Renal Osteodystrophy: -Heremans Schmid Glycoprotein/Fetuin-A, Matrix GLA Protein Serum Levels, and Bone Histomorphometry

Transcription

1 Renal Osteodystrophy: -Heremans Schmid Glycoprotein/Fetuin-A, Matrix GLA Protein Serum Levels, and Bone Histomorphometry Giorgio Coen, MD, Paola Ballanti, PhD, Alessandro Balducci, MD, Fabio Grandi, PhD, Micaela Manni, MD, Daniela Mantella, MD, Andrea Pierantozzi, PhD, Maria Ruggeri, PhD, Daniela Sardella, PhD, Giovanni Sorbo, MD, and Ermanno Bonucci, MD Background: Fetuin-A of hepatic origin circulates in large amounts in serum, but also is expressed in bone, where it is an inhibitor of transforming growth factor (TGF- )/bone morphogenetic protein (BMP) proteins. Together with matrix GLA protein (MGP), fetuin-a is able to make up a complex with calcium and phosphate that is more soluble than calcium and phosphate alone, preventing its deposition in extraskeletal tissues. Experimental results suggested that this complex is made at bone tissue level. The aim of this study is to evaluate whether serum fetuin-a and MGP are influenced by type of renal osteodystrophy, they correlate with bone histomorphometric and histodynamic parameters, and/or serum levels may influence bone turnover. Methods: Thirty-eight hemodialysis patients who volunteered to undergo a bone biopsy were studied. Patients (27 men, 11 women) had a mean age of years and dialysis vintage of months. They were not administered vitamin D or drugs connected with mineral metabolism. They underwent transiliac bone biopsy after tetracycline labeling. Biopsies were performed for histological, histomorphometric, and histodynamic evaluation and aluminum histochemistry. Serum fetuin-a and MGP were measured by using enzyme-linked immunosorbent assay kits. Results: Serum fetuin-a levels were significantly less than normal, whereas MGP levels were less than the normal average. Fetuin-A levels in patients with hyperparathyroidism, mixed osteodystrophy, and low-turnover osteodystrophy were , , and ng/ml, respectively (P not significant). Fetuin-A level significantly correlated inversely with values for several histomorphometric parameters, such as osteoid volume (OV/BV), osteoblastic surface (Ob.S/BS), osteoid surface (OS/BS), and osteoclastic surface (Oc.S/BS). Logistic regression showed odds ratios of 5.3 and 4.9 for the association of high fetuin-a levels with low values for OS/BS and Ob.S/BS, respectively. Results of multiple regression analysis with intact parathyroid hormone and fetuin-a levels as independent variables and OV/BV and Ob.S/BS as dependent variables showed that independent variables correlated significantly with dependent variables, positively for intact parathyroid hormone levels and inversely for fetuin-a levels. MGP levels in patients with hyperparathyroidism, mixed osteodystrophy, and low-turnover osteodystrophy were not significantly different ( , , and nmol/l, respectively). By dividing MGP serum values into tertiles, mean values for OV/BV were different (analysis of variance, P < 0.04), with a greater value in the higher MGP tertile. By exclusion of 3 extravariant cases (>3 SDs greater than the mean), 1 case for each type of osteodystrophy, a significant correlation between bone formation rate and MGP serum level was found (P < 0.05). In addition, a significant correlation was found between MGP level and trabecular thickness. Conclusion: Fetuin-A and MGP levels correlated with bone formation parameters. This association could be caused by an effect of these proteins on bone formation, presumably mediated by the TGF- /BMP system. Fetuin-A, as opposed to MGP, is known to inhibit the TGF- /BMP complex, a protein-cytokine system that appears to be an important regulator of bone formation and probably a factor with an important role in renal osteodystrophy. Am J Kidney Dis 48: by the National Kidney Foundation, Inc. INDEX WORDS: Renal osteodystrophy; fetuin-a; matrix GLA protein; bone biopsy; histomorphometry; transforming growth factor /bone morphogenetic protein. FETUIN-A, or -Heremans Schmid glycoprotein, is a protein synthesized by the liver and circulating in large amounts in serum. 1,2 It is a negative acute-phase reactant protein with decreased levels after an inflammatory insult, and it accumulates in bone, where it has not-yet well- From the Ospedale Israelitico; Department of Experimental Medicine and Pathology, La Sapienza University; Departments of Nephrology and Laboratory Medicine, S. Giovanni-Addolorata Hospital; Department of Nephrology, San Camillo Hospital; Department of Laboratory Medicine, Policlinico TorVergata, Rome; and Hospal SpA, Bologna, Italy. Received January 10, 2006; accepted in revised form March 31, Originally published online as doi: /j.ajkd on May 30, Support: This study has been made possible by funds of Ospedale Israelitico, Rome, and of Hospal SpA, Bologna. Potential conflicts of interest: None. Address reprint requests to Giorgio Coen, MD, Via Dandolo 75, Rome, Italy. giorgio.coen@fastwebnet.it 2006 by the National Kidney Foundation, Inc /06/ $32.00/0 doi: /j.ajkd American Journal of Kidney Diseases, Vol 48, No 1 (July), 2006: pp

2 FETUIN-A, MATRIX GLA PROTEIN, AND BONE HISTOMORPHOMETRY 107 defined functions. In addition, it can be expressed in cartilage 3 and is considered an inhibitor of the transforming growth factor (TGF- )/ bone morphogenetic protein (BMP) family. 4 Experimental studies of fetuin-a knockout mice showed impairment of growth plate chondrocyte maturation and retardation in the longitudinal growth of femurs. 5 There also was accelerated trabecular bone remodeling with an increase in number of osteoblastic cells. In the absence of fetuin-a, results were in favor of increased activity of the TGF- /BMP system, a protein system with osteoinductive properties. Matrix GLA protein (MGP) is a vitamin K dependent protein containing carboxyglutamic acid 6 that is like osteocalcin and some coagulation factors, namely prothrombin. It is a 10-kd molecular-weight secreted protein expressed by a wide variety of tissues and cells. Cells shown to express MGP messenger RNA include chondrocytes, vascular smooth muscle cells, kidney cells, and fibroblasts, but also osteoblasts. 7 MGP seems to have a role in the regulation of the process of cartilage calcification. 8 In experimental conditions, fetuin-a, together with MGP, is able to make up a complex with calcium and phosphate that is more soluble than calcium and phosphate alone, preventing deposition of the salt in extraskeletal tissues. Price and Lim 9 showed that the fetuin-a MGP calcium phosphate complex was formed experimentally in animals treated with etidronate, a bisphosphonate able to block the mineralization of osteoid. The overflow of calcium-phosphate, caused by mineralization impairment, could be buffered, and extraskeletal calcium deposition could be prevented by formation of the complex. Therefore, circulating fetuin-a and MGP levels, although derived from other organs and cell systems, could be influenced by bone turnover in patients with renal osteodystrophy. The aim of this study is to evaluate whether levels of these proteins are linked to histological type of renal osteodystrophy, they correlate with histomorphometric and histodynamic parameters, and/or serum levels may influence bone turnover. METHODS The study was carried out with 38 hemodialysis patients with a wide range of serum intact parathyroid hormone (ipth) levels who volunteered to undergo bone biopsy. Patients (27 men, 11 women) had a mean age of years and a mean dialysis age of months. Causes of terminal renal failure were chronic glomerulonephritis in 9 patients, hypertension/ischemic nephropathy in 6 patients, tubulointerstitial nephropathy in 2 patients, polycystic kidney disease in 5 patients, diabetic nephropathy in 3 patients, obstructive nephropathy in 3 patients, and unknown in 10 patients. Patients were not administered vitamin D compound. However, most patients had been treated previously with relatively limited doses of calcitriol administered orally or intravenously (1.5 to 3.5 g/wk), but had discontinued this treatment at least 3 months before bone biopsy in 2 patients and more than 5 months in the other patients. They were not administered anticoagulants, corticosteroids, nonsteroidal anti-inflammatory drugs, antiepileptic drugs, or estrogens. Phosphate binders were mainly calcium salts and sevelamer. In the majority of patients, regular erythropoietin intravenous treatment was underway. All patients were treated with standard hemodialysis, 12 hours a week, divided into 3 sessions. Patients underwent transiliac bone biopsy with a Bordier trocar after a double-labeling course with oral tetracycline, with a 12-day interval. The biopsy, performed 3 to 5 days after the end of tetracycline administration, was for histological, histomorphometric, and histodynamic evaluation and aluminum histochemistry. At the same time, a blood sample was drawn for the following assays: calcium, phosphate, ipth, total alkaline phosphatase, serum fetuin-a, and MGP. Serum fetuin-a was measured by using an enzyme-linked immunosorbent assay kit (Epitope Diagnostics Inc, San Diego, CA); normal values are g/l (25 healthy control subjects of the same age as the hemodialysis cohort). Intra-assay and interassay coefficients of variation were 6.8% and 8.2%, respectively. MGP was measured by using an enzyme-linked immunosorbent assay provided by Biomedica (Wien, Austria); normal values for 25 subjects were nmol/l. Intra-assay and interassay coefficients of variation were 5.5% and 8%, respectively. Serum PTH was measured by means of a commercial (Nichols Institute Diagnostics, San Juan Capistrano, CA) immunoradiometric assay, directed to the intact molecule, based on a double-antibody technique. Both 1-84 and 7-84 PTH species are measured together. Normal range of values is 10 to 65 pg/ml. Serum total calcium was determined by means of a spectrophotometric assay using cresolphthalein as substrate. Serum phosphate and alkaline phosphatase measurements were performed spectrophotometrically (DU-65; Beckman, Fullerton, CA), using molybdate and p-nitrophenyl-phosphate as respective substrates. Normal ranges for the adult population are 2.5 to 4.5 mg/dl and 60 to 170 U/L, respectively. Bone specimens were fixed by using 4% paraformaldehyde in 0.1 mol/l of phosphate buffer (ph 7.2). They then were longitudinally halved, dehydrated in acetone, and embedded without decalcification in glycol-methacrylate, as previously described. 10 Sections were cut with a Reichert-Jung Autocut microtome (Leica Microsystems, Wetzlar, Germany) equipped with a tungsten carbide

3 108 COEN ET AL Table 1. Histomorphometric Variables Bone volume (BV/TV, %) Trabecular thickness (Tb.Th, m) Osteoid volume (OV/BV, %) Osteoid thickness (O.Th, m) Osteoid surface (OS/BS, %) Osteoblast surface (Ob.S/BS, %) Eroded surface (ES/BS, %) Osteoclast surface (Oc.S/BS, %) Osteoclast number (Oc.N/BS, no./mm 2 ) Single-labeled surface (sls/bs, %) Double-labeled surface (dls/bs, %) Mineralizing surface (MS/BS, %) Mineralizing surface (MS/OS, %) Mineral apposition rate (MAR, m/d) Adjusted apposition rate (Aj.AR, m/day) Bone formation rate (BFR/BS, m 3 / m 2 /d) Mineralization lag-time (Mlt, d) Percent of whole trabecular bone volume occupied by calcified and uncalcified bone tissue Thickness of bone trabeculae Percent of bone volume consisting of osteoid Thickness of osteoid seams Percent of trabecular surface covered by osteoid tissue Percent of trabecular surface covered by osteoid lined by active osteoblasts Percent of trabecular surface with Howship lacunae Percent of trabecular surface covered by osteoclasts No. of osteoclasts per mm 2 of trabecular surface Extent of tetracycline single-labeled surface in percentage of trabecular surface Extent of tetracycline double-labeled surface in percentage of trabecular surface Extent of double-labeled plus half the extent of single-labeled surface in percentage of trabecular surface Extent of double-labeled plus half the extent of single-labeled surface in percentage of osteoid surface Distance between the midpoints of 2 consecutive labels, divided by the time between the midpoints of the 2 labeling periods Mineral apposition rate corrected for osteoid surface Volumetric amount of new mineralized bone per unit of trabecular bone surface per day Mean time between deposition of osteoid matrix and its mineralization knife. Three- to 4- m thick sections of the specimens, together with positive controls, were stained by using the aluminum histochemical staining technique. 11 Alternate sections, 1 to 2 m thick, were stained with azure II-methylene blue for histomorphometric measurement of structural and static variables. Alternate sections, about 5 m thick, also were prepared unstained to be examined under UV light for histodynamic evaluation of tetracycline fluorescent labels. Histomorphometric and histodynamic measurements, performed in 35 patients (in 3 cases, bone sample was not adequate for quantification), were obtained by using an interactive color video-based Image Analysis System (IAS 2000; Delta Sistemi, Rome, Italy) with personalized software developed for bone histomorphometry. Measured variables 12 are listed in Table 1. Bone volume (BV/TV) was evaluated at objective magnification 2.5, and static and dynamic variables were evaluated at objective magnification Normal values (Table 2) for histomorphometric 13 and histodynamic 10 parameters were obtained in our own histomorphometric laboratory. Bone pathological states were classified as hyperparathyroidism (HP), osteomalacia (OM), mixed osteodystrophy (MO), and adynamic bone disease (ABD). Classification was made on the basis of morphological criteria, as already reported. 10,14 The designation of HP implied a general increase in bone turnover rate, and OM was characterized by a decrease in bone turnover rate associated with an increase in both osteoid surface (OS/BS) and thickness (O.Th). MO included all intermediate features between pure HP and OM, 15 and ABD was characterized by reduced bone turnover associated with thin osteoid seams, bone cell paucity, and a decrease in tetracycline uptake. Statistical evaluation was carried out with a statistical package (version 12.0; SPSS Inc, Chicago, IL) on a personal computer. Correlation analysis was made with Pearson coefficients. We also used regression models. For multiple regression analysis, we used the forward method, with a probability level of P less than 0.05 for entering the model, whereas for logistic regression, we used the mean of the sample as the cutoff value. To compare different groups, parametric tests (analysis of variance and t-test) were used. P less than 0.05 is considered significant. Values are expressed as mean SD. RESULTS Mean values for clinical and biochemical variables and histomorphometric and histodynamic results of patients are listed in Table 2. Histological diagnoses were HP in 21 patients, MO in 11 patients, and low-turnover osteodystrophy (LTO) in 6 patients (3 patients, ABD; 3 patients, OM). No patient had a positive aluminum histochemistry result. Fetuin-A serum levels were less than normal ( mg/ml; normal, mg/ml). In addition, MGP serum levels ( nmol/l) were less than the normal average value ( nmol/l). There was a

4 FETUIN-A, MATRIX GLA PROTEIN, AND BONE HISTOMORPHOMETRY 109 Table 2. Clinical and Biochemical Data and Histomorphometric Variables of Patients and Normal Reference Values positive nonsignificant correlation between the 2 protein levels (P 0.1). Fetuin-A levels in patients with HP, MO, and LTO were , , and ng/ml, respectively; this did not reach statistical difference. Serum fetuin-a levels correlated inversely with several histomorphometric parameters. Considering the entire sample, only the correlation of fetuin-a versus osteoblast surface (Ob.S/BS) was significant (r 0.340; P 0.05). With the exclusion of patients with the diagnosis of LTO, an increase in correlation All Normal Values No. of patients 38 Age (y) Hemodialysis age (mo) Body mass index (kg/m 2 ) Calcium (mg/dl) Phosphate (mg/dl) ipth (pg/ml) Alkaline phosphatase (U/L) Hydroxychole calciferol (ng/ml) Fetuin-A (g/l) Matrix GLA protein (nmol/l) BV/TV (%) OV/BV (%) O.Th (m) OS/BS (%) Ob.S/BS (%) ES/BS (%) Oc.S/BS (%) Oc.N/BS (no./mm 2 ) sls/bs (%) dls/bs (%) MS/BS (%) MS/OS (%) MAR ( m/d) BFR/BS ( m 3 / m 2 /d) Aj.AR ( m/d) Mlt (d) Tb.Th ( m) NOTE. Values expressed as mean SD. To convert calcium in mg/dl to mmol/l, multiply by ; phosphate in mg/dl to mmol/l, multiply by ; PTH in pg/ml to ng/l, multiply by 1; 25-hydroxychole calciferol in ng/ml to nmol/l, multiply by coefficients and significance levels of some correlations was observed. Fetuin-A levels correlated significantly with osteoid volume (OV/ BV; r 0.466; P 0.01), Ob.S/BS (r 0.498; P 0.01), OS/BS (r 0.477; P 0.01), and osteoclast surface (Oc.S/BS; r 0.396; P 0.01; Table 3). The significance level also was increased by log transform of fetuin-a levels (fetuin-a versus Ob.S/BS, r 0.538; P 0.005). Scatterplots of fetuin-a and log transform of fetuin-a versus Ob.S/BS are shown in Figs 1 and 2. Fetuin-A level also was related inversely to bone formation rate (BFR/BS); however, this did not reach a significance level. Considering mean values of the variables of the sample as cutoff values, logistic regression of fetuin-a versus the parameters OS/BS and Ob.S/BS show odds ratios of 5.3 (95% confidence interval, 1.02 to 27.7; P 0.047) and 4.9 (95% confidence interval, 1.01 to 23.8; P 0.048) for the association of high fetuin-a levels with low levels of OS/BS and Ob.S/BS, respectively. Results of multiple regression analysis with ipth and fetuin-a levels as independent variables and OV/BV, Ob.S/BS, OS/BS, Oc.S/BS, and osteoclast number (Oc.N/ BS) as dependent variables are listed in Table 4. Both independent variables correlated sig- Table 3. Correlation Coefficients of Fetuin-A and ipth With Histomorphometric Variables Histomorphometric Variables Fetuin-A PTH BV/TV OV/BV 0.466* 0.458* OS/BS 0.477* 0.420* Ob.S/BS 0.498* 0.544* ES/BS * Oc.S/BS 0.396* 0.393* Oc.N/BS 0.386* 0.497* sls/bs dls/bs * MS/BS * MS/OS BFR/BS * O.Th * MAR * Aj.AR * Mlt Tb.Th *P 0.01.

5 110 COEN ET AL Fig 1. Scatterplot of fetuin-a versus the parameter Ob.S/BS. nificantly with the dependent variables OV/BV and Ob.S/BS, positively for ipth level and inversely for fetuin-a level, whereas the other dependent variables were excluded by the model. MGP serum levels in patients with HP, MO, and LTO were not significantly different, with values of , , and nmol/l, respectively. By dividing MGP serum values into tertiles, mean values of OV/BV were different (analysis of variance, P 0.04), with a higher value of the variable in the higher MGP tertile (Fig 3). By exclusion of 3 extravariant cases ( 3 SDs greater than the mean), 1 for each type of osteodystrophy, a significant correlation between BFR/BS and MGP serum values was found (r 0.348; P 0.05; Fig 4). In addition, a significant correlation was found between MGP level and the variable trabecular thickness (Tb.Th), available in 30 patients, as shown in Fig 5. DISCUSSION Results of this cross-sectional study favor the hypothesis that fetuin-a serum levels are associated with bone cell activity. Fetuin-A levels correlated inversely with several histomorphometric parameters connected with bone formation (Ob.S/BS) and bone resorption (Oc.S/BS). Exclusion of cases of LTO increased significance levels of the correlations, probably because of the extremely low values of their pertinent histomorphometric and histodynamic parameters. Probably being low for different, still relatively unknown, factors, these parameters cannot be influenced further by the supposed suppressive effect of fetuin-a. For Fig 2. Scatterplot of ln fetuin-a versus the parameter Ob.S/BS.

6 FETUIN-A, MATRIX GLA PROTEIN, AND BONE HISTOMORPHOMETRY 111 Table 4. Multiple Regression Analysis (forward method) of Fetuin-A and ipth as Independent Variables Dependent Variable Fetuin-A ipth OV/BV (P) (0.017) (0.029) Z R Ob.S/BS (P) (0.008) (0.004) Z R MGP level, it showed a direct relationship with BFR/BS after exclusion of 3 extravariant cases. The data do not allow us to conclude whether bone turnover influences serum fetuin-a level or the reverse. However, considering that high turnover renal osteodystrophy and LTO were not associated with significantly different fetuin-a levels, bone tissue and turnover do not seem to influence serum fetuin-a levels. Conversely, it is possible that this protein is a factor directly or indirectly influencing bone cell activity, with an effect inversely related to PTH level. Several studies in the past emphasized the relation between bone fetuin-a level and osteoblastic activity. In fetuin-a knockout mice, bone formation was increased, indicated by greater cortical thickness, accelerated trabecular bone remodeling, and increased osteoblast number on bone surfaces. 5 This phenomenon was considered related to the interaction between fetuin-a level and the TGF- /BMP complex, with fetuin-a acting as an inhibitor of the osteoinductive effect of the TGF- /BMP protein family. Similar conclusions were reached by other investigators 16 in an in vitro study of osteoblastic cells of bone marrow cultures of rats treated with dexamethasone. Fetuin-A was found to inhibit osteogenesis, probably by binding to the TGF- /BMP cytokines, blocking their osteogenic activity. 4 For MGP serum levels, no difference was found in patients with LTO and high turnover osteodystrophy. However, a significant positive correlation was found between MGP level and BFR/BS. OV/BV was significantly greater in the higher tertile of MGP values. MGP also correlated positively with Tb.Th (Fig 5). Therefore, for MGP as for fetuin-a, it is possible that levels influence bone turnover, whereas serum levels do not seem to be influenced by bone in patients with renal osteodystrophy. How- Fig 3. OV/BV average values ( SD) for different tertiles of MGP serum levels (analysis of variance, P < 0.04).

7 112 COEN ET AL Fig 4. Scatterplot of serum MGP versus the parameter BFR/BS. ever, MGP appears to show an opposite effect compared with fetuin-a. In in vitro studies, MGP enhanced osteogenic and chondrogenic BMP-2 activity. 17 MGP also stimulated vascular endothelial growth factor A through increased TGF- 1 activity in endothelial cells. 18 In addition, the increase in OV/BV at greater MGP levels is in line with the experimental calcification defect of osteoid, with appearance of osteoidosis observed in animals producing MGP at osteoblastic levels. 19 This study has intrinsic limitations because of its cross-sectional nature and the association of rather static bone histomorphometric and histodynamic parameters with relatively unstable humoral parameters. In addition, as frequently occurs for studies with bone biopsies, the size of the patient cohort and statistical strength of the model obviously are limited. Caution also should be used in the interpretation of presently available MGP measurements because they do not distinguish between active and uncarboxylated molecules. Immunohistochemical evaluations showing the relation between serum levels and local expression of fetuin-a and MGP probably could contribute to the understanding of these preliminary results. However, based on the present findings, we may conclude that fetuin-a and MGP serum levels are associated with bone formation. If we assume that this association is the result of an action of these proteins on bone, we might think that it possibly could be mediated by the TGF- / BMP system. Fetuin-A, as opposed to MGP, is known to inhibit bone turnover and bone cell activity by inhibiting the TGF- /BMP complex, 4,5 a protein-cytokine system that appears to be an important regulator of bone formation and Fig 5. Scatterplot of MGP serum levels versus the parameter Tb.Th.

8 FETUIN-A, MATRIX GLA PROTEIN, AND BONE HISTOMORPHOMETRY 113 probably a factor with an important role in renal osteodystrophy. 20 REFERENCES 1. Triffitt JT, Gebauer U, Ashton BA, Owen ME, Reynolds JJ: Origin of plasma alpha2hs-glycoprotein and its accumulation in bone. Nature 262: , Dickson IR, Poole AR, Veis A: Localization of plasma alpha2hs glycoprotein in mineralising human bone. Nature 256: , Yang F, Schwartz Z, Swain LD, Lee CC, Bowman BH, Boyan BD: Alpha 2-HS-glycoprotein: Expression in chondrocytes and augmentation of alkaline phosphatase and phospholipase A2 activity. Bone 12:7-15, Demetriou M, Binkert C, Sukhu B, Tenenbaum HC, Dennis JW: Fetuin/ 2-HS glycoprotein is a transforming growth factor- type II receptor mimic and cytokine antagonist. J Biol Chem 271: , Szweras M, Liu D, Partridge EA, et al: 2-HS Glycoprotein/fetuin, a transforming growth factor- /bone morphogenetic protein antagonist, regulates postnatal bone growth and remodeling. J Biol Chem 22: , Price PA, Williamson MK: Primary structure of bovine matrix Gla protein, a new vitamin K-dependent bone protein. J Biol Chem 2: , Sato M, Yasui N, Nakase T, et al: Expression of bone matrix proteins mrna during distraction osteogenesis. J Bone Miner Res 13: , Yagami K, Suh JY, Enomoto-Iwamoto M, et al: Matrix GLA protein is a developmental regulator of chondrocyte mineralization and, when constitutively expressed, blocks endochondral and intramembranous ossification in the limb. J Cell Biol 147: , Price PA, Lim JE: The inhibition of calcium phosphate precipitation by fetuin is accompanied by the formation of a fetuin-mineral complex. J Biol Chem 278: , Coen G, Ballanti P, Bonucci E, et al: Bone markers in the diagnosis of low turnover osteodystrophy in haemodialysis patients. Nephrol Dial Transplant 13: , Maloney NAQ, Ott SM, Alfrey AC, et al: Histological quantification of aluminium in iliac bone from patients with renal failure. J Lab Clin Med 99: , Parfitt AM, Drezner MK, Glorieux FH, et al: Bone histomorphometry standardization of nomenclature, symbols and units. Bone Miner 2: , Ballanti P, Bonucci E, Della Rocca C, et al: Bone histomorphometric reference values in 88 normal Italian subjects. Bone Miner 11: , Malluche HH, Faugere MC: Renal bone disease 1990: An unmet challenge for the nephrologist. Kidney Int 38: , Fournier A, Ghitu S, Bako G, et al: Bone markers in the diagnosis of low turnover osteodystrophy in haemodialysis patients. Nephrol Dial Transplant 13: , 1999 (letter; reply) 16. Binkert C, Demetriou M, Sukhu B, Szweras M, Tenenbaum HC, Dennis JW: Regulation of osteogenesis by fetuin. J Biol Chem 274: , Bostrom K, Tsai D, Shen S, Wang Y, Demer L: Matrix GLA protein modulates differentiation induced by bone morphogenetic protein-2 in C3H10T1/2 cells. J Biol Chem 276: , Bostrom K, Zebboudj AF, Yao Y, Lin TS, Torres A: Matrix GLA protein stimulates VEGF expression through increased transforming growth factor- 1 activity in endothelial cells. J Biol Chem 279: , Murshed M, Schinke T, McKee MD, Karsenty G: Extracellular matrix mineralization is regulated locally; Different roles of two Gla-containing proteins. J Cell Biol 165: , Li T, Surendran K, Zawaideh MA, Mathew S, Hruska KA: Bone morphogenetic protein 7: A novel treatment for chronic renal and bone disease. Curr Opin Nephrol Hypertens 13: , 2004