–45.[CrossRef]
10. Guo W, Morrisett JD, Lawrie GM, DeBakey ME, Hamilton JA. Identification of different lipid phases and calcium phosphate deposits in human carotid artery plaques by MAS NMR spectroscopy. Magn Reson Med. 1998; 39: 184–189.[Medline] [Order article via Infotrieve]
11. Shanahan CM, Cary NR, Metcalfe JC, Weissberg PL. High expression of genes for calcification-regulating proteins in human atherosclerotic plaques. J Clin Invest. 1994; 93: 2393–2402.[Medline] [Order article via Infotrieve]
12. Groisman GM, Amar M, Polak-Charcon S. Microcalcifications in the anterior pituitary gland of the fetus and the newborn: a histochemical and immunohistochemical study. Hum Pathol. 1999; 30: 199–202.[CrossRef][Medline] [Order article via Infotrieve]
13. Kagayama M, Sasano Y, Mizoguchi I, Kamo N, Takahashi I, Mitani H. Localization of glycosaminoglycans in periodontal ligament during physiological and experimental tooth movement. J Periodontal Res. 1996; 31: 229–234.[CrossRef][Medline] [Order article via Infotrieve]
14. Mizoguchi I, Takahashi I, Sasano Y, Kagayama M, Mitani H. Localization of types I, II and III collagen and glycosaminoglycans in the mandibular condyle of growing monkeys: an immunohistochemical study. Anat Embryol (Berl). 1997; 195: 127–135.[CrossRef][Medline] [Order article via Infotrieve]
15. Takahashi I, Mizoguchi I, Sasano Y, Saitoh S, Ishida M, Kagayama M, Mitani H. Age-related changes in the localization of glycosaminoglycans in condylar cartilage of the mandible in rats. Anat Embryol (Berl). 1996; 194: 489–500.[Medline] [Order article via Infotrieve]
16. Muller B, Prante C, Gastens M, Kuhn J, Kleesiek K, Gotting C. Increased levels of xylosyltransferase I correlate with the mineralization of the extracellular matrix during osteogenic differentiation of mesenchymal stem cells. Matrix Biol. 2008; 27: 139–149.[CrossRef][Medline] [Order article via Infotrieve]
17. Wu LN, Genge BR, Wuthier RE. Association between proteoglycans and matrix vesicles in the extracellular matrix of growth plate cartilage. J Biol Chem. 1991; 266: 1187–1194.
18. Lasser EC, Berry C, Kortman K. Diminished atherosclerotic arterial calcifications in asthma. A possible role for elevated endogenous heparin-like material. Allergy. 1987; 42: 549–552.[Medline] [Order article via Infotrieve]
19. Ross R. Atherosclerosis-an inflammatory disease. N Engl J Med. 1999; 340: 115–126.
20. Srinivasan SR, Vijayagopal P, Dalferes ER Jr, Abbate B, Radhakrishnamurthy B, Berenson GS. Dynamics of lipoprotein-glycosaminoglycan interactions in the atherosclerotic rabbit aorta in vivo. Biochim Biophys Acta. 1984; 793: 157–168.[Medline] [Order article via Infotrieve]
21. Aikawa E, Nahrendorf M, Figueiredo JL, Swirski FK, Shtatland T, Kohler RH, Jaffer FA, Aikawa M, Weissleder R. Osteogenesis associates with inflammation in early-stage atherosclerosis evaluated by molecular imaging in vivo. Circulation. 2007; 116: 2841–2850.
22. Hsu HH, Camacho NP, Sun F, Tawfik O, Aono H. Isolation of calcifiable vesicles from aortas of rabbits fed with high cholesterol diets. Atherosclerosis. 2000; 153: 337–348.[CrossRef][Medline] [Order article via Infotrieve]
23. Okura Y, Brink M, Itabe H, Scheidegger KJ, Kalangos A, Delafontaine P. Oxidized low-density lipoprotein is associated with apoptosis of vascular smooth muscle cells in human atherosclerotic plaques. Circulation. 2000; 102: 2680–2686.
24. Yin J, Chaufour X, McLachlan C, McGuire M, White G, King N, Hambly B. Apoptosis of vascular smooth muscle cells induced by cholesterol and its oxides in vitro and in vivo. Atherosclerosis. 2000; 148: 365–374.[CrossRef][Medline] [Order article via Infotrieve]
25. Morgan MP, McCarthy GM. Signaling mechanisms involved in crystal-induced tissue damage. Curr Opin Rheumatol. 2002; 14: 292–297.[CrossRef][Medline] [Order article via Infotrieve]
26. Aigner T, Neureiter D, Campean V, Soder S, Amann K. Expression of cartilage-specific markers in calcified and non-calcified atherosclerotic lesions. Atherosclerosis. 2008; 196: 37–41.[CrossRef][Medline] [Order article via Infotrieve]
27. Jono S, Shioi A, Ikari Y, Nishizawa Y. Vascular calcification in chronic kidney disease. J Bone Miner Metab. 2006; 24: 176–181.[CrossRef][Medline] [Order article via Infotrieve]
28. Fischer JW, Steitz SA, Johnson PY, Burke A, Kolodgie F, Virmani R, Giachelli C, Wight TN. Decorin promotes aortic smooth muscle cell calcification and colocalizes to calcified regions in human atherosclerotic lesions. Arterioscler Thromb Vasc Biol. 2004; 24: 2391–2396.
29. Christensen B, Nielsen MS, Haselmann KF, Petersen TE, Sorensen ES. Post-translationally modified residues of native human osteopontin are located in clusters: identification of 36 phosphorylation and five O-glycosylation sites and their biological implications. Biochem J. 2005; 390: 285–292.[CrossRef][Medline] [Order article via Infotrieve]
30. Batge B, Winter C, Notbohm H, Acil Y, Brinckmann J, Muller PK. Glycosylation of human bone collagen I in relation to lysylhydroxylation and fibril diameter. J Biochem (Tokyo). 1997; 122: 109–115.
Related Article:
Arterioscler Thromb Vasc Biol 2008 28: 1882-1884.
This article has been cited by other articles:
 |
|
 |
|
  L. L. Demer, A. P. Sage, and Y. Tintut Nanoscale Architecture in Atherosclerotic Calcification Arterioscler Thromb Vasc Biol, November 1, 2008; 28(11): 1882 - 1884. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||