Degree of Mineralization-related Collagen Crosslinking in ... - CiteSeerX

Calcif Tissue Int (2006) 79:160 168 DOI: 10.1007/s00223-006-0035-1

Degree of Mineralization-related Collagen Crosslinking in the Femoral Neck Cancellous Bone in Cases of Hip Fracture and Controls Mitsuru Saito, Katsuyuki Fujii, Keishi Marumo Department of Orthopaedic Surgery, Jikei University School of Medicine, 3-25-8, Nishi-Shinbashi, Minato-ku, Tokyo, Japan

Received: 14 February 2006 / Accepted: 4 May 2006 / Online publication: 11 September 2006

Abstract. Based on the present definition of osteoporosis, both bone density and quality are important factors in the determination of bone strength. Collagen crosslinking is a determinant of bone quality. Crosslinks can form enzymatically by the action of lysyl oxidase or non-enzymatically, resulting in advanced glycation end products. Collagen crosslinking is affected by tissue maturation as well as the degree of mineralization. Homocysteine and vitamin B6 (pyridoxal) are also regulatory factors of collagen crosslinking. We elucidate the relationship between the degree of mineralization and collagen cross-links in cancellous bone from hip fracture cases. We also determined plasma levels of homocysteine and pyridoxal. Twenty-five female intracapsular hip fracture cases (78 ± 6 years) and 25 agematched postmortem controls (77 ± 6 years) were included in this study. Collagen crosslinking was analyzed after each bone specimen was fractionated into low (1.7 2.0 g/ml) and high (>2.0 g/ml) density fractions. The content of enzymatic (immature reducible and mature nonreducible cross-links) and nonenzymatic cross-link (pentosidine) were determined. In the controls, there was no difference in total enzymatic crosslinks between low and high density bone, while pentosidine content was significantly higher in high density bone. In the fracture cases, not only reduced enzymatic cross-links in high density bone and increased pentosidine in both low and high density bone, but also higher plasma homocysteine and lower pyridoxal levels were evident compared with the controls. These results indicate that detrimental crosslinking in both low and high mineralized bone result in impaired bone quality in osteoporotic patients. Key words: Cross-links — Degree of mineralization — Homocysteine — Vitamin B6 — Trabecular bone

Based on the present definition of osteoporosis, both bone density and quality are important factors in the determination of bone strength [1 3]. Although the definition of bone quality remains controversial, it is thought to encompass both structural and material All authors have no Conflict of interest policy. Correspondence to: M. Saito; E-mail: [email protected]

properties of bone. The proposed determinants of the material properties of bone are the degree of secondary mineralization of basic structure units (BSUs), microdamage accumulation, and collagen cross-link formation that is affected by rate turnover [4 7]. Collagen cross-links play important roles in the expression of bone strength [7 11] and the proper biological function of bone [12 17]. The cross-links of collagen can be roughly divided into two types: lysyl oxidase mediated cross-links (enzymatic cross-links) and advanced glycation end-products (AGEs; nonenzymatic cross-links). These two types vary by both the mechanism of formation and by functional differences. Enzymatic cross-links form in a two-stage process. Initially, immature cross-links, keto-imines, are formed via the action of lysyl oxidase and lysyl hydroxylases [14, 16]. The keto-imines then undergo a spontaneous reaction to form a trivalent mature pyridinium cross-link. It is well known that enzymatic crosslinking is indispensable for in the expression of proper strength of both collagen fibers and bone [8]. Therefore, reduction in enzymatic cross-link formation results in reduced bone strength [9, 10]. Enzymatic crosslink formation is contributed to have a positive effect on mechanical strength of bone within a beneficial level [7 10]. Several investigations concerning enzymatic cross-links have focused only on mature cross-links (pyridinoline; Pyr, deoxypyridinoline; Dpyr) without accounting for the presence of immature cross-links (dehydro-dihydroxylysinonorleucine; deH-DHLNL, dehydro-hydroxylysinonorleucine; deH-HLNL, dehydro-lysinonorleucine; deH-LNL). However, immature as well as mature cross-links may affect the physiological function of bone [10]. Immature cross-links are the most frequently observed forms in bone because the conversion rate of immature forms into mature forms is lower than that of other connective tissues, such as tendons, ligaments and articular cartilage [18 20]. These observations have led to the proposal that a simultaneous estimation of both immature and mature cross-links is important for elucidating the actual dynamic state of enzymatic cross-link formation.

M. Saito et al.: Collagen Cross-Links and Homocysteine

Advanced glycation end products (AGEs) such as pentosidine (Pen) are formed by nonenzymatic glycation [21] or oxidation reactions [22] resulting in tissue maturation [20, 23, 24]. In contrast to the positive effects bestowed by enzymatic cross-links, nonenzymatic crosslinks have detrimental effects on the mechanical [7, 21] and biological [22] functions of bone. Accumulation of Pen in bone is thought to deteriorate the mechanical properties of bone, particularly post-yield properties and toughness, consequently making collagen fibers brittle [7, 21]. Thus, AGE nonenzymatic cross-links are considered to be disadvantageous in bone. There is now evidence that a mildly elevated homocysteine plasma level in the general population is a common condition and this mild hyperhomocysteinemia may be a risk factor for hip fracture. This is independent to a reduction in bone mineral density [25, 26], indicating that hyperhomocysteinemia may have a detrimental effect on bone quality. Reynolds et al. [27] showed that hip fracture patients may be vitamin B6 (pyridoxal; PLP) deficient. Interestingly, homocysteine is known as an inhibitor of lysyl oxidase [28, 29], whereas vitamin B6 acts as an essential coenzyme of lysyl oxidase [30]. Thus, high levels of plasma homocysteine and low levels of pyridoxal may affect collagen cross-link formation in bone. To date, there has been little research into the relationship between plasma levels of homocysteine and vitamin B6 and collagen crosslinking in human bone. Not only collagen crosslinking, but also the degree of mineralization is an important determinant factor of bone quality [1, 5, 31, 32]. The mineralization process is comprised of primary mineral deposition of collagenous matrix on the calcification front and subsequent slow and progressive deposition called secondary mineralization [33]. The degree of secondary mineralization of each BSU depends on the lifespan regulated by rate of turnover [34]. Depending on the rate of turnover, bone tissue exhibits a heterogeneous degree of mineralization in different areas [33]. The material properties of newly formed bone may differ in tissue maturity from older bone. Because bone cell function and extracellular environment differ in osteoporotic patients from those in non-osteoporotic subjects, newly synthesized bone in osteoporotic patients may not necessarily be as well made as in non-osteoporotic subjects [22, 35]. Mineralization areas should be analyzed separately, because areas with different degrees of mineralization have distinctive tissue material properties. However, classical biochemical analyses of whole bone cannot be utilized to estimate the different degree of mineralization in certain areas. The density fractionation technique developed by Russell et al. [36] and refined by Grynpas et al. [37] is suitable to isolate BSUs of different mineralization stages to narrow the heterogeneity of bone samples. This technique may be useful to estimate distinctive collagen quantitative and qualitative changes in

161

areas with different degree of mineralization. Because both enzymatic and nonenzymatic cross-link formation events are post-translational processes, cross-link formation may be influenced by both tissue maturation time and secondary mineralization processes. Thus, we should assess separate fractions from each area, each fraction containing a sample with a different degree of mineralization. The aim of the present study is to evaluate correlations between the degree of mineralization and collagen cross-link formation in cancellous bone from intracapsular hip fracture cases compared to age- and gendermatched controls. Furthermore, plasma homocysteine and vitamin B6 (pyridoxal) concentrations were measured in order to understand the role of collagen crosslinks in determining bone quality. Materials and Methods Bone Specimens We used specimens from 25 female intracapsular hip fracture cases, aged 72 86 years (mean; 78 ± 6 years), and age- and gender-matched non-fractured control samples from 25 female subjects aged 69 89 years (mean; 77 ± 6 years). Specimens were obtained either from women who had undergone hemi-arthroplasty for intracapsular hip fracture (the fracture group), or from female cadavers within 10 hours after death. In order to exclude the effect of fracture healing processes, cases were operated within 2 4 days (mean, 2.4 days) and small cylindrical cancellous central portions of the femoral neck were sampled 5 mm away from the bone surface, which had been sawn during hemiarthroplasty. Neither the patients nor the control subjects were considered as being emaciated or obese, all with a mass index (BMI) between 21 to 25 kg/m2. The cases with hip fracture and control subjects were excluded who had medications (estrogen, bisphosphonates, and vitamins) and had osteoarthritis or preexisting conditions known to affect bone metabolism, such as diabetes, renal failure, and rheumatoid arthritis. Informed consent to the use of material for research was given at the time of surgery by the patients themselves, according to procedures required by an independent ethics committee and the hospital trust. Bone specimens of control subjects were also used in our previous study into age-related changes in the biochemical characteristics of collagen from human subjects [38] approved by an independent ethics committee with consent obtained from relatives. Density Fractionation Bone specimens were powdered in liquid nitrogen and sieved in a sonic sifter to isolate bone particle sizes below 20 lm. The recovery after grinding and sieving was 95% by weight. The bone powder was defatted with chloroform/methanol (2:1, v/v) for 48 hr then with methanol for 12 hr before being air-dried. The bone powder was then fractionated into three fractions corresponding to the osteoid (2.0 g/mL) density fractions in a bromoformtoluene mixture by the density fractionation method of Grynpas [39]. The lowest density fractionations (2.0 g/mL) and low

162

density (1.7 2.0 g/mL) bone fractions. We defined the boundary line between the low and high density fractions at 2.0 g/ml based on the reports by Sodek et al. [40]. They showed that the distribution of bone particles after density fractionation showed a peak at a density of 2.0 g/ml. By separating bone at that boundary line of 2.0 g/ml, we obtained sufficient bone from both low and high density fractions for triplicate analyses of collagen cross-links. Approximately 200 mg of sifted bone powder was added to a polyallomer tube containing 30 ml of 2.0 g/m density solution (calibrated with sink float). The bone powder solution was centrifuged at 10,000 g for 30 min. The density of the supernatant was then modified to 1.7 g/ml by the addition of toluene, and the new solution was recentrifuged. The contents of each fraction were recovered after adding 10 ml of toluene followed by centrifugation for 30 min at 6,000g. The amount of material in the fraction was weighed and then the relative proportion of low density fraction to total bone dry weight was calculated by weight of the fraction per total bone (w/w %). In an attempt to investigate whether toluene and bromoform affected the biochemical nature of bone collagen, two equal amounts of bone powder were compared; one being exposed to bromoform-toluene, the other not. 100 mg of human bone powder was kept in a bromoform-toluene mixture at 5C for 6 hr then subsequently centrifuged after addition of toluene, leaving a clear supernatant. The pellet was washed in toluene then lyophilized. The quantitative assessment of cross-links was performed as described below. It was found that the sample was identical to another 100 mg of bone powder treated in the same way which had not been treated with bromoform-toluene. Calcium and Phosphorus Content in Bone The calcium and phosphorus content of each fractionated bone specimen was analyzed by atomic emission spectroscopy (ICP-AES, Nippon Jarrell-Ash Co., Ltd, Kyoto, Japan) [42] after the organic matrix was removed by ignition and oxidation in a mixture of perchloric and nitric acid. Approximately 100 mg of each sample of fractionated bone powder was dried at 105C for 24 hr, weighed, then ashed at 700C for 5 hr before calcium and phosphorus concentrations were determined. The amount of calcium and phosphorus was expressed as percent of dry bone weight. Collagen Content in Bone Aliquots of bone specimens were weighed and hydrolyzed in 6 M hydrochloric acid at 110C for 24 h. The amount of collagen in bone was calculated by measuring hydroxyproline (Hyp) by high-performance liquid chromatography (HPLC), assuming that collagen weighed 7.5 times the measured Hyp weight, with a molecular weight of 300,000 [43]. Total collagen content was expressed as dry weight percentage from the original powdered bone tissue. Characterization of enzymatic and nonenzymatic collagen cross-links The reduction of collagen in bone with sodium borohydride (NaBH4) (Sigma-Aldrich, St Louis, MO, USA) and subsequent quantification of cross-links were carried out as previously described [20]. Briefly, each sample of bone powder was demineralized twice with 0.5 M EDTA in 50 mM Tris buffer, pH 7.4, for 96 hr at 4C. The demineralized bone residues were then sequentially suspended in potassium phosphate buffer, pH 7.6 (ionic strength = 0.15), and reduced at 37C with non-radioactive NaBH4. The reduced specimens were hydrolyzed in 6 M hydrochloric acid at 110C for 24 hr. Hydrolysates were then analyzed for cross-links using Shimadzu LC9 high-performance liquid chromatography (HPLC) fitted with a cation exchange column (0.9 · 10 cm, Aa pack-Na, JASCO, Ltd., Tokyo, Japan) linked to an on-line


Table 1. Calcium, phosphorus, and Ca/P distribution pattern in bone Calcium

Phosphorus

(mg/g of tissue dry weight) Controls Low density High density Fracture cases Low density High density

228.9 ± 8.5 248.1 ± 6.6b

Ca/P

104.8 ± 2.5 2.180 ± 0.028 112.7 ± 2.7b 2.204 ± 0.010b

200.0 ± 13.1a 91.4 ± 5.8a 2.188 ± 0.024 248.1 ± 8.6b 108.9 ± 5.3b 2.216 ± 0.031b

Values are expressed as mean ± SD a Indicates significant differences between the controls and the fracture cases within the same density (low or high) at P < 0.0001 b Indicates significant differences between low and high density fractions within the same individuals at P < 0.0001 fluorescence flow monitor (RF10AXL, Shimadzu, Shizuoka, Japan). We determined the content of enzymatic cross-links, such as immature reducible and mature nonreducible cross-links, and nonenzymatic cross-links (advanced glycation end-products). Lysyl oxidase mediated reducible immature cross-links (dehydro-dihydroxylysinonorleucine; deH-DHLNL, dehydro-hydroxylysinonorleucine; deH-HLNL, dehydro-lysinonorleucine; deH-LNL) were identified and quantified according to their reduced forms (DHLNL, HLNL, and LNL, respectively). Reducible immature cross-links and common amino acids, such as hydroxyproline (Hyp), were detected by O-phthalaldehyde derivatization using a post-column method, whereas enzymatic nonreducible mature crosslinks, such as pyridinoline (Pyr), deoxypyridinoline (Dpyr), and nonenzymatic glycation-induced cross-link, Pen, were detected by natural fluorescence. Contents of each cross-link were expressed as mmol/mol of collagen. Plasma Levels of Homocysteine and Vitamin B6 Blood examinations were done from subjects after a 12 hour fast and promptly centrifuged and stored at )80C. Serum homocysteine was measured as a fluorescence derivative by HPLC [44]. The levels of serum vitamin B6, pyridoxal, were determined by HPLC [45, 46]. Statistical analyses All values presented are means with their standard deviation (mean ± SD) in the text and the tables. Box-and-whiskers plots to show the distribution of variables were obtained for the parameter of cross-links in controls and fracture cases. Where the data were normally distributed, statistical significance was assessed using two-way analysis of variance (density vs. fracture group) with post hoc analyses adjustment for multiple comparisons using the JPM statistical software package (V.3.1.6, SAS Institute, Cary, NC). Where the data were not normally distributed, normalization by log transformation preceded the above and was in each case successful. Differences in the parameters of cross-links between low and high density fractions within each individual were analyzed using a paired t-test. Results Density Fractionation and Mineral Contents in Bone

There was no difference in the relative proportion of the low density fraction (1.7 2.0 g/ml) per total bone (w/w,


163

Table 2. Comparison of the contents of enzymatic and nonenzymatic cross-links between the controls and the fracture cases Enzymatic cross-links

Controls Low density High density Fracture cases Low density High density

Mature forms

Non-enzymatic cross-link

LNL

Pyr

Dpyr

Pen

233 ± 47 254 ± 27

65 ± 7 74 ± 18

170 ± 20 189 ± 31

56 ± 4 58 ± 7

251 ± 32 158 ± 36b

34 ± 8b 45 ± 9b

149 ± 33 132 ± 36a

36 ± 12b 28 ± 17b

Immature forms

Collagen contents (% of tissue dry weight)

DHLNL

HLNL

32.2 ± 2.2 29.8 ± 2.6

372 ± 84 344 ± 47

29.9 ± 3.0 31.4 ± 2.9

335 ± 124 320 ± 102

4.696 ± 1.772 7.723 ± 2.385c 19.134 ± 12.507b 15.801 ± 10.284a

Values are expressed as mean ± SD (mmol/mol of collagen) a,b Indicates significant differences between the controls and the fracture cases within the same density (low or high) at a P < 0.01, b P < 0.0001 c Indicates significant differences between low and high density fractions within the same individuals at P < 0.01

Fig. 1. Collagen cross-links in human femoral neck in the cases of intracapsular hip fracture and postmortem controls: (A) total enzymatic cross-link contents: the sum of immature (DHLNL, HLNL, and LNL) and mature (Pyr and Dpyr) cross-links, (B) nonenzymatic cross-link (Pen). Data are depicted as box-and-whisker plots showing medians, 25th and 75th quartiles and complete data range.

%) between the controls and the fracture cases (cases: 44.9 ± 15.6%, controls: 51.7 ± 24.0%; P = 0.5379). The calcium and phosphorus contents and the ratio Ca/P of the low density fraction were significantly lower than those of the high density fraction in both the controls (P < 0.0001) and the fracture cases (P < 0.0001) (Table 1). Thus, the high density fraction is considered to be more highly mineralized bone than the low density fraction, suggesting that fractionation by the density gradient method of Grynpas et al. [39] also reflects the degree of mineralization. The contents of calcium and phosphorus of the low density fraction from the fracture cases were significantly lower than that from the controls (P < 0.0001), while there were no significant differences in the calcium and phosphorus contents of the high density fraction between the controls and the fracture cases. Cross-links in Bone

There was no difference in collagen content between each bone fraction (mean values; 29.8 to 32.2%) (Table 2). In the controls, there was no significant difference in the content of the total enzymatic cross-links (the sum of DHLNL, HLNL, LNL, Pyr, and Dpyr) between low

(892 ± 138 mmol/mol of collagen) and high (917 ± 25 mmol/mol of collagen) mineralized bone (Table 2 and Fig. 1A). However, the content of nonenzymatic crosslink, Pen, was significantly higher in high mineralized bone than low mineralized bone (P = 0.0071) (Table 2 and Fig. 1B). In the fracture cases, a trend was observed (P = 0.070) of lower enzymatic cross-link content in high mineralized bone (684 ± 171 mmol/mol of collagen) compared to that of low mineralized bone (804 ± 142 mmol/mol of collagen) (Table 2 and Fig. 1A). Interestingly, in the fracture cases, even in low mineralized bone, the Pen content was elevated markedly to a similar extent to that of high mineralized bone (Table 2 and Fig. 1B). Comparing controls and the fracture cases, there was no significant difference in the total enzymatic crosslinks in low mineralized bone. In contrast, a significant reduction in enzymatic cross-links in high mineralized bone (P < 0.0001) from the fracture cases was observed compared to the controls (Fig. 1A). The Pen content of both low and high mineralized bone from the fracture cases was significantly higher than that of the controls (low; P < 0.0001, high; P = 0.0067, respectively) (Fig. 1B).

164


Fig. 2. Collagen maturation index: mature cross-links (Pyr+Dpyr) / immature cross-links (DHLNL+HLNL+LNL) (A), The ratio of nonenzymatic cross-link, Pen to total enzymatic cross-links (DHLNL+HLNL+LNL+Pyr+ Dpyr) (B) in human femoral neck in the cases of intracapsular hip fracture and postmortem controls. Data are depicted as box-andwhisker plots showing medians, 25th and 75th quartiles and complete data range.

Fig. 3. Plasma levels of homocysteine (A) and pyridoxal (B) in the cases of intracapsular hip fracture and controls. Data are depicted as box-and-whisker plots showing medians, 25th and 75th quartiles and complete data range.

In this study, we calculated the ratio of enzymatic total mature cross-links (Pyr+Dpyr) to total immature cross-links (DHLNL+HLNL+LNL), defined as the biochemical collagen maturation index [3, 13 15, 47] (Fig. 2A). The maturation index showed no significant difference between low and high mineralized bone fractions in the controls (P = 0.44) and the fracture cases (P = 0.84) (Fig. 2A). While enzymatic and nonenzymatic cross-links have distinctive and opposing functions in bone [7, 10, 23], we propose that the biochemical index of collagen crosslink profile should take both enzymatic and nonenzymatic cross-links into consideration. Enzymatic crosslinks are considered to be beneficial in bone, while they have a positive effect on the expression of proper strength of bone [8 10]. In contrast to the positive effects bestowed by enzymatic cross-links, nonenzymatic cross-links have detrimental effects on the mechanical [7, 21] and biological [22] functions of bone. Therefore, we calculated the relative ratio of nonenzymatic cross-link (Pen) content to total enzymatic cross-link content (the sum of DHLNL, HLNL, LNL, Pyr, and Dpyr) defined as the index of disadvantageous cross-link profile for mechanical properties of bone (Fig. 2B). This index was significantly higher in high mineralized bone fraction

than in low mineralized bone fraction (P = 0.0027) in the controls (Fig. 2B). Interestingly, the fracture cases showed a different trend from that of the controls for this index - this index was increased in low mineralized bone to a similar extent to that of high mineralized bone (P = 0.93) (Fig. 2B). When the controls and fracture cases were compared within the same density fraction, the ratio is significantly higher for fracture cases compared to controls in both the low (P = < 0.0001) and high (P = 0.006) mineralized bone (Fig. 2B). Plasma Homocysteine and Pyridoxal Concentration

As shown in Fig. 3, plasma homocysteine concentration was significantly higher (P = 0.037) in fracture cases (12.942 ± 3.381 nmol/ml) compared to controls (10.843 ± 1.853 nmol/ml). In contrast, a significantly (P =0.035) lower concentration of plasma pyridoxal was observed in fracture cases (6.522 ± 4.240 ng/ml) compared to controls (12.309 ± 5.750 ng/ml). Discussion

Our results support several previous reports that hypothesize such a relationship following in vitro or


animal studies [28, 30, 48, 49], a recent human case control study [27], and cohort studies [25, 26]. Density Fractionation of Human Bone

Our results of the calcium and phosphorus contents in low and high density fractions show that density dependent fractionation of human cancellous bone also reflect differences in the degree of mineralization (Table 1). These results are in agreement with previous reports in human cancellous bone [50, 51] and animals such as rat cancellous [52] and cortical bone [39, 50], pig cortical bone [40], and rabbit cortical bone [37]. In the fracture cases, cancellous bone is less mineralized than controls. The degree of mineralization is thought to be dependent on rate of turnover [1, 4, 5]. Therefore, the decrease in the degree of mineralization may be due to elevated bone turnover in the fracture cases. The other possible explanation for the reduction in degree of mineralization may be due to prolonged reversal phase of remodeling, or to a longer formation period, although further study will be required to elucidate the correlation between the degree of mineralization and bone remodeling status by bone histomorphometry. In the process of primary mineralization, enzymatic cross-link profile plays important roles in proper calcium deposition in collagenous matrix. Stereochemical and X-ray diffraction studies have revealed that differences in the molecular packing of collagen within fibrils are associated with differences in cross-link profiles [54]. Three-dimensional structures supporting effective mineralization may be created between intrafibrillar collagen molecules via the formation of specific cross-link formation that support proper mineralization [12 16]. For instance, the nucleation of calcium apatite crystals starts in gap regions, that is, in areas adjacent to crosslinking sites [55]. Thus, aberrant mineralization may be due to alterations in cross-linking patterns associated with changes in molecular packing. In contrast, AGEs formation is thought to induce the osteoblastic function via the interaction with receptor of AGEs [56]. Consequently, AGEs has an adverse effect on primary mineralization process. However, to date, it is not clear how enzymatic and nonenzymatic cross-link profile affects the time dependent mineral deposition in the process of the secondary mineralization. Thus, further study will be needed to elucidate the correlation between the collagen cross-link profiles and the secondary mineralization process in human bone. Enzymatic and Glycation Induced Cross-links

An excessive formation of enzymatic cross-links does not occur in the physiological mineralization process [14, 17] because the total quantity of enzymatic crosslinking in bone is controlled by the expression of lysyl oxidase. However, little is known regarding the spatial

165

distribution of enzymatic cross-links in bone. The series of recent reports by Paschalis and Boskey et al. [3, 47] investigated the spatial variation of collagen maturity. It was found that the spatial variation of collagen maturity within 50 lm of forming surfaces of cancellous bone was increased but, in the deeper zone away from the forming surface, values become distinguishable. These results indicate that collagen maturity reaches a maximum within the bone forming area. This evidence is supported by reports that conversion of immature to mature cross-links reaches a maximum during the primary mineralization process [14, 17]. However, a ‘‘deeper’’ zone away from the bone forming surface gradually increases its degree of secondary mineralization according to further maturation [5]. Thus, it is difficult to estimate maturity using the enzymatic crosslink ratio in the deeper zone. In this study, we showed that total collagen maturity estimated by the ratio of total mature to total immature cross-links (Fig. 2A) was found in a similar level in both low and high mineralized bone fraction in the controls. This suggests that crosslink maturation of the bone in the speculated deeper zone defined as having a bone density >1.7 g/mL may reach a maximum even if the degree of mineralization is different. Advanced glycation end products (AGEs) accumulate by nonenzymatic glycation with tissue maturation [20, 21, 23, 24]. It is known that Pen, one type of glycation induced cross-link in bone, accumulates with advancing age [7, 20, 38], whereas enzymatic crosslinking reaches a maximum during adolescence and subsequently stays in a similar range throughout adult life [18, 20, 38, 57]. The biochemical data reported here is the first evidence from human subjects that shows Pen accumulates more in high mineralized bone, which represents older bone, compared to low mineralized bone, which represents younger bone in the control subjects in spite of no difference in collagen maturity (Fig. 1B, 2A). It seems reasonable to assume that nonenzymatic cross-links accumulate in highly mineralized older bone in the controls because nonenzymatic glycation [21] and oxidation [22, 23] occur spontaneously with tissue maturation. Interestingly, the Pen content was higher in low-mineralized bone than that in highmineralized bone in fracture cases despite the fact that low-mineralized bone consists predominantly of younger bone. The mechanism of acceleration of non-enzymatic cross-link formation even in the younger bone from the fracture cases is not known at the present time. The collagen maturation index estimated by the ratio of enzymatic cross-links (Fig. 2A) did not differ when compared not only between the low and high mineralized bone in the fracture cases, but also between the fracture cases and the controls. This would suggest that enzymatic cross-link maturation is not impaired, but subsequent spontaneous nonenzymatic cross-link for-

166

mation is increased even in younger bone in the fracture cases. It is known that oxidation reactions are required at some stage in the formation of Pen [58]. Therefore, Pen is considered to be a glycoxidation product [58]. The fracture cases did not show hyperglycemia in this study, so the high accumulation of AGEs, Pen, even in the younger bone may be predominantly attributed to increased oxidation reactions rather than glycation induced by hyperglycemia. Interestingly, not only estrogen deficiency [35], but also hyperhomocysteinemia [59, 60] is thought to increase in oxidative stress. Thus, further work will be required to evaluate the correlation between the formation of AGEs such as Pen in bone collagen and the level of oxidative stress in bone matrix. An excessive formation of Pen in bone is thought to deteriorate the mechanical properties of bone, particularly post-yield properties and toughness, consequently making collagen fibers brittle [7]. On the other hand, Oxlund et al. [9] demonstrated in an animal study that impaired enzymatic cross-link formation in bone by the treatment of beta-aminopropionitrile, which is irreversibly inhibits the enzyme lysyl oxidase, results in the significant decrease in bone strength down to approximately 30% of the control animals. Thus, the enzymatic crosslinking has beneficial effects on proper bone strength without excessive brittleness. In this study, we also found impaired enzymatic crosslink formation in the fracture cases (Fig. 1A, Table 2). The mechanisms of reduction in both enzymatic immature and mature cross-links are not clear at the present time. In this study, a decrease in enzymatic cross-links was observed to a similar extent in the fracture cases without any change in collagen maturity (Fig. 1A, 2A). This finding indicates the initial step of crosslinking may be inhibited without disturbance of the conversion of immature cross-links to mature forms. This result is in accordance with the previous reports by Oxlund et al. [9] and Bailey et al. [61]. They showed a 20 44% reduction of enzymatic cross-links in osteoporotic cancellous bone. The first step of enzymatic cross-link formation is oxidative deamination of -amino groups of lysine and hydroxylysine residues located in telopeptides [17, 62]. Previous reports [9, 61] and our present data indicate that this process is inhibited by a certain negative regulator of lysyl oxidase. To date, the mechanism underlying the inhibition of collagen cross-linking in osteoporotic bone remains unknown. One candidate is AGEs formation. AGEs are thought to be formed between sugar and lysine residues, which are essential sites of enzymatic crosslinking on collagen, resulting in competitively inhibiting the formation of enzymatic cross-links [63]. Thus, the reduction in enzymatic cross-links in the fracture cases in our findings may be attributable to the accumulation of AGEs. Other candidates for inhibitors of lysyl oxidase have been proposed. Vitamin B6 (pyridoxal) is known to act as an indispensable co-factor of


lysyl oxidase [30]. Reynolds et al. [27] reported that hip fracture patients may be vitamin B6 (pyridoxal 5 phosphate; PLP) deficient. In an earlier study, we demonstrated that vitamin B6 deficiency in normal rats led to impaired enzymatic immature reducible cross-link formation [48]. Furthermore, Masse et al. [49] showed that vitamin B6 deficiency in chick resulted in reduction of bone strength. Thus, latent vitamin B6 deficiency may account for impaired enzymatic cross-link formation and consequent reduced bone strength in osteoporosis. Homocysteine also acts as a negative regulator of lysyl oxidase [28, 29]. People with mild hyperhomocysteinemia in the general population have an increased susceptibility to fracture, which is independent of bone mineral density [25, 26], indicating that hyperhomocysteinemia may reduce bone quality via impaired enzymatic cross-link formation in bone by inhibiting the action of lysyl oxidase. In this study, we measured the plasma concentration of pyridoxal and homocysteine. There was a reduction of pyridoxal concentration and an elevated homocysteine level for fracture cases compared to age-matched controls, although the possibility of a relationship between plasma pyridoxal and homocysteine levels and cross-link formation remains hypothetical at the present time from our results. In the future, the exact relationship between lysyl oxidase regulators and cross-link formation in human bone needs to be further elucidated. In conclusion, the degree of mineralization correlated with distinctive patterns of enzymatic and nonenzymatic cross-links in human bone. In fracture cases, reduced mineralization, impaired enzymatic cross-linking, and excessive non-enzymatic crosslinking may have led to increased fragility in the bone. Acknowledgments. The authors are grateful to Takaaki Tanaka Ph.D (NHO, National Utsunomiya Hospital, Tochigi, Japan), Ryuichi Fujisawa Ph.D (Hokkaido University, Japan) and Kazumi Hirakawa (Jikei University School of Medicine) for technical support. References 1. Burr BB (2002) Bone material properties and mineral matrix contributions to fracture risk or age in women and men. J Musculoskelet Neuronal Interact 2:201 204 2. Gearnero P, Delmas PD (2004) Contribution of bone mineral density and bone turnover markers to the estimation of risk of Osteoporotic fracture in postmenopausal wemen. J Musculoskelet Neuronal Interact 4:50 63 3. Paschalis EP, Shne E, Lyritis D, Skaeantavos G, Mandelsohn R, Boskey AL (2004) Bone fragility and collagen cross-links. J Bone Miner Res 19:2000 2004 4. Boivin GY, Meunier PJ (2003) The mineralization of bone tissue: a forgotten dimension in osteoporosis research. Osteopor Int 14 supple 3:S19 S24 5. Mashiba T, Mori S, Burr DB, Komatsubara S, Cao Y, Manabe T, Norimatsu H (2005) The effects of suppressed bone remodeling by bisphosphonates on microdamage accumulation and degree of mineralization in the cortical bone of dog rib. J Bone Miner Metab 23 Supple:36 42


6. Mori S, Harruff R, Ammrosius W, Burr DB (1997) Trabecular bone volume and microdamage accumulation on the femoral heads of women with and without femoral neck fractures. Bone 21:521 526 7. Wang X, Shen X, Li X, Agrawal CM (2002) Age-related changes in the collagen network and toughness of bone. Bone 31:1 7 8. Banse X, Sims TJ, Bailey AJ (2002) Mechanical properties of adult vertebral cancellous bone: correlation with collagen intermolecular cross-links. J Bone Miner Res 17:1621 1628 9. Oxlund H, Barckman M, Ortoft G, Andreassen TT (1995) Reduced concentrations of collagen cross-links are associated with reduced strength of bone. Bone 17:365S 371S 10. Oxlund H, Mosekilde L, Ortoft G (1996) Reduced concentration of collagen reducible crosslinks in human trabecular bone with respect to age and osteoporosis. Bone 19:479 484 11. Zioupos P, Currey JD, Hamer AJ (1999) The role of collagen in the declining mechanical properties of aging human cortical bone. J Biomed Mater Res 45:108 116 12. Kuboki Y, Kudo A, Mizuno M, Kawamura M (1992) Time-dependent changes of collagen cross-links and their precursors in the culture of osteogenic cells. Calcif Tissue Int 50:473 480 13. Saito M, Fujii K, Tanaka T, Soshi S (2004) Effect of lowand high-intensity pulsed ultrasound on collagen posttranslational modifications in MC3T3-E1 osteoblasts. Calcif Tissue Int 75:384 385 14. Saito M, Soshi S, Tanaka T, Fujii K (2004) Intensityrelated differences in collagen post-traslational modification in MC3T3-E1 osteoblasts after exposure to low and high intensity pulsed ultrasound. Bone 35:644 655 15. Saito M, Soshi S, Fujii K (2003) Effect of hyper- and microgravity on collagen post-translatonal controls of MC3T3-E1 osteoblasts. J Bone Miner Res 18:1695 1705 16. Uzawa K, Grzesik WJ, Nishiura T, Kuznetsov SA, Robey PG, Brenner DA, Yamauchi M (1999) Differential expression of human lysyl hydroxylase genes, lysyl hydroxylation, and cross-linking of type I collagen during osteoblastic differentiation in vitro. J Bone Miner Res 14:1272 1280 17. Yamauchi M, Katz EP (1993) The post-translational chemistry and molecular packing of mineralizing tendon collagens. Connect Tissue Res 29:81 98 18. Eyre DR, Dickson IR, Ness KV (1988) Collagen crosslinking in human bone and articular cartilage. Age-related changes in the content of mature hydroxypyridinium residues. Biochem J 252:495 500 19. Marumo K, Saito M, Yamagishi M, Fujii K (2005) The ‘‘ligamentization’’ process in human anterior cruciate ligament reconstruction with autogenous patellar and hamstring tendons. Am J Sports Med 33:1166 1173 20. Saito M, Marumo K, Fujii K, Ishioka N (1997) Single column high performance liquid chromatographic fluorescence detection of immature, mature and senescent cross-links of collagen. Anal Biochem 253:26 32 21. Vashishth D, Gibson GJ, Khoury JI, Schaffler MB, Kimura J, Fyhrie DP (2001) Influence of nonenzymatic glycation on biomechanical properties of cortical bone. Bone 28:195 201 22. McCarthy AD, Etcheverry SB, Bruzzone L, Lettieri G, Barrio DA, Cortizo AM (2001) Non-enzymatic glycation of a type I collagen matrix: effect on osteoblastic development and oxidative stress. BMC Cell Biol 2:16 23. Paul RG, Bailey AJ (1996) Glycation of collagen: the basis of its central role in the late complications of ageing and diabetes. Int. J Biochem Cell Biol 28:1297 1310 24. Sell DR, Monnier VM (1989) Structure elucidation of a senescence cross-link from human extracellular matrix, Implication of pentoses in the aging process. J Biol Chem 264:21597 21602

167

25. McLean RR, Jacques PF, Selhub J, Tucker KL, Samelson EJ, Broe KE, Hannan MT, Cupples LA, Kiel DP (2004) Homocysteine as a predictive factor for hip fracture in older persons. N Engl J Med 350:2042 2049 26. van Meurs JB, Dhonukshe-Rutten RA, Pluijm SM, van der Klift M, de Jonge R, Lindemans J, de Groot LC, Hofman A, Witteman JC, van Leeuwen JP, Breteler MM, Lips P, Pols HA, Uitterlinden AG (2004) Uitterlinden, Homocysteine levels and the risk of osteoporotic fracture. N Engl J Med 35:2033 2041 27. Reynolds TM, Marshall PD, Brain AM (1992) Hip fracture patients may be vitamin B6 deficient. Controlled study of serum pyridoxal 5 phosphate. Acta Orthop Scand 63:635 638 28. Khan M, Yamauchi M, Srisawasdi S, Stiner D, Doty S, Paschalis EP, Boskey AL (2001) Homocysteine decrease chondrocyte-mediated matrix mineralization in differentiating limb-bud mesenchymal cell micro-mass cultures. Bone 28:387 398 29. Lubec B, Fang-Kircher S, Lubec T, Blom HJ, Boers GHJ (1996) Evidence for McKusickÕs hypothesis of deficient collagen cross-linking in patients with homocystinuria. Biochim Biophys Acta 13:159 162 30. Bird TA, Levene CI (1982) Lysyl oxidase: evidence that pyridoxal phosphate is a co-factor. Biochem Biophys Res Commun 108:1172 1180 31. Grynpas MD (1993) Age and disease-related changes in the mineral of bone. Calcif Tissue Int 53 supple1:S57 S64 32. Turner CH (2002) Biomechanics of bone: determinants of skeletal fragility and bone quality. Osteoporos Int 13(2002):97 104 33. Jowsey J (1964) Variation in bone mineralization with age and disease. In: Bone Biodynamics, Little Brown, Boston, pp. 461 479 34. Meunier PJ, Boivin G (1997) Bone mineral density reflects bone mass but also the degree of mineralization of bone: Therapeutic implications. Bone 21:373 377 35. Muthusami S, Ramachandran I, Muthusamy B, Vasudevan G, Prabhu V, Subramaniam V, Jagadeesan A, Narasimhan S (2005) Ovariectomy induces oxidative stress and impairs bone antioxidant system in adult rats. Clin Chim Acta 360:81 86 36. Russell JE, Avioli LA (1972) Effect of experimental chronic renal insufficiency on bone mineral and collagen maturation. J Clin Invest 51:3072 3079 37. Grynpas MD, Hunter GK (1988) Bone mineral and glycosaminoglycans in newborn and mature rabbits. J Bone Miner Res 3:159 164 38. Saito M (1999) Age-related changes in biochemical characteristics of collagen from human weight-bearing and non-weight-bearing bone. Tokyo Jikeikai Medical Journal 114:327 337 (in Japanese) 39. Grynpas MD, Marie PJ (1990) Effects of low doses of strontium on bone quality and quantity in rats. Bone 11:313 319 40. Sodek KL, Tupy LH, Sodek J, Grynpas MD (2000) Relationship between bone protein and mineral developing porcine long bone and calvaria. Bone 26:189 198 41. Termine JD, Eanes ED, Greenfield DJ, Nylen MU (1973) Hydrazine deproteinated bone mineral. Calcif Tissue Res 12:73 90 42. Vukicevic S, Stavljenic A, Boll T, Vujicic G, Degenhardt C, Cervar M, Krempien B (1989) Effects of parathyroidectomy on tissue calcium, phosphorus, magnesium, and copper concentrations in aluminum-loaded uremic rats. Biol Trace Elem Res 22:45 53 43. Eyre DR, Koob TJ, van Ness KP (1984) Quantification of hydroxypyridinium crosslinks in collagen by High-Performance Liquid Chromatography. Anal Biochem 137:380 388 44. Araki A, Sako Y (1987) Detremination of free and total homocysteine in human plasma by high-performance liquid chromatography with fluorescence detection. J Chromatogr 422:43 52

168

45. Tani Y (1983) Assay methods of vitamin B6 1. Separation and determination of vitamin B6. Vitamins 57:263 271 (in Japanese) 46. Yoshida T, Yunoki N, Nakajima Y, Kaito T, Anmo T (1978) Simultaneous determination of vitamin B6 group in blood by high-performance liquid chromatography (HPLC). Yakuzai Zassi 98:1319 1326 (in Japanese) 47. Paschalis EP, Recker R, Dicarlo E, Doty SB, Atti E, Boskey AL (2003) Distribution of collagen cross-links in normal human trabecular bone. J Bone Miner Res 18:1942 1946 48. Fujii K, Kajiwara T, Kurosu H (1979) Effect of vitamin B6 deficiency on the crosslink formation of collagen. FEBS Lett 97:193 195 49. Masse PG, Rimnac CM, Yamauchi M, Coburn SP, Rucker RB, Howell DS, Boskey AL (1996) Pyridoxine deficiency affects biomechanical properties of chick tibial bone. Bone 18:567 574 50. Grynpas M (1993) Age and disease-related changes in the mineral of bone. Calcif Tissue Int 53 supple 1:S57 S64 51. Mbyu-Muamba JM, Dequeker J, Gevers G (1989) Collagen and non-collagenous proteins in different mineralization stages of human femur. Acta Anat 134:265 268 52. Simmons DJ, Russel JE, Geynpas MD (1986) Bone maturation and quality of bone material in rats flown on the space shuttle Spacelab-3Mission. Bone and Mineral 1:485 493 53. Boskey AL, Marks SC (1985) Mineral and matrix alterations in the bones of incisors-absent (ia/ia) osteopetrotic rats. Calcif Tissue Int 37:287 292 54. Paschalis EP, Paschalis V EP, Verdelis K, Doty SB, Boskey AL, Mendelsohn R, Yamauchi M (2001) Spectro-


55. 56.

57. 58. 59. 60.

61.

62. 63.

scopic characterization of collagen cross-links in bone. J Bone Miner Res 16:1821 1828 Yamauchi M, Katz EP (1993) The post-translational chemistry and molecular packing of mineralizing tendon collagens. Connect Tissue Res 29:81 98 Katayama Y, Akatsu T, Yamamoto M, Kugai N, Nagata N (1996) Role of nonenzymatic glycosylation of type I collagen in diabetic osteopenia. J Bone Miner Res 11:931 937 Fujii K, Kuboki Y, Sasaki S (1976) Aging of human bone and cartilage collagen: changes in the reducible cross-links and their precursors. Gerontology 22:363 370 Baynes JW (1991) Role of oxidative stress in development of complications in diabetes. Diabetes 40:405 412 Starkebaum G, Harlan JM (1986) Endothelial cell injury due to copper-catalyzed hydrogen peroxide generation from homocysteine. J Clin Invest 77:1370 1376 Eberhardt RT, Forgione MA, Cap A, Leopold JA, Rudd MA, Trolliet M, Heydrick S, Stark R, Klings ES, Moldovan NI, Yaghoubi M, Goldschmidt-Clermont PJ, Farber HW, Cohen R, Loscalzo J (2000) Endothelial dysfunction in a murine model of mild hyper homocyst(e)inemia. J Clin Invest 106:483 491 Bailey AJ, Wotton SF, Simes TJ, Thompson PW (1992) Post-translational modification in the collagen human osteoporotic femoral head. Biochem Biophys Res Commun 185:801 805 Robins SP (1982) Analysis of the crosslinking components in the collagen and elastin. Methods Biochem Anal 28:329 379 Reiser KM (1991) Nonenzymatic glycation of collagen in aging and diabetes. Proc Soc Exp Biol Med 196:17 29