TIEG-Null Mice Display an Osteopenic Gender-Specific Phenotype

Abstract

TGFβ Inducible Early Gene-1 (TIEG) was originally cloned from human osteoblasts (OB) and has been shown to play an important role in TGFβ/Smad signaling, regulation of gene expression and OB growth and differentiation. To better understand the biological role of TIEG in the skeleton, we have generated congenic TIEG-null (TIEG^-/-) mice in a pure C57BL/6 background. Through the use of DXA and pQCT analysis, we have demonstrated that the femurs and tibias of two-month-old female TIEG^-/- mice display significant decreases in total bone mineral content, density, and area relative to wild-type (WT) littermates. However, no differences were observed for any of these bone parameters in male mice. Further characterization of the bone phenotype of female TIEG^-/- mice involved mechanical 3-point bending tests, micro-CT, and histomorphometric analyses of bone. The 3-point bending tests revealed that the femurs of female TIEG^-/- mice have reduced strength with increased flexibility compared to WT littermates. Micro-CT analysis of femurs of two-month-old female TIEG^-/- mice revealed significant decreases in cortical bone parameters compared to WT littermates. Histomorphometric evaluation of the distal femur revealed that female TIEG^-/- mice also display a 31% decrease in cancellous bone area, which is primarily due to a decrease in trabecular number. At the cellular level, female TIEG^-/- mice exhibit a 42% reduction in bone formation rate which is almost entirely due to a reduction in double labeled perimeter. Differences in mineral apposition rate were not detected between WT and TIEG^-/- mice. Taken together, these findings suggest that female TIEG^-/- mice are osteopenic mainly due to a decrease in the total number of functional/mature OBs.

Introduction

TGFβ inducible early gene-1 (TIEG) was originally discovered in our laboratory as an early response gene following TGFβ treatment of human osteoblasts (OB) [1]. TIEG is expressed in numerous tissues [2-7] and is a member of the Krüppel-like family of transcription factors (KLF-10) which are known to be involved in antiproliferative and apoptotic inducing functions similar to those initiated by TGFβ action [8-10]. These factors bind to Sp-1-GC rich DNA elements via their zinc fingers and regulate the expression of genes involved in cell growth, differentiation and apoptosis [11-13]. The TIEG gene encodes a 480 amino acid (72 kDa) protein and has previously been shown to both activate and repress the transcription of a number of genes [14-16].

Previous studies by our laboratory have demonstrated that TIEG plays an important role in mediating TGFβ signaling. TIEG directly represses the expression of the inhibitory Smad 7 gene by binding to a Sp-1-like GC-rich sequence in the proximal promoter [14]. Additionally, TIEG induces the expression and activity of Smad 2 [17]. Further, we have demonstrated that overexpression of TIEG enhances the TGFβ activation of a Smad binding element reporter construct and increases the transcription of known TGFβ target genes including p21, PAI-1, and Smad 2 [14]. Overall, TIEG expression results in activation of the Smad signaling pathway and mimics the functions of TGFβ in multiple cell types.

TGFβ is known to be an important signaling molecule in bone and has significant effects on OB growth and differentiation. Despite conflicting results, the majority of studies indicate that TGFβ increases bone formation by recruiting OB progenitors and stimulating their proliferation resulting in an increased number of cells committed to the OB lineage. More specifically, two studies revealed that TGFβ promotes early stages of OB differentiation, while it blocks later stages of differentiation and mineralization [18,19]. The TGFβ family consists of three closely related isoforms: TGFβ1, β2, and β3, and all three isoforms have been shown to induce TIEG expression [20]. The major isoform found in bone is TGFβ1 which accounts for about 90% of all TGFβ in the bone microenvironment. The in vivo functions of the three TGFβ isoforms are highly divergent as determined by gene knockouts [21].

Since TGFβ is known to play important roles in multiple facets of bone biology, and since we have shown that TIEG is an important regulator of TGFβ signaling in OBs, and can mimic TGFβ action when overexpressed, we developed a TIEG knockout mouse model (TIEG^-/-) [22]. These animals were originally developed in a C57BL/6:129SVJ mixed breed genetic background. Calvarial OBs, isolated from the TIEG^-/- mice revealed decreased expression of important OB marker genes, including osteocalcin, osterix, and alkaline phosphatase [22]. Co-culture studies with osteoclast (OC) precursor cells demonstrated that fewer mature OCs developed when TIEG^-/- OBs were used to support OC differentiation compared to that of wild-type (WT) OBs [22]. Gene expression analysis of TIEG^-/- OBs revealed a decrease in the expression of RANKL and an increase in the expression of OPG relative to WT OBs, which could well explain the decreased support of OC differentiation by TIEG^-/- OBs [22].

Further characterization of these animals, through the use of pQCT and micro-CT, revealed a number of skeletal defects [23]. However, since the skeletons of mixed breed knockout mice can mask a number of bone defects, and since there are significant baseline differences between the skeletons of C57BL/6 and 129SVJ mouse strains, as well as quantitative differences in their skeletal responses to ovariectomy [24], we sought to create a congenic line of TIEG^-/- mice before further analyzing the basis for this bone phenotype. Our original mixed-breed (C57BL/6:129SVJ) TIEG^-/- mice have now been crossed with C57BL/6 WT mice for more than 10 generations. This manuscript describes a comprehensive analysis of the bone defects resulting from loss of TIEG expression in these congenic animals using micro-CT, DXA, pQCT, and histomorphometric techniques. These studies demonstrate that TIEG^-/- mice display a gender-specific osteopenic phenotype. These studies further define the biological role of TIEG in bone and demonstrate that TIEG is critical for normal bone formation and/or maintenance.

Materials and Methods

Animals

TIEG^-/- mice were originally developed in a C57BL/6:129SVJ mixed breed background as described previously [22]. To derive TIEG^-/- mice in a congenic background, we bred our mixed breed TIEG^-/- animals against C57BL/6 WT mice for at least 10 generations. Ten WT, male and female, and ten TIEG^-/-, male and female, 3-month-old mice were used for the 3-point bending studies. Twelve WT, male and female, and twelve TIEG^-/-, male and female, 2 month-old mice were used for the DXA and pQCT studies. Ten WT and ten TIEG^-/- 2-month-old female mice were used for micro-CT and histomorphometric analysis. All mice were housed in a temperature controlled room (22 ± 2°C) with a light/dark cycle of 12 hours. All mice had free access to water and were fed standard laboratory chow (Laboratory Rodent Diet 5001; PMI Feeds, Richmond, VA) ad libitum. To reduce variability among experiments, WT and TIEG^-/- littermates were utilized in all of the experiments performed in these studies. The Institutional Animal Care and Use Committee (IACUC) approved all animal care and experimental procedures.

Peripheral quantitative computed tomography (pQCT)

pQCT measurements were performed on tibias of 12 WT and 12 TIEG^-/- male and female mice at 2 months of age. The mice were anesthetized and placed in a supine position on a gantry using the Stratec XCT Research SA Plus using software version 5.40 (Norland Medical Systems, Fort Atkinson, WI). Slice images were measured at 1.9 mm (corresponding to the proximal tibial metaphysis) and at 9 mm (corresponding to the tibial diaphysis) from the proximal end of the tibia as described previously [23].

Dual-energy x-ray absorptiometry (DXA)

The same mice that were used for pQCT analysis were used in the DXA scans. Mice were anesthetized and placed on the Lunar PIXImus densitometer (software version 1.44.005; Lunar Corp, Madison, WI). Calibration of the machine was performed before scanning with the hydroxylapatite phantom provided by the manufacturer.

Micro-CT

Ten WT female mice and ten TIEG^-/- female mice were sacrificed using CO₂ and whole femurs were removed. Entire femurs were scanned at a voxel size of 12 × 12 × 12 μm using a Scanco Medical μCT 40 machine and data were processed on an HP AlphaStation DS15 (Scanco Medical AG, Basserdorf, Switzerland). Entire femurs (cancellous + cortical bone) were evaluated followed by site specific evaluation of cortical bone at the femoral midshaft and cancellous bone in the distal femur metaphysis. The threshold for analysis was determined empirically and set at 235 (grey scale 0-1,000). For the femoral midshaft, 10 slices (120 μm) of bone were evaluated and total cross-sectional tissue volume (cortical and marrow volume, mm³), cortical volume (mm³), marrow volume (mm³), and cortical thickness (μm) were measured. For the femoral metaphysis, 125 slices (1.5 mm) of bone were evaluated and included secondary spongiosa only. Direct cancellous bone measurements in the distal femur included: 1) tissue volume (mm³), cancellous bone volume (bone volume unadjusted for tissue volume, mm³), 2) cancellous bone volume/tissue volume (volume of total tissue occupied by cancellous bone, %), 3) trabecular number (number of trabeculae within the tissue sample, 1/mm), 4) trabecular thickness (mean thickness of individual trabeculae, μm), and 5) trabecular separation (the distance between trabeculae, μm).

Mechanical testing

Three-point bending tests were performed on the left femurs of 10 WT, male and female, and 10 TIEG^-/-, male and female, mice at 3 months of age. The mice were sacrificed using CO₂ and the femurs dissected out and cleaned to be free of ligaments, muscle, and tendons. The three-point bending tests were performed on cleaned femurs as previously described [23]. Briefly, femurs were placed in the anteroposterior direction on a metal support and a load was applied in the middle of the femoral shaft. The ultimate force, stiffness and work to failure were measured for each femur.

Calcein labeling

Ten WT and ten TIEG^-/- 2-month-old female mice were injected with Calcein (15 mg/kg) subcutaneously 5 days and 2 days before sacrificing. Mice were sacrificed using CO₂ and femurs were harvested, as described above, and fixed in 10% neutral formalin overnight. Femurs were switched to 70% ethanol and stored at 4°C until time for histomorphometric analysis.

Histomorphometric analysis

After fixation in 70% ethanol, distal femurs were dehydrated in graded ethanols and xylene, and embedded, undecalcified, in modified methyl methacrylate for histomorphometric evaluation of cancellous bone. Longitudinal sections (4 μm thick) were cut with a vertical bed microtome (Leica 2065) and affixed to slides pre-coated with a 1% gelatin solution. One femoral section per animal was stained according to the von Kossa method with a Tetrachrome counterstain (Polysciences, Warrington, PA) and used for determining cancellous bone area and cellular endpoints. A second femoral section was left unstained and used for assessing fluorochrome labeling and dynamic measurements of bone formation. Histomorphometric data were collected with the OsteoMeasure System (OsteoMetrics, Inc., Atlanta, GA). The measurement area consisted of secondary spongiosa at distances greater than 0.25 mm from the growth plate. On average, 1.25 mm² of tissue (marrow and bone) was evaluated in each section. Cancellous bone area was measured and expressed as a percentage of tissue area. Osteoblast and osteoclast perimeters were also measured and expressed as a percentage of cancellous bone perimeter. Trabecular thickness, trabecular number, and trabecular separation were calculated based on measures of perimeter and area [25]. Measured fluorochrome-based indices of bone formation included double labeled perimeter, expressed as a percentage of total perimeter, and mineral apposition rate. Bone formation rate was calculated by multiplying double labeled perimeter by mineral apposition rate and normalized to bone perimeter, bone area, or tissue area. All histomorphometric data are reported in accordance with standard bone nomenclature [26].

Statistical Analysis

Data were analyzed by unpaired t-tests. Differences were considered significant at p-values < 0.05. All data are expressed as mean ± standard error.

Results

pQCT DXA and micro-CT analysis of WT and TIEG^-/- mice

As an initial approach to determine if significant differences in specific bone parameters exist between WT and TIEG^-/- mice, we performed pQCT analysis on the left tibias of twelve WT, male and female mice, and twelve TIEG^-/-, male and female mice at 2 months of age. As shown in Table 1, the tibias of 2-month-old female TIEG^-/- mice exhibited significant decreases in total bone content, density, area; cortical bone content, density and area; and cortical bone thickness relative to WT littermates. The differences in these parameters ranged between 3% and 11%. In contrast, the tibias of male TIEG^-/- mice did not show any significant difference in these bone parameters compared to WT littermates (Table 1). To determine the total body bone mineral density and content, excluding the head, we performed DXA analysis on ten WT, male and female mice, and ten TIEG^-/-, male and female mice. As shown in Table 1, only the female TIEG^-/- mice exhibited statistically significant decreases in total, lumbar, femur and tibia bone mineral density as well as total bone mineral content, when compared to WT littermates. Since multiple bone parameters exhibited statistically significant decreases in the bones of female TIEG^-/- mice, we performed micro-CT analysis on ten femurs isolated from WT and TIEG^-/- female animals. As shown in Table 2, femoral bone mass (cortical and cancellous bone) was significantly reduced by 10% in TIEG^-/- female mice relative to WT controls. Analysis of a subsample of cortical bone in the midshaft of the femur revealed that cortical volume was decreased by 8% in female TIEG^-/- mice compared to WT controls. There was also a tendency for decreased cross-sectional volume (P<0. 053) and cortical thickness (P<0.072) in female TIEG^-/- mice relative to WT mice. Tissue volume for evaluation of cancellous bone in the distal femur was lower in female TIEG^-/- mice than in WT mice, indicating that the cancellous compartment is smaller in TIEG^-/- mice compared to WT mice. There were no significant differences in cancellous bone mass or architecture (bone volume, bone volume/tissue volume, trabecular number, thickness, or spacing) between TIEG^-/- and WT mice as detected by micro-CT.

Table 1.

2 Month Congenic Diaphysis Measurements

Parameter	Female TIEG+/+	Female TIEG -/- *=P<0.05	Female % Difference	Male TIEG+/+	Male TIEG -/- *=P<0.05	Male % Difference
pQCT: Total Content (mg)	0.81 ± 0.01	0.73 ± 0.02*	9.90%	1.02 ± 0.03	0.99 ± 0.03	2.94%
pQCT: Total Density (mg/cm3)	669.33 ± 7.28	643.45 ± 5.66*	3.87%	683.89 ± 7.81	688.05 ± 6.37	0.60%
pQCT: Total Area (mm2)	1.21 ± 0.02	1.14 ± 0.02*	5.79%	1.48 ± 0.04	1.44 ± 0.03	2.70%
pQCT: Cortical Content (mg)	0.59 ± 0.01	0.52 ± 0.01*	11.86%	0.76 ± 0.03	0.75 ± 0.02	1.32%
pQCT: Cortical Density (mg/cm3)	1044.82 ± 8.12	1012.65 ± 4.92*	3.08%	1078.02 ± 9.15	1079.67 ± 7.49	0.15%
pQCT: Cortical Area (mm2)	0.56 ± 0.01	0.51 ± 0.01*	8.93%	0.70 ± 0.02	0.69 ± 0.02	1.43%
pQCT: Cortical Thickness (mm)	0.17 ± 0.003	0.15 ± 0.003*	11.76%	0.188 ± 0.004	0.188 ± 0.003	0.00%
DXA: Total BMD (gm/cm2)	0.043 ± 0.0004	0.041 ± 0.0005*	4.65%	0.045 ± 0.0005	0.045 ± 0.0004	0.00%
DXA: Lumbar BMD (gm/cm2)	0.049 ± 0.001	0.046 ± 0.001*	6.12%	0.049 ± 0.001	0.051 ± 0.0005	3.92%
DXA: Femur BMD (gm/cm2)	0.055 ± 0.001	0.052 ± 0.001*	5.45%	0.062 ± 0.0015	0.064 ± 0.001	3.13%
DXA: Tibial BMD (gm/cm2)	0.044 ± 0.0005	0.041 ± 0.0005*	6.80%	0.047 ± 0.0004	0.045 ± 0.0005	4.25%
DXA: Total BMC (gm)	0.334 ± 0.005	0.303 ± 0.006*	9.20%	0.367 ± 0.01	0.371 ± 0.006	1.08%

Table 2.

Micro-CT Analysis of 2 Month Old Female Congenic Mouse Femurs

Parameter	Female TIEG+/+	Female TIEG -/- *=P<0.05	Female % Difference
Total Femur (Cortical + Cancellous)
Bone Volume (mm3)	17.3 ± 0.3	15.6 ± 0.3*	9.83%
Midshaft Femur (Cortical)
Cross-sectional volume (mm3)	0.161 ± 0.002	0.150 ± 0.003	6.83%
Cortical Volume (mm3)	0.077 ± 0.002	0.071 ± 0.001*	7.79%
Marrow Volume (mm3)	0.084 ± 0.001	0.079 ± 0.002	5.95%
Cortical Thickness (μm)	207.0 ± 4.0	197.0 ± 3.0	4.83%
Distal Femur (Cancellous)
Tissue Volume (mm3)	1.995 ± 0.052	1.795 ± 0.076*	10.03%
Bone Volume (mm3)	0.051 ± 0.009	0.035 ± 0.005	31.37%
Bone Volume/Total Volume (%)	2.54 ± 0.37	1.93 ± 0.24	24.02%
Trabecular Number (1/mm)	3.09 ± 0.08	2.71 ± 0.20	12.30%
Trabecular Thickness (μm)	35.0 ± 2.0	37.0 ± 2.0	5.41%
Trabecular Spacing (μm)	325.0 ± 8.0	389.0 ± 33.0	16.45%

Mechanical testing

Since a significant decrease in bone mass was detected in the femurs of female TIEG^-/- mice, 3-point bending assays were used to measure femoral strength. This analysis was performed on ten WT, male and female mice, and ten TIEG^-/-, male and female mice, at 3 months of age. As shown in Figure 1, and consistent with the pQCT, DXA and micro-CT results, the force to failure and stiffness were decreased only in female TIEG^-/- mice relative to WT littermates. Interestingly, the work to failure is actually greater for femurs isolated from female TIEG^-/- mice suggesting that these bones may be more flexible than those of WT animals, possibly as a result of decreased bone mineral content and bone mineral density. In any event, these data indicate that the mechanical properties of bones isolated from TIEG^-/- female, but not male, mice differ greatly from those of WT animals.

Figure 1.

Bone strength and stiffness as determined by 3-point bending tests. The femurs of ten WT, male and female mice, and ten TIEG^-/-, male and female mice, were subjected to 3-point bending tests at 3-months of age. The force to failure, stiffness and work to failure were determined. Error bars represent standard error of the mean. Asterisks denote significance at the p < 0.05 level.

Histomorphometric analysis of 2-month-old WT and TIEG^-/- female mice

Based on the observation of decreased bone content, density, and strength only in female TIEG^-/- mice, it was of interest to further examine this bone phenotype at the cellular level. Histomorphometric analysis was performed on femurs isolated from ten WT female mice and ten TIEG^-/- female mice at 2 months of age. The histological evaluation, as shown in Figure 2A, clearly demonstrates that there is a significant defect in the cancellous bone of TIEG^-/- mice relative to WT littermates. As shown in Figure 2B, TIEG^-/- mice display a 31% decrease in cancellous bone area (B.Ar/T.Ar) which is primarily due to a 22% decrease in trabecular number (Tb.N). There is also a trend toward decreased trabecular thickness (Tb.Th) in female TIEG^-/- mice (10%, p ≈ 0.1). In accordance with a reduction in trabecular number in female TIEG^-/- mice, there is a 32% increase in the trabecular separation (Tb.Sp) in the femurs of these animals (Figure 2B).

Figure 2.

Histological analysis of the distal femur metaphysis of WT and TIEG^-/- female mice. A). Four micron thick sections were stained using the von Kossa method with a tetrachrome counterstain and viewed using a 10× objective. Note the decrease in trabecular number (black staining) in the TIEG^-/- female mice relative to WT littermates. B). Static histomorphometric analysis of WT and TIEG^-/- female mouse femurs. The femurs of ten 2-month old female WT and TIEG^-/- littermates were processed for static histomorphometric analysis of cancellous bone. Bones were analyzed for indicated parameters and the percent difference between WT and TIEG^-/- mice is depicted below each graph. Error bars represent standard error of the mean. Asterisks denote significance at the p < 0.05 level.

In light of the striking defects observed in the femurs of female TIEG^-/- mice with regard to trabecular number, it was of interest to determine if this defect was at least in part due to differences in the number of osteoclasts or osteoblasts lining the bone surface of these animals (Figure 3A and B). As shown in Figure 3C, no significant differences in osteoclast perimeter/bone perimeter were detected with genotype. However, there was a highly significant decrease (p < 0.01) in the osteoblast perimeter/bone perimeter of TIEG^-/- mice (12.4 ± 2.5%) relative to WT littermates (30.0 ± 5.6%) (Figure 3D).

Figure 3.

Histological analysis of the distal femur metaphysis of WT and TIEG^-/- female mice. A). Four micron thick sections were stained using the von Kossa method with a tetrachrome counterstain and viewed using a 56× objective. Note the decrease in osteoblast perimeter/bone perimeter of TIEG^-/- mice relative to WT littermates. B). For enlarged images, indicated areas from (A) were digitally magnified (3×). Osteoclast perimeter/bone perimeter (C) and osteoblast perimeter/bone perimeter (D) were determined by histomorphometric analysis of ten WT and TIEG^-/- female mouse femurs at 2 months of age. The percent difference between WT and TIEG^-/- mice is depicted below each graph. Error bars represent standard error of the mean. Asterisks denote significance at the p < 0.05 level.

To further analyze the basis for the observed bone phenotype, bone formation and mineralization rates were determined by dynamic histomorphometry. Female TIEG^-/- mice displayed a reduction in bone formation rate (BFR/B.Pm, BFR/B.Ar, BFR/T.Ar) which was almost entirely due to a 40% reduction in double labeled perimeter (dL.Pm/B.Pm) (Figure 4). Mineral apposition rate (MAR) in the female TIEG^-/- mice did not differ from WT mice. Overall, these finding suggest that female TIEG^-/- mice are osteopenic mainly due to a significant decrease in total osteoblast number and not likely the result of large differences in osteoblast activity.

Figure 4.

Dynamic histomorphometric analysis of WT and TIEG^-/- female mouse femurs. The femurs of ten 2-month old female WT and TIEG^-/- littermates were processed for dynamic histomorphometric analysis of cancellous bone. Bones were analyzed for indicated parameters and the percent difference between WT and TIEG^-/- mice is depicted below each graph. Error bars represent standard error of the mean. Asterisks denote significance at the p < 0.05 level.

Discussion

Our original TIEG^-/- mice were developed in a C57BL/6:129SVJ mixed breed background. Characterization of bones isolated from these mixed breed animals revealed several bone defects including decreases in bone content, density, and size [23]. Micro-CT analysis of the femoral head and vertebrae indicated increases in trabecular separation and decreases in cortical bone thickness and vertebral bone volume in the TIEG^-/- animals [23]. Additionally, transmission electron microscopy indicated a significant decrease in osteocyte number in the femurs of TIEG^-/- mice implying a potential role for TIEG in OB maturation/differentiation [23]. However, since the skeletons of mixed breed knockout mice can mask bone defects, and since there are significant baseline differences between the skeletons of C57BL/6 and 129SVJ mouse strains, complete characterization of the TIEG^-/- mouse bone phenotype was not carried out in these animals.

In order to further elucidate the biological functions of TIEG in bone in vivo, and to better characterize the bone phenotype of TIEG^-/- mice, we have now developed a purebred line of animals in a pure C57BL/6 genetic background. Compared to our earlier work using only female mice with a mixed background, analysis of congenic TIEG^-/- animals, as described in this paper, has revealed an even more striking bone defect. Interestingly, this defect is only observed in female mice as no changes in any of the bone parameters analyzed to date are present in male animals. Through the use of static and dynamic histomorphometry, we have now characterized this bone defect at the cellular level. Specifically, femurs and tibias isolated from 2-month-old female TIEG^-/- mice display significant decreases in multiple bone parameters relative to WT littermates including decreased bone content, density, and thickness. Three-point bending tests revealed that the femurs of female TIEG^-/- mice are much weaker compared to WT animals. Histomorphometric analysis of the distal femur revealed that female TIEG^-/- mice display a decrease in cancellous bone area associated with a decrease in trabecular number. Alterations in cancellous bone architecture were not clearly evident using micro-CT. This is probably due to differences in resolution between the microscope (∼1 micron) and the micro-CT (∼30 microns for the 12 micron voxel size used in these studies). At the cellular level, female TIEG^-/- mice exhibit a reduction in OB perimeter and bone formation rate. These findings suggest that female TIEG^-/- mice are osteopenic mainly due to decreases in total OB number and not in OB activity.

As described in the Introduction, we have demonstrated that OBs isolated from mixed background TIEG^-/- mice exhibit decreased expression levels of important OB marker genes including osteocalcin, osterix, and alkaline phosphatase [22]. Additionally, these TIEG^-/- OBs exhibit a delayed rate of mineralization relative to WT controls[22]. The reduced expression of these bone related genes in TIEG^-/- cells could in part explain the osteopenic phenotype observed in the present studies.

We have previously demonstrated that TIEG plays an important role in TGFβ mediated Smad signaling by inducing the expression of Smad 2 and repressing the expression of Smad 7 [14,17]. Additionally, we have shown that overexpression of TIEG in OBs mimics the actions of TGFβ1 by decreasing cellular proliferation and inducing the expression of alkaline phosphatase [4]. Thus, we hypothesized that TIEG plays an important role in OB and skeletal biology since: 1) TGFβ is a major regulator of OB growth and differentiation and plays a critical role in normal bone formation and maintenance; 2) TGFβ increases bone formation by recruiting OB progenitors and stimulating their proliferation, resulting in an increased number of cells committed to the OB lineage and; 3) TGFβ promotes early stages of OB differentiation (bone matrix production), while it blocks later stages of differentiation and mineralization [18,19].

The TGFβ family consists of three highly homologous isoforms classified as TGFβ1, β2, and β3. Mice deficient for TGFβ1, β2, and β3 have all been shown to develop severe bone defects [21]. Specifically, TGFβ1 KO mice display decreased bone mineral content, with a near absence of OBs in cancellous bone, resulting in an osteopenic phenotype [27]. TGFβ2 null mice contain numerous bone defects (bone loss) in the rib, sternum, vertebrae and long bones [28]. TGFβ3 null mice also exhibit loss of bone [29,30]. There are currently no data available concerning the effects of TGFβ1 or TGFβ3 overexpression in bone. However, overexpression of TGFβ2, under the control of the osteocalcin promoter, results in an age-dependent loss of bone mass resembling osteoporosis [31]. These results are likely explained by the fact that osteocalcin is expressed at late stages of differentiation and the fact that TGFβ inhibits OB differentiation at later stages.

In addition to TGFβ KO mice, disruption of TGFβ signaling has also been shown to have a significant impact on bone. Overexpression of a truncated TGFβ type II receptor, which is incapable of mediating TGFβ signaling in mice, under the control of the osteocalcin promoter, results in an age-dependent increase in trabecular bone mass [32]. As discussed above, it is expected that the disruption of TGFβ signaling during late stages of OB differentiation would result in increased bone formation. Finally, disruption of TGFβ signaling through deletion of Smad 3 results in an osteopenic phenotype [33]. The present study demonstrates that female TIEG^-/- mice exhibit a bone phenotype similar to that observed in TGFβ1 knockout animals, further implying an important role for TIEG in mediating the actions of TGFβ1 in bone.

Based on the observed gender specific osteopenic phenotype (females only), it is possible that E₂ is involved in mediating this phenomenon. The skeleton is one of the main targets of E₂ action in the body as it regulates bone growth and remodeling. Decreased E₂ levels are known to be one of the major causes of osteoporosis and play a critical role in regulating bone metabolism and homeostasis [34-36]. Interestingly, studies from our laboratory indicate that E₂ induces the expression of TIEG in OB [7]. Estrogen receptor α and β double knockout mice display profound decreases in trabecular bone volume only in female animals with no differences in male mice [37]. The bone phenotype of these mice closely resembles that of our TIEG^-/- mice suggesting that TIEG may play an important role in mediating the effects of E₂ in bone. There are two probable causes, each of which could create the gender specific phenotype identified in our TIEG^-/- mice. First, it is possible that E₂ is enhancing bone loss in female TIEG^-/- mice or, second, that testosterone is protecting male TIEG^-/- mice from bone loss. Detailed studies involving ovariectomy and orchiectomy, in combination with hormone replacement therapy, are currently underway to address the basis for the observed gender specific phenotype.

In summary, histomorphometric analysis of femurs isolated from congenic TIEG^-/- female mice reveals defects in multiple bone parameters. It appears that the major basis for this osteopenic phenotype is primarily due to a reduction in OB number and not OB activity. Taken together, these data demonstrate that loss of TIEG expression results in an osteopenic phenotype only in female mice, suggesting a potential role for TIEG in mediating E₂ (as well as TGFβ) actions in bone.