Full text of DOI:10.1359/jbmr.2001.16.2.206

JOURNAL OF BONE AND MINERAL RESEARCH
Volume 16, Number 2, 2001
© 2001 American Society for Bone and Mineral Research
Variation in Bone Biomechanical Properties,
Microstructure, and Density in BXH Recombinant
Inbred Mice
1
1
2
2
¨
CHARLES H. TURNER, YEOU-FANG HSIEH, RALPH MULLER, MARY L. BOUXSEIN,
CLIFFORD J. ROSEN,3 MARGARET E. McCRANN,4 LEAH RAE DONAHUE,4 and WESLEY G. BEAMER4
ABSTRACT
To test the hypothesis that factors associated with bone strength (i.e., volumetric bone mineral density
[vBMD], geometry, and microstructure) have heritable components, we exploited the 12 BXH recombinant
inbred (RI) strains of mice derived from C57BL/6J (B6; low bone mass) and C3H/HeJ (C3H; high bone mass)
progenitor strains. The femurs and lumbar vertebrae from each BXH RI strain were characterized for
phenotypes of vBMD, microstructural, biomechanical, and geometrical properties. Methods included bending
(femur) and compression (vertebra) testing, peripheral quantitative computed tomography (pQCT), and
microcomputed tomography (␮CT). Segregation patterns of femoral and vertebral biomechanical properties
among the BXH RI strains suggested polygenic regulation. Femoral biomechanical properties were strongly
associated with femoral width in the anteroposterior (AP) direction and cortical thickness—geometric
properties with complex genetic regulation. Vertebral vBMD and biomechanical properties measured in BXH
RI strains showed a greater variability than either B6 or C3H progenitors, suggesting both progenitor strains
have independent subsets of genes that yield similar vBMD and strength. The ␮CT and pQCT data suggested
that the distribution of vertebral mineral into cortical and trabecular compartments is regulated genetically.
Although the B6 and C3H progenitors had similar vertebral strength, their vertebral structures were
markedly different: B6 had good trabecular bone structure and modest cortical bone mineral content (BMC),
whereas C3H had high cortical BMC combined with a deficiency in trabecular structure. These structural
traits segregated independently in the BXH RI strains. Finally, vertebral strength was not correlated
consistently with femoral strength among the BXH RI strains, suggesting genetic regulation of bone strength
is site specific. (J Bone Miner Res 2001;16:206–213)
Key words: biomechanics, bone density, osteoporosis, genetics
factors contributing to the prevalence of osteoporosis is re-
duced skeletal density.(1) Studies in twins have determined that
approximately 70% of the variability in bone density is genet-
INTRODUCTION
STEOPOROSIS IS primarily a disease of bone fragility
O
resulting from decreased bone mass, altered microar- ically based.(2–4) This variation in bone density probably is
chitecture, and possibly impaired bone quality. Key among the influenced by multiple genes, which have not yet been identi-
1Biomechanics and Biomaterials Research Center, Indiana University, Indianapolis, Indiana, USA.
2Orthopedic Biomechanics Laboratory, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts,
USA.
3Maine Center for Osteoporosis Research and Education, Bangor, Maine, USA.
4The Jackson Laboratory, Bar Harbor, Maine, USA.
206

BONE STRENGTH IN RECOMBINANT INBRED MICE
207
fied.(5) Beamer et al.(6) showed a large difference in the femoral density is achieved at about 4 months of age in mice and
volumetric bone mineral density (vBMD) and bone mineral maintained until about 8–12 months of age.(6) Conse-
content (BMC) between the inbred mouse strains C3H/HeJ quently, the mice in this study are mature adults with peak
(C3H) and C57BL/6J (B6). Thus, these strains appear to be bone mass. These mice were group-housed in polycarbonate
very good models for the analyses of high and low bone mass, cages at The Jackson Laboratory (Bar Harbor, ME, USA).
respectively. However, the genetic influence on skeletal bone Water was acidified with HCl to achieve a pH of 2.8–3.2 for
microstructure and strength in these mice is site specific and suppression of bacteria and was freely available. The diet
complex. For example, C3H mice have thicker femoral corti- used for all mice was pasteurized National Institutes of
ces and stronger femoral shafts when compared with B6 but Health (NIH) 31 (6% fat diet, vitamin and mineral fortified;
are deficient in trabecular bone structure in the proximal femur PMI, Richmond, IN, USA) and was freely available. Use of
and lumbar spine.(7) Although C3H mice have significantly mice in this research project was reviewed and approved by
stronger femurs compared with B6, their lumbar vertebrae are the Institutional Animal Care and Use Committee of The
not stronger, but instead are more brittle,(7) suggesting that the Jackson Laboratory.
genes contributing to improved femoral strength have no effect
or even a negative effect on trabecular bone structure in the
spine. These findings led us to hypothesize that multiple genes
BMD measurements
contribute independently to cortical and trabecular bone mi-
crostructures and bone fragility in the C3H and B6 strains.
Isolated femurs and vertebrae were assessed using periph-
eral quantitative computed tomography (pQCT; Stratec
In the current study, we evaluated the bone microstruc- XCT 960M; Norland Medical Systems, Ft. Atkinson, WI,
ture, vBMD, and biomechanical properties of femurs and USA) as described previously.(6) Briefly, bones were iso-
vertebrae in BXH recombinant inbred (RI) mouse strains. lated and stored in 95% EtOH until assessed by pQCT for
RI strains were created by first crossbreeding two different vBMD. Thresholds of 1.300 attenuation units differentiated
inbred strains, intercrossing the resulting F1 hybrids and mineralized bone from water, adipose tissue, muscle, and
then inbreeding pairs of F2 progeny for 20 subsequent tendon; a threshold of 2.000 differentiated high-density
generations.(8) This breeding strategy creates a set of RI cortical bone from bone of lower density. Precision of the
strains. The mice within each RI strain are “twins” and thus pQCT for femoral, tibial, and vertebral vBMD were 1.2,
contain the same genetic alleles, half of these alleles come 1.5, and 1.4%, respectively. Isolated femurs were scanned at
from each of the two progenitor strains. A set of 12 RI 2-mm intervals over their entire lengths. Total vBMD was
strains (denoted BXH) were previously created from B6 and calculated by dividing the total mineral content by the total
C3H progenitor strains. Each of the BXH RI strains contains bone volume. Femoral cortical thickness was calculated at
a unique combination of genetic alleles from the B6 and the midpoint of the diaphysis. Isolated lumbar vertebrae
C3H strains that can be phenotyped for genetically based (L5) were scanned at their measured midpoints and vBMD
polymorphic traits that differ between the B6 and C3H was calculated as described previously. Cortical BMC val-
progenitors. Comparison among strains of the distribution ues also are presented for the L5 vertebrae.
pattern for a quantitative trait can be used to gain chromo-
somal location of genes responsible for the trait of inter-
Bone microstructure measurements
est.(9) Originally, RI strains were developed for mapping
single genes with major effects; mapping precision rapidly
The L5 vertebra from BXH RI strains with the highest
degenerates with increasing genetic complexity of a trait.(10) and lowest bone strength were analyzed using a desktop
Nevertheless, such strains can be exploited to discriminate micro-CT (␮CT 20; Scanco Medical AG, Bassersdorf,
between single versus polygenic inheritance, as well as to Switzerland). A microfocus X-ray tube with a focal spot of
gain clues about subsets of regulatory genes and possible 10 ␮m was used as a source. To perform a measurement, the
locations of major genes.(11)
specimen was mounted on a turntable that could be shifted
We hypothesized that factors associated with bone automatically in the axial direction. Six hundred projections
strength (i.e., BMD, geometry, and microstructure) have were taken over 216° (180° plus half the fan angle on either
heritable components. To test this hypothesis, we measured side). A standard convolution-backprojection procedure
femoral and vertebral bone strength in the 12 BXH RI with a Shepp and Logan filter was used to reconstruct the
strains and the progenitor B6 and C3H strains. In addition, CT images in 1024 pixel ϫ 1024 pixel matrices. For each
we measured femoral and vertebral geometry, vBMD, and sample, a total of 100–200 microtomographic slices, with a
trabecular microstructure in the lumbar spine. The resultant slice increment of 17 ␮m, were acquired. Measurements
phenotypic segregation patterns differed from each other, were stored in three-dimensional (3D) image arrays with an
indicating distinct inheritance patterns and a complexity of isotropic voxel size of 17 ␮m. A constrained 3D Gaussian
genetic regulation for these traits.
filter was used to partly suppress the noise in the volumes.
The bone tissue was segmented from marrow using a global
thresholding procedure.(12) In addition to the visual assess-
ment of structural images, morphometric indices were de-
termined from the microtomographic data sets. Cortical and
trabecular bone were separated using a semiautomated con-
MATERIALS AND METHODS
Animal care
The study involved 12 BXH RI strains of mice (n ϭ tour tracking algorithm to detect the outer and inner bound-
7–13) and the B6 and C3H progenitor strains. All mice used aries of the cortex. In trabecular bone, basic structural
in the study were female and 8 months of age. Adult femoral metrics were measured using direct 3D morphometry.(13)

208
TURNER ET AL.
TABLE 1. BODY WEIGHT AND FEMUR GEOMETRY, INCLUDING LENGTH, WIDTH IN AP AND ML DIRECTIONS, AND
MIDDIAPHYSEAL THICKNESS, FOR B6, C3H, AND BXH RI STRAINS
Inbred
strain
Body
weight (g)
Femur length
(mm)
Width ML
(mm)
Width AP
(mm)
Cortical thickness
(mm)
n
B6
C3H
8
10
28.4 Ϯ 1.5
27.7 Ϯ 1.0
16.1 Ϯ 0.11
16.0 Ϯ 0.08
1.80 Ϯ 0.05a
1.60 Ϯ 0.01b
1.23 Ϯ 0.02a
1.31 Ϯ 0.02b
0.40 Ϯ 0.01a
0.56 Ϯ 0.01b
BXH-2c
BXH-3
BXH-4
BXH-6
BXH-7
BXH-8
BXH-9
BXH-10
BXH-11
BXH-12
BXH-14
BXH-19
10
11
13
8
10
8
10
7
10
9
25.6 Ϯ 1.0
31.4 Ϯ 1.3
25.7 Ϯ 1.3
26.9 Ϯ 2.3
24.3 Ϯ 1.3
31.9 Ϯ 1.9
26.6 Ϯ 0.7
24.5 Ϯ 1.8
28.9 Ϯ 2.5
26.2 Ϯ 1.4
21.7 Ϯ 0.4a,b
16.0 Ϯ 0.09
17.2 ؎ 0.07a,b
16.1 Ϯ 0.04
15.9 Ϯ 0.08
16.1 Ϯ 0.09
16.4 ؎ 0.05a,b
15.1 Ϯ 0.06a,b
16.3 Ϯ 0.14a
16.0 Ϯ 0.09
16.6 ؎ 0.10a,b
16.2 Ϯ 0.07
1.54 Ϯ 0.04b
1.81 Ϯ 0.02a
1.73 Ϯ 0.03a
1.90 ؎ 0.05a,b
1.84 Ϯ 0.03a
1.82 Ϯ 0.04a
1.52 Ϯ 0.02b
1.84 Ϯ 0.02a
1.60 Ϯ 0.02b
1.66 Ϯ 0.03b
1.72 Ϯ 0.03a
1.11 Ϯ 0.02a,b
1.28 Ϯ 0.02b
1.02 Ϯ 0.01a,b
1.16 Ϯ 0.03a,b
1.20 Ϯ 0.01a
1.26 Ϯ 0.02a
1.07 Ϯ 0.01a,b
1.27 Ϯ 0.03
0.36 Ϯ 0.01a
0.49 Ϯ 0.01a,b
0.45 Ϯ 0.01a,b
0.37 Ϯ 0.01a
0.44 Ϯ 0.02a,b
0.46 Ϯ 0.02a,b
0.37 Ϯ 0.01a
0.41 Ϯ 0.01a
0.44 Ϯ 0.01a,b
0.49 Ϯ 0.02a,b
0.46 Ϯ 0.02a,b
1.15 Ϯ 0.02a,b
1.10 Ϯ 0.01a,b
1.08 Ϯ 0.01a,b
8
Data presented are mean Ϯ SEM.
Boldface indicates significantly greater than both C3H and B6; underline indicates significantly less than both C3H and B6.
a Significantly different from C3H.
b Significantly different from B6.
c BXH-2 developed early onset leukemia and were excluded from the study.
The measurements included bone volume density (BV/TV), specimen submerged in 37°C saline using a displacement
bone surface density (BS/TV), trabecular number (Tb.N), rate of 1 mm/s. The standard parameters of Fu, S, and U
trabecular thickness (Tb.Th), and trabecular spacing were calculated from the resulting load-displacement
(Tb.Sp). Previous studies have shown that trabecular struc- curves. Variation was noted in the specimen height resulting
tural metrics measured using ␮CT closely correlate with from both genetic differences and effects of making the
those measured using standard histomorphometry.(14,15) Av- endplates parallel. Because stiffness and work to failure are
erage height and width of the L5 vertebrae were measured affected by specimen height, these parameters were normal-
directly on midcoronal ␮CT images.
ized by dividing U by specimen height and by multiplying
S by specimen height.
Biomechanical tests
Statistical tests
We measured bone strength of the femur and the L5
lumbar vertebrae. Femurs were tested at the mid-shaft by
three-point bending at room temperature. Load was applied
in the anteroposterior (AP) direction midway between two
supports that were 5 mm apart. Load-displacement curves
were recorded at a crosshead speed of 0.5 mm/s using a
microforce materials testing machine (Vitrodyne 2000).
Data were stored on a microcomputer. Ultimate force (Fu),
stiffness (S), and work to failure (U) were calculated from
the load-displacement curve as described elsewhere.(16) Fu
reflects the strength of the bone, while S reflects the rigidity
and U is the energy necessary to cause a fracture. Widths of
the cortical midshaft in the mediolateral (ML) and antero-
Comparisons among the BXH RI strains and the B6 and
C3H progenitors were made using analysis of variance
(ANOVA) implemented by StatView software (Abacus
Concepts, Berkeley, CA, USA). Post hoc comparisons be-
tween groups were accomplished using a Fisher’s protected
least significant difference test. Statistical significance was
assumed at p Ͻ 0.05.
RESULTS
The average body weights for BXH RI strains were not
posterior (AP; a.k.a. cranial-caudal) directions were mea- significantly different from the B6 and C3H progenitor
sured using digital calipers accurate to 0.01 mm, with a strains, with the exception of BXH-19 mice, which were
precision of Ϯ0.005 mm (Mitutoyo, Aurora, IL, USA).
significantly lighter than both progenitor strains (Table 1).
The L5 vertebrae were dissected free and the posterior The C3H mice had greater femoral strength (Fu), stiffness
elements were removed using a small clipper. The endplates (S), and work to failure (U) compared with B6, as previ-
of the vertebral body were cut parallel using a diamond ously reported.(7) For most RI strains, femoral biomechani-
wafering saw (Isomet, Buehler, Lake Bluff, IL, USA). Me- cal properties fell between those of the progenitor strains
chanical tests were performed in compression using a ser- (Fig. 1, note that mice from the BXH-2 RI strain developed
vohydraulic materials testing machine (810; MTS Corp., early onset leukemia and were excluded from the study).
Minneapolis, MN, USA). All tests were done with the However, five RI strains (BXH-3, -6, -7, -10, and -11) had

BONE STRENGTH IN RECOMBINANT INBRED MICE
209
(2.0 Ϯ 0.2 mJ/mm3) was significantly lower than all other
RI strains (p Ͻ 0.001) and considerably less than the tough-
ness of B6 (8.1 Ϯ 0.5 mJ/mm3; p Ͻ 0.0001) or C3H (17.2 Ϯ
0.6 mJ/mm3; p Ͻ 0.0001) progenitors.
There was variation among the strains in femoral geom-
etry. Femoral length was not significantly different between
C3H and B6 progenitor strains but varied greatly among
BXH RI strains (Table 1). In particular, mice from BXH-4,
-9, and -14 strains had femoral lengths that exceeded those
from either progenitor strain, whereas mice from the
BXH-10 strain had significantly shorter femurs. Femoral
width varied greatly among the progenitor and RI strains,
particularly in the ML direction as presented in Table 1.
Femurs from the B6 progenitor mice were 15% wider in the
ML dimension compared with C3H measurements. Among
the RI strains, ML width values distributed rather uniformly
as intermediate, B6-like, or C3H-like.
Total vBMD values for femurs were higher in C3H mice,
compared with B6 (Table 2). The total femoral vBMD
values for BXH RI strains fell between those of the progen-
itor strains and in all instances were significantly different
from both progenitors. The femoral cortical BMC values for
the RI strains tended to segregate into low B6-like levels
(BXH-3, -7, and -10), intermediate levels (BXH-6, -8, -11,
and -12), or high C3H-like levels (BXH-4, -9, -14, and -19).
There were no differences in vertebral bone strength (Fu)
between these 8-month-old C3H and B6 mice, yet there was
considerable variation in bone strength among the BXH RI
strains. Four RI strains (BXH-3, -7, -10, and -12) had
significantly weaker vertebrae than either B6 or C3H pro-
genitors, while one RI strain—BXH-4—had significantly
greater vertebral bone strength than both progenitors (Fig.
2). Normalized stiffness (S*h) and normalized work to
failure (U/h) tended to distribute more normally with values
similar to, as well as between, those of the B6 and C3H
progenitors.
The total vBMD values for L5 vertebrae distributed dif-
ferently from vBMD in femurs: five of the BXH RI strains
(BXH-3, -7, -8, -10, and -11) were statistically below both
progenitor strains, whereas BXH-4, -6, -9, and -19 were like
FIG. 1. Femoral biomechanical properties for B6, C3H, and BXH RI
inbred strains of mice. Data presented are mean Ϯ SEM (n ϭ 7–13). B the B6 and C3H progenitors, and only BXH-14 exceeded
indicates a significant difference from B6 and H indicates a significant
difference from C3H; Fu is ultimate force, S is stiffness, and U is work
to failure.
the high-density C3H value. C3H mice had substantially
more vertebral cortical BMC when compared with B6 val-
ues. The vertebral cortical BMC (Table 2) showed that six
of the RI strains (BXH-3, -6, -7, -8, 10, and -11) had
significantly lower BMC values than both progenitor strains
significantly reduced U compared with either of the progen- and two RI strains (BXH-9 and -12) were statistically iden-
itor strains. Three of these strains (BXH-3, -10, and -11) tical with B6 progenitors, whereas three RI strains (BXH-4,
also had significantly reduced Fu compared with progeni- -14, and -19) showed BMC values like that of the C3H
tors. This indicates that these strains had femurs that are progenitors. None of the BXH RI strains had vertebral BMC
more fragile. For the most part, the increased femoral fra- values statistically greater than C3H mice.
gility was caused by a combination of reduced cortical
␮CT analyses revealed marked differences between the
thickness and reduced femoral width in the AP direction vertebrae with highest (BXH-4) and lowest (BXH-3) bone
(Table 1). A remarkable exception was BXH-11, where strength (Fig. 3). BXH-4 had superior bone microstructure
cortical thickness and AP femoral width were not different compared with BXH-3 and progenitor strains, including
from B6, yet Fu and U were only 60% and 25% of B6, significantly higher trabecular bone volume and thickness
respectively. These results suggest that BXH-11 mice had (Table 3). Both BXH-3 and C3H mice had poor trabecular
diminished femoral bone quality. We explored this issue bone structure with reduced Tb.N and increased Tb.Sp
further by calculating femoral toughness, which is a mea- compared with B6 or BXH-4 mice. There were no signifi-
sure of the bone tissue integrity corrected for the size and cant differences in trabecular structure between BXH-3 and
shape of the bone.(16) Femoral toughness in BXH-11 mice C3H; however, the vertebral bone strength for BXH-3 was

210
TURNER ET AL.
TABLE 2. BONE PARAMETERS DETERMINED BY PQCT FOR THE FEMUR AND LUMBAR VERTEBRAE FOR B6, C3H,
AND BXH RI STRAINS OF MICE
Midvertebral
total vBMD
(mg/mm3)
Femoral total
Femoral cortical
BMC (mg)
Midvertebral cortical
BMC (mg)
Strain
n
vBMD (mg/mm3)
B6
C3H
8
10
0.459 Ϯ 0.009a
0.680 Ϯ 0.010b
7.77 Ϯ 0.25a
11.9 Ϯ 0.83b
0.228 Ϯ 0.007a
0.240 Ϯ 0.007b
0.45 Ϯ 0.07a
0.69 Ϯ 0.06b
BXH-2c
BXH-3
BXH-4
BXH-6
BXH-7
BXH-8
BXH-9
BXH-10
BXH-11
BXH-12
BXH-14
BXH-19
10
11
13
8
10
8
10
7
10
9
0.511 Ϯ 0.013a,b
0.601 Ϯ 0.006a,b
0.602 Ϯ 0.012a,b
0.524 Ϯ 0.012a,b
0.548 Ϯ 0.016a,b
0.585 Ϯ 0.007a,b
0.551 Ϯ 0.004a,b
0.522 Ϯ 0.008a,b
0.594 Ϯ 0.008a,b
0.613 Ϯ 0.013a,b
0.604 Ϯ 0.008a,b
6.78 Ϯ 0.43a
12.7 Ϯ 0.16b
9.69 Ϯ 0.41a,b
7.61 Ϯ 0.38a
9.95 Ϯ 0.58a,b
11.0 Ϯ 0.46b
7.01 Ϯ 0.11a
9.66 Ϯ 0.46a,b
9.50 Ϯ 0.36a,b
11.2 Ϯ 0.40b
10.9 Ϯ 0.33b
0.189 Ϯ 0.008a,b
0.246 Ϯ 0.008
0.220 Ϯ 0.006
0.200 Ϯ 0.009a,b
0.213 Ϯ 0.007a,b
0.229 Ϯ 0.006
0.188 Ϯ 0.004a,b
0.189 Ϯ 0.005a,b
0.226 Ϯ 0.007
0.279 ؎ 0.010a,b
0.256 Ϯ 0.010b
0.09 Ϯ 0.02a,b
0.57 Ϯ 0.07
0.29 Ϯ 0.05a,b
0.19 Ϯ 0.03a,b
0.20 Ϯ 0.04a,b
0.39 Ϯ 0.05a
0.10 Ϯ 0.01a,b
0.20 Ϯ 0.04a,b
0.32 Ϯ 0.05a
0.82 Ϯ 0.11b
0.67 Ϯ 0.07b
8
Data are presented as mean Ϯ SEM.
Boldface indicates significantly greater than both C3H and B6; underline indicates significantly less than both C3H and B6.
a Significantly different from C3H.
b Significantly different from B6.
c BXH-2 developed early onset leukemia and were excluded from the study.
only 54% of C3H. These results apparently are caused by the traits that differ between the B6 and C3H progenitors,
the very low vertebral cortical BMC of BXH-3, compared analyses of resultant patterns yield insights about genetic
with C3H (Table 2). B6 vertebrae were shorter than those
from C3H, whereas C3H vertebrae were more narrow. The
selected BXH RI strains tended to retain the height of C3H
and the width of B6 (Table 3).
regulation underlying each trait. The evidence reported here
shows a remarkable complexity to the biological basis for
differences in bone strength between B6 and C3H mice.
The segregation patterns of values of biomechanical
properties for BXH RI femurs lead to the following conclu-
sions. First, none of the investigated properties segregated
the RI strains into subsets that simply resembled either the
B6 or the C3H progenitors. In the absence of such a simple
segregation pattern, we conclude that each of the pheno-
types represented by Fu, S, and U of these skeletal sites is
regulated by more than one gene. The modest numbers of RI
strains available in this BXH set will not support linkage
analyses of traits with more than one gene. Therefore, we
have not compared the BXH RI strain distribution patterns
recorded for any of the phenotypes reported here with
published data in search of single genes. Second, several
BXH RI had significantly lower femoral Fu and U values
than those found in the low-strength B6 strain and none of
the BXH RI strains achieved values that approached those
observed in the high-strength C3H strain. This phenomenon
of transgenesis suggests that (a) the C3H strain carries
genetic alleles contributing to low bone strength and (b) the
high bone strength of C3H may result from the interaction
of specific genes in C3H but, by chance, none of the BXH
RI strains contain the correct combination of C3H genes
supporting high femoral strength.
Because of the independent variation of cortical and
trabecular bone microstructures among progenitor and RI
strains, the variance in vertebral strength was not reflected
consistently in the femoral strength (Fig. 4). The solid
symbols in Fig. 4 denote the inbred strains that are partic-
ularly useful to illustrate the point that genetic regulation of
vertebral strength is sometimes independent of regulation of
femoral strength. The BXH-12 and BXH-4 strains showed a
difference in vertebral strength of almost 2-fold with no
corresponding difference in femoral strength. Likewise, the
B6 and C3H strains showed a 2.2-fold difference in femoral
strength with no difference in vertebral strength.
DISCUSSION
It is widely recognized that loss of bone strength is
accompanied by a corresponding rise in risk for osteopo-
rotic fracture. Critical insight to osteoporotic fracture will
follow from a better understanding of factors regulating
biomechanical properties. In this report, we have hypothe-
sized that factors associated with bone strength (i.e., BMD,
size, and microstructure) have heritable components. To test
this concept, we exploited a unique genetic model system in
Femoral geometry (overall length and ML and AP
widths) also showed complex genetic regulation. Although
the form of the BXH RI strains derived from B6 and C3H the length of the B6 and C3H femurs were not different, half
progenitors known to differ in many aspects of bone of the BXH RI strains had femurs that were either shorter or
biology.(6,7,17–22) By phenotyping each BXH RI strain for longer than the progenitor strains. Middiaphyseal ML

BONE STRENGTH IN RECOMBINANT INBRED MICE
211
FIG. 3. ␮CT section of the L5 vertebral body for progenitor strains
(B6 and C3H) and the BXH RI strains with maximum (BXH-4) and
minimum (BXH-3) vertebral strength. The images were measured 3-D
providing a 17-␮m isotropic voxel size.
FIG. 2. Vertebral biomechanical properties for B6, C3H, and BXH
RI inbred strains of mice. Data presented are mean Ϯ SEM (n ϭ 7–13).
B indicates a significant difference from B6 and H indicates a signif-
icant difference from C3H; Fu is ultimate force, S*h is normalized
stiffness, and U/h is normalized work to failure.
ation in the BXH-11 femoral tissue caused the deficit in
toughness; nevertheless, the BXH RI strains may be useful
for studying genetic influences on bone quality.
As was true for biomechanical and morphology proper-
widths distributed into either progenitor and intermediate ties, segregation of total vBMD and cortical BMC for fe-
categories, while the majority of the RI strain AP widths murs was consistent with conclusions of polygenic regula-
were less that those of the smaller B6 progenitor. As was tion. Femoral vBMD values from the BXH RI strains were
noted for the femoral strength properties, the thicker fem- distributed between those of the progenitor strains and no
oral cortex of the C3H femurs was not observed in the BXH BXH RI replicated either progenitor strain value, again
RI strains. Collectively, these data argue cogently for com- arguing for the fact that combinations of specific genes are
plex genetic regulation of parameters for femoral size.
required to achieve the respective densities of the progenitor
The RI strain BXH-11 was remarkable because appar- strains.
ently it had robust femoral shaft geometry, yet with weaker
The biomechanical properties of the L5 vertebrae, partic-
and more fragile femurs. The femoral toughness of the ularly Fu, in BXH RI strains, showed a greater range of
BXH-11 mice was substantially lower than other strains, variation than either B6 or C3H progenitors, suggesting
indicating a deficit in bone quality. These results suggest regulation by multiple genes. Likewise, patterns for verte-
that one of the B6 and C3H progenitor strains contains a bral vBMD values were noteworthy in that the B6 and C3H
genetic allele(s) that influences some aspect of bone quality. progenitors were marginally different from each other,
It is unclear, based on the present data, exactly what alter- whereas the BXH RI strains showed a much greater range of

212
TURNER ET AL.
TABLE 3. ␮CT ANALYSIS OF LUMBAR VERTEBRAE FOR B6, C3H, AND BXH RI STRAINS WITH THE GREATEST (BXH-4)
AND LEAST (BXH-3) VERTEBRAL STRENGTH (MEAN Ϯ SEM)
Inbred strain
Variable
B6
C3H
BXH-3
BXH-4
Vertebral geometry
Height (mm)
Width (mm)
2.60 Ϯ 0.05a
1.95 Ϯ 0.13a
3.22 Ϯ 0.10b
1.66 Ϯ 0.08b
3.51 Ϯ 0.11b
2.03 Ϯ 0.06a
3.45 Ϯ 0.22b
1.98 Ϯ 0.06a
Trabecular structure
BV/TV (%)
24.1 Ϯ 1.3a
8.5 Ϯ 0.2a
5.1 Ϯ 0.2a
61 Ϯ 2
16.0 Ϯ 1.0b
4.4 Ϯ 0.2b
2.8 Ϯ 0.1b
69 Ϯ 4
10.8 Ϯ 1.3a,b
3.3 Ϯ 0.4a,b
2.6 Ϯ 0.2b
73 Ϯ 1
30.0 ؎ 3.6a,b
6.7 Ϯ 0.2a,b
4.1 Ϯ 0.2a,b
93 ؎ 8a,b
BS/TV (mmϪ1
)
Tb.N (mmϪ1
Tb.Th (␮m)
Tb.Sp (␮m)
)
199 Ϯ 9a
388 Ϯ 11b
415 Ϯ 38b
255 Ϯ 13a,b
Boldface indicates significantly greater than both C3H and B6; underline indicates significantly less than both C3H and B6.
a Significantly different from C3H.
b Significantly different from B6.
strains, B6 had good trabecular bone structure and poorer
cortical BMC, whereas C3H had the opposite combination,
resulting in intermediate biomechanical properties. How-
ever, the correlation between vertebral structure and biome-
chanical properties was not perfect because the RI strain
with the highest total vertebral vBMD and cortical BMC
(BXH-14) did not have the greatest vertebral strength. This
observation suggests that although the measured densitom-
etry and geometric variables explain much of the variation
in vertebral strength, other important structural variables
that contribute to vertebral integrity remain.
It was noteworthy that the variation in vertebral strength
was not correlated consistently with femoral strength. In
particular, the B6 and C3H strains indicated over a 2-fold
difference in femoral strength with no difference in verte-
bral strength while BXH-12 and BXH-4 strains were vastly
different in vertebral strength with no corresponding differ-
ence in femoral strength. These results strongly suggesting
site-specific genetic regulation of bone strength.
Although the ultimate goal of our studies is to better
understand the pathogenesis of human osteoporosis, the
mouse model used does not mimic osteoporotic humans.
FIG. 4. Relationship between femoral strength (ultimate force) and
vertebral strength for B6, C3H, and BXH RI strains. The progenitor
strains (B6 and C3H) show a large variation in femoral strength with no
difference in vertebral strength. Likewise, selected BXH RI strains
(BXH-3, BXH-4, and BXH-12) show large variation in vertebral
strength with little or no variation in femoral strength.
We chose 8-month-old female mice, which model middle-
values. This distribution of values supports the conclusion
aged, premenopausal women. Our aim was to study the
variation in bone microstructure and biomechanical proper-
ties in animals with peak bone density. In women, the peak
bone density achieved during childhood, adolescence, and
young adulthood protects against osteoporosis in later
years.(23) A further limitation of mice in the study of skeletal
biology is their lack of osteonal remodeling in compact
bone tissue. This key difference between mice and humans
can lead to confounding results. These limitations not with-
standing, mice remain one of the best animal models for
studying genetic influences on human diseases and disor-
ders, skeletal or otherwise.
that both progenitor strains have independent subsets of
genes that yield very similar mean values. Recombination of
those genes in the BXH RI strains yields a broader range of
values.
Not only is vertebral vBMD under genetic control, but it
appears that the distribution of vertebral mineral into corti-
cal and trabecular compartments also is regulated geneti-
cally. Trabecular structure in the lumbar vertebra tended to
vary independently of cortical BMC of the vertebra, and
both trabecular and cortical bone contributed to vertebral
strength. BXH-4 mice had alleles that contributed to supe-
rior vertebral strength by increasing Tb.N, reducing Tb.Sp,
improving Tb.Th, and increasing cortical BMC, while
BXH-3 vertebrae were osteopenic with diminished trabec-
In conclusion, we found that the regulation of femoral and
vertebral biomechanical properties in BXH RI mice in-
ular and cortical bone structure. As for the progenitor volved multiple genes and was both site specific and com-

BONE STRENGTH IN RECOMBINANT INBRED MICE
213
12. Mu¨ller R, Ru¨egsegger P 1997 Micro-tomographic imaging for
the nondestructive evaluation of trabecular bone architecture.
In: Lowet G, Ru¨egsegger P, Weinans H, Meunier A (eds.)
Bone Research in Biomechanics. IOS Press, Amsterdam, The
Netherlands, pp. 61–79.
13. Hildebrand T, Mu¨ller R, Laib A, Dequeker J, Ru¨egsegger P
1999 Direct 3-D morphometric analysis of human cancellous
bone: Microstructural data from spine, femur, iliac crest, and
calcaneus. J Bone Miner Res 14:1167–1174.
partment specific. Independent genetic regulation of femo-
ral geometry, cortical BMC, and vertebral trabecular
microstructure contributed to the variation in biomechanical
properties among the strains. The genetic control of bone
strength appears to be rich and complex with many puzzles
to be solved before the genetic mechanisms are understood.
14. Mu¨ller R, Van Campenhout H, Van Damme B, Van der Perre
G, Dequeker J, Hildebrand T, Ru¨egsegger P 1998 Morpho-
metric analysis of human bone biopsies: A quantitative struc-
tural comparison of histological sections and micro-computed
tomography. Bone 23:59–66.
15. Li XJ, Mu¨ller R, Stevens ML, Golden J, Seeherman H, Boux-
sein ML 1998 Evaluation of trabecular bone morphometry:
Comparison of micro-computed tomography and histomor-
phometry in the rat proximal tibia. Bone 23:S635.
ACKNOWLEDGMENTS
This work was supported in part by the United States
Department of Health and Human Services National Insti-
tutes of Health grants AR43618 (W.G.B.), CA34196 CORE
(The Jackson Laboratory), and AR43730 (C.H.T.). M.E.M.
received support from the Alumni Fund of The Jackson
Laboratory.
16. Turner CH, Burr DB 1993 Basic biomechanical measurements
of bone: A tutorial. Bone 14:595–608.
17. Dimai HP, Linkhart TA, Linkhart SG, Donahue LR, Beamer
WG, Rosen CJ, Farley JR, Baylink DJ 1998 Alkaline phos-
phatase levels and osteoprogenitor cell numbers suggest bone
formation may contribute to peak bone density differences
between two inbred strains of mice. Bone 22:211–216.
18. Kodama Y, Dimai HP, Wergedal J, Sheng M, Malpe R,
Kutilek S, Beamer W, Donahue LR, Rosen C, Baylink DJ,
Farley J 1999 Cortical tibial bone volume in two strains of
mice: Effects of sciatic neurectomy and genetic regulation of
bone response to mechanical loading. Bone 25:183–190.
19. Chen C, Kalu DN 1999 Strain differences in bone density and
calcium metabolism between C3H/HeJ and C57BL/6J mice.
Bone 25:413–420.
REFERENCES
1. Cummings SR, Black DM, Nevitt MC, Browner W, Cauley J,
Ensrud K, Genant HK, Palermo L, Scott J, Vogt TM 1993
Bone density at various sites for prediction of hip fractures.
The Study of Osteoporotic Fractures Research Group. Lancet
341:72–75.
2. Pocock NA, Eisman JA, Hopper JL, Yeates MG, Sambrook
PN, Eberl S 1987 Genetic determinants of bone mass in adults:
A twin study. J Clin Invest 80:706–710.
3. Slemenda CW, Christian JC, Williams CJ, Norton JA,
Johnston CC 1991 Genetic determinants of bone mass in adult
women: A reevaluation of the twin model and the potential
importance of gene interaction on heritability estimates.
J Bone Miner Res 6:561–567.
4. Smith DM, Nance WE, Kang KW, Christian JC, Johnston CC
Jr 1973 Genetic factors in determining bone mass. J Clin
Invest 52:2800–2808.
20. Linkhart TA, Linkhart SG, Kodama Y, Farley JR, Dimai HP,
Wright KR, Wergedal JE, Sheng M, Beamer WG, Donahue
LR, Rosen CJ, Baylink DJ 1999 Osteoclast formation in bone
marrow cultures from two inbred strains of mice with different
bone densities. J Bone Miner Res 14:39–46.
5. Rogers J, Mahaney MC, Beamer WG, Donahue LR, Rosen CJ
1997 Beyond one gene-one disease: Alternative strategies for
deciphering genetic determinants of osteoporosis. Calcif Tis-
sue Int 60:225–228.
6. Beamer WG, Donahue LR, Rosen CJ, Baylink DJ 1996 Ge-
netic variability in adult bone density among inbred strains of
mice. Bone 18:397–403.
7. Turner CH, Hsieh Y-F, Mu¨ller R, Bouxsein ML, Baylink DJ,
Rosen CJ, Grynpas MD, Donahue LR, Beamer WG 2000
Genetic regulation of cortical and trabecular bone strength and
microstructure in inbred strains of mice. J Bone Miner Res
15:1126–1131.
8. Taylor BA 1979 Established and incipient recombinant inbred
strains of mice. In: Green MC (ed.) Genetic Variants and
Strains of the Laboratory Mouse. Fischer, Stuttgart, Germany,
p. 397.
21. Rosen CJ, Dimai HP, Vereault D, Donahue LR, Beamer WG,
Farley J, Linkhart S, Linkhart T, Mohan S, Baylink DJ 1997
Circulating and skeletal insulin-like growth factor-I (IGF-I)
concentrations in two inbred strains of mice with different
bone mineral densities. Bone 21:217–223.
22. Sheng MH, Baylink DJ, Beamer WG, Donahue LR, Rosen CJ,
Lau KH, Wergedal JE 1999 Histomorphometric studies show
that bone formation and bone mineral apposition rates are
greater in C3H/HeJ (high-density) than C57BL/6J (low-
density) mice during growth. Bone 25:421–429.
23. Riis BJ, Hansen MA, Jensen AM, Overgaard K, Christiansen
C 1996 Low bone mass and fast rate of bone loss at meno-
pause: Equal risk factors for future fracture: A 15-year
follow-up study. Bone 19:9–12.
Address reprint requests to:
Charles H. Turner, Ph.D.
Director of Orthopedic Research
Indiana University
9. Bailey DW 1981 Recombinant inbred strains and bilinear
congenic strains. In: Foster HS, Small JK, Fox JG (eds.) The
Mouse in Biomedical Research. Academic Press, New York,
NY, USA, pp. 223–239.
10. Silver LM 1995 Mouse Genetics: Concepts and Applications.
Oxford University Press. New York, NY, USA, pp. 207–227.
11. Beamer WG, Shultz KL, Tennent BJ, Nadeau JH, Churchill
GA, Eicher EM 1998 Multigenic and imprinting control of
541 Clinical Drive, Room 600
Indianapolis, IN 46202, USA
ovarian granulosa cell tumorigenesis in mice. Cancer Res Received in original form May 16, 2000; in revised form August
58:3694–9. 10, 2000; accepted September 5, 2000.