Age- and genotype-dependence of bone material properties in the osteogenesis imperfecta murine model (oim)

Bone Vol. 29, No. 5 November 2001:453– 457

Age- and Genotype-dependence of Bone Material Properties in the Osteogenesis Imperfecta Murine Model (oim) B. GRABNER,1,2 W. J. LANDIS,3 P. ROSCHGER,1 S. RINNERTHALER,1,2 H. PETERLIK,4 K. KLAUSHOFER,1 and P. FRATZL2 1

Ludwig Boltzmann Institute of Osteology, Fourth Medical Department, Hanusch Hospital and UKH Meidling, Vienna, Austria Erich Schmid Institute of Materials Science, Austrian Academy of Sciences and University of Leoben, Leoben, Austria 3 Northeastern Ohio Universities College of Medicine, Rootstown, OH, USA 4 Materials Physics Institute, University of Vienna, Vienna, Austria 2

Introduction

Cortical mineralization of long bones was studied in collagen ␣2(I)-deficient mice (oim) used as a model for human osteogenesis imperfecta. Aspects of the age development of the mice were characterized by combining nanometer- to micrometer-scale structural analysis with microhardness measurements. Bone structure was determined from homozygous (oim/oim) and heterozygous (oim/ⴙ) mice and their normal (ⴙ/ⴙ) littermates as a function of animal age by small-angle X-ray scattering (SAXS) and quantitative backscattered electron imaging (qBEI) measurements. SAXS studies found anomalies in the size and arrangement of bone mineral crystals in both homozygous and heterozygous mice aged 1–14 months. Generally, the crystals were smaller in thickness and less well aligned in these mice compared with control animals. An increase in the mean crystal thickness of the bone was found within all three genotypes up to an age of 3 months. Vicker’s hardness measurements were significantly enhanced for oim bone (homozygotes and heterozygotes) compared with controls. The microhardness values were correlated directly with increased mineral content of homozygous and heterozygous compared with control bone, as determined by qBEI analysis. There was also a significant increase of mineral content with age. Two possibilities for collagen-mineral association are discussed for explaining the increased hardness and mineral content of oim/oim bone, together with its decreased toughness and thinner mineral crystals. As a consequence of the present measurements, one model for oim bone could incorporate small and densely packed mineral crystals. A second model for possible collagen-mineral association in oim material would consist of two families of mineral crystals, one being smaller and the other being much larger than the crystals found in normal mouse long bones. (Bone 29:453– 457; 2001) © 2001 by Elsevier Science Inc. All rights reserved.

The murine model (oim) of human osteogenesis imperfecta (OI) has a genetic mutation in type I collagen that results in the exclusive production of ␣1(I) collagen homotrimers in the homozygous (oim/oim) mouse.5 As a paradigm for the human disease, the oim homozygotes represent a moderate to severe form and heterozygotes a mild form of osteogenesis imperfecta.3,5,22,28 These alterations likely underlie a reduced torsional strength of the whole bone, which has been observed to decrease by about 40% for the bone of the mutants.3,22 The homozygous animals are smaller in body mass compared with heterozygous and normal littermates,22 and morphometric examination of the femurs has revealed lower area moments of inertia for the oim mice.3 The morphometric considerations alone, however, cannot explain the decreased strength of oim bone, suggesting a deficiency of the quality of bone material. Collagen molecular structure altered through a decreased axial order and a loss of crystalline lateral packing has been demonstrated by X-ray diffraction, ascribed to the genetic defect in ␣2(I) collagen in the oim mouse and correlated with human OI.21 This modified collagen was found to have a reduced resistance to tensile stress when compared with normal collagen fibers.23 Electron microscopic investigations of the Achilles tendon of such animals revealed unusual mineral crystals different in size, shape, and orientation from those of normal tissue. These crystals were described as large, dense blocks having poor alignment with the long axis of collagen, rather than small thin platelets highly aligned with the protein.14,16 Small-angle X-ray scattering (SAXS) has provided evidence that the mineral crystals comprising cortical bone of oim mice are thinner and less well aligned than those in normal bone.10 Changes in mineralization of the oim mouse have also been reported in work utilizing Fourier transform infrared (FTIR) microscopy in which a higher mineral:collagen-matrix ratio was measured for oim/oim compared with oim/⫹ and ⫹/⫹ bone.4 The present work reports an age-dependence study on mineralization and microhardness, in which several methods applied to the same specimens provide new information from nanometer to micrometer scales for three different age groups and genotypes. We propose collagen-mineral associations that relate the influence of both collagen and mineral to the decreased mechanical properties of oim bone.

Key Words: Osteogenesis imperfecta (OI); Murine model, Small-angle X-ray scattering (SAXS); Quantitative backscattered electron imaging (qBEI), Mineral density distribution; Microhardness.

Address for correspondence and reprints: Dr. Peter Fratzl, Erich Schmid ¨ sterreichischen Akademie der Wissenschaften, Jahnstrasse Institut der O 12, 8700 Leoben, Austria. E-mail: [email protected] © 2001 by Elsevier Science Inc. All rights reserved.

453

8756-3282/01/$20.00 PII S8756-3282(01)00594-4

454

B. Grabner et al. Age-dependency of oim bone material properties

Materials and Methods Bone specimens from a total of 62 mice of three genotypes, homozygous (oim/oim, 23 samples), heterozygous (oim/⫹, 25 samples), and littermate controls (⫹/⫹, 14 samples), were investigated by SAXS. Microhardness and electron microscopy measurements were carried out on 43 and 51 of these 62 specimens, respectively. The animals were also grouped by age in addition to genotype: 13 were 1 month old; 16 were 2–3 months old; and 15 were 6 – 8 months old. In addition, 18 samples from homozygous and heterozygous oim mice aged 10 –14 months were examined. Undecalcified samples of tibiae and femurs of the 62 mice were treated with ethylene glycol under vacuum and embedded in 100% Epon. The method is known to limit potential artifacts in the mineral of calcifying tissues prepared for microscopic analysis.20 Longitudinal sections of the bone samples were cut and ground to a thickness of 200 ␮m for SAXS.8,9,11 For quantitative backscattered electron microscopic analysis (qBEI) of the bone mineral density distribution (BMDD), the specimens were polished and carbon coated. Small-angle X-ray Scattering Studies SAXS measurements were performed within the corticalis of the diaphysis of the 62 tibiae and femurs of the mice to obtain the structure and alignment of composite bone mineral crystals. Cu-K␣ radiation of a rotating-anode X-ray generator (Bruker, AXS, Karlsruhe, Germany), operating at 36 kV and 100 mA, was used for SAXS analysis. Data were collected with an area detector (Bruker) and background-corrected. The two-dimensional scattering patterns obtained were analyzed as described previously11 and two parameters, T and ␳, were determined. The mean crystal thickness, T ⫽ 4 ␾ (1 ⫺ ␾) / ␴, was obtained from radially averaged SAXS patterns by a mathematical procedure based on Porod’s law,25 provided in detail by Fratzl.7 ␾ is the volume fraction of mineral in the bone tissue and ␴ the total surface of the mineral crystals per unit volume. This approach to such calculation for T assumes only that a two-phase system differing in electron density exists. This criterion is satisfied in a tissue such as bone where both organic and mineral components are present. For both needleand plate-like crystals, T is a measure of the smallest dimension of the crystals, and is therefore considered the thickness of the mineral crystals. The second parameter, ␳, is a measure of the degree of alignment of the crystals. It is equal to 1 if all crystals are exactly parallel and equal to 0 if the crystals are completely randomly oriented. The mathematical formulas used for its determination have been given in the appendix of the study by Fratzl et al.11 Microhardness Measurements and Quantitative Backscattered Electron Imaging Studies Vicker’s hardness measurements were performed by forcing a diamond pyramid (with an apex of 136°) into the bone sample surface. For this purpose the Epon-embedded and polished bone samples were mounted on a glass slide and the indented contact area was viewed and measured microscopically. At least five indentations, with a typical diagonal length ranging from 20 to 30 ␮m, at different sites in the central cortical bone area of each sample were made. The contact area was used to determine Vicker’s hardness (expressed in the SI-unit pascals [Pa]): Hv ⫽ P/contact area ⫽ 1.854 P/d2, where d is the mean diagonal length of the diamond-shaped indentation on the surface (meters), and P

Bone Vol. 29, No. 5 November 2001:453– 457

is the load (newtons). A mean standard deviation of 14% of microhardness within each sample was measured. In addition, the specimens were analyzed for BMDD within cortical bone areas by qBEI using a method described in detail previously.1,27,29 The samples were examined in a digital scanning electron microscope equipped with a four-quadrant semiconductor backscattered electron detector (Model DSM 962, Zeiss, Oberkochen, Germany). Five calibrated digital backscattered electron images of different areas (1 ⫻ 1.25 mm, including the regions of microindentations) of the corticalis of each sample were taken using electron beam energy of 20 keV, a beam current of 110 pA, a scanning speed of 100 sec/image, and a spatial pixel resolution of 2 ␮m. The gray levels were calibrated using carbon and aluminum as reference standards (atomic number contrast method) and were converted into calcium concentration values. Resulting histograms of different gray levels thereby represent the BMDDs of the specimens.26,27 The Ca concentration value of the histogram peak position (Camax) and the full width at halfmaximum (FWHM) of this peak were determined. Camax gives the typical calcium concentration (in wt%) (degree of mineralization) and FWHM the typical variation of the Ca content (homogeneity of mineralization). Statistics For statistical information, the mean values within each genotype and age group of animals were compared by a two-way analysis of variance (ANOVA) test, followed by pairwise comparison using a Tukey test. The statistical software package used was SIGMASTAT (Jandel Scientific, San Rafael, CA). Results Specimens from tibiae or femurs of the three genetic types of mice, the homozygous mutant, the heterozygous animal, and the cross-bred normal, all aged 1– 8 months, and additional samples from oim/oim and oim/⫹ animals, aged 10 –14 months (data from the latter animals not shown in the figures), were analyzed in the corticalis of the midshaft region. Values for mean thickness of the mineral crystals, their degree of alignment, the microhardness, and the calcium content and its variation are summarized in Figures 1–3. These data are presented in three age groups, 1 month, 2–3 months, and 6 – 8 months (see left diagram in Figures 1–3), and in the three aforementioned genotypic groups (see right diagram in Figures 1–3) for the animals. Results of statistical analysis are included in each figure. Two-way ANOVA was used to test for significance as a function of genotype (not differentiating between the age groups) and for significance correlated with age (not differentiating between the genotypic groups). With respect to animal age, Figure 1 shows that mean crystal thickness, T, increased significantly from the first (1 month) to either the second or third age grouping. There was no further statistical change in T for mice older than 2–3 months. The bone samples from homozygous and heterozygous oim mice aged 10 –14 months (data not shown in Figure 1) confirm this result. In addition, a genotype dependence for T could be detected: The mean mineral thickness was observed to be significantly smaller for oim/oim bone compared with oim/⫹ and ⫹/⫹. The degree of alignment, ␳, was found to be dependent on genotype. It increased significantly from a more random to a more parallel crystal arrangement in comparing oim/oim, oim/⫹, and ⫹/⫹ animals. No change with age could be detected for this parameter. A significant dependence of microhardness values on genotype, but not on age group, was observed (Figure 2). The

Bone Vol. 29, No. 5 November 2001:453– 457

Figure 1. SAXS measurements. The mineral thickness (a) and the alignment of the mineral crystals (b) are grouped by age (left) and genotype (right) (data expressed as mean ⫾ SE). Values at the top of each bar indicate the number of samples within a group. *p ⬍ 0.05; n.s., not significant in this and in Figures 2 and 3 based on statistical comparisons among groups of mice.

oim/oim and oim/⫹ bone was found to be harder than control bone at all ages; values for heterozygous animals lay between those of the other mice at all age groupings (with the exception of the 10 –14 month group, which showed a slightly but not significantly higher value for oim/⫹). Microhardness of oim/oim bone had an approximately constant value of about 1000 MPa. Backscattered imaging studies showed significantly increased values for calcium content in bone from homozygous mice compared with heterozygous and control animals (Figure 3). Camax (the typical calcium content in each sample) was highest for oim/oim and lowest for ⫹/⫹ mice, regardless of animal age. The values for the heterozygotes lay between those of the other

Figure 2. Microhardness measurements. Data (mean ⫾ SE) are grouped by age (left) and genotype (right). For sample data see Figure 1. *p ⬍ 0.05.

B. Grabner et al. Age-dependency of oim bone material properties

455

Figure 3. qBEI measurements. Typical calcium content Camax (a) and FWHM (b) (mean ⫾ SE). For sample data see Figure 1. *p ⬍ 0.05.

groups at all ages. Bone mineral content increased significantly from 1 month to 2–3 months. The corresponding Camax values generally increased over the whole age range for all genotypes. The width of the calcium peak (FWHM), which measures the variation of calcium concentration, did not change significantly with age or genotype. When microhardness was plotted as a function of Camax (Figure 4), there was significant correlation (R2 ⫽ 0.62) between these parameters (p ⫽ 0.012) throughout all age groups and genotypes, despite the fact that there was little variation of microhardness with age in the homozygous and heterozygous oim. Discussion Although the genetic defect in osteogenesis imperfecta concerns collagen-encoding genes,5 there are many indications that, in addition to defects in molecular packing21 and mechanical properties of tendon collagen,23 the bone material in the oim model has reduced biomechanical quality.3,4,10,22 In particular, indicators of lower maturity of oim bone have been reported, including reduced crystallinity and mineral carbonate as measured by FTIR3 or thinner and less-well-aligned mineral crystals, as obtained from X-ray scattering.10 Moreover, whereas the collagen content was found to be significantly reduced, the mineral:matrix ratio remained generally unchanged in the aforementioned FTIR study.3 In the present work, both age- and genotype-dependency were found for microhardness, mineral content, and microstructure of murine bone specimens. Microhardness and mineral content in homozygous animals were found to be significantly increased compared with their heterozygous and control counterparts regardless of age. The enhanced degree of mineralization was in good agreement with previous FTIR measurements of oim bone samples.4 On the other hand, the present results for mineral content are somewhat in contrast to data published for ash-

456

B. Grabner et al. Age-dependency of oim bone material properties

Figure 4. Microhardness as a function of the typical mineral content Camax. Results for the three age groups at 1 month, 2–3 months, and 6 – 8 months. R2 ⫽ 0.62 for linear regression; and p ⫽ 0.012. Circles indicate mean values of three different age groups of oim/oim (black), oim/⫹ (gray), and ⫹/⫹ (white). Data expressed as mean ⫾ SE.

weight analysis mentioned earlier.3 This difference between ash-weight and backscattered microscopic analysis could possibly arise from the fact that the two methods average over quite different volumes of material,31 and this may be a problem if the mineralization pattern is inhomogeneous. It is also of interest to note that classical “bone mineral density” (BMD) is decreased for oim bone,24 whereas the tissue mineral content is increased. This can only be explained by a considerably reduced bone volume fraction in oim. The SAXS measurements reproduced the same trend for the genotype-dependency reported previously for samples of the same age.10 In this study, the scattering experiments were performed to follow the age-dependence of thickness, T, and degree of alignment, ␳, of the mineral crystals. Both parameters effectively increased with animal age in each genotype sampled and decreased in the order from normal controls to heterozygotes to homozygotes within each age group of mice. As is the observation within each genetic group, T and ␳ typically increase with age from 1 to 2–3 months.8 Thus, the crystal thickness and its alignment with respect to collagen generally increase in the homozygous mutants with age, but these parameters lag behind those in their counterparts at all ages examined. However, for the heterozygous animals the difference in crystal size compared with controls was not significant at any age and the difference in the alignment showed a tendency to become smaller with age. It is most interesting that, on the one hand, smaller (T) and less-well-aligned (␳) crystals were measured in oim/oim bone, and, on the other hand, the mineral content (Camax) and microhardness were clearly greater in the same samples compared with heterozygotes and controls at each age grouping. Two different possibilities for collagen-mineral association in oim bone would be in accordance with these experimental results. First, one population of intrafibrillar crystals must be both more densely packed (to account for the BEI9 results) as well as thinner and less well aligned (to account for the SAXS results) than in normal murine bone. A second more likely possibility is that two families of mineral crystals contribute to the mineralization of

Bone Vol. 29, No. 5 November 2001:453– 457

oim bone. One family, presumably located principally within the collagen fibrils, would be smaller and less well aligned than normal because of defects in the collagen fibril structure.14,15,21 In addition to this type of intrafibrillar mineralization, electron microscopy has also provided evidence of different forms of extrafibrillar crystal aggregates in many calcifying tissues.6,15,18,30 In particular, large crystals or aggregates of small crystals were found in the mineralizing oim Achilles tendon, with such structures being absent in tendon from normal animals.14,16 These presumably larger and poorly aligned extrafibrillar crystals or crystal aggregates, which may be similar to the mineral found in vesicles of tendon,14,16 –18 putatively increase both mineral content and microhardness of oim/oim bone, which the present data show are significantly correlated. Because of the size of the large crystals (dimensions on the order of 400 ⫻ 125 ⫻ 60 nm as found in oim/oim tendon,14 they would not be detected by SAXS, a method that is insensitive to dimensions larger than ⬇50 nm. Hence, this second model, considering two populations of crystals (small intrafibrillar crystals and large extrafibrillar crystals or crystal aggregates being composed of small crystals), is also consistent with all experimental observations. Smaller crystals and aggregates of crystals have also been reported for human OI bone.30 There remains the question of why the oim/oim bone shows a greater degree of mineralization. Here one can only speculate: It is known that mechanical stimulation is an important factor for the formation of new collagen matrix and for its subsequent mineralization.12,19 In this context, ␣1(I) homotrimer is being produced by mutant animals in their usual ambulation but this collagen has inherently different structural and biomechanical characteristics compared to normal collagen. Such changes result in a reduced quality of mineralization associated with intrafibrillar collagen of oim/oim bone, which is perhaps related to similarly impaired abnormal mineralization of oim/oim tendon.14,15 On the other hand, bone, tendon, and other vertebrate calcifying tissues mineralize their extrafibrillar collagen volumes.6,18,20 As compensation for the decreased quality of ␣1(I) collagen homotrimers and their intrafibrillar calcification in mutant mice, the degree of overall bone mineralization may be enhanced through deposition of densely packed small, intra- and/or extrafibrillar crystals or, alternatively, by extrafibrillar deposition of larger, block-like crystals. This additional mineral deposited would increase the hardness and Young’s modulus of the oim bone. As a consequence, the resulting strains, caused by forces acting on the oim bone, are reduced and the oim bone collagen is protected against large strains. The advantage of such increased mineralization would be offset by an enhanced brittleness of oim bone. Although the oim mouse is a useful model for some forms of human OI, its relevance for describing the great variety of triple-helical mutations in human OI is probably limited. Remarkably, an increased mineral density has now been found for oim as well as for human OI,2,13 and has been linked in both cases to the brittleness of these tissues. Increased mineralization may be a common response for bone tissues with inherent weakness attributable to defective collagen. Indeed, the additional mineral stabilizes the tissue, but it does so at the cost of increased brittleness. Further studies of the origin of both intraand extrafibrillar mineral formation and of the complex interrelationships between mechanical stresses, collagen, mineral, and bone remodeling are needed.

Acknowledgments: The authors thank G. Dinst and J. Thorvig (Ludwig Boltzmann Institute of Osteology, Vienna, Austria) for preparing the bone samples.

Bone Vol. 29, No. 5 November 2001:453– 457

B. Grabner et al. Age-dependency of oim bone material properties

References 1. Boyde, A., Maconnachie, E., Reid, S. A., Delling, G., and Mundy, G. R. SEM in bone pathology: Review of methods, potential and applications. Scann Electron Microsc IV:1537–1554; 1986. 2. Boyde, A., Travers, R., Glorieux, F. H., and Jones, S. J. The mineral density of iliac crest bone from children with osteogenesis imperfecta. Calcif Tissue Int 64:185–190; 1999. 3. Camacho, N. P., Hou, L., Toledano, T. R., Ilg, W. A., Brayton, C. F., Raggio, C. L., Root, L., and Boskey, A. L. The material basis for reduced mechanical properties in oim mice bone. J Bone Miner Res 14:264 –272; 1999. 4. Camacho, N. P., Landis, W. J., and Boskey, A. L. Mineral changes in a mouse model of osteogenesis imperfecta detected by Fourier transform infrared microscopy. Connect Tissue Res 35:259 –265; 1996. 5. Chipman, S. D., Sweet, O. H., McBride, D. J., Davission, M. T., Marks, S. C., Schuldiner, A. R., Wenstrup, R. J., Rowe, D. W., and Shapiro, J. R. Defective pro␣2(I) collagen synthesis in a recessive mutation in mice: A model of human osteogenesis imperfecta. Proc Natl Acad Sci USA 90:1701–1705; 1993. 6. Christoffersen, J. and Landis, W. J. A contribution with review to the description of mineralization of bone and other calcified tissues in vivo. Anat Rec 230:435– 450; 1991. 7. Fratzl, P. Statistical model for the habit and arrangement of mineral crystals in the collagen of bone. J Stat Phys 77:125–143; 1994. 8. Fratzl, P., Fratzl-Zelman, N., Klaushofer, K., Vogl, G., and Koller, K. Nucleation and growth of mineral crystals in bone studied by small-angle X-ray scattering. Calcif Tissue Int 48:407– 413; 1991. 9. Fratzl, P., Groschner, M., Vogl, G., Plenk, H., Jr., Eschberger, J., FratzlZelman, N., Koller, K., and Klaushofer, K. Mineral crystals in calcified tissues: a comparative study by small-angle X-ray scattering. J Bone Miner Res 7:329 –334; 1992. 10. Fratzl, P., Paris, O., Klaushofer, K., and Landis, W. J. Bone mineralization in an osteogenesis imperfecta mouse model studied by small-angle X-ray scattering. J Clin Invest 97:396 – 402; 1996. 11. Fratzl, P., Schreiber, S., and Klaushofer, K. Bone and mineralization as studied by small angle X-ray scattering. Connect Tissue Res 34:247–254; 1996. 12. Frost, H. M. Bone “mass” and the “mechanostat”: A proposal. Anat Rec 219:1–9; 1987. 13. Jones, S. J., Glorieux, F. H., Travers, R., and Boyde, A. The microscopic structure of bone in normal children and patients with osteogenesis imperfecta: A survey using backscattered electron imaging. Calcif Tissue Int 64:8 –17; 1999. 14. Landis, W. J. The strength of a calcified tissue depends in part on the molecular structure and organization of its constituent mineral crystals in their organic matrix. Bone 16:533–544; 1995. 15. Landis, W. J. Tomographic imaging of collagen-mineral interaction: Implications for osteogenesis imperfecta. Connect Tissue Res 31:287–290; 1995. 16. Landis, W. J. Mineral characterization in calcifying tissues: Atomic, molecular and macromolecular perspectives. Connect Tissue Res 34:239 –246; 1996. 17. Landis, W. J., Hodgens, K. J., McKee, M. D., Nanci, A., Song, M. J., Kiyonaga, S., Arena, J., and McEwen, B. Extracellular vesicles of calcifying turkey leg tendon characterized by immunocytochemistry and high voltage electron mi-

18.

19.

20.

21.

22.

23.

24.

25. 26.

27.

28.

29.

30. 31.

457

croscopic tomography and 3-D graphic image reconstruction. Bone Miner 17:237–241; 1992. Landis, W. J., Hodgens, K. J., Song, M. J., Arena, J., Kiyonaga, S., Marko, M., Owen, C., and McEwen, B. F. Mineralization of collagen may occur on fibril surfaces: Evidence from conventional and high voltage microscopy and threedimensional imaging. J Struct Biol 117:24 –35; 1996. Landis, W. J., Librizzi, J. J., Dunn, M. G., and Silver, F. H. A study of the relationship between mineral content and mechanical properties of turkey gastronemicus tendon. J Bone Miner Res 10:859 – 867; 1995. Landis, W. J., Paine, M. C., and Glimcher, M. J. Electron microscopic observations of bone tissue prepared anhydrously in organic solvents. J Ultrastruct Res 59:1–30; 1977. McBride, D. J., Jr., Choe, V., Shapiro, J. R., and Brodsky, B. Altered collagen structure in mouse tail tendon lacking the alpha 2(I) chain. J Mol Biol 270:275–284; 1997. McBride, D. J., Jr., Shapiro, J. R., and Dunn, M. G. Bone geometry and strength measurements in aging mice with the oim mutation. Calcif Tissue Int 62:172–176; 1998. Misof, K., Landis, W. J., Klaushofer, K., and Fratzl, P. Collagen from the osteogenesis imperfecta mouse model (oim) shows reduced resistance against tensile stress. J Clin Invest 100:40 – 45; 1997. Phillips, C. L., Bradley, D. A., Schlotzhauer, C. L., Bergfeld, M., LibrerosMinotta, C., Gawenis, L. R., Morris, J. S., Clarke, L. L., and Hillman, L. S. Oim mice exhibit altered femur and incisor mineral composition and decreased bone mineral density. Bone 27:219 –226; 2000. Porod, G. Die Ro¨ ntgenkleinwinkelstreuung von dichtgepackten kolloiden Systemen. Kolloid Zeitschrift 124:83–114; 1951. Roschger, P., Fratzl, P., Eschberger, J., and Klaushofer, K. Validation of quantitative backscattered electron imaging for the measurement of mineral density distribution in human bone biopsies. Bone 23:319 –326; 1998. Roschger, P., Fratzl, P., Klaushofer, K., and Eschberger, J. A new scanning electron microscopy approach to the quantification of bone mineral distribution: Backscattered electron image grey-levels correlated to calcium K␣-line intensities. Scann Microsc 9:75– 88; 1995. Saban, J., Zussman, M. A., Havey, R., Patwardhan, A. G., Schneider, G. B., and King, D. Heterozygous oim mice exhibit a mild form of osteogenesis imperfecta. Bone 19:575–579; 1996. Skedros, J. G., Bloebaum, R. D., Bacus, K. N., and Boyce, T. M. The meaning of graylevels in backscattered electron images of bone. J Biomed Mater Res 27:47–56; 1993. Traub, W., Arad, T., Vetter, U., and Weiner, S. Ultrastructural studies of bones from patients with osteogenesis imperfecta. Matrix Biol 14:337–345; 1994. Vajda, E. G., Skedros, J. G., and Bloebaum, R. D. Errors in quantitative backscattered electron analysis of bone standardized by energy-dispersive X-ray spectrometry. Scanning 20:527–535; 1998.

Date Received: March 15, 2000 Date Revised: January 22, 2001 Date Accepted: May 4, 2001