6.1.1. Measuring bone resorption

The method used to monitor bone resorption is based on urinary excretion of ³H radioactivity from [³H];tetracycline-prelabeled rats. This method has been described by Klein and Jackman (1976), and further developed by Mühlbauer and Fleisch (1990). Tetracycline forms strong complexes with bone mineral once it incorporated into the bone (Ibsen & Urist 1964), and is removed only during resorption (Frost 1965). The use of multiple prelabeling of rapidly growing rats permits homogenous distribution of [³H];tetracycline throughout the bones (Klein & Jackman 1976).

The validity of the method used is dependent on the stability of tritium in [³H];tetracycline. Kelly and Buyske (1960) have demonstrated that [³H];tetracycline does not undergo metabolic transformation in rats. Furthermore, Klein and Reilly (1976) have found that the molecular exchange with nonradioactive tetracycline in vitro does not induce any significant loss of tritium as water from the [³H];tetracycline of bones of [³H];tetracycline-prelabeled rats.

The efficiency of renal excretion of [³H];tetracycline that is removed during bone resorption assures that the ³H, unlike isotopes of calcium, will be only minimally reused at new sites of bone formation (Klein et al. 1985). Furthermore, the [³H];tetracycline released from bone is unbound or poorly bound to apatite, and is thus not reincorporated into the newly formed bone (Klein & Wong 1986, Mühlbauer & Fleisch 1990).

The use of this method was preferred also because, unlike pyridinium crosslinks which are continuously generated in bone, the excretion of [³H];tetracycline stems only from the bone formed during the labeling period, without a possible confusing effect of subsequent bone formation (Egger et al. 1994).

The possible confusing effect of tetracycline in the study protocol must not be ignored, because tetracyclines are known to affect bone metabolism in many ways. For example, minocycline given 10-20 mg/day orally has shown to protect against cancellous bone loss and to maintain a normal bone formation rate in ovariectomized aged rats (Williams et al. 1996) and in streptozotocin-diabetic rats (Bain et al. 1997). Doxycycline has shown to reduce extracellular matrix breakdown, including bone loss, in adult periodontitis during long-term clinical trials (for review see Ryan et al. 1996). Furthermore, chemically modified tetracyclines have been discovered to be potent inhibitors of severe classes of matrix metalloproteinases, preventing collagen breakdown and bone loss in a variety of animal models (for review see Ryan et al. 1996). In the [³H];tetracycline studies, however, the risk for the tetracycline-caused confusion is minimized by using very low doses of tetracycline. Furthermore, in the present studies, all the groups were given identical tetracycline doses. The possible influence of tetracycline should thus be similar in all of the groups, and should not affect the comparisions made between the different groups.

In the present studies, the initial ³H measurement of each rat served as a baseline value for their individualized patterns of ³H excretion. This was done, instead of using mean values at each time point, to avoid possible confusing effects caused by the biological variations between the animals. The urine collection period was always 24 hours, to avoid possible confusion caused by the diurnal rhythm, and the possible confusing effect of different urine volumes between the animals was eliminated by calculating the amount of ³H radioactivity in proportion to the whole daily volume of urine. Furthermore, the bone resorption values were confirmed by measuring the amount of ³H radioactivity preserved in bones at the end of the experiment.

6.1.2. Measuring bone biomechanics

Of the biomechanical methods used in the present studies (IV and V), the three-point bending and torsional tests of long bones have been used as indicators of cortical bone strength (Saville 1969, Strömberg & Dalen 1976, Paavolainen 1978, Danielsen et al. 1992). The loading test of femoral neck measures also the properties of trabecular bone (Hou et al. 1991, Peng et al. 1994b), although the proportion of cortical bone in the femoral neck in rats is much higher than in humans (Bagi et al. 1997).

The results of the present studies, concerning bone strength properties of healthy and ovariectomized rats, were in accordance with previous studies performed either with the same testing machine (Lepola et al. 1993, Peng et al. 1994b), or like equipment (Strömberg & Dalen 1976, Paavolainen 1978), indicating valitidy and repeatability of the present measurements. The precision, accuracy and reproducibility of the testing machines used have been evaluated by Jämsä et al. (1996, 1998).

It should be noted that it is not possible to fully account for geometric effects in whole bone tests. However, unlike the values of strength and stiffness, the values of stress, strain, Young"s modulus, shear stress and shear modulus of elasticity take into account bone geometric properties, and thus represent the intrinsic strength of bone materia. In the present study the accuracy of derivations was improved by measuring the actual cross-sectional areas of the bones by a computerized planimeter.

The bone biomechanical parameters are influenced by the histological structure of cortical and trabecular bone, as well as by the biochemical composition of bone mineral and organic structures. As a consequence, the biomechanical properties of bone can not be entirely understood without knowledge of the existing background variables. To get a more reliable overall view, in the present studies, measurements were made also concerning the effects of dietary xylitol supplementation on bone weight, bone density, bone concentrations of calcium and phosphorus, trabecular bone volume, and on the amounts of bone collagen and its pyridinoline crosslinks.