SEARCH

SEARCH BY CITATION

Keywords:

  • bone fractures;
  • bone densitometry;
  • polymorphisms;
  • single nucleotide polymorphism;
  • group-specific component;
  • vitamin D binding protein;
  • Gc-MAF;
  • soluble CD163

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

The phenotype of the vitamin D binding and macrophage activating protein, Gc, is a predictor of premenopausal bone fracture risk, possibly mediated through activation of osteoclasts. This was concluded from a study on 595 Danish perimenopausal women 45-58 years of age (30,040 person years).

Introduction: The multifunctional plasma protein Gc, also known as group-specific component, Gc globulin, or vitamin D binding protein (DBP), has two functions with relation to bone tissue: it is the major carrier protein of vitamin D in the circulation, and deglycosylation converts it into a very potent macrophage- and osteoclast-activating factor (Gc-MAF). There are several phenotypes of Gc, and in this study, we examined the relation between Gc phenotype and bone fragility.

Materials and Methods: By isoelectric focusing we identified the Gc phenotype of 595 white recent postmenopausal women enrolled into the Danish Osteoporosis Prevention Study (DOPS) and identified three groups: Gc1-1 (n = 323), Gc1-2 (n = 230), and Gc2-2 (n = 42). Differences between the three groups were examined with respect to number of fractures before enrollment, BMC and BMD, and various biochemical and clinical parameters, including the concentration of Gc measured by immunonephelometry and the concentration of the macrophage marker soluble CD163 measured by ELISA.

Results and Conclusions: The risk of having at least one premenopausal bone fracture (total number of women with fracture = 179) differed significantly (p = 0.017) in women with phenotype Gc1-1 (110/323 = 0.34), Gc1-2 (63/230 = 0.27), and Gc2-2 (6/42 = 0.14). The differences were even more striking (p = 0.005) for fractures caused by low-energy traumas. Using logistic regression, we found the relative risk of premenopausal fracture to be 0.32 (0.13-0.80) in Gc2-2 compared with Gc1-1. We propose that the Gc phenotypes cause differences in osteoclast activity, a theory supported by our finding of lower levels of Gc and of soluble CD163 in women with Gc2-2 compared with Gc1-1.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

IN EVERY TRAUMA, THE risk of bone fracture depends on the interaction between the velocity and direction of the force and the biomechanical competence of the skeleton. In osteoporosis, low bone mass and microarchitectural deterioration lead to enhanced bone fragility and a consequent increase in fracture risk. It is well documented that inherited factors play an important role in the pathogenesis of bone fragility, but the search for the exact genes is still ongoing.

The Gc gene is among the possible candidates. The gene encodes for the multifunctional 50- to 58-kDa plasma protein Gc, also known as group-specific component, Gc-globulin, and vitamin D binding protein (DBP). This protein has important functions related to the skeleton: First, Gc is the major carrier protein of vitamin D metabolites in the circulation and is important for the preservation of the vitamin.(1) Second, deglycosylation converts Gc into a very potent macrophage-activating factor (Gc-MAF), which is also capable of activating osteoclasts.(2,3)

The Gc gene is located on chromosome 4 (4q11-13). There are three common co-dominant phenotype alleles known as Gc1s, Gc1f, and Gc2, differing by amino acid substitutions as well as glycosylation. Amino acid 416 is aspartic acid (codon GAT) in Gc1f and Gc2 and glutamic acid (codon GAG) in Gc1s. At position 420, threonine (codon ACG) is found in Gc1f and Gc1s and lysine in Gc2.(4) Gc2 is glycosylated with a terminal galactose, whereas Gc1s and Gc1f contain both galactose and sialic acid.(5) In addition to the three common alleles, >120 variants have been described.(6) From widespread use in forensic medicine and population genetics, the frequency of Gc2 is known to be highest among whites and lowest among black Africans, whereas the opposite is true for Gc1f.(7) The plasma concentration of Gc depends on Gc phenotype being highest in Gc1-1 (Gc1s-1s, Gc1s-1f, Gc1f-1f), intermediate in Gc1-2 (Gc1s-2, Gc1f-2), and lowest in Gc2-2.(8)

The aim of this cross-sectional study was to investigate the possible influence of Gc phenotype and Gc concentration on BMD and fracture risk in women.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Study population

We enrolled 595 recent postmenopausal women participating in the Danish Osteoporosis Prevention Study (DOPS) at the Osteoporosis Clinic in Aarhus in the present substudy. DOPS is an ongoing 20-year, partly randomized population-based multicenter study of osteoporotic fracture prevention in postmenopausal white Danish women through the use of hormonal replacement therapy.(9) The women were recruited by direct mailing to a random sample of 45- to 58-year-old women drawn from the Danish Civil Registration System. For women with an intact uterus, inclusion criteria were an age of 45-58 years and 3-24 months since last menstrual bleeding or perimenopausal symptoms and elevated serum follicle stimulating hormone (FSH). For hysterectomized women, inclusion criteria were an age of 45-52 years and an elevated FSH. Exclusion criteria were metabolic bone disease (including nontraumatic vertebral fractures), current estrogen use, being treated with glucocorticoids for >6 months, current or past malignancy, chronic disease if newly diagnosed or out of control, and any hospitalization because of alcohol or drug addiction. Furthermore, all women enrolled had plasma concentration of calcium, alanine amino transferase, and thyroid stimulating hormone (TSH) within normal ranges. At present, the DOPS database includes information on biochemical variables, BMD, and clinical data at baseline and after 1, 2, and 5 years of follow-up. In this study, we used only the baseline data.

Bone fractures and bone scans

Bone fracture:

We obtained information on previous bone fractures as part of a structural interview (30,040 person years at risk). We categorized the energy of the traumas causing the fractures as low (for example falls from standing level), high (for example traffic accidents and falls from higher levels), or unknown. We registered the location of the fractures as forearm, proximal humerus, columna, costae, pelvis, femoral neck, or “other.”

Bone scans:

We used carefully cross-calibrated Hologic QDR 1000 W and 2000 W DXA scanners (Hologic, Waltham, MA, USA) to determine total body and regional BMC and BMD. We scanned the forearm (ultradistal and proximal one-third), lumbar spine (L2-L4), hip region (trochanter, femoral neck, and total), and total body. CVs for BMD were in the range of 0.7-2.1%.(10)

Biochemical methods

Gc phenotyping (genotyping):

The Gc genotype can be deduced from the phenotype because the alleles are co-dominant. The phenotypes can be obtained by isoelectric focusing. We used Immobiline Dryplate gels (pH 4.5-5.4) from Amersham Biosciences. The running conditions were 3500 V, 5 mA, 15 W, and 18-25 h, and the gels were stained with Amido black.

Gc concentration:

We measured Gc concentration in plasma by an immunonephelometric method as described.(8) The interassay CV was 2.6-4.6%.

CD163:

We measured the plasma form of the macrophage receptor for clearance of haptoglobin-hemoglobin complexes, soluble CD163, by ELISA(11) (interassay CV, 4.3-6.7%) in a subset of available samples with as many as possible of the infrequent phenotypes (Gc1f-1f, Gc1f-2, Gc2-2), and 50 of each of the other types.

Other biochemical variables:

We measured several other biochemical quantities employing commercial kits or standard laboratory methods. Below we indicate the applied methods with their commercial source (if any) and the interassay CV of their analytical performance. We measured plasma 25-hydroxyvitamin D by a radio assay (CV 13.5%) and calculated the 25-hydroxy-vitamin D index as the vitamin concentration divided by the Gc concentration. We measured plasma intact parathyroid hormone (PTH) using an Immulite automated analyzer (DPC Immulite; CV 11%), plasma alkaline phosphatase activity by spectrophotometry with p-nitrophenylphosphate as substrate (CV 5%), bone isoenzyme alkaline phosphatase activity in plasma by lectin precipitation (CV 6.8%),(12) plasma osteocalcin (BGP) by radioimmunoassay (CV 10%), and plasma estradiol using an AutoDelfia flouroimmunoassay (Wallac OY; CV 8.5%). We measured calcium, phosphate, creatinine, TSH, and FSH levels by standard laboratory techniques (CV < 5%). We determined urine hydroxyproline (U-OHP) spectrophotometrically with p-dimethylaminobenzaldehyde as substrate (Organon Teknika; CV 11%) and urine pyridinolines (pyridinoline and deoxypyridinoline) by HPLC (CV 11%). We calculated urinary bone markers as creatinine ratios on fasting second void morning spot urine. Finally, we measured plasma carboxyterminal peptide of procollagen type 1 (1CPT), plasma C-terminal propeptide of type 1 collagen (P1CP), and plasma N-terminal propeptide of type 1 collagen (P1NP) by radioimmunoassay (Orion Diagnostica, Espoo, Finland; CV 8%, 14%, and 7%, respectively).

Other variables

We measured height and weight in light indoor clothes to the nearest 0.5 cm and 0.5 kg, respectively. By means of structural interviews, we got information on lifetime use of alcohol and tobacco, current exercise patterns (hours per week), current intake of vitamin supplements, and current consumption of tea and coffee expressed in number of standard cups per day.

Ethics

The Danish Osteoporosis Prevention Study was approved by the regional ethics committees (1990/1821). The women were enrolled after written informed consent (Helsinki II).

Statistics

We evaluated the association between Gc phenotype and the risk of having at least one bone fracture before enrollment by χ2 or Fisher's exact test as appropriate and by logistic regression. As a test for trend, we calculated Spearman's nonparametric correlation coefficient. We used ANOVA, Kruskal-Wallis test, t-test, and Mann-Whitney test as appropriate for comparison of BMD values and biochemical variables among Gc phenotypes and Pearson's or Spearman's correlation coefficient as appropriate to look for significant correlations. We applied a significance level of 0.05.

In the logistic regression, we used the BMD values, biochemical variables, height, weight, body mass index (BMI), fat mass, lean body mass, age at enrollment, age at menarche, current physical activity, and consumption of tea and coffee as continuous variables, whereas we dichotomized former and present smoking and current intake of vitamin supplements into “yes” or “no.” We categorized the Gc phenotypes into three groups: Gc1-1, Gc1-2, and Gc2-2. To look for nonlinear relations, we also performed logistic regression analysis with BMI, weight, fat mass, and lean body mass categorized into quartiles. We first performed simple logistic regression with occurrence of one or more bone fractures before menopause as the dependent variable and the independent variables entered one at a time. After exclusion of variables with p values >0.15, we checked for effect modification and confounding among those with p values <0.15. In the final logistic regression analysis, we used forward stepwise method (Wald test and log likelihood ratio) and entered the following: Gc phenotype, BMD in the ultradistal forearm, 25-hydroxyvitamin D index, total alkaline phosphatase, height, weight, age at enrollment (corresponding to observation time and reflecting the age at menopause), and current smoking.

We performed all calculations using SPSS version 10.0.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

We identified the Gc phenotypes as Gc1s-1s (n = 191), Gc1s-1f (n = 115), Gc1f-1f (n = 17), Gc1s-2 (n = 199), Gc1f-2 (n = 31), and Gc2-2 (n = 42). However, in our data presentation, we group the phenotypes in three groups: Gc1-1 (Gc1s-1s, Gc1s-1f, Gc1f-1f), Gc1-2 (Gc1s-2, Gc1f-2), and Gc2-2 for the following reasons: (1) Gc1s and Gc1f are alike with regard to the glycosylation of importance for Gc-MAF generation, and (2) the plasma concentration of Gc depends on Gc phenotype with no significant difference for individuals bearing one or the other of the Gc1s or Gc1f types. As reported earlier, the mean plasma Gc level (mean ± SE) was 272 ± 2 mg/liter in women with Gc1-1, 249 ± 2 mg/liter in Gc1-2, and 226 ± 3 mg/liter in Gc2-2 (p < 0.001), and the 95% central reference intervals in the three groups were 218-346, 210-310, and 183-268 mg/liter, respectively. We found no differences among women in the three Gc groups with respect to age, age at menarche, body size, exercise patterns, smoking status, or intake of vitamin supplements (Table 1).

Table Table 1.. Baseline Characteristics of the Study Population
Thumbnail image of

Premenopausal bone fracture frequency and Gc phenotype

One hundred seventy-nine (30%) of the 595 women had a history of one or more bone fractures before enrollment into the DOPS (242 fractures altogether). Seventy-five of these were forearm fractures, 1 was a hip fracture, 3 were rib fractures, 5 were traumatic spine fractures, and the remaining 158 fractures were other types. Figure 1 shows the number of fractures and fracture types at different ages.

thumbnail image

Figure FIG. 1.. The number of bone fractures among 595 females at different ages from birth until inclusion into the study (30,040 person-years).

Download figure to PowerPoint

In our data analysis, we only include the first fracture experienced by each of the 179 women. The fracture was a low-energy fracture in 67 and a high-energy fracture in 110 of the women. Fracture energy was unknown in two. The risk of fracture varied significantly (p = 0.017) according to Gc phenotype, being highest for Gc1-1 (110/323 = 0.34), intermediate for Gc1-2 (63/230 = 0.27), and lowest for Gc2-2 (6/42 = 0.14; Fig. 2A). The differences were even more striking (p = 0.005) when only low-energy fractures were included in the analysis. None of the Gc2-2 women had experienced a low-energy fracture (Fig. 2B). The test for trend in fracture frequency with number of Gc2 alleles was highly significant for all fractures (p = 0.009) and for low-energy fractures (p = 0.005); in other words, the data support that the risk of fracture decreases as the number of Gc2 alleles increases. The fracture frequencies after stratification of the data according to trauma energy, age intervals, and fracture type (forearm or other) are shown in Fig. 3. In all groups, except for high-energy forearm fractures, the frequency was highest in Gc1-1, intermediate in Gc1-2, and lowest in Gc2-2. However, because of the reduced number of events, the results were not significant.

thumbnail image

Figure FIG. 2.. Frequency of one or more bone fractures in women (n = 595) with different Gc phenotypes. (A) All fractures (n = 179). (B) Fractures caused by low-energy trauma (n = 67).

Download figure to PowerPoint

thumbnail image

Figure FIG. 3.. Frequency of fracture in women (n = 595) with different Gc phenotypes after dividing the data according to fracture site (white bars, forearm; patterned bars, other), trauma energy, and age interval. (A-C) Fractures caused by low-energy trauma. (D-F) Fractures caused by high-energy traumas. Age range: A and D, 0-19 years; B and E, 20-39 years; C and F, 40-59 years.

Download figure to PowerPoint

BMC, BMD, and markers of bone resorption and bone formation

We found no significant differences in BMC (p = 0.07-0.81) or BMD (p = 0.21-0.97) among Gc phenotypes in any of the scanned bone areas (forearm, lumbar spine, hip region, and total body). Neither did we find any significant differences among the Gc phenotypes with regard to the measured resorptive and formative bone markers. The correlation between Gc concentration and bone markers was significant only with regard to plasma alkaline phosphatase (r = 0.17, p < 0.001) and bone specific alkaline phosphatase (r = 0.12, p = 0.003). Looking at the entire study population we found no significant correlation between Gc concentration and BMD values (p = 0.53-0.96); however, in the women with fractures, and especially in those with low-energy fractures, Gc concentration correlated negatively with BMD in all regions. Correlations were significant in women with low-energy fractures at lumbar spine (r = −0.28, p = 0.02), total hip (r = −0.24, p = 0.04), femoral neck (r = −0.33, p = 0.005), and total body (r = −0.34, p = 0.004). The women with low-energy fractures had lower BMD in all areas compared with women without low-energy fractures; these were significantly lower at the ultradistal forearm (p = 0.004) and total body (p = 0.047).

Logistic regression

In the simple logistic regression with occurrence of one or more bone fractures before menopause as the dependent variable, we found the following independent variables to have p values below 0.15 (listed according to growing p value): BMD in ultradistal forearm, Gc phenotype, Gc concentration, total alkaline phosphatase, fat mass, age at enrollment, weight, total body BMD, bone specific alkaline phosphatase, current smoking, height, lean body mass, and 25-hydroxyvitamin D index. For fat mass (kg; OR = 1.02, p = 0.043 when treated as a continuous variable), a nonlinear relation was detected with the lowest fracture risk in those with fat mass in the quartile just below the median (OR = 0.48, p = 0.005 compared with those with fat mass in the highest quartile). We found no significant effect modification but various degrees of confounding. For example, the Gc concentration (OR = 1.006, p = 0.029 in simple logistic regression) became insignificant when analyzed together with Gc phenotype because the concentration depends on Gc type, and when analyzing BMD at different sites in the same model, only BMD of the ultradistal forearm remained significant because the BMD values are highly correlated. The final model is presented in Table 2. Compared with women with Gc phenotype Gc1-1, the OR for having a least one bone fracture before menopause is 0.32 (0.13-0.80) for women with Gc2-2 and 0.72 (0.49-1.05) for women with Gc1-2.

Table Table 2.. Final Logistic Regression Model Based on Data From 595 Women With Occurrence of One or More Bone Fractures Before Menopause as the Dependent Variable (Forward Stepwise Inclusion Using Wald Test and Log Likelihood Ratio)
Thumbnail image of

Soluble CD163

This newly identified marker of macrophage activity was measured in plasma of 194 women with different Gc phenotypes. The mean soluble CD163 concentration was highest in women with Gc1-1 (2.91 ± 0.19 mg/liter), intermediate in Gc1-2 (2.79 ± 0.13 mg/liter), and lowest in Gc2-2 (2.14 ± 0.15 mg/liter; p = 0.03). There was no significant correlation with Gc concentration or BMD values.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

We report a highly significant difference in premenopausal bone fracture risk among women with different Gc phenotype. An obvious question is whether the three times lower fracture risk in women with Gc2-2 compared with Gc1-1 can be explained biologically by our current knowledge about the Gc protein. There is at present no clear answer to this question, but existing knowledge concerning the Gc isoproteins combined with the results presented in this paper strongly suggest that the relation between fracture risk and Gc phenotype is more than a chance observation. The Gc phenotype remained a significant risk factor even after inclusion of several possible confounders in the logistic regression analysis, and we exclude the possibility of recall bias because we interviewed the women before any knowledge of their Gc phenotype. We propose that the activated Gc in the form of Gc-MAF plays a role in bone modeling/remodeling and thereby has an effect on the risk of bone fractures.

Gc-MAF was first discovered as a very potent activator of macrophages, but as demonstrated both in vivo and in vitro, Gc-MAF is now recognized as an activator of osteoclasts as well. Defective Gc-MAF generation has been connected with osteopetrosis through studies of blood from patients with osteopetrosis(13) and through animal models. Administration of Gc-MAF to osteopetrotic rodents reversed their bone defects,(2) and in vitro osteoclast activity determined by pit formation capacity rose in a dose-dependent manner when incubated with Gc-MAF.(14)

Differences between the Gc phenotypes concerning Gc-MAF might be either qualitative or quantitative. The variation in amino acid sequence among Gc types could theoretically result in a qualitative difference between Gc1-MAF and Gc2-MAF. Quantitative differences could be caused by two mechanisms. First, the fact that Gc1 is glycosylated with both galactose and sialic acid, whereas Gc2 carries only the galactose residue,(5) is of interest, because the conversion of Gc into the Gc-MAF involves removal of galactose and sialic acid residues from Gc by the actions of β-galactosidase and sialidase enzymes associated with the membranes of B- and T-lymphocytes, respectively.(3) Second, the higher mean plasma level of Gc in women with Gc1-1 (272 ± 2 mg/liter) compared with Gc2-2 (226 ± 3 mg/liter) results in a larger quantity of substrate for Gc-MAF synthesis in women with Gc1-1, and in that way, perhaps a higher basal level of Gc-MAF. Our finding of a significant increase in fracture risk per 1-mg/liter rise in Gc concentration (OR = 1.006, p = 0.029) actually supports the substrate theory, but at the moment, we can only guess about the mechanisms.

Based on current knowledge about Gc-MAF combined with our present findings, we wanted to explore whether persons with Gc2-2 and low risk of bone fractures displayed a lower macrophage/osteoclast activity than did persons with the Gc1-1 phenotype. We looked for a marker of activated macrophages/osteoclasts and chose the newly described marker of macrophages, soluble CD163. Soluble CD163 is the plasma form of the macrophage receptor CD163 for clearance of haptoglobin-hemoglobin complexes.(15,16) To our knowledge soluble CD163 has never been evaluated as a marker of osteoclast activity but is regarded as a marker of macrophage activity/proliferation.(17) Highly elevated levels were found in patients with Gaucher disease, which is characterized by accumulation of tissue macrophages. Increased levels were also seen in patients with sepsis, myeloid leukemia, and arthritis.(16,18) Our finding of a significantly lower CD163 concentration in women with Gc2-2 compared with Gc1-1 could very well be because of differences in macrophage activity without any relation to osteoclasts, but because the influence of Gc phenotype is mediated by Gc-MAF, capable of activating both macrophages and osteoclasts, we assume the two cell types to be activated in parallel and therefore interpret the results to reflect lower osteoclast activity in Gc2-2 (those with the lowest fracture risk).

The inclusion of “age at enrollment” in the logistic regression analysis had only a minor effect on the ORs and p values of the other variables; however, “age at enrollment” turned out to have a significant protective effect on premenopausal fracture risk (OR = 0.90; 0.84-0.96). The protective effect is most likely because “age at enrollment” reflects “age at menopause.” We did not include the latter for two reasons: “age at menopause” is unknown in hysterectomized women (17.3% of the women), and “age at enrollment” and “age at menopause” are highly correlated and therefore cannot be in the same regression model. The time of menopause is determined by genetic and environmental factors. For instance, genetically inferior ovaries might result in lower and more inconsistent hormone production, causing weaker bones in the premenopausal period as well as early menopause. Smoking is an environmental factor affecting both bone tissue and the time of menopause. Smoking is thought to have a direct negative effect on osteoblasts and bone formation, but it also enhances the degradation of estrogen, thereby reducing plasma levels of estrogen and leading to early menopause. In our study, there was a tendency toward increased fracture risk in smokers compared with nonsmokers (OR = 1.34, p = 0.11). Perhaps other environmental factors can, like smoking, induce both weak bones and early menopause.

The very significant protective effect of forearm BMD found in our logistic regression model (OR = 0.41, p < 0.001) is in agreement with the fact that the risk of fracture is known to increase with decreasing BMD. However, despite highly significant differences in premenopausal fracture risk depending on Gc phenotypes, we found no differences among Gc types regarding BMD or bone markers measured at enrollment into the study, 3-24 months after last menstrual bleeding. At first this seems conflicting, but there are several possible explanations. First, the measurements were all cross-sectional, and the mean BMD might have been higher for women with Gc2-2 compared with Gc1-1 earlier in life. Second, the influence of Gc phenotype on premenopausal fracture risk could be independent of BMD. In a study of white American women, a history of hip fracture in their mother doubled the risk of hip fracture, after adjustment for BMD, indicating involvement of hereditary factors that are independent of bone mass.(19) In studies of the efficacy of various therapeutic agents used for treatment of osteoporosis, very different relations between BMD change and fracture risk have been found. Fluoride therapy resulted in large increases in BMD but only a small effect on fracture rates,(20) whereas raloxifene-treated patients had significantly lower vertebral fracture risk compared with placebo-treated patients for any percentage change (decrease or increase) in femoral neck or lumbar spine BMD.(21) In other words, the measured BMD changes were poor predictors of fracture risk reduction. The marked antifracture effects of antiresorptive drugs without major changes in BMD are believed to result from reducing the number and depth of resorption lacunae in trabecular bone, thereby preventing local stress zones with reduced biomechanical competence.

Besides BMD, the bone structure is also important for bone strength. For instance, Parfitt et al.(22) have demonstrated fewer trabeculae and increased separation between trabeculae in patients with fractures than in subjects of the same age without bone fractures. The theory is that the structural changes are because of trabecular perforations caused by active osteoclasts eroding thinner trabecular structures. Possibly, the Gc phenotype relates to bone structure. A paper supporting such a correlation was recently published. In a genome screen of 309 white sister pairs for quantitative trait loci underlying variation in femoral bone structure, Koller et al.(23) found evidence of linkage to chromosome 4q11-13, with a LOD score of 3.9 for femur neck axis length and a LOD score of 3.5 for femur midshaft width. The gene coding for Gc is located on chromosome 4q11-13. Unfortunately, they did not repeat the evidence in a larger sample, and with regard to chromosome 4, they conclude that “no definitive statement can be made about the presence or absence of femoral structure genes.”(24) In a very recent paper, Ezura et al.(25) suggested a combined effect of several single nucleotide polymorphisms (SNPs) within the Gc gene when testing for contribution to postmenopausal radial BMD in Japanese women.

In conclusion, our study reports that Gc phenotype may be an important predictor of premenopausal bone fracture in whites and presents data to support the hypothesis that Gc plays an active pathophysiological role in the activity of osteoclasts. Further studies are required to confirm the relationship to premenopausal fracture risk in other groups, including black Africans and other populations with different frequencies of the Gc phenotypes. Further studies are also needed to reveal whether the Gc phenotype is a predictor of postmenopausal fractures as well, and whether Gc phenotyping may identify subgroups of individuals in special need for preventive measures to diminish the risk of bone fractures.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Our study was supported by research grants from Th Maigaards Eftf Fru Lily Benthine Lunds Fond af 1/6-78, PA Messerschmidt og Hustrus Fond til udforskning og bekempelse af sygdomme, Beckett-fonden, Aarhus Universitetshospitals Forskningsinitiativ, and Institute for Experimental Clinical Research (Aarhus University, Aarhus, Denmark). We thank Mariann Thymann (Institute of Forensic Science, University of Copenhagen, Copenhagen, Denmark) for the gift of control serum samples of each Gc phenotype, Soren K Moestrup (Department of Medical Biochemistry, University of Aarhus, Aarhus, Denmark) for the gift of purified CD163, and Asger R Pedersen (Department of Biostatistics, University of Aarhus, Aarhus, Denmark) for statistical advice. We thank Irene Molbo, Inger Nygaard, Inger Marie Jensen, and Kirsten Bank for technical assistance.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES
  • 1
    Schmidtgayk H, Grawunder C, Tschope W, Schmitt W, Ritz E, Pietsch V, Andrassy K, Bouillon R 1977 25-hydroxy-vitamin-D in nephrotic syndrome. Lancet 2: 105108.
  • 2
    Schneider GB, Benis KA, Flay NW, Ireland RA, Popoff SN 1995 Effects of vitamin D binding protein-macrophage activating factor (DBP-MAF) infusion on bone resorption in two osteopetrotic mutations. Bone 16: 657662.
  • 3
    Yamamoto N, Kumashiro R 1993 Conversion of vitamin D3 binding protein (group-specific component) to a macrophage activating factor by the stepwise action of beta-galactosidase of B cells and sialidase of T cells. J Immunol 151: 27942802.
  • 4
    Braun A, Bichlmaier R, Cleve H 1992 Molecular analysis of the gene for the human vitamin-D-binding protein (group-specific component): Allelic differences of the common genetic GC types. Hum Genet 89: 401406.
  • 5
    Yamamoto N, Homma S 1991 Vitamin D3 binding protein (group-specific component) is a precursor for the macrophage-activating signal factor from lysophosphatidylcholine-treated lymphocytes. Proc Natl Acad Sci USA 88: 85398543.
  • 6
    Cleve H, Constans J 1988 The mutants of the vitamin-D-binding protein: More than 120 variants of the GC/DBP system. Vox Sang 54: 215225.
  • 7
    Kamboh MI, Ferrell RE 1986 Ethnic variation in vitamin D-binding protein (GC): A review of isoelectric focusing studies in human populations. Hum Genet 72: 281293.
  • 8
    Lauridsen AL, Vestergaard P, Nexo E 2001 Mean serum concentration of vitamin D-binding protein (Gc globulin) is related to the Gc phenotype in women. Clin Chem 47: 753756.
  • 9
    Mosekilde L, Hermann AP, Beck-Nielsen H, Charles P, Nielsen SP, Sorensen OH 1999 The Danish Osteoporosis Prevention Study (DOPS): Project design and inclusion of 2000 normal perimenopausal women. Maturitas 31: 207219.
  • 10
    Abrahamsen B, Gram J, Hansen TB, Beck-Nielsen H 1995 Cross calibration of QDR-2000 and QDR-1000 dual-energy X-ray densitometers for bone mineral and soft-tissue measurements. Bone 16: 385390.
  • 11
    Moller HJ, Hald K, Moestrup SK 2002 Characterization of an enzyme-linked immunosorbent assay for soluble CD163. Scand J Clin Lab Invest 62: 293299.
  • 12
    Brixen K, Nielsen HK, Eriksen EF, Charles P, Mosekilde L 1989 Efficacy of wheat germ lectin-precipitated alkaline phosphatase in serum as an estimator of bone mineralization rate: Comparison to serum total alkaline phosphatase and serum bone Gla-protein. Calcif Tissue Int 44: 9398.
  • 13
    Yamamoto N, Naraparaju VR, Orchard PJ 1996 Defective lymphocyte glycosidases in the macrophage activation cascade of juvenile osteopetrosis. Blood 88: 14731478.
  • 14
    Swamy N, Ghosh S, Schneider GB, Ray R 2001 Baculovirus-expressed vitamin D-binding protein-macrophage activating factor (DBP-maf) activates osteoclasts and binding of 25-hydroxyvitamin D(3) does not influence this activity. J Cell Biochem 81: 535546.
  • 15
    Kristiansen M, Graversen JH, Jacobsen C, Sonne O, Hoffman HJ, Law SK, Moestrup SK 2001 Identification of the haemoglobin scavenger receptor. Nature 409: 198201.
  • 16
    Moller HJ, Peterslund NA, Graversen JH, Moestrup SK 2002 Identification of the hemoglobin scavenger receptor/CD163 as a natural soluble protein in plasma. Blood 99: 378380.
  • 17
    Moller HJ, Aerts H, Gronbaek H, Peterslund NA, Hyltoft PP, Hornung N, Rejnmark L, Jabbarpour E, Moestrup SK 2002 Soluble CD163: A marker molecule for monocyte/macrophage activity in disease. Scand J Clin Lab Invest Suppl 62 (237): 2933.
  • 18
    Matsushita N, Kashiwagi M, Wait R, Nagayoshi R, Nakamura M, Matsuda T, Hogger P, Guyre PM, Nagase H, Matsuyama T 2002 Elevated levels of soluble CD163 in sera and fluids from rheumatoid arthritis patients and inhibition of the shedding of CD163 by TIMP-3. Clin Exp Immunol 130: 156161.
  • 19
    Cummings SR, Nevitt MC, Browner WS, Stone K, Fox KM, Ensrud KE, Cauley J, Black D, Vogt TM 1995 Risk factors for hip fracture in white women. Study of Osteoporotic Fractures Research Group. N Engl J Med 332: 767773.
  • 20
    Riggs BL, Hodgson SF, O'Fallon WM, Chao EY, Wahner HW, Muhs JM, Cedel SL, Melton LJ III 1990 Effect of fluoride treatment on the fracture rate in postmenopausal women with osteoporosis. N Engl J Med 322: 802809.
  • 21
    Sarkar S, Mitlak BH, Wong M, Stock JL, Black DM, Harper KD 2002 Relationships between bone mineral density and incident vertebral fracture risk with raloxifene therapy. J Bone Miner Res 17: 110.
  • 22
    Parfitt AM, Mathews CH, Villanueva AR, Kleerekoper M, Frame B, Rao DS 1983 Relationships between surface, volume, and thickness of iliac trabecular bone in aging and in osteoporosis. Implications for the microanatomic and cellular mechanisms of bone loss. J Clin Invest 72: 13961409.
  • 23
    Koller DL, Liu G, Econs MJ, Hui SL, Morin PA, Joslyn G, Rodriguez LA, Conneally PM, Christian JC, Johnston CC Jr, Foroud T, Peacock M 2001 Genome screen for quantitative trait loci underlying normal variation in femoral structure. J Bone Miner Res 16: 985991.
  • 24
    Koller DL, White KE, Liu G, Hui SL, Conneally PM, Johnston CC, Econs MJ, Foroud T, Peacock M 2003 Linkage of structure at the proximal femur to chromosomes 3, 7, 8, and 19. J Bone Miner Res 18: 10571065.
  • 25
    Ezura Y, Nakajima T, Kajita M, Ishida R, Inoue S, Yoshida H, Suzuki T, Shiraki M, Hosoi T, Orimo H, Emi M 2003 Association of molecular variants, haplotypes, and linkage disequilibrium within the human vitamin D-binding protein (DBP) gene with postmenopausal bone mineral density. J Bone Miner Res 18: 16421649.