Journal of Orthopardie RtTsrurch 10:774-783 Raven Press, Ltd.. New York 0 1992 Orthopacdic Research Society

Effect of Short-Term Hypomagnesemia on the Chemical and Mechanical Properties of Rat Bone Adele L. Boskey, Clare M. Rimnac, Manjula Bansal, *Micheline Federman, ?Jane Lian, and $Barbara D. Boyan The Hospitul for Special Surgery, New, York, New Yorh; *Neil' England Deaconess Horpital, Rorton; /University of Massachusetts, Amhersi, M a ~ s u c h u ~ e tund t ~ , i T h e University oj Texas Health Science Center ut Sun Antonio, San Antonio, Texas, U.S.A.

Summary: Magnesium is known to have an essential role in determining the properties oi'bone. but the way in which Mg exerts its actions remains unclear. Although long-term Mg deficiency is known to produce osteopenia, the effects of short-term Mg deficiency have not been established. To test the hypothesis that Mg deficiency results in an altered pattern of initial mineralization and concomitant altered bone properties, the radiographic, histologic, chemical, and mechanical properties of the bones of rats given a Mg-deficient diet were compared to those of rats pair-fed the same diet supplemented with Mg. Shortterm Mg-deficiency in the diet of growing rats produced a significant decrease in both the trabecular bone volume and the mineral content of the newly formed metaphysis, a significant increase in the Ca:P ratio, and a slight, but significant increase in hydroxyapatite crystallite size and/or perfection in the metaphysis. Comparable, but not significant, trends were found in the diaphyses. Metaphyseal bone osteocalcin levels were reduced in the Mg-deficient rats and lipid was more easily extracted from their bones. No detectable alterations in radiographic microstructure were noted. Mechanically, a significant decrease in the maximum three-point bend strength of the femurs of Mg-deficient rats was observed. These data support the hypothesis that short-term Mg deficiency affects the pattern of bone mineral formation. Key Words: Hypomagnesemia-Bone mechanical behavior-Bone lipids-Bone mineralOsteocalcin.

most often affected by Mg deficiency is bone, the major storage site for Mg in the body (1,14,30,47). Several investigators have documented a role for Mg in bone development by describing the effects of chronic Mg deficiency on animal bones. Both adult and young rodents fed Mg-deficient diets for extended periods exhibited marked osteopenia and complete cessation of growth in the proximal end of their tibia (5,28,33,38). In contrast, short-term Mg deficiency in growing animals was reported to result in increased bone mineral and Ca contents (21,36, 4 9 , increased radiographic density (16), and in-

Magnesium, the fourth most abundant element in the body, and the second most prevalent intracellular cation, is essential for normal metabolism, growth, and development (1). Deficiency of Mg (hypomagnesemia) may occur in response to a wide variety of nutritional deficiencies, malabsorption syndromes, hormonal imbalances, renal dysfunction, and chronic alcoholism (43). One of the tissues Received March 18, 1992; accepted May 28. 1992. Address correspondence and reprint requests to Dr. A. L. Boskey at The Hospital For Special Surgery. 535 East 70th Street, New York, NY 10021. U.S.A.

774

HYPOMAGNESEMIA AND BONE PROPERTIES

creased serum alkaline phosphatase activity (29). Mg deficiency in the rat also is reported to affect the activity of osteoblasts (17) and growth plate chondrocytes (16,19). Direct effects of Mg on the process of mineralization have been suggested from both in vitro and in vivo studies. In solution, Mg binds to the surface of apatite crystals (34), retarding both their formation and their growth (3,8,9,34,40,52). When rats are fed excess Mg, the mineral crystals in their bones are smaller than those in the bones of control pair-fed animals (15). Mg deficiency in the rat was found to be associated with calcium phosphate mineral within kidney tubulcs (26). This is an indication of the alterations in mineral homeostasis in thesc animals that is perhaps attributable to the deficit of a crystal growth inhibitor, i.e., Mg. Studies on the effect of short-term Mg deficiency on bone have been limited to measurement of bone weight and Ca and Mg contents (36,38,45). The composition and size of bone mineral crystals have not been described in either chronic or acute Mg deficiency. The process by which mineral crystals are deposited within the bone matrix involves a regulated series of events which includes both extracellular matrix formation and modification, crystal nucleation, and crystal growth. In rapidly growing bone tissues such as the metaphysis, nucleation may occur in association with lipid-rich matrix vesicles (2,53,54). This nucleation occurs via the interaction of Ca and inorganic phosphate with phospholipids to form Ca-acidic phospholipid-phosphate complexec (6). lntegral membrane proteins called proteolipids, enriched in the matrix vesicle membranes, are associated with these Ca-phospholipid-phosphatc complexes (11,13). Under physiologic conditions, nucleation may, in part, bc regulated by the presence of high concentrations of Mg in the matrix vesicles (1 l), since in solution Mg inhibits hydroxyapatite formation induced by either proteolipids or by preformed complexed acidic phospholipids (9,25). These observations suggest that short-term hypomagnesemia may markedly affect bone mineral formation in the growing animal by altering (a) the deposition of new mineral, (b) the interactions between matrix constituents and the mineral, and (c) the crystal size and perfection of the existing mineral. The objective of this study was to examine the effects of short-term Mg deficiency on initial mineralization in rat bone by the analysis of the composition of the mineral and of the organic matrix, the microstructure, and the mechanical behavior.

775

MATERIALS AND METHODS Experimental Animals

Twenty male Sprague-Dawley rats with initial weights of 155 -+ 3 g were allowed to acclimate to vivarium conditions for 4 days and were given water and standard lab Rat Chow (Wayne Rodent Blox) ad libitum. The animals, randomized into two equivalent groups, were housed in a 12 x 14 square foot room under constant ventilation at 25 ? 2°C with fluorescent illumination using a 14:lO h light: dark cycle. Animals were caged individually in polyurethane cages and fed either a Mg-deficient diet [Tcklad, Madison WI, U.S.A.; Diet no. 170490 (26)] containing a background level of 20 mg elemental Mgikg diet, ad libitum; or were pair-fed the Tcklad Mg-deficient diet to which 3.22 g MgSOJkg diet (650 ppm Mg) was added. Pair-feeding was essential since published studies (29) indicated that rats eat less of Mg-deficient diets and tend to be anorexic. Dextrose was adjusted in both diets so that animals would receive equivalent amounts. Animals in both groups received a complete vitamin mix. Dietary iron was supplied as ferric citrate. Animal weights and dietary intakes were recorded each morning. Animals were killed by decapitation on day 17 of the test diets. Seventeen days corresponds to the earliest time point after which serum Mg levels are reported to reach equilibrium in the Mg-deficient growing rat (36). Sample Preparation

Prior to death, blood was collected by cardiac puncture for analyses of the Ca (50), Mg (511, and Zn (37) contents of the serum. The lumbar vertebrae and all long bones were dissected and stored on ice. The vertebral bodies from all animals and five randomly selected right femora from the animal$ in each group were immediately fixed in 70% ethanol for histologic evaluation. The left femora were reserved for bone densitometry, fine focus radiography. and mechanical testing. All bones were stripped of soft tissue, including bone marrow, by sharp dissection and copious rinsing in physiological saline (pH 7.6). The cleaned bones were rinsed in ultrapure water, and divided into diaphyses, metaphyses, and epiphyses under a dissecting microscope. The metaphyses and diaphyses were cut into 2 X 4 mm bone chips, frozen in liquid nitrogen, lyophilized, and thcn powdered using a Spex

J Orrhop Res. Vol. 10, N o . 6 , 1992

A . L . BOSKEY ET AL.

776

freezer mill (Spex Industries, Metuchen NJ, U.S.A.). Except for samples for histology, all bones were stored at - 20°C before analysis. Bone Mineral Parameters To determine bone mineral content, aliquots of bone powder (10 mg) were dried at 110°C for 24 h. The dried samples were then ashed at 600°C for 24 h and the gravimetric yield of the ash determined. The ash was then dissolved in 0.1 ml 12 N ultrapure nitric acid and diluted to 2 ml with ultrapure water. Mg (51), Ca (50), and Zn (39) concentrations in the bone ash were determined by atomic absorption spectrophotometry. A second sample of ashed bone dissolved in 0.1 N HCI was used for analysis of Ca:P ratio. For these analyses, the phosphate content of the ash was determined colorimetncally (24) and the calcium content was determined by atomic absorption (50). Hydroxyapatite crystallite size and perfection were measured by wide angle x-ray diffraction using Cu K, radiation. Each 10-mg bone powder sample, mounted on a quartz crystal, was scanned using a Siemens powder diffractometer to demonstrate the absence of mineral phases other than hydroxyapatite. The line broadening measurement of the hydroxyapatite 002 (c-axis) reflection of each specimen was repeated five times using 30-min scans. Line broadening measurements were repeated on three separate aliquots of the same bone sample. Using the Scherrer equation (20), the line broadening parameter was used to estimate crystallite size in the c-axis direction, noting that crystal perfection also influences line broadening. Values reported are the means from triplicate determinations for 7-10 animals. Organic Matrix Parameters To test the effect of short-term Mg deficiency on the newly deposited bone, the chemical composition of the metaphysis was evaluated for osteocalcin [a specific Mg-binding bone matrix protein (49)] and for the presence of lipids. The content of the bone y-carboxyglutamic acid containing protein (osteocalcin) was measured by radioimmunoassay (22). Proteins were extracted from 2-mg aliquots of the ground metaphyseal bones, and assayed using an antibody specific for rat osteocalcin. Values reported are the means of

J Orrltop Res, Vol. 10, No. 6 , 1992

duplicate determinations from the tibia1 metaphyses of 10 animals. Lipids were extracted from the metaphyses using an adaptation (13) of the method described by Odutuga and Prout (35). This method was chosen to allow differentiation of cell-associated and mineralassociated lipids. Weighed and lyophilized bone powders (diameter, 50-450 pm)were extracted with 200 ml ch1oroform:methanol 2:1, v h , for 24 h at 4°C. This extract was separated from the bones by filtration through a Whatman 50 H filter and washed with 40 ml0.1 M NaCl overnight at 4°C. The resultant lower phase was collected, concentrated to near dryness by rotary evaporation, and then recovered in 5 ml ch1oroform:methanol 2: I , viv, containing 0.05% butylated hydroxytoluene (BHT). The samples were reextracted with 200 ml ch1oroform:methano1:concentrated HC1 200: 100:1, v/v, and washed with 40 ml 0.2 M NaHCO,. This resultant lower phase was concentrated, redissolved in 5 ml chloroform:methanol 2: 1 with BHT and stored. The samples were then demineralized by dialysis against 2 N formic acid until no Ca was present in the dialysate (50). After deacidification by copious rinsing with ultrapure water, the samples were extracted with ch1oroform:methanol 2: 1 , v/v, and ch1oroform:methanol:HCI 200: 100:1, v/v. The demineralized bones were analyzed for Ca (SO) to verify that decalcification was complete. Lipid yields at each step were determined gravimetrically. Histologic Analysis Histologic evaluation was conducted on femora and vertebral bodies fixed in 70% ethanol, embedded in methylmethacrylate, and sectioned u9ing a Jung sliding microtome. Sections were stained, alternatively, with hematoxylin and eosin, the Goldner stain, or the von Kossa stain (4). Sections from five animals in each group were examined. Growth plate thickness was measured in both the distal epiphyses and the vertebral bodies. Histomorphometric analyses of the trabecular bone volume of lumbar vertebrae were performed using an Image Processor (Image Technologies, Deer Park, NY, U.S.A.). The number of multinucleated giant cells/ section were counted to estimate osteoclast number. Bone Density The density of the midshaft of the left femur was measured using a Norland 278 single photon bone

777

HYPOMAGNESEMlA A N D BONE PROPERTIES

densitometer. For these analyses, the bones were placed in a polystyrene holder in a fixed orientation perpendicular to the source, with 1 cm of epiphysis extending beyond the holder. To account for variation in holder placement, bones were removed and repositioned in the holder and the measurements repeated three times. The mean value for each specimen was used to calculate the bone density for each group of 10 animals. Whole Bone Mechanical Behavior To examine the effect of Mg deficiency on whole bone mechanical behavior, three-point bend tests were conducted on the left femur from each animal. The tests were performed in ambient air on an MTS closed-loop servohydraulic test apparatus at a displacement rate of 12.5 mmis. Testing was performed in the mediolateral (ML) plane with a support span of 12.7 mm across the diaphysis. Load versus displacement of the loading ram was recorded on an oscilloscope. Specimens were kept wet throughout testing and were tested in random order. From the load versus deflection curves the following parameters were determined for each specimen: maximum strength (N), maximum deflection (mm), stiffness of the initial linear region of the load versus deflection curve (Nim), and energy absorption at maximum strength (Nam). To consider the influence of geometrical variation on whole bone mechanical properties, all specimens were radiographed before testing for measurement of endosteal and periosteal diameters in the ML plane. Measurements were taken in the diaphysis at the midpoint between the tip of the greater trochanter and the intercondylar notch. Comparison of the geometrical cross-sections of the control and Mg-deficient femora were then made. Data Analysis All parameters were evaluated for correlation with animal weight. Statistical comparison was

based on a two-tailed Student's t test. A p value of 0.05 was chosen for significance.

RESULTS Animal Weights The Mg-deficient animals exhibited a significantly smaller weight gain, when compared with the Mg-supplemented animals, even though these groups were pair-fed (Table 1). Serum Chemistry The animals fed the Mg-deficient diet exhibited a threefold reduction in serum Mg content. Diet had no effect on serum Ca content (Table 1). Serum Zn content was slightly, but not significantly, decreased in the Mg-deficient animal. Bone Mineral Parameters Mg-deficient animals showed significant alterations in metaphyseal bone mineral content and composition (Table 2). The total mineral content of the Mg-deficient metaphyses was decreased by 10% as compared with control metaphyses. Crystallite size and perfection of the Mg-deficient metaphyseal bone was significantly increased, and the Ca:P molar ratios tended to be higher. The Mg content of the Mg-deficient metaphyses was significantly reduced. Diaphyseal bone mineral content and composition, although, in general, showing trends similar to the metaphyseal bone, were not significantly affected by short-term Mg deficiency. Organic Matrix Parameters The osteocalcin content of the metaphyseal bone of the Mg-deficient animals was significantly less than that for the metaphyseal bone of the Mg-

TABLE 1. Eflect of hypomagnesemia on weight gain and serum cation content in growing rats Serum chemistry, mgidl

Group

Final weight, g

Weight gain, g

Mg + Mg

224 5 10" 244 ? 18

72 2 9" 94 i 7

~

Mg

0.6 2.0

2 ?

0.1" 0.3

Ca

Zn

10.2 f 1.0 10.4 -+ 0.7

1.2 ? 0.3 1.6 f 0.3

Each group consisted of 10 animals; values are means t SD; comparisons are based on Student's t test; - Mg; mg-deficient diet; + Mg, pair-fed Mg-replete diet. p < 0.05.

J

Orthop RL>s,Vol. 10, N O . 6,1992

778

A . L . BOSKEY ET AL.

TABLE 2. Effect of hypomagnesemin on mineral parameters o j metaphyseal and diaphyseal bone in gronVng rats Mg-deficient Mineral content (5%) Metaphysis Didphy sis CrystalIite size Metaphysis Diaphysis

53 2 4 66 t_ 4

Mg-replete 59 2 4 67 k 2

(A) 164 2 11" 152 ? 17

152 147

t_

7

*7

Ca:P (molar ratio) Metaphysis Diaphysis

1.59 5 0.05 1.60 0.06

*

1.50 2 0.02 1.63 0.06

Ca content (kg/mg bone) Metaphysis Diaphysis

226 2 17 332 2 40

229 2 13 295 I 30

Mg content (pgimg bone) Metaphysis Diaphysis

2.9 -C 0.9" 4.0 2 1.0

5.9 i 0.7 5.7 I 2.0

*

Values are means t_ SD for 7-10 determinations; statistical comparisons are based on Student's t test; for Ca:P analyses, the synthetic apatite standard was I ..59 0.04. p < 0.05.

*

replete animals (726 t 133 ngimg vs. 975 -+ 233 ng/rng bone, respectively). Prior to decalcification, more lipids were extracted from the Mg-deficient metaphyses (1 87 % 63 pgimg demineralized dry weight) than from the Mgreplete metaphyqes (163 5 44 pgimg demineralized dry weight). The lipid soluble extracts following decalcification contained 522 2 226 bglmg and 117 440 t.Lg/mg, respectively, for the Mg-deficient and Mg-replete bones. Based on percentage of total lipid in each of the four types of extracts (Table 3), there did not appear to be any significant differences in the overall distribution [undecalcified vs. mineral associated (decalcified)] of lipid-soluble materials in the bones. However, the lipid extracts from undecalcified Mg-deficient bones had a significantly higher percentage of lipids soluble in acidified as opposed to neutral solvents than did the undecalcified Mg-replete bones.

*

from the osteoblasts at a comparable site in a Mgreplete animal. Although the Goldner- and von Kossa-stained sections from vertebral bodies of Mgdeficient animals appeared comparable to those in Mg-replete animals (Fig. 2), they showed a significant decrease in trabecular bone volume when evaluated morphometrically. Based on comparison of multiple sections from vertebral bodies from five animals in each group, the trabecular bone volume in the Mg-deficient rats' lumbar vertebrae was 21.5 2 5.796, while that in the control animals was 32.6 5 1.5%. In the trabecular bone. the number of osteoclasts per unit volume was not significantly different when five sections from each of five animals in each group were compared. N o significant differences were noted when growth plate thickness in comparable sections was measured. Bone Density Radiographic diaphyseal bone density was significantly higher in the Mg-deficient animals (Table 4), although these bones did not differ significantly in either diaphyseal diameter or mineral content as determined by ash weight (Table 2). Although not significant, the more dense Mg-deficient diaphyses had a higher Ca content, and contained crystals of larger size than those in the Mg-replete animals (Table 2). Whole Bone Mechanical Behavior The femora from the Mg-deficient animals exhibited a significant decrease in maximum three-point bend strength as compared with the femora from the Mg-replete animals. No significant differences were found for maximum deflection, stiffness, or energy absorption. There was no significant differTABLE 3. Eflect of hypomagnesrmiu on lipids of metaphyseal bone of growing ruts Total lipid extract (96)

Histologic Analysis Few histologic differences among dietary groups were visible in hematoxylin and eosin- or Goldnerstained sections. Femoral and vertebral bone in the Mg-deficient animals appeared normal, with no histologic evidence of altered osteoblastic activities. Figure 1 shows the osteoblasts in the femur of a Mg-deficient animal. These did not appear different

J Orthop R e & , "01. 10. N o . 6, 1992

Lipid extract Undecalcified Ch1oroform:methanol Ch1oroform:rnethanol:HCI Total: Decalcified Ch1oroform:methanol Ch1oroform:rnethanol:HCl Values are means ' p < 0.05.

I SD

Mg-deficient

Mg-replete

28.8 1 7 . 0 29.5 t 12.5" 58.3 I 8.4

33.9 I14.3 22.3 t 3.6 56.2 -t 15.9

*

19.2 5.4 22.5 2 11.5

for 7-10 determinations

19.6 I 6.9 24.1 -+ 14.5

HYPOMAGNESEMIA AND BONE PROPERTIES

779

FIG. 1. Light micrograph (magnification ~ 2 5 showing ) the abundant osteoblasts lining trabecular bone in the Mg-deficient rat tibia. Goldner Stain. Reduced 35% for reproduction.

ence in measurements of either the endosteal or periosteal diameters (Table 4).

DISCUSSION Long-term Mg deficiency leads to severe osteopenia and growth retardation in rats (13,15,2729,46). Thc effccts of short-term Mg deficiency are less well documented (27,36,38). In the 1930s, investigators concerned with the effects of acute Mg deficiency reported that dietary depletion of Mg resulted in a decrease in the Mg contcnt of bone and an increase in bone weight and Ca content (27,36, 48), but techniques for more detailed analyses were not available. Later, animals fed a diet low in Mg and phosphate were noted to exhibit a slower weight gain than pair-fed controls (28), attributed to the reported requirement of Mg for protein synthesis (32). The effect of Mg deficiency on bone mineral composition, relative to control animals, was similar to that in the present study, with invariant ash weight, increased Ca content, and slightly, but not significantly, increased Ca:P ratios in the metaphyses (28). Although no formal mechanical evaluation was performed, the authors noted that the bones of the Mg-deficient animals were “grossly more brittle” than those of the control animals. The current study not only confirms the abovementioned changes in physiologic parameters and in bone mineral content seen in Mg-deficient animals, but also provides additional quantitative parameters (decreased mechanical strength, decreased osteocalcin, altered lipid composition)

which may explain the qualitative changes in bone fragility observed by Lai et al. ( 2 8 ) . Analysis of the mechanism by which chronic Mg deficiency causes osteopenia is complicated by the accompanying anorexia, hypoparathyroidism, and significant loss of organ function (18,30,40), which compromises any analysis of the effect of Mgrestriction alone on bone mineral formation. The 2.5-week period of Mg-deficiency was selected so that the effects on new mineral deposition in the metaphysis (41), as well as on diaphyseal bone properties, could be examined in the absence of the complications seen in chronically Mg-deficient rats. The analysis of serum mineral ion concentration suggests that the intent of the experimental design was met since only serum Mg content, and not serum Ca content, was significantly reduced. Bone Mg content was decreased but there were no concomitant changes in bone Ca or Zn. Carpenter et al. (17) have also noted that at 8 days, when Mg levels are low, and Ca levels are normal, there is no alteration in the parathyroid hormone (PTH) and vitamin D axis of Mg-deficient rats. Histologic analysis of the vertebrae and radiographic analyses of the long bones showed no evidence of the severe osteopenia expected from hypoparathyroidism or osteoporosis, which have previously been associated with long term Mg-deficiency. Mineral characteristics were, altered in the metaphyses of Mg-deficient rats in a manner analogous to that seen in animals in which bone turnover is accelerated (7,lO). Specifically, crystallite size and perfection were increased, mineral content was de-

J Orthop R E $ .Vol. 10, N O . 6 , 1992

780

A. L . BOSKEY ET A L .

A,B

FIG. 2. Light micrograph of von Kossa-stained transverse sections of trabecular bone in the vertebral body. Each micrograph was photographed in the center of the section, directly under the growth plate. A: Mg-deficient animal. B: Mg-replete animal. (Magnification x16; reduced 30% for reproduction.)

creased, and the Ca:P ratio was increased. This suggests that because of the Mg deficiency, bone mineral is being mobilized, probably in an attempt to raise serum Mg levels. The finding of similar trends for the mineral in the diaphysis supports this suggestion. However, the failure to find an increase in the number of osteoclasts indicates that mechanisms other than osteoclastic resorption may be important. The crystal size data could also be interpreted on a physicochemical basis. Because Mg is known to inhibit hydroxyapatite crystal growth in solution ( 3 , 34,40,52), crystal growth may be more extensive in a Mg-deficient milieu due to the absence of this crystal growth inhibitor. The presence of calcium phosphate deposits in the kidneys of rats chronically fed a Mg-deficient diet was interpreted on this basis (26). Similarly, the smaller crystal size of the mineral in rats given a high Mg-diet for 6 weeks was attributed to the excessive amount of crystal growth inhibitor present (15). One could predict, however,

J Orthop Res, Vol. 10. No. 6, 1992

that in the absence of the crystal growth inhibitor, more, rather than less, mineral would form. Yet the metaphyses of Mg-deficient animals in this study contained less mineral. The decreased amount of TABLE 4. Ejyect of hypomagnesemia on bone densitometvy and mechanical properties of the midshaft of femora from growing rats

Bone density (BMCibw) Max strength (N) Max deflection (mm) Stiffness (Nh) Energy absorption (N.m) ML periosteal diameter (mmj ML endosteal diameter (rnrnj

Mg-deficient

Mg-replete

0.252 2 0.009" 7 9 3 i 12.9" 4.5 1.1 31.2 t 7.4

*

0.211 k 0.002 105.0 2 19.3 4.4 i 1.2 36.8 6.0

0.020 2 0.007

0.016 i 0.007

*

3.68

2

0.19

3.70 i 0.25

2.62

2

0.21

2.60

* 0.24

Values are means 2 SD for 8-10 determinations; statistical comparisons are based on Student's t test; BMC, bone mineral content; bw, bone weight; ML, mediolateral . a p < 0.05.

HYPOMAGNESEMIA AND BONE PROPERTIES mineral in the metaphysis might arise from a failure of the cells to produce Mg-regulated hydroxyapatite nucleators, or from the impaired synthesis of matrix proteins that regulate mineral deposition. Thus, while in the absence of Mg, preformed crystals would grow more easily, the number of crystal nuclei might be reduced. The decreased bone osteocalcin content of the Mg-deficient rats reflects an altered bone matrix composition. This may be interpreted as being due to decreased new bone matrix synthesis. A recent study (17) of serum osteocalcin levels in growing rats showed decreased levels of osteocalcin in Mgdeficient, as compared with Mg-replete, animals. Decreased osteocalcin mRNA was also found in osteoblasts of Mg-deficient animals (17). Thus, the decreased osteocalcin accumulation in bone found in the present study is likely related to decreased synthesis. Alternatively, the decreased osteocalcin may be related to the increased crystal size, and consequent decreased surface area available for protein binding (23,49). Notably, this decrease occurred in the absence of severe osteopenia, as is seen in bones of animals on Mg-deficient diets for longer periods of time (33). There was no detectable histologic evidence of altered osteoblastic number or morphology in the present study. However, osteoblastic function may have been compromised, as indicated by the altered lipid composition as well as the decreased osteocalcin content. Other osteoblast products and matrix components may have similarly been altered. Histomorphometric evaluation of the vertebral bodies demonstrated that short-term Mg deficiency produced a significant decrease in trabecular bone volume. From the available data, it could not be determined whether this decrease in trabecular bone volume was due only to decreased matrix synthesis or was also a reflection of increased resorption. The number of osteoclasts per unit area in the bones of both the Mg-deficient animals and the Mgreplete animals was similar ( 3 4 per field), and at day 8 Mg-deficient animals were reported to have normal PTH and I ,25-dihydroxyvitamin D levels (17). Thus. decreased synthesis of matrix proteins may have been more prevalent than increased bone re sow tion. Furthermore, the alterations in lipid composition in the bones of the Mg-deficient rats suggests an alteration in cellular activity. Odutuga and Prout (35) and Shapiro (42) have indicated that cellular lipids are more easily extractable prior to decalcifi-

781

cation, whereas lipids associated with the mineral phase requiire extensive demineralization for extraction. The trend toward increased amounts of cell-associated lipid in the Mg-deficient metaphysea1 bones suggests that Mg might have some effect on lipid production. These changes may affect the formation of hydroxyapatite because alterations in lipid cornpotsition in the presence of Mg have been shown to modify the extent and distribution of mineral deposition in vitro and in vivo (6,11,13). A significant decrease in the maximum threepoint bend strength was found for the femoral diaphyses of the Mg-deficient animals. Jt is not clear why the other measured mechanical parameters, i.e., stiffness, maximum deflection, and energy absorption at maximum strength, were not also significantly affeci ed. However, in a preliminary study of the short-term effect of Mg deficiency on the torsional mechanical behavior of rat bone (12), the Mgdeficient rat femora also demonstrated a significant decrease in maximum torque, while torsional stiffness and angular displacement were not significantly affected. Simple geometrical measurements of the diaphyses were made to elucidate whether the effect of the Mg-deficient diet on the mechanical behavior of the rat femora was due to material or structural changes. No significant difference was found in mediolaterd plane measurements of the diaphyseal endosteal and periosteal diameters. While these measurements w,ere only taken in one plane and at one position, they suggest that the diaphyses of the femora from each group were geometrically similar. Therefore, the significant decrease in maximum bending strength of the Mg-deficient rat femora may be primarily due to material, and not structural changes. A discrepancy between changes in mechanical properties and changes in bone mineral composition have been noted by Martin (31). Martin examined the effect of short-term (1 and 2 week) simulated weightlessness via tail suspension on the threepoint bend mechanical behavior of rat femurs. In this model of osteopenia due to hypogravity, failure load and work to failure were significantly decreased after 2 weeks, while stiffness was not significantly changed. Comparable to the present study, Martin found no significant difference in percent mineralization (of the entire femur) between control and suspended rat femur groups. He concluded: “The effects of disuse on the mineralization of bone matrix, and the effects of the apatite-

J O i t h o p Rrs, Vol. 10,No. 6 . 1992

A . L . BOSKEY ET AL.

782

collagen microstructure on mechanical properties, are apparently both subtle and critical” (3 1). In the present study, there was a significant increase in mineral crystallite size in the metaphyses, and a trend toward increase in crystallite size in the diaphyses of the Mg-deficient rat femora. As in Martin’s study (311, it appears that the mechanical behavior of the diaphyses of the rat femora were also affected by subtle microstructural changes. Based on the significant changes in mineral properties found in the metaphyyes of the Mg-deficient rat femora, it could be predicted that the mechanical strength of the metaphyses were weakened even more than those of the diaphyses due to Mg depletion. The mechanism for the action of Mg on both new (metaphyseal) and older (diaphyseal) bone is likely to be an abnormal pattern of bone remodeling. However, from the data presented in this study, the effects of decreased new mineral deposition and increased resorption cannot be separated. Magnesium levels affect PTH activity in man and animals (181, and Mg has been shown to have an effect on the formation of vitamin D metabolites (39,44). Mg deficiency also affects osteoblastic activity (17). The aberrant bone turnover found in short term Mgdepletion is, therefore. not surprising. It is suggested that matrix synthesis, mineral production, cell membrane composition, as well as bone turnover, may be affected by Mg deficiency. These changes in turn, result in increased mineral crystal size, decreased trabecular bone volume, and as a result of these modifications, decreased mechanical strength . Acknowledgment:The research was supported by PHS Grants DE-05937, DE-05932, DE-04141, and AR-38007 and the Clark Foundation. The authors would like to acknowledge the technical assistance of Mr. William UIrich, Ms. Virginia Ramirez, and Mr. Luis DeLaCruz, as well as the support of Sister Mary Daniel Healey, former Director of the MBRS program at lncarnate Word College, San Antonio, Texas.

REFERENCES 1. Anast CS, Gardner DW: Magnesium metabolism. In: Disorders of Mineral Metabolism: Pathophysiology of Culciurn,

Phosphorus und Magnesium, ed by F Bronner, JW Coburn, New York, Academic Press, 1981, pp 423-522 2. Anderson HC: Matrix vesicle calcification: review and update. In: Bone and Minerul Research. ed by WA Peck, Amsterdam, Excerpta Medica. 1985, pp 109-149 3. Bachra BN, Fischer HKA: The effect of some inhibitors on

. I Orihop Rrs, Voi. 10, N U . 6 , 1992

the nucleation and crystal growth of apatite. Calcif Tissue Res 3:348-357. 1969 4. Baron R, Vignery A, Neff L, Silverglate A. Sanla Maria A: Processing of undecalcified bone specimens for bone histomorphometry. In: Bone Histomorphometry: Techniques and Internretation, ed by RR Recker, Boca Katon, CRC Press. 1983: pp 17-76 5 . Rernick DL, Hungerford GF: Effect of dietary magnesium deficiencv on the bones and teeth of rats. J Dent Res 44: 1317-1324, 1965 6. Roskey AL: phospholipids and calcification: an overview. In: Cell Mediated Calcification id Mairix Vesicles. ed by

s

Yousuf Ali, Amsterdam, Elsevier Science Publishers, 1986, pp 175-179 7. Boskey AL: Lipid changes in the hones of the healing vitamin D deficient, phosphate deficient rat. Merub Bone Dis Re/ Res 6 : 173-178. 1Y85 8. Boskey AL, Posner AS: Magnesium stabilization of amorphous calcium phosphate: a kinetic study. M a f Re.5 Bull 9: 907-9 16. 1974 9. Roskey A L , Posner AS: Effect of magnesium on lipidinduced calcification: an in vitro model for bone mineralization. Calcif Tissue Int 32:139-143, 1980 10. Boskey AL, Wientroub S: The effect of Vitamin D deficiency on rat hone lipid composition. Bone 7:277-281, 1986 1 I. Boyan B: Proteolipid-dependent calcification. In: Chemistry and Biology of Minernlized Tissues, ed by WF Butler, Birmingham. EBSCO Media, 1985, pp 12s-131 12. Boyan BD. Roskey AL. Rimnac C, DeLaCruz L: Effect of magnesium deficiency on the mechanical behavior of rat hone. Tritn.~32nd Annual ORS 11:295, 1986 13. Royan BD, Schwartz Z, Swain L, Khare A: Role of lipids in cartilage. Anut Rec 224211-219. 1989 14. Brommage R, Neuman KWF: Passive accumulation of magnesium, sodium. and potassium by chick calvaria. Calcif Tissrre Int 2857-63, 1979 15. Burnell JM, Liu C, Miller AG. Teubner E: Effects of dietary alteration of bicarbonate and magnesium on rat bone. Am J Ph\)siol 250:f302-f307, 1986 16. Bussell M, Vogel JJ, Meharg LS, Levy RM: Magnesium deficiency. I. The femur in acute magnesium deficiency. Pror Soc Exp B i d Med 139:387-389, 1972 17. Carpenter TO; Mackowiak SJ, Troiano N. Gundberg CM: Effects of brief magnesium deprivation on osteocalcin and osteocalcin message in the rat: relationship to abnormal development of the growing skeleton. Am J Physiol, in press 18. Cholst IN, Steinberg SF, Tropper PJ? Fox HE, Segre GV, Bilezikian JP: The influence of hypermagnesemia on serum calcium and parathyroid hormone levels in human subjects. New Engl J Med 310:1221-1225, 1984 19. Clark I. Belanger I,: The effects of alteration in dietary magnesium on calcium, phosphate and skeletal metabolism. CuC cif Tissue Res 1:204-218, 1965 20. Cullity BI): Elements of X-Ray DQJruction. Addison Wesley, Reading Mass. 1967, pp 261-263 21. Duckworth J. Godden W , WarnockGM: The effect of acute magnesium deficiency on hone formation in rats. Biochem J 39:97-108, 1940 22. Gundberg CM, Hauschka PV. Lian JB, Gallop MP: Osteocalcin: isolation. characterization. and detection. In: Methods qfEnzymology, Vol 107, ed by K Molddve, New York. Academic Press, 1980, 516544 23. Hauschka PV, Wians FH: Osteocalcin-hydroxyapatite interaction in the extracellular organic matrix of hone. Anal Rec 224:18%188, 1989 24. Heinonen JK, Lahti RJ: A new and convenient colorimetric determination of inorganic orthophosphate and its application to lhe assay of inorganic pyrophosphates. Anal Biochem 113:313-317. 1981

HYPOMAGNESEMIA AND BONE PRUPERlIES 25. Killian WF, Ennever J: Effect of magnesium on Bacterionernu mutruchoiti calcification. J Dent Res 54:185> 1975 26. KO KW, Fellers FX, Craig JM: Observations on magnesium deficiency in the Rat. Lab Invest 11:294-305, 1962 27. Kruse HD?Orent ER, McCollum EV: Studies on magnesium deficiency in animals. 1. Symptomatology resulting from magnesium deprivation. J B i d Chem 96:519-539, 1933 28. Lai CC, Singer L, Armstrong WD: Bone composition and phosphatase activity in magnesium deficiency in rats. J Bone Joint Surg 57A:516521, 1975 29. Loveless BW, Heaton FW: Changes in the alkaline phosphatase (EC 3.1.3.1) and inorganic pyrophosphatase (EC 3.6.1.1) activities of rat tissues during magnesium deficiency. The importance of controlling feeding pattern. Br J Nutr 36:487-495. 1976 30. Massry SG: Pharmacology of magnesium. Annu Rev Pharmacol Toxicol 17:67-82, 1977 31. Martin RB: Effects of simulated weightlessness on bone properties in rats. J Biomech 10:1021-1029. 1990 32. Menaker W, Kleiner IS: Effect of deficiency of magnesium and other minerals on protein synthesis. Proc Soc Exp Biol Med 81377-378, 1952. 33. Mirra JM, Alcock N W ? Shils ME, Tannenbaum P: Effects of calcium and magnesium deficiencies on rat skeletal development and parathyroid gland area. Magnesium 1:1633, 1982 34. Neuman V. Mulryan BJ: Synthetic hydroxyapatite crystals IV. Magnesium incorporation. C'alcif Tissue Res 7: 133-138, 1971 35. Odutuga A, Prout R: Lipid analysis of human enamel and dentine. Arch Orul Biol 19:729-731, 1974 36. Orenl ER, Kruse HI), McCollum EV: Studies on magnesium deficiency in animals. J B i d Chem 106:573-593, 1934 37. Parker MM, Humoller FL, Mahler DJ: Determination of copper and zinc in biological material. Clin Chem 33:4W7, 1967 38. Rayssignier Y . Lawar P: Mineral bone composition and some elements of calcium metabolism in magnesiumdeficient growing rats. Ann Biol Anirn Biochim Biophys 18: 57-166, 1978 39. Rude RK, Oldham SB, Sharp CF Jr, Singer FR: Parathyroid hormone secretion in magnesium deficiency. J Clin Endocrinol Metab 47:800-806. 1978

783

40. Salimi MH, Heughebaert JC, Nancollas GH: Crystal growth of calcium phosphates in the presence of magnesium ions. Lanpmrir. I : ] 19-122, 1985 41. Schenk KK: Basic histomorphology and physiology of skeletal growth. In: Treatment ofFrarturrs in Children and Adolescenls, ed by BG Weber, C Bruner, F Freuler, Berlin, Springer-Verlag, 1980, p 3 42. Shapiro IM: The association of phospholipids with inorganic bone. Cukif Tissue Int 5: 13-20. 1970 43. Shils ME: Magnesium in health and disease. Adv Rev Nutr 8:429460, 1985 44. Slatopolsky E , Mercado A , Morrison JA, Y a k s J, Klahr S: lnhibitory effects of hypermagnesemia on the renal action of parathyroid hormone. J Clin Zniwr 58:1273-1279, 1976 45. Smith BSW, Field AC: Effect of age on magnesium deficiency in rats. Br J Nutr 17:591-600, 1969 46. Tongyai S , Rayssiguier Y. Motta C, Gueux E, Maurois P, Heaton FW: Mechanism of increased erythrocyte membrane fluidity during magnesium deficiency in weanling rats. Arn .I Physiol 257:C27@C276, 1989 47. Wallach S: Effects of magnesium on skeletal metabolism. Mugnevium 9:l-14, 1990 48. Watchorn E, McCance KA: Subacute magnesium deficiency in rats. Eiochem J 31:1379-1390. 1937 49. Wians F H Jr, Strickland DM. Hankins GDV, Snyder RR: The effect of hypermagnesemia on serum levels of osteocalcin in an animal model. Magnesium 928-35, 1990 SO. Willis JB: The determination of metals in blood serum by atomic absorption spectroscopy. I. Calcium. Spectrochim Aciu 16:259-272, 1960 51. Willis JB: The determination of metals in blood serum by atomic absorption spectroscopy. 11. Magnesium. Spectrochim Aciu 16:273-283, 1960 52. Wilson JW. Werness PG. Smith LH: Inhibitors of crystal growth of hydroxyapatite: a constant composition appriach. d Urol 134:1255-1258. 1986 53. Wuthier RE: Electrolytes of isolated epiphyseal chondrocytes. matrix vesicles and extracellular fluid. Calcif Tissue Int 23:125-133, 1977 54. Wuthicr RE: Mechanism of de novo mineral formation by matrix vesicles, Connect Tissue Res 22:27-32, 1989

J Orthop Res, Vol. 10, N O . 6 , I992