This article is mentioned in:

Introduction

Postmenopausal osteoporosis is characterized by a decrease in bone mass and a deterioration in bone architecture. The lifetime risk for women to have an osteoporotic fracture is 30–40% worldwide (1), which has a notable social, physical and economic impact (2–5).

Experimental animal models have contributed tremendously to knowledge of the pathophysiology and treatment targets of postmenopausal osteoporosis. The ovariectomy (OVX) rat is the most commonly used animal model for evaluating the mechanisms underlying postmenopausal osteoporosis and therapeutic strategies for treating this disease (6–9). However, compared with rats, mice may be a more effective model as they have a more easily manipulated genome and lower drug doses are required for treatment. Several strains of mice have been used in postmenopausal osteoporosis research (10–13). However, there are controversies regarding the use of C57BL/6J mice as an animal model for postmenopausal osteoporosis (14–16). The starting age of C57BL/6J mice subjected to ovariectomy varies from 4 to 30 weeks old (10,15–18). Certain combinations of age, skeletal site and time post-surgery may lead to varying levels of bone alterations in response to estrogen deficiency in female rats (6). It is reasonable to hypothesize that the starting age may result in varying levels of bone deterioration upon estrogen deficiency in C57BL/6J mice. Therefore, to improve the understanding of disease pathogenesis and to develop novel therapies, further characterization of C57BL/6J mice as an animal model for postmenopausal osteoporosis is required.

The current study investigated whether C57BL/6J mice were a valid model for postmenopausal osteoporosis. Mice were used in three different age groups (8, 12 and 16 weeks old) to evaluate the extent of bone loss in response to OVX by micro-computed tomography (µCT). To further evaluate bone mass changes of C57BL/6J mice in response to OVX and estrogen replacement, the inbred strain (BALB/c) and outbred strains (ICR and Kunming) of mice were used in this study, and OVX was performed in those mice at 8 weeks of age.

Materials and methods

Animals

Female C57BL/6J mice (8, 12 and 16 weeks old). The total number of 8-, 12- and 16-week-old mice was 26, 21 and 22, respectively. The range of body weight of 8-, 12- and 16-week-old mice was 16.3–20.3, 15.9–24.7 and 20.7–25.6 g, respectively), and female BALB/c, ICR and Kunming mice (all 8 weeks old). The total number of BALB/c, ICR and Kunming was 24, 28 and 28, respectively. The range of body weight of BALB/c, ICR and Kunming mice was 20.9–27.5, 26.5–32 and 32.4–41.8 g, respectively) were purchased from Cavens Biological Technology Co., Ltd., Nanjing, China, https://www.biomart.cn/56079/index.htm. All mice were maintained under a 12-h light/dark cycle at room temperature (22±2°C) with a humidity of 45% and allowed ad libitum access to water and standard rodent chow. All animal procedures were approved by the Animal Care and Use Committee of the Model Animal Research Center of Nanjing University (Nanjing, China).

Ovariectomy and estrogen supplement

Ovariectomy was conducted in 8-, 12- and 16-week-old female C57BL/6J mice and 8-week-old female BALB/c, ICR and Kunming mice. Mice were assigned to the following groups (Table I): Baseline, sacrificed at the corresponding age; SHAM, were subjected to sham operation; OVX, received bilateral ovariectomy; and OVX + 17β-estradiol (E2), at 1 week following surgery administered E2 (30 µg/kg; Sigma-Aldrich; Merck KGaA, Darmstadt, Germany; cat no. E2758) subcutaneously five times per week for 7 weeks. The dose of E2 used was similar to the levels used in a previous study where 20 µg/kg/day resulted in a serum E2 concentration that was similar to the levels in SHAM-operated mice (19). C57BL/6J mice were divided into 8-, 12- and 16-week-old groups (OVX8, OVX12 and OVX16, respectively) from the beginning of OVX. BALB/c, ICR and Kunming mice were subject to surgery at the age of 8 weeks old (OVX8). Surgery was conducted at the Model Animal Research Center of Nanjing University by two operators (SZ and GHW), who were skilled in OVX. In brief, animals were anesthetized (100 mg/kg ketamine, purchased from Fujian Gutian Pharmaceutical Co., Ltd., Fujian, China, http://www.fjgtyy.com/; and 5 mg/kg xylazine, purchased from Jilin Huamu Animal Health Product Co., Ltd., Jilin, China) for ~1 h. For the OVX group, anterior uterine horns were cut to remove the ovaries; for the SHAM group dorsoventral incisions were made through the skin, muscles and periosteum without removal of the ovaries. At 8 weeks post-surgery, animals were sacrificed by inhalation of CO2 and left femurs were harvested for µaT analysis. The uteri were excised and weighed.

Table I. Number of mice in each group.

Table I.

Number of mice in each group.

Group	Baseline mice (n)	SHAM mice (n)	OVX mice (n)	OVX + E2 mice (n)
8-week old C57BL/6	6	6	7	7
12-week old C57BL/6	8	6	7	–
16-week old C57BL/6	8	6	8	–
8-week old BALB/c	6	6	6	6
8-week old ICR	7	8	7	6
8-week old Kunming	7	7	7	7

[i] E2, 17β-estradiol; OVX, ovariectomized.

µCT analysis

Isolated left femurs were fixed with 4% paraformaldehyde for 24 h at 4°C. Then, femurs were placed in plastic tubes and stored within the animal bed inside the SkyScan1176 (Bruker Corporation, Billerica, MA, USA). Femurs were scanned using an 18 µ. resolution protocol: (45 kV, 556 µA, 0.1-mm Cu filter, and 0.2° rotation step, 7-min scan). Volumetric reconstruction software NRecon version 1.5 (Bruker Corporation) was used to reconstruct CT images. Quantification of bone mineral density and trabecular morphometric parameters was performed in a hand-picked cancellous bone area within the primary ossification center. The analyzed area was 0–2 mm above the distal growth plate (Fig. 1). The analysis was performed using scanner software (CTAn, Version 1.13, Bruker Corporation, Billerica, MA, USA). Trabecular parameters were as follows: Trabecular volumetric bone mineral density [vBMD; mg hydroxyapatite (HA)/cm3], bone volume fraction (BV/TV; %), trabecular thickness (Tb.Th; mm), trabecular number (Tb.N; 1/mm), trabecular separation (Tb.Sp; mm), structure model index (SMI) and connectivity density (Conn.D; 1/mm3). The present study evaluated an area 1 mm above and below the midline of the femur to measure the bone mineral density (cortical BMD; mg HA/cm3) and cross-sectional thickness (Cs.Th; mm) of cortical bone (Fig. 1).

Figure 1. A three-dimensional model of the distal femur is shown. Brackets define the trabecular and cortical regions chosen for analysis. Scale bar=1 mm.

Statistical analysis

The effect of age on the body weight, trabecular and cortical bone properties in C57BL/6J were identified using one-way analysis of variance and followed by Fisher's least significant difference (LSD) if the variance was equal, otherwise Dunnett's post hoc test was performed. To ascertain whether bone loss in C57BL/6J following OVX was influenced by age, a two-way analysis of variance followed by Bonferroni's post hoc test was used with treatment (SHAM vs. OVX) and age. An unpaired Student's t-test was used to compare body weight and bone morphology between SHAM and OVX to detect OVX effects in three different age C57BL/6J groups.

To compare the body mass and skeletal parameter differences among the inbred and outbred mice strains, the values of baseline were compared using a one-way analysis of variance followed by an LSD post hoc test if the variance was equal, otherwise Dunnett's post hoc test was used. To evaluate whether bone loss following OVX or bone gain following subsequent E2 treatment was influenced by genetic factors, a two-way analysis of variance followed by Bonferroni's post hoc test was used with treatment (SHAM vs. OVX, or OVX vs. OVX+ E2) and strain. An unpaired Student's t-test was used to compare body weight and bone morphology between SHAM and OVX, or OVX and OVX+E2 to detect OVX or estrogen supplement effects among inbred and outbred mice.

All statistical tests were performed using SPSS version 17.0 software (SPSS, Inc., Chicago, IL, USA) and P<0.05 was considered to indicate a statistically significant difference. Data are presented as the mean ± standard error.

Results

Age-associated alterations in body weight, distal femur trabecular and femoral mid-shaft cortical bone density and architecture in C57BL/6J female mice by µCT

Age-associated alterations of body weight and distal femur metaphyseal trabecular architecture in C57BL/6J female mice are presented in Table II. The body weight of C57BL/6J mice increased steadily with age from 8–24 weeks old by 43.9% (P<0.05). Trabecular vBMD plateaued from 8–16 weeks and declined thereafter. Cancellous BV/TV decreased continuously from 12–24 weeks of age. A similar trend was found in Tb.Th, Tb.N and Conn.D (a measure of trabecular connectedness). Correspondingly, Tb.Sp followed an opposite pattern of alterations. SMI remained relatively constant from 8–16 weeks of age, but were increased significantly at 20 (P<0.05 vs. 8, 12 and 16 weeks) and 24 weeks (P<0.05 vs. 12 and 16 weeks), indicative of a shift to a more rod-like architecture.

Table II. Body weight, bone mass and microarchitecture in 8-, 12-, 16-, 20- and 24-week old female C57BL/6 mice.

Table II.

Body weight, bone mass and microarchitecture in 8-, 12-, 16-, 20- and 24-week old female C57BL/6 mice.

Parameter	8-week old	12-week old	16-week old	20-week old	24-week old
Body weight (g)	17.49±0.30c,e	20.55±1.00e	23.26±0.59a	22.92±1.19	25.16±0.902a,b
Distal femur: Trabecular
vBMD (mgHA/cm3)	0.137±0.004	0.137±0.006	0.142±0.002e	0.129±0.004	0.123±0.005c
BV/TV (%)	9.342±0.634	14.361±2.151e	12.983±0.912e	8.118±1.402	6.467±0.813b,c
Tb.Th (mm)	0.074±0.002	0.080±0.003d	0.077±0.001d	0.070±0.002b,c	0.073±0.002
Tb.N (mm−1)	1.253±0.074b,c	1.826±0.202a,d,e	1.681±0.099a,d,e	1.137±0.182a–c	0.875±0.096b,c
Tb.Sp (mm)	0.359±0.029	0.259±0.013e	0.260±0.006e	0.303±0.028	0.329±0.014b,c
SMI	2.559±0.033e	2.547±0.088d,e	2.537±0.057d,e	2.751±0.076b,c	2.779±0.067a–c
Conn.D (mm−3)	41.546±4.465e	69.722±1.015e	55.104±0.508e	41.531±0.781	19.872±2.994a–c
Femoral midshaft: Cortical
BMD (mgHA/cm3)	0.848±0.027b–e	1.064±0.008a,c–e	1.107±0.008a,b,d	1.163±0.009a–c	1.136±0.009a,b
Cs.Th (mm)	0.138±0.002b–e	0.170±0.003a,c–e	0.183±0.004a,b	0.182±0.006a,b	0.190±0.003a,b

{ label (or @symbol) needed for fn[@id='tfn2-etm-0-0-5839'] } Data are presented as the mean ± the standard error of the mean.

a P<0.05 vs. 8-week old

b P<0.05 vs. 12-week old

c P<0.05 vs. 16-week old

d P<0.05 vs. 20-week old

e P<0.05 vs. 24-week old. vBMD, volumetric bone mineral density; BV/TV, bone volume fraction; Tb.Th, trabecular thickness; Tb.N, trabecular number; Tb.Sp, trabecular separation; SMI, structure model index; Conn.D, connectivity density; Cs.Th, cross-sectional thickness; BMD, bone mineral density; HA, hydroxyapatite.

Age-related changes in the mid-femur diaphysis are also presented in Table II. Cortical BMD and Cs.Th increased continuously with age; the largest increase occurred between 8 and 12 weeks, and measurements plateaued thereafter.

Influence of age on the sensitivity to estrogen deprivation in C57BL/6J female mice

When mice were sacrificed, the abdominal cavity was explored. The ovaries could not be located and the uteruses were very small, which indicated that OVX mice had their ovaries removed. As expected, OVX8, OVX12 and OVX16 were also heavier than mice in the SHAM group, although there was not a significant difference between OVX16 and SHAM groups (P<0.01, P<0.05 and P=0.057, respectively; Table III).

Table III. Effect of estrogen deprivation on body weight, bone mass and microarchitecture in OVX8, OVX12 and OVX16 female C57BL/6J mice.

Table III.

Effect of estrogen deprivation on body weight, bone mass and microarchitecture in OVX8, OVX12 and OVX16 female C57BL/6J mice.

	8-week-old			12-week-old			16-week-old
Parameter	SHAM	OVX	Difference between SHAM and OVX (%)	SHAM	OVX	Difference between SHAM and OVX (%)	SHAM	OVX	Difference between SHAM and OVX (%)	P-valuetreatment*age
Body weight (g)	22.00±0.80	26.81±0.46	21.9b	22.92±1.19	27.51±1.34	20.0a	25.16±0.90	28.10±1.05	11.7	NS
Distal femur: trabecular
vBMD (mgHA/cm3)	0.143±0.006	0.092±0.003	−35.1c	0.129±0.004	0.100±0.005	−22.6b	0.123±0.005	0.104±0.005	−15.8a	0.016
BV/TV (%)	8.005±1.513	2.862±0.318	−64.2a	8.118±1.402	4.292±1.068	−47.1a	6.467±1.154	4.452±1.068	−31.2	NS
Tb.Th (mm)	0.069±0.003	0.068±0.002	−1.4	0.070±0.002	0.072±0.005	2.9	0.073±0.002	0.075±0.004	2.7	NS
Tb.N (mm−1)	1.119±0.160	0.415±0.039	−62.9a	1.137±0.182	0.568±0.110	−50.0a	0.875±0.096	0.571±0.121	−34.7	NS
Tb.Sp (mm)	0.307±0.013	0.630±0.025	100.0c	0.303±0.028	0.416±0.015	37.3b	0.329±0.014	0.402±0.024	22.2a	<0.001
SMI	2.684±0.071	2.725±0.060	1.5	2.751±0.076	2.956±0.153	7.5	2.779±0.067	3.005±0.084	8.1	NS
Conn.D (mm−3)	33.369±0.418	12.722±1.129	−61.9c	41.531±0.781	18.015±2.523	−56.6a	19.872±2.994	16.511±3.189	−16.9	0.035
Femoral midshaft: Cortical
BMD (mgHA/cm3)	1.136±0.004	1.132±0.008	−0.4	1.163±0.009	1.123±0.006	−3.5b	1.136±0.009	1.102±0.008	−2.9a	NS
Cs.Th (mm)	0.177±0.003	0.165±0.002	−6.8b	0.182±0.006	0.179±0.003	−1.6	0.190±0.003	0.184±0.005	−3.2	NS

{ label (or @symbol) needed for fn[@id='tfn8-etm-0-0-5839'] } Data are presented as the mean ± standard error of the mean, unless otherwise stated. Difference between SHAM and OVX=(OVX_mean-SHAM_mean)/SHAM_mean.

a P<0.05

b P<0.01

c P<0.001 between SHAM and OVX groups. OVX, ovariectomized; vBMD, volumetric bone mineral density; BV/TV, bone volume fraction; Tb.Th, trabecular thickness; Tb.N, trabecular number; Tb.Sp, trabecular separation; SMI, structure model index; Conn.D, connectivity density; Cs.Th, cross-sectional thickness; BMD, bone mineral density; HA, hydroxyapatite; NS, not significant.

The patterns of microarchitectural deterioration in response to OVX varied among the three age groups (Table III). In general, microarchitectural deterioration was severe in OVX8, intermediate in OVX12 and lowest in OVX16. This was highlighted by significant differences between OVX and SHAM across the parameters examined in OVX8, with the exception of Tb.Th and SMI. For trabecular vBMD, Tb.Sp and Conn.D in the distal femur, the response to OVX depended on the age, as determined by two-way analysis of variance analyses between treatment and age (Ptreatment*age=0.016, Ptreatment*age<0.001, P treatment*age=0.035, respectively).

Upon ovariectomy, trabecular vBMD in all OVX groups was significantly reduced relative to their SHAM counterparts with OVX8 demonstrating the greatest reduction (−35.1%). BV/TV declined significantly in OVX8 (−64.2%, P=0.0078 vs. SHAM) and OVX12 (−47.1%, P=0.0496 vs. SHAM) and declined non-significantly in OVX16 (−31.2%, P=0.1834 vs. SHAM). For Tb.Th, there were no marked alterations in all age groups. All age groups exhibited a decline in Tb.N, ranging from −34.7 to −62.9%, but this decline was only statistically significant in OVX8 (−62.9%, P=0.0017 vs. SHAM) and OVX12 (−37.3%, P=0.0184 vs. SHAM). No significant alterations in SMI existed among the three groups. Conn.D decreased significantly in OVX8 (−61.9%, P=0.0008 vs. SHAM) and OVX12 (−56.6%, P=0.0108 vs. SHAM), but OVX16 exhibited a non-significant decrease.

The effect of OVX on mid-femoral cortical bone did not appear to depend on age (Table III). Cortical BMD exhibited a statistically significant decline in OVX12 (−3.5%, P=0.028 vs. SHAM) and OVX16 (−2.9%, P=0.02 vs. SHAM), except for OVX8. By contrast, Cs.Th only declined significantly in OVX8 (−6.8%, P=0.0114).

Differences in body mass and bone morphology among inbred and outbred strains of mice

To further explore the bone mass alterations in C57BL/6J mice in response to OVX and estrogen replacement, the above studies were extended by evaluating the inbred and outbred strains, and using an estrogen supplement.

At 8 weeks (baseline), body weight was highest in Kunming, intermediate in ICR and BALB/C, and lowest in C57BL/6J mice (Table IV). Femur trabecular bone in C57BL/6J mice was characterized by the lowest trabecular vBMD, BV/TV, Tb.Th, Tb.N and Conn.D, whereas these mice also exhibited the highest Tb.Sp and SMI. Kunming mice had the most trabeculae and consequently the least Tb.Sp. In trabecular regions, Tb.Th was significantly higher in BALB/c compared with ICR mice (P<0.05).

Table IV. Body weight, bone mass and microarchitecture in four inbred and outbred strains.

Table IV.

Body weight, bone mass and microarchitecture in four inbred and outbred strains.

Parameter	C57BL/6J (n=6)	BALB/c (n=6)	ICR (n=7)	Kunming (n=6)
Body weight (g)	17.49±0.30b–d	23.16±0.85a	29.20±0.43a	36.29±0.82a
Distal femur: Trabecular
vBMD (mgHA/cm3)	0.137±0.004b–d	0.239±0.012a	0.230±0.008a	0.267±0.018a
BV/TV (%)	9.342±0.634b–d	26.236±1.920a	20.161±1.844a	29.749±3.233a
Tb.Th (mm)	0.074±0.002b–d	0.104±0.003a,c	0.091±0.003a,b,d	0.102±0.003a,c
Tb.N (mm−1)	1.253±0.074b–d	2.535±0.175a	2.182±0.144a	2.881±0.236a
Tb.Sp (mm)	0.359±0.286b,d	0.270±0.017a	0.301±0.025d	0.218±0.014a,c
SMI	2.559±0.033b–d	1.646±0.108a	1.984±0.063a	1.725±0.133a
Conn.D (mm−3)	41.546±4.465b–d	81.648±5.149a	73.165±4.898a,d	92.052±9.051a,c
Femoral midshaft: cortical
BMD (mgHA/cm3)	0.848±0.027b–d	1.119±0.011a	1.151±0.006a	1.129±0.009a
Cs.Th (mm)	0.138±0.002b–d	0.182±0.005a,c,d	0.215±0.006a,b	0.210±0.009a,b

{ label (or @symbol) needed for fn[@id='tfn12-etm-0-0-5839'] } Data are presented as the mean ± standard error of the mean.

a P<0.05 vs. C57BL/6J

b P<0.05 vs. BALB/c

c P<0.05 vs. ICR

d P<0.05 vs. Kunming. vBMD, volumetric bone mineral density; BV/TV, bone volume fraction; Tb.Th, trabecular thickness; Tb.N, trabecular number; Tb.Sp, trabecular separation; SMI, structure model index; Conn.D, connectivity density; Cs.Th, cross-sectional thickness; BMD, bone mineral density; HA, hydroxyapatite.

In the femoral cortical bone, BMD was significantly lower in C57BL/6J than the other three strains of mice, whereas Cs.Th was highest in ICR and Kunming, intermediate in BALB/c and lowest in C57BL/6J (Table IV).

Influence of strain on the sensitivity to estrogen deprivation and supplement

C57BL/6J and BALB/c mice subjected to OVX exhibited significant increases 21.9% (P=0.0001) and 6.5% (P=0.0353) of their body mass compared with SHAM counterparts (Tables III and V). Only C57BL/6J and Kunming mice treated with estrogen exhibited significantly decreased body masses (−8.3%, P=0.0066; −4.8%, P=0.0465; respectively; Table VI).

Table V. Effect of estrogen deprivation on body weight, bone mass and microarchitecture in three inbred and outbred mice.

Table V.

Effect of estrogen deprivation on body weight, bone mass and microarchitecture in three inbred and outbred mice.

	BALB/c			ICR			Kunming
Parameter	SHAM	OVX	Difference between SHAM and OVX (%)	SHAM	OVX	Difference between SHAM and OVX (%)	SHAM	OVX	Difference between SHAM and OVX (%)	P-valuetreatment*age
Body weight (g)	24.30±0.31	25.88±0.66	6.5a	36.65±0.59	33.46±0.91	−0.6	44.80±1.55	48.09±1.93	7.5	NS
Distal femur: trabecular
vBMD (mg HA/cm3)	0.276±0.006	0.198±0.010	−28.4c,d	0.231±0.016	0.159±0.014	−31.2b,e	0.234±0.020	0.183±0.010	−22.0a,d	NS
BV/TV (%)	21.590±0.561	9.740±1.842	−54.9c	19.827±2.447	11.052±1.858	−44.3b	21.719±3.144	12.025±1.358	−44.6	NS
Tb.Th (mm)	0.091±0.001	0.077±0.002	−15.4c	0.097±0.005	0.082±0.004	−15.7a	0.104±0.002	0.091±0.002	−12.7c	NS
Tb.N (mm−1)	2.384±0.053	1.249±0.211	−47.6c	2.013±0.165	1.313±0.185	−34.8a	2.067±0.279	1.322±0.139	−36.0a	NS
Tb.Sp (mm)	0.262±0.004	0.438±0.043	67.2a	0.300±0.025	0.417±0.045	39.0a	0.288±0.025	0.423±0.033	47.2a	0.007
SMI	1.798±0.043	2.274±0.070	26.5c	1.997±0.087	2.299±0.085	15.1a	2.013±0.184	2.253±0.081	11.9	NS
Conn.D (mm−3)	61.169±1.637	40.762±7.984	−33.4	62.382±0.657	42.713±5.481	−31.5a	45.978±0.814	29.597±0.430	−35.6	0.035
Femoral midshaft: Cortical
BMD (mg HA/cm3)	1.218±0.009	1.213±0.008	−0.4	1.258±0.008	1.253±0.008	−0.4	1.209±0.007	1.220±0.009	0.9	NS
Cs.Th (mm)	0.222±0.002	0.209±0.007	−5.9	0.255±0.006	0.255±0.007	0.4	0.212±0.004	0.207±0.011	−2.6	NS

{ label (or @symbol) needed for fn[@id='tfn17-etm-0-0-5839'] } Data are presented as the mean ± standard error of the mean, unless otherwise stated. Difference between SHAM and OVX=(OVX_mean-SHAM_mean)/SHAM_mean.

a P<0.05

b P<0.01

c P<0.001 between SHAM and OVX groups.

d P<0.001

e P<0.01 vs. C57BL/6J. vBMD, volumetric bone mineral density; BV/TV, bone volume fraction; Tb.Th, trabecular thickness; Tb.N, trabecular number; Tb.Sp, trabecular separation; SMI, structure model index; Conn.D, connectivity density; Cs.Th, cross-sectional thickness; BMD, bone mineral density; HA, hydroxyapatite; NS, not significant.

Table VI. Effect of estrogen replacement on body weight, bone mass and microarchitecture in four inbred and outbred mice.

Table VI.

Effect of estrogen replacement on body weight, bone mass and microarchitecture in four inbred and outbred mice.

	C57BL/6J		BALB/c		ICR		Kunming
Parameter	OVX+E2	Difference between OVX and OVX+E2 (%)	OVX+E2	Difference between OVX and OVX+E2 (%)	OVX+E2	Difference between OVX and OVX+E2 (%)	OVX+E2	Difference between OVX and OVX+E2 (%)	P-valuetreatment*strain
Body weight (g)	24.60±0.491	−8.3a	25.59±0.42	−1.1	31.167±0.83	−6.8	43.257±1.008	−4.8a	NS
Distal femur: trabecular
vBMD (mgHA/cm3)	0.118±0.005	26.8b	0.218±0.009	10.2	0.179±0.009	12.5	0.176±0.013	−3.4	NS
BV/TV (%)	4.394±0.451	53.5	11.570±1.009	18.8	12.421±1.772	12.4	11.433±1.284	−4.9	NS
Tb.Th (mm)	0.078±0.003	14.2a	0.087±0.002	13.7b	0.088±0.002	7.2	0.104±0.002	14.9c	NS
Tb.N (mm−1)	0.559±0.049	34.6a	1.323±0.102	5.9	1.420±0.198	8.2	1.096±0.115	−17.1	NS
Tb.Sp (mm)	0.429±0.015	−31.9c	0.360±0.015	−17.7d	0.357±0.031	−14.4d	0.457±0.054	8.0e	0.02
SMI	2.947±0.036	8.1a	2.350±0.074	3.3	2.358±0.083	2.6	2.481±0.034	10.2	NS
Conn.D (mm−3)	15.932±2.532	25.2	30.063±3.984	−33.4	44.729±7.660	4.7	21.349±3.498	−27.9	NS
Femoral midshaft: Cortical
BMD (mgHA/cm3)	1.154±0.005	2.0a	1.224±0.009	0.8	1.262±0.007	0.7	1.229±0.008	0.8	NS
Cs.Th (mm)	0.181±0.002	9.5c	0.217±0.002	4.0	0.266±0.007	4.1	0.213±0.002	3.4	NS

{ label (or @symbol) needed for fn[@id='tfn23-etm-0-0-5839'] } OVX+E2 data in this table are compared with OVX data from Table V. Data are presented as the mean ± standard error of the mean, unless otherwise stated. Difference between OVX+E2 and OVX=(OVX + E2_mean-OVX_mean)/OVX_mean.

a P<0.05

b P<0.01

c P<0.001 between OVX+E2 and OVX groups.

d P<0.01

e P<0.05 vs. C57BL/6J. vBMD, volumetric bone mineral density; BV/TV, bone volume fraction; Tb.Th, trabecular thickness; Tb.N, trabecular number; Tb.Sp, trabecular separation; SMI, structure model index; Conn.D, connectivity density; Cs.Th, cross-sectional thickness; BMD, bone mineral density; HA, hydroxyapatite; OVX, ovariectomized; NS, not significant.

All four strains of mice lost cancellous bone in the femur following OVX, with no treatment-strain interaction observed, except in Tb.Sp and Conn.D. (Ptreatment*age =0.007; Ptreatment*age =0.035, respectively; Tables III and V). Femur trabecular vBMD was lower in OVX than the SHAM group in all four strains of mice (Tables III and V). The effect of OVX on trabecular vBMD in C57BL/6J (−35.1%, P<0.001 vs. SHAM) differed significantly from the effect in BALB/c (−28.4%, P<0.001 vs. SHAM; P<0.001 vs. C57BL/6J), ICR (−31.2%, P=0.002 vs. SHAM; P=0.001 vs. C57BL/6J) and Kunming (−22.0%, P=0.024 vs. SHAM; P<0.001 vs. C57BL/6J). Similar to femur trabecular vBMD, C57BL/6J exhibited the greatest decline in BV/TV (−64.2%, P=0.001 vs. SHAM; Tables III and V). The other strains also exhibited a decline in femur BV/TV, ranging from −44.3 to −54.9%, but this decline was statistically significant only in BALB/c (P<0.001 vs. SHAM) and ICR (P=0.008 vs. SHAM). A significant difference in Tb.Th between SHAM and OVX group was not detected in the C57BL/6J strain, whereas BALB/c exhibited a −15.4% decrease (P<0.001 vs. SHAM), ICR −15.7% (P=0.016 vs. SHAM) and Kunming −12.7% (P<0.001 vs. SHAM). Tb.N decreased in C57BL/6J (−62.9%, P=0.016 vs. SHAM), BALB/c (−47.6%, P<0.001 vs. SHAM), ICR (−34.8%, P=0.011 vs. SHAM) and Kunming (−36.0%, P=0.013 vs. SHAM). Tb.Sp increased in C57BL/6J (100%, P<0.001 vs. SHAM), BALB/c (67.2%, P=0.025 vs. SHAM), ICR (39.0%, P=0.023 vs. SHAM) and Kunming (47.2%, P=0.026 vs. SHAM). SMI increased significantly in BALB/c (26.5%, P<0.001 vs. SHAM) and ICR (15.1%, P=0.020 vs. SHAM) whereas Conn.D declined in C57BL/6J (−61.9%, P<0.001 vs. SHAM) and ICR (−31.5%, P=0.042 vs. SHAM). With regard to cortical bone indices, significant alterations were only found in C57BL/6J (Tables III and V). The difference in Cs.Th between SHAM and OVX mice was −6.8% in C57BL/6J mice (P=0.002; Tables III and V).

For Tb.Sp, the response to estrogen supplement was dependent on the strain (Ptreatment*strain=0.02; Table VI). The alteration in this index following estrogen supplement differed significantly between C57BL/6J and BALB/c (P=0.002), ICR (P=0.001) and Kunming (P=0.027). Notably, C57BL/6J mice were the only strain markedly responsive to estrogen supplement, demonstrating significant alterations in all parameters except for BV/TV and Conn.D. Whereas estrogen decreased Tb.Sp by 31.9% in the metaphyseal trabecular bone of C57BL/6J mice (P<0.001 vs. OVX), it increased trabecular vBMD by 26.8% (P=0.003 vs. OVX; Table VI), BV/TV by 53.5% (P=0.059 vs. OVX; Table VI), Tb.Th by 14.2% (P=0.045 vs. OVX) and SMI by 8.1% (P=0.014 vs. OVX). In the mid-femoral cortical bone of C57BL/6J mice, estrogen increased cortical BMD by 2.0% (P=0.023 vs. OVX; Table VI) and Cs.Th by 9.5% (P<0.001 vs. OVX). In contrast with C57BL/6J, estrogen treatment failed to significantly alter any above indices in BALB/c and Kunming mice except Tb.Th (13.7%, P=0.001 vs. OVX; 14.9%, P<0.001; respectively). ICR was the least sensitive to estrogen treatment showing no significant alterations across all the parameters examined.

Discussion

OVX model mice have been widely used in the study of postmenopausal osteoporosis. Multiple lines of evidence have revealed that genetic factors serve a profound role in regulating the balance of bone metabolism in mice (14,20,21). In the present study, it was demonstrated that the peak of cancellous mass in C57BL/6J female mice was present at 8 weeks old. Mice subjected to OVX at 8 weeks old had marked cancellous bone loss, and OVX-induced cancellous bone loss was restored by estrogen treatment. The data also suggested that OVX-induced loss of cancellous bone may gradually decline with age in C57BL/6J female mice.

The BV/TV at distal femur in C57BL/6J mice has been shown to be greatest at 6-weeks old and sequentially decline with age (22). The present study demonstrated an age-associated reduction in cancellous bone in C57BL/6J female mice, as trabecular vBMD and BV/TV were greatest at 8 weeks old. Notably, it was revealed that OVX-induced loss of cancellous bone was associated with age, and the data indicated that the rate of trabecular bone loss in mice subjected to OVX at 8 weeks old was higher than that at 12 weeks old. In addition, there were no significant differences in cancellous bone alterations in mice subjected to OVX at 16 weeks old when compared with the SHAM group. These findings indicated that the effect of estrogen on regulating turnover of cancellous bone in C57BL/6J female mice may gradually decline with age. Further studies are necessary to explore cellular mechanisms underlying the reduced response to estrogen with age.

In contrast with cancellous bone, the results showed that an age-associated increase of cortical bone persisted throughout 8–16 weeks of age in C57BL/6J female mice. However, only a slight reduction in cortical bone, in an age-independent manner, was observed in those mice in response to OVX, suggesting that cortical bone was not sensitive to estrogen deprivation in C57BL/6J female mice.

C57BL/6J mice are typically used to produce gene knockout and transgenic animals. A number of previous studies have identified bone loss when the ovariectomy timing was 2 months old in this strain (23,24). However, C57BL/6J mice have been questioned as an ideal model for the study of postmenopausal osteoporosis as they have been found to be resistant to OVX-induced loss of bone. Bouxsein et al (14) previously discovered that 4-month-old C57BL/6J mice were resistant to OVX-induced trabecular bone deterioration in the proximal tibia. Iwaniec et al (15) demonstrated that no significant difference was observed between C57BL/6J OVX and SHAM groups in the distal femur cancellous BV/TV 3 months post-surgery. Klinck and Boyd (16) demonstrated that the majority of morphology parameters in the C57BL/6J femur and tibia were not statistically significant at 5 weeks post-ovariectomy. This data from previous research were obtained from mice subjected to OVX at the age of 12 or 16 weeks, which is consistent with the findings of the present study. The present study also revealed that alterations of cancellous bone in C57BL/6J mice at the age of 8 weeks were greatest.

Previous studies have demonstrated that the skeletal response to OVX varied among inbred mice (14–16). This study extended the above by adding one inbred (BALB/c) and two outbred strains (ICR and Kunming), and all strains of mice were subjected to OVX at 8 weeks old. It was demonstrated that C57BL/6J mice were most sensitive to OVX-induced loss of bone among the four strains, exhibiting greater alterations in body weight, trabecular vBMD, BV/TV, Tb.N, Conn.D and Cs.Th. In contrast with previous results (14–16), the present study revealed that C57BL/6J mice were the most sensitive to estrogen deficiency among the four strains. It should be noted that Cs.Th, reflecting the cortical bone, decreased only in C57BL/6J mice. Contradicting results may be due to the age of the animals at OVX (previous studies using no less than 3-month-old C57BL/6J mice) and the duration of time elapsed following OVX (8 weeks). The various skeletal alterations during the 2–4 months old period caused by differing post-pubertal architectural development patterns among different strains may account for discrepancies arising from the different starting ages (25). It would be interesting to determine the underlying cellular mechanisms that determine the different strain-associated responses to estrogen deficiency by analysis of histomorphometry or biochemical markers.

E2 has been documented to improve bone quantity and miscrostructure both in mice and rats following OVX (19,26). With respect to humans, estrogen therapy prevents postmenopausal bone loss (27). In the present study, it was revealed that skeletal responses to estrogen were strain-dependent where C57BL/6J improved most in regard to trabecular vBMD, BV/TV, Tb.Th, cortical BMD and Cs.Th. As for Conn.D, estrogen failed to significantly increase this as this index was thought to be irreversible when bone loss has occurred (28), as it was unclear whether the anabolic treatment parathyroid hormone restored any connectivity (29,30).

The findings of the present study raise notable issues for selecting an animal model to study the mechanism of certain genes in postmenopausal osteopenia and evaluate the effectiveness of a novel candidate agent for treatment. Following OVX at 4 months old, C57BL/6J prostaglandin E2 receptor (one of the four prostanoid receptors) knockout mice exhibited protection against bone loss in femur and L4 vertebrae (31). However, when OVX was performed on 30-week-old female Col.1-PPARγ mice, ovariectomy-induced bone loss was accelerated (10). As the present results revealed that the factor age of C57BL/6J influenced the sensitivity of bone to OVX, the conclusions of a specific gene on osteoporosis must be identified with care. In addition, BALB/c, ICR and Kunming mice have been exploited in searching for anti-osteoporotic medicine, and drug testing results may be negative as the present study revealed that C57BL/6J was the most sensitive to the anti-osteoporotic drug estrogen.

There are certain limitations associated with this study. First, only one bone (femur) was assessed to evaluate the extent of OVX-induced bone loss and response to estrogen. Second, only a single time point (8 weeks) after OVX was assessed and therefore the rate of bone loss was not assessed. Third, no histomorphometry or biochemical markers were performed, and thus it is impossible to observe the underlying cellular mechanisms that determine the different strain-associated responses to estrogen deficiency. In spite of these shortcomings, conclusions were made by ex vivo analysis of femoral bone microarchitecture by µCT, as the bone (femur), the postsurgery time (8 weeks) and method (µCT) were the most commonly used in previous literature.

In conclusion, the results revealed that OVX-induced bone loss may be age- and strain-specific, which emphasizes the importance of appropriate selection of mouse strains and their age in postmenopausal osteoporosis research. It was demonstrated that C57BL/6J female mice subjected to OVX at 8 weeks old resulted in pronounced loss of bone that was sensitive to estrogen. Therefore, the present data may give novel insights into the age- and strain-associated effect of OVX in regulating turnover of bone in female mice, and this must be considered when using C57BL/6J mice as an animal model of postmenopausal osteoporosis.

Acknowledgements

The authors would like to thank Jason Kang for editing the manuscript. The present study was supported by the National Science Foundation for Distinguished Young Scholars of China (grant no. 81125013) and National Natural Science Foundation of China (grant no. 81472116).

References