Osteocalcin: An Endocrine Link Between Bone and Glucose Metabolism

Bu B Yeap

Expert Rev Endocrinol Metab. 2011;6(2):177-185.

Feasibility of Measuring ucOC in Large Cohort Studies

Measurement of TOC in large cohort studies can be conveniently achieved using automated immunoassays capable of detecting both intact osteocalcin (amino acids 1–49) and the major N-terminal fragment (amino acids 1–43) (Figure 2).^[35] This is preferable to an intact osteocalcin assay which may underestimate osteocalcin levels due to lability of the C-terminus and loss of amino acids 44–49. However, measurement of ucOC is less straightforward, contributing to the relative lack of ucOC data. A commercial antibody for ucOC is available (Takara Biomedical Group, Japan) but it is known to overestimate the large 1–43 and 4–43 fragments of ucOC.^[20] Therefore, hydroxyapatite binding has been the standard method for measurement of ucOC. For this, aliquots of serum are treated with hydroxyapatite, which binds carboxylated osteocalcin, allowing it to be precipitated by centrifugation, leaving ucOC from the supernatant available for detection (Figure 2). This method is technically demanding and labor intensive, hence, it is not widely available.^[20] Recently, Ferron et al. reported an ELISA-based assay for detection of osteocalcin carboxylation in mice.^[36] Murine osteocalcin has 46 amino acids and the glutamic acid residues (GLU) that can be γ-carboxylated (GLA) are at positions 13, 17 and 20.^[19,36] Using specific antibodies, Ferron et al. demonstrated in cultured osteoblasts and in mice that inhibition of γ-carboxylation by warfarin therapy resulted in a decrease in serum 13-GLA osteocalcin and a corresponding increase in serum 13-GLU osteocalcin.^[36] Therefore, 13-GLU osteocalcin, which comprised the large majority of ucOC, may be the biologically active form corresponding to ucOC measured by the hydroxyapatite-binding assay. However, assay properties of 13-GLU and 13-GLA osteocalcin-specific antibodies have not been validated in human sera. Establishment of a robust assay for human ucOC, either by standardization of the hydroxyapatite-binding assay or creation of immunoassays based on 17-GLU and 17-GLA osteocalcin-specific antibodies, would greatly improve the ability of researchers to assess the relevance of ucOC to human metabolism and health. Important questions to address are whether 17-GLU osteocalcin is the main form of ucOC in humans, and the degree to which carboxylation status of GLU at positions 21 and 24 modulate any metabolic effects of ucOC.

1 2 3 4 5 6 7 8 9 10 11

Next Section

Cite this: Osteocalcin: An Endocrine Link Between Bone and Glucose Metabolism - Medscape - Mar 01, 2011.

Table 1. Studies of total osteocalcin level in relation to insulin resistance, diabetes and metabolic syndrome in humans.
Study (year)	Study (type)	Results	Ref.
Im et al. (2008)	339 postmenopausal women (31 with Type 2 diabetes; X)	Serum TOC was lower in women with Type 2 diabetes vs controls (17.5 vs 22.2 µg/l; p < 0.005), and correlated inversely with HbA1c (r = −0.22, p < 0.001) and IR (r = −16, p = 0.003)	[23]
Zhou et al. (2009)	254 men (128 newly diagnosed Type 2 diabetes) and 180 postmenopausal women (92 with diabetes; X)	Serum TOC was lower in adults with Type 2 diabetes vs controls (15.1 vs 16.8 µg/l; p = 0.003)	[24]
Kindblom et al. (2009)	857 nondiabetic and 153 diabetic men (X)	Diabetic men had lower TOC (21.7 vs 27.8 µg/l; p < 0.001), TOC was inversely related to BMI, fat mass and fasting glucose (p < 0.001)	[25]
Kanazawa et al. (2009)	179 men and 149 postmenopausal women with Type 2 diabetes (X)	TOC correlated negatively with fasting plasma glucose (r = −0.24 for men, p = 0.001; r = −0.19 for women, p = 0.03) and HbA1c (r = −0.16, p = 0.03; r = −0.27, p =0.002). TOC correlated with total adiponectin in women (r = 0.30, p = 0.0003)	[26]
Fernandez-Real et al. (2009)	149 nondiabetic men (X) and 46 nondiabetic men and women (I)	Serum TOC correlated with insulin sensitivity (r = 0.23, p = 0.006) and total adiponectin (r = 0.19, p = 0.02). TOC was increased by dietary weight loss (16.8% of bodyweight) or weight loss (8.7%) and exercise	[27]
Pittas et al. (2009)	380 men and women (X), 198 (L), 5% with diabetes	Serum TOC inversely correlated with fasting glucose (β = −1.16, p = 0.01), insulin (β = −0.32, p = 0.006) and IR (β = −0.12, p = 0.002). Higher TOC was associated with a lower rise in fasting glucose over 3 years (β = −0.89; p = 0.03)	[28]
Saleem et al. (2010)	2493 men and women (X)	Serum TOC inversely correlated with BMI, fasting glucose, IR and leptin, and positively correlated with adiponectin (p < 0.001 for each variable). Higher TOC was associated with reduced odds of metabolic syndrome (highest vs lowest quartile OR: 0.33, 95% CI: 0.23–0.46 in African–Americans; OR: 0.43, 95% CI: 0.31–0.63 in non-Hispanic white individuals)	[29]
Yeap et al. (2010)	2765 older men with metabolic syndrome present in 797 (28.8%; X)	TOC level was inversely associated with waist circumference, glucose, triglyceride levels and IR (all p < 0.001), and was lower in men with metabolic syndrome (20.1 vs 21.4 µg/l; p = 0.002). OR (95% CI) for metabolic syndrome was 2.39 (1.71–3.35) for men with TOC <13.25 µg/l and 1.51 (1.07–2.13) for men with TOC 13.25–16.55 µg/l compared with men with TOC ≥30 µg/l	[30]

CI: Confidence interval; HbA1c: Glycated hemoglobin; I: Interventional study; IR: Insulin resistance; L: Longitudinal study; OR: Odds ratio; TOC: Total osteocalcin; X: Cross-sectional analysis.

Table 2. Studies of undercarboxylated osteocalcin, insulin sensitivity and diabetes control in humans.
Study (year)	Study (type)	Results	Ref.
Hwang et al. (2009)	199 men (X)	Higher ucOC level associated with greater insulin sensitivity (highest vs lowest tertile: HOMA-B 81 vs 65%, p < 0.05)	[31]
Shea et al. (2009)	348 nondiabetic men and women (M = 142, F = 206; X, L)	Lower total and carboxylated osteocalcin levels (not ucOC) associated with IR (p = 0.006 and p = 0.02, respectively), association attenuated by adjustment for adiponectin. Higher carboxylated osteocalcin level at baseline predicted less change in IR at 3 years, lower % ucOC predicted greater increase in IR	[32]
Kanazawa et al. (2009)	50 men and women with poorly controlled Type 2 diabetes (L)	After 1 month of improved glycemic control, TOC level increased (+1.9 µg/l, p < 0.001) and the ratio of ucOC/TOC decreased (-0.15, p < 0.01)	[33]
Kanazawa et al. (2011)	180 men and 109 postmenopausal women with Type 2 diabetes (X)	In adjusted analyses, log (ucOC) was inversely correlated with fasting glucose (β = −0.32, p < 0.001) and HbA1c (β = −0.22, p = 0.17) in men. No corresponding associations of ucOC were seen in postmenopausal women	[34]

B: β-cell function; F: Female; HbA1c: Glycated hemoglobin; HOMA: Homeostasis model assessment; IR: Insulin resistance; L: Longitudinal study; M: Male; ucOC: Undercarboxylated osteocalcin; X: Cross-sectional.

Comments

Commenting is limited to medical professionals. To comment please Log-in.

Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.