Both hPTH(1–34) and bFGF Increase Trabecular Bone Mass in Osteopenic Rats but They Have Different Effects on Trabecular Bone Architecture


  • The authors have no conflict of interest.


Osteoporosis is a syndrome of excessive skeletal fragility that results from both the loss of trabecular bone mass and trabecular bone connectivity. Recently, bFGF has been found to increase trabecular bone mass in osteoporotic rats. The purpose of this study was to compare how trabecular bone architecture, bone cell activity, and strength are altered by two different bone anabolic agents, bFGF and hPTH(1–34), in an osteopenic rat model.

Materials and Methods: Six-month-old female Sprague-Dawley rats (n = 74) were ovariectomized (OVX) or sham-operated (sham) and maintained untreated for 2 months. Then OVX rats were subcutaneously injected with basic fibroblast factor (bFGF; 1 mg/kg, 5 days/week), human parathyroid hormone [hPTH(1–34); 40 μg/kg, 5 days/week], or vehicle for 60 days (days 60–120 ). Sham-operated and one group of OVX animals were injected with vehicle. Biochemical markers of bone turnover (urinary deoxypyridinoline cross-links; Quidel Corp., San Diego, CA, USA) and serum osteocalcin (Biomedical Technologies, Stroughton, MA, USA) were obtained at study days 0, 60, 90, and 120 and analyzed by ELISA. At death, the right proximal tibial metaphysis was removed, and microcomputed tomography was performed for trabecular bone structure and processed for histomorphometry to assess bone cell activity. The left proximal tibia was used for nanoindentation/mechanical testing of individual trabeculae. The data were analyzed with Kruskal Wallis and post hoc testing as needed.

Results: Ovariectomy at day 60 resulted in about a 50% loss of trabecular bone volume compared with sham-treated animals. By day 120 post-OVX, OVX + vehicle treated animals had decreased trabecular bone volume, connectivity, number, and high bone turnover compared with sham-operated animals [p < 0.05 from sham-, hPTH(1–34)-, and bFGF-treated groups]. Treatment of OVX animals with bFGF and hPTH(1–34) both increased trabecular bone mass, but hPTH(1–34) increased trabecular thickness and bFGF increased trabecular number and connectivity. Histomorphometry revealed increased mineralizing surface and bone formation rate in both bFGF and hPTH(1–34) animals. However, osteoid volume was greater in bFGF-treated animals compared with both the hPTH(1–34) and OVX + vehicle animals (p < 0.05). Nanoindentation by atomic force microscope was performed on approximately 20 individual trabeculae per animal (three animals per group) and demonstrated that elastic modulus and hardness of the trabeculae in bFGF-treated animals were similar to that of the hPTH-treated and sham + vehicle-treated animals.

Conclusion: Both hPTH(1–34) and bFGF are anabolic agents in the osteopenic female rat. However, hPTH(1–34) increases trabecular bone volume primarily by thickening existing trabeculae, whereas bFGF adds trabecular bone mass through increasing trabecular number and trabecular connectivity. These results suggest the possibility of sequential treatment paradigms for severe osteoporosis.


OSTEOPOROSIS IS A SYNDROME of reduced bone mass that results from both estrogen deficiency and aging,(1,2) and accounts for some, but not all of the increased fracture risk. Deterioration of the trabecular bone with estrogen loss leads to a reduction in both trabecular bone mass and an irreversible alteration of trabecular bone structure. Antiresorptive agents effectively suppress trabecular bone degeneration because of estrogen deficiency.(3–8) These agents work by reducing the activation of new bone remodeling units while still allowing bone to form on already activated bone remodeling units (BMUs). However, inhibition of bone turnover alone may be insufficient to reverse skeletal fragility in severe osteoporosis. Instead, it may be necessary to establish a scaffold of new trabeculae. Under these circumstances, bone anabolic agents may be needed to increase bone mass by stimulating bone formation. Parathyroid hormone [PTH(1–34)] is one of the most studied bone anabolic agents. Its primary action is to increase bone mass by adding new bone to the pre-existing trabecular surface. By thickening the remaining trabeculae, hPTH(1–34) may be able to reconnect the old trabeculae together and restore bone strength.(9–14) However, PTH may not be as effective in severe osteoporosis, where there may be an insufficient number of remaining trabeculae to be able to significantly increase bone strength through thickening alone.(15–17)

Basic fibroblast factor (bFGF) and other skeletal growth factors have been found to stimulate uncommitted bone marrow stromal cells to differentiate into osteoblasts and form osteoid.(18–22) By using bFGF in severely osteopenic rats, Liang et al.(18) found that it markedly increased osteoblast surface and osteoid formation after 14 days of daily treatment at a dose of 200 μg/kg. In addition, they found evidence that bFGF formed new trabeculae de novo and provided two-dimensional evidence that connectivity or bridging of the trabeculae occurred.(18,21,22) Results from our own research group found that bFGF treatment for 14 days can create new trabecular elements, but with withdrawal of bFGF injections, the new trabeculae were rapidly lost through accelerated resorption.(23) However, after bFGF treatment with either estrogen or hPTH(1–34) the newly formed trabeculae produced by bFGF were maintained. These findings suggest that bFGF may be especially useful for the treatment of severe osteoporosis.

Based on these earlier findings, we hypothesize that both systemic bFGF and hPTH treatment will increase trabecular bone mass in osteopenic animals. However, the main action of bFGF will be to form new trabecular spicules and increase connectivity, whereas hPTH will primarily increase trabecular bone volume by thickening existing trabeculae. We further hypothesized that the new trabecular elements formed by bFGF treatment will have mechanical properties equivalent to the pre-existing bone.



Seventy-four 6-month-old retired breeder female Sprague-Dawley rats were obtained from Charles River, Inc. (San Jose, CA, USA). The rats were maintained on commercial rat chow (22/5 Rodent Diet; Teklad, Madison, WI, USA) available ad libitum with 0.95% calcium and 0.67% phosphate. Rats were housed in a room that was maintained at 70°F with a 12-h light/dark cycle. All animals were treated according to the U.S. Department of Agriculture animal care guidelines with the approval of the University of California San Francisco Committee on Animal Research.

At 6 months of age, the rats were randomized by body weight into six groups with 10–14 animals each. Group 1 (n = 10) and group 3 (n = 10) were subjected to sham surgery, during which time the ovaries were exteriorized and replaced intact. Groups 2 (n = 12), 4 (n = 12), 5 (n = 14), and 6 (n = 14) had bilateral ovariectomies using the dorsal approach. After the surgeries, the food consumption of the OVX rats was restricted to that of the control rats (pair-feeding) to minimize the increase in body weights associated with ovariectomy. Approximately 18 g chow/day was fed to each rat, and tap water was supplied ad libitum. All rats were untreated for 2 months (60 days) after the surgery to allow for the development of moderate to severe cancellous osteopenia. Rat weight was measured every week to assess changes.


Lyophilized hPTH(1–34) (Bachem Inc., Torrance, CA, USA) was dissolved with a vehicle composed of 0.01N HCl acid saline and 2% heat-inactivated rat serum. The bFGF was a 146 amino acid, nonglycerate monomeric 16.5-kDa protein that was produced in genetically engineered yeast (Chiron Corp., Emeryville, CA, USA). It was diluted with a vehicle of PBS and stored at −80°C till the day of use. For labeling of bone forming surfaces, calcein (10 mg/kg ip) was given 14 and 4 days before death. For surgery, serum sample collection, and necropsy, the animals were anesthetized with intraperitoneal injections of ketamine (10 mg/kg) and xylazine (5 mg/kg).

Experimental protocol

All rats were untreated for 60 days after surgery to allow for the development of cancellous osteopenia. At day 60, approximately 2 months after ovariectomy, animals in group 1 (pretreatment sham control, n = 10) and group 2 (pretreatment OVX control, n = 12) were killed. Treatments were initiated at day 60 in groups 3–6 as follows: group 3, sham + vehicle, 5 days/week (killed at day 120, n = 10); group 4, OVX + vehicle, 5 days/week (killed at day 120, n = 12); group 5, OVX + PTH, 40 μg/kg subcutaneously 5 days/week (killed at day 120, n = 14); and group 6, OVX + bFGF, 1 mg/kg subcutaneously 5 days/week (killed at day 120, n = 14).

After 60 days of sham or ovariectomy, the animals were dosed 5 times/week with a vehicle composed of 0.01N HCl acid saline and 2% heat-inactivated rat serum, 40 μg/kg PTH(1–34) (Bachem Inc.), or 1 mg/kg bFGF (Chiron Corp.) for 60 days. The bFGF was diluted with PBS and stored at −80°C till the day of use. Demeclocycline (20 mg/kg ip) and calcein (10 mg/kg ip) were given 14 and 4 days before killing. The rats were weighed every week to assess the effects of diet and dosing. For surgery and necropsy, the animals were anesthetized with intraperitoneal injections of ketamine (10 mg/kg) and xylazine (5 mg/kg). At days 0 (before surgery), 60, 90, and 120, rats were housed in individual metabolic cages, and a fasting 24-h urine sample was collected. Serum samples were obtained on the same day (between 10:00 a.m. and 3:00 p.m.) as the urine samples. For collection of the serum samples, animals were anesthetized, and blood was obtained from the tail vein. On day 60 or 120, all animals were anesthetized with ketamine/xylazine anesthesia and killed by exsanguination from the abdominal aorta. Additional blood samples were obtained at the time of death from the aorta. The urine and serum samples were stored at −80°C before assessment of bone markers. At necropsy, successful removal of the ovaries was confirmed by failure to detect ovarian tissue and by marked atrophy of the uterine horns. The right tibias were placed in 10% phosphate-buffered formalin for 24 h and then transferred to 70% ethanol for microcomputed tomography (μCT) and bone histomorphometry measurements.

Biochemical assays

Urinary levels of deoxypyridinoline cross-links and creatinine (DPD and Cr, respectively) were analyzed in duplicate using rat ELISA kits from Metrobiosystems (Mountain View, CA, USA). Serum levels of osteocalcin (OSC) were measured using a rat sandwich ELISA kit from Biomedical Technologies (Stroughton, MA, USA). The manufacturer's protocols were followed, and all samples were assayed in duplicate. A standard curve was generated from each kit, and the absolute concentrations were extrapolated from the standard curve. The CVs were less than 8% for DPD and Cr and less than 10% for OSC in our laboratory, which is similar to the manufacturer's reference.(23)


The right proximal tibial metaphyses were imaged without further sample preparation with a desktop μCT (μCT-20; Scanco Medical, Bassersdorf, Switzerland), with a resolution of 26 μm in all three spatial dimensions.(24) The scans were initiated from the growth plate distally in 26-μm sections, for a total of 120 slices per scan. From this region, 60 slices starting at a distance of 1 mm distal from the lower end of the growth plate(25) and encompassing a volume of 1.56 mm length were chosen for the evaluation. The trabecular and the cortical regions were separated with semi-automatically drawn contours.(24,25)

The complete secondary spongiosa of the proximal tibia was evaluated, thereby completely avoiding sampling errors incurred by random deviations of a single section. The resulting gray-scale images were segmented using a low-pass filter to remove noise, and a fixed threshold was used to extract the mineralized bone phase.(25,26) From the binarized images, structural indices were assessed with three-dimensional (3D) techniques for trabecular bone.

Relative bone volume (BV/TV), trabecular number (Tb.N), thickness (Tb.Th), and separation (Tb.Sp) were calculated by measuring 3D distances directly in the trabecular network(26) and taking the mean over all voxels. The diameter of spheres filling the structure was taken as Tb.Th, the thickness of the marrow spaces as Tb.Sp, and the inverse of the mean distances of the skeletonized structure was calculated as Tb.N. Because the complete structure is distance-transformed, thickness and separation for every single point was calculated, and the histograms of these indices were determined.

Bone surface (BS) was calculated from a tetrahedron meshing technique.(27) By displacing the surface of the structure in infinitesimal amounts (dr), the structure model index (SMI) was calculated.(28) The SMI quantifies the plate versus rod characteristics of trabecular bone, in which an SMI of 0 pertains to a purely plate-shaped bone, an SMI of 3 designates a purely rod-like bone, and values between stand for mixtures of plates and rods.(28) Furthermore, connectivity density based on the Euler number was determined.(25,29,30) In addition, a 3D cubical voxel model of bone was built, and cortical thickness (Ct.Th) was measured.(25)

Bone histomorphometry

The right proximal tibias were dehydrated in ethanol, embedded undecalcified in methylmethacrylate, and sectioned longitudinally with a Leica/Jung 2065 microtome in 4- and 8-μm-thick sections. The 4-μm sections were stained with von Kossa and Toluidine blue for collection of bone mass and architecture data with the light microscope, whereas the 8-μm sections were left unstained for measurements of fluorochrome-based indices. Static and dynamic histomorphometry were performed using a semi-automatic image analysis OsteoMeasure System (OsteoMetrics Inc., Decatur, GA, USA) linked to a microscope equipped with transmitted and fluorescence light.(15,16)

A counting window, measuring 8 mm2 and containing only cancellous bone and bone marrow, was created for the histomorphometric analysis.(31) Static measurements included total tissue area (T.Ar), bone area (B.Ar), and bone perimeter (B.Pm). Dynamic measurements included single- (sL.Pm) and double-labeled perimeter (dL.Pm), osteoid perimeter (O.Pm), and interlabel width (Ir.L.Wi). These indices were used to calculate bone volume (BV/TV), trabecular number (Tb.N), trabecular thickness (Tb.Th) and trabecular separation (Tb.Sp), osteoid surface (OS/BS), mineralizing surface (MS/BS), and mineral apposition rate (MAR). Osteoid volume (OV/TV) was measured separately and was not included in the volume for cancellous bone. Mineralization lag time in days (MLT) was calculated as osteoid thickness/MAR. Finally, surface-based bone formation rate (BFR/BS) was calculated by multiplying mineralizing surface (single-labeled surface/2 + double-labeled surface) with MAR according to Parfitt et al.(32,33)

Mechanical property testing

For topographic imaging and discrete mechanical properties determination of individual trabeculae, a modified atomic force microscope (AFM; Nanoscope IIIa; Digital Instruments, Santa Barbara, CA, USA) was used. The modification consisted of replacing the cantilever/tip assembly of the microscope with a transducer driven head and tip (Triboscope; Hysitron, Minneapolis, MN, USA) that allowed the microscope to operate both as an imaging and an indentation instrument. The detailed modifications for this discrete indentation have been described in detail elsewhere.(34) All indentations were performed with a triangular load profile of 0.3mm/s in time to 300-μN maximum load. Elastic modulus and hardness were calculated from the unloading force/displacement slope at maximum load and the projected contact area at this load. The instrument was then further modified to perform dynamic stiffness imaging that allows simultaneous determination of surface topography and both storage and loss moduli by applying a small sinusoidal force on the AFM tip in contact mode and measuring the resulting displacement amplitude and its phase lag with respect the force. These quantities were used to determine the viscoelastic properties, pixel-by-pixel, as the tip scanned over the surface of the bone.(35) In the present work, the loss modulus was found to be less than 5% of storage modulus; therefore, we considered the storage modulus to be roughly equivalent to the elastic modulus (small viscoelastic effect).

The methylmetacrylate-embedded right proximal tibial metaphyses samples (approximately 3 mm thick) that had been used for bone histomorphometry were further polished on one side with progressively finer grades of diamond paste (down to 0.1 μm) until a smooth bone surface was exposed (approximate nanometer roughness). The AFM measurements were performed on different trabeculae on each specimen in both longitudinal as well as transverse orientations. Three right proximal tibial metaphyseal bone samples from each of the four treatment groups (sham, OVX, PTH, and bFGF) were tested (approximately 20 trabeculae per bone specimen). The elastic modulus (E) and hardness (H) were obtained by indentation along a line crossing the edge of the samples with an interval of 2 mm, covering a length of at least 30 mm for each trabeculae measured.


The mean and SD are reported at each time point for animal weights, biochemical markers of bone turnover (DPD cross-links and OSC), trabecular bone structure by μCT, bone turnover by histomorphometry, and mechanical properties. To determine significant differences at separate time points for animals weights and biochemical markers of bone turnover, repeated measures of variance was used with Tukey's post hoc tests (p < 0.05). To determine significant differences between groups for histomorphometry and μCT evaluation of trabecular bone structural variables, the non-parametric Kruskal-Wallis test was used. Data from groups 1, 2, and 3–6 were analyzed separately. When Kruskal-Wallis testing showed overall significant differences among groups, we applied Ryan's post hoc test to identify groups that were significantly different (SPSS Version 10; SPSS Inc., Chicago, IL, USA). Differences were considered significant at p < 0.05.(16,23)


General observations

In general, the animals tolerated the surgery and experimental treatments without complications. Animals in all treatment groups had modest body weight decreases of 5–7% from days 0 to 60 because both sham-operated and OVX animals were pair-fed only 1 g of food per day. Body weights stabilized from days 60 to 120 in all animals groups. All animals were significantly lighter than baseline at days 60 and 120. However, there were no differences between the study groups at any of the time points (Table 1).

Table Table 1. Body Weights During the Experiment (gj mean ± SD)
original image

Animals in the bFGF-treated group tolerated treatment well until day 40; they developed orbital pallor. This most likely resulted from an bFGF-induced reduction in red cell production. At day 120, the mean hematocrit of the bFGF-treated animals was 25% lower than sham-operated, OVX + vehicle, and PTH-treated groups.

Analysis of biochemical markers in response to treatments

Bone resorption, measured by serial determination of DPD/Cr cross-links (Table 2) was significantly increased by 60% to nearly 100% at day 60 in all OVX animals (groups 2, 4, 5, and 6) compared with baseline pre-OVX day 0 levels and with sham-operated animals (groups 1 and 3) at day 60 (p < 0.05). By day 90, DPD/Cr cross-link excretion in the OVX + vehicle-treated animals remained stable, whereas in the animals in the OVX + hPTH(1–34) or OVX + bFGF groups, DPD/Cr cross-link excretion increased about 25% from day 60. Although these changes within groups were not significant from day 60 values, they were significantly higher than sham-operated animals at day 90 (p < 0.05) and the within-day 0 baseline values. By day 120, in all OVX groups (groups 4, 5, and 6), DPD/Cr cross-links were decreased from 24% to 28% from day 90, although all OVX groups had DPD/Cr cross-link levels significantly higher than the sham-operated group at day 120 and 0 within-group baseline values (p < 0.05).

Table Table 2. Changes in Biochemical Markers of Bone Turnover From Day 0 to Day 120 (Mean ± SD)
original image

Bone formation, measured by serial determination of serum osteocalcin (Table 2), was significantly increased in all OVX animals compared with the sham-operated animals at days 60, 90, and 120. OSC levels in the OVX + vehicle-treated group increased approximately 40% over baseline levels by day 60 and remained elevated at this level until day 120 (Table 2). The OSC levels in the hPTH(1–34)-treated animals increased 32% from days 0 to 60, remained unchanged from days 60 to 90, and increased an additional 35% from days 90 to 120, which was significantly different from day 0 values (p < 0.05). The OSC levels in the bFGF-treated animals also increased by 32% from baseline, was unchanged from days 60 to 90, and was further increased by 34% from days 90 to 120, which was significantly different from day 0 values (p < 0.05).

Trabecular and cortical bone structural changes measured by μCT

At day 60, when compared with the sham animals, total trabecular volume significantly decreased by 53%, trabecular number decreased by 35%, and connectivity decreased by 60%, whereas trabecular and cortical thickness did not change in the OVX group (Table 3). The SMI reflects the pattern of trabeculae (combination of rods and plates). The SMI in the sham group was 1.9, reflecting about 40% rods and 60% plate-like trabeculae, whereas in the OVX group, it was 2.7, reflecting a ratio of about 90% rod-like trabeculae and 10% plate-like trabeculae.

Table Table 3. Trabecular Bone Structural Variables of the Proximal Tibia by μCT (Mean ± SD)
original image

At day 120, when compared with the sham + vehicle animals, total trabecular bone volume (BV/TV) decreased by 60% in the OVX + vehicle group (Table 3; Figs. 1 and 2). Trabecular bone connectivity density and trabecular number were also significantly decreased 75% and 27%, respectively. The SMI increased by 40% in the OVX + vehicle group compared with the sham + vehicle group. Trabecular and cortical thickness did not change in the OVX + vehicle group compared with sham + vehicle group (Figs. 1 and 2).

Figure FIG. 1..

(A) Connectivity density and (B) trabecular number measured by μCT in sham control and OVX + vehicle-treated rats. hPTH(1–34) or bFGF at day 120.ap < 0.05 from sham + vehicle and OVX + bFGF;bp < 0.05 from sham + vehicle.

Figure FIG. 2..

One proximal tibia per group was 3D imaged by XTM.(29,30) At day 120, when sham + vehicle was compared with OVX + vehicle, there was less trabecular bone and trabecular connectivity. OVX + hPTH(1–34) had increased trabecular thickness compared with OVX + vehicle. OVX + bFGF had increased trabecular number and connectivity compared with the OVX + vehicle group.

At day 120, hPTH(1–34) increased trabecular bone volume by 187% compared with the OVX + vehicle group, and this bone volume was similar to that of the sham + vehicle groups. Trabecular thickness increased by 62% compared with the OVX + vehicle group and increased by 36% compared with the sham + vehicle group. SMI of the hPTH(1–34)-treated rats was 1.3, reflecting that it had a mix of rod- and plate-like structures that was 54% lower than the OVX + vehicle rats and was 35% lower than the sham + vehicle rats. Trabecular connectivity and trabecular number did not increase significantly from the OVX + vehicle group, and trabecular connectivity with hPTH(1–34) treatment was significantly lower than in the sham + vehicle group. Cortical thickness was 30% greater in OVX + hPTH(1–34) (p < 0.05) and 16% greater in OVX + bFGF compared with sham-operated and OVX + vehicle-treated animals.

At day 120, trabecular bone volume was 170% higher than the OVX + vehicle group in the bFGF-treated group. Trabecular connectivity and trabecular number increased by 165% and 32%, respectively, in the bFGF-treated animals compared with the OVX + vehicle animals and were not significantly different from the sham + vehicle values. SMI in the OVX + bFGF group was 32% lower than that of the OVX + vehicle group and was similar to that of the Sham + vehicle group. Although there was a 33% increase in trabecular thickness and a 14% increase in cortical thickness in the OVX + bFGF animals compared with the OVX + vehicle animals, these were not statistically significant.

Histomorphometric measurements

Trabecular bone volume and structure in the proximal tibial metaphyses:

Trabecular bone volume (BV/TV) was maintained in the sham animals. However, compared with pretreatment sham animals, bone volume in the OVX + vehicle group decreased by 40% at day 60 and 52% at day 120. Both OVX + PTH and OVX + bFGF groups had bone volumes similar to the sham + vehicle levels. Compared with baseline animals, Tb.N was decreased by 30% at day 60 and by 50% at day 120 in OVX + vehicle animals, and it was increased to a level similar to the sham + vehicle group in the OVX + bFGF group. No significant decrease of Tb.Th was seen in the OVX + vehicle group. However, it increased by 54% in the OVX + PTH animals from the pretreatment sham level. Tb.Sp was increased by 47% at day 60 and by 101% at day 120 in OVX + vehicle animals compared with the sham + vehicle animals. hPTH(1–34) did not significantly lower Tb.Sp compared with OVX + vehicle animals (Table 4; Fig. 3).

Table Table 4. Trabecular Bone Variables of the Proximal Tibia by Histomorphometry (Mean ± SD)
original image
Figure FIG. 3..

Proximal tibial metaphyses from (A, a, a′) a vehicle-treated sham rat, (B, b, b′) a vehicle-treated OVX rat, (C, c, c′) an hPTH-treated OVX rat, and (D, d, d′) a bFGF-treated OVX rat at day 120. (B) Bone area was reduced significantly in the vehicle-treated OVX rat. Treatment with PTH increased trabecular bone area, (C) trabecular width, and (c) double-labeled surface and interlabel width. (D) Treatment with bFGF increased trabecular bone area and number, (d) maintained elevated double-labeled surface and interlabel width, (d′) and increased osteoid area (arrows). (A-D) von Kossa-stained sections, ×40. (a-d) Unstained sections, ×100. (a′-d′) Toluidine blue-stained sections, ×200.

Bone formation and resorption:

The mineralizing surface (MS/BS) increased by 73% at day 60 after OVX and remained at this level at day 120. It was significantly increased by treatments with both hPTH(1–34) and bFGF. Surface based bone formation rate (BFR/BS) increased 133% at day 60 and remained at this level at day 120 after OVX, and it was further increased significantly after treatment with both hPTH(1–34) and bFGF. The changes in MAR, bone formation rate/bone volume, and bone formation rate/tissue volume were similar to those for BFR/BS. Osteoid volume (OV/TV) did not change significantly after OVX or treatment with hPTH(1–34), but was increased by 30-fold in bFGF-treated animals (Table 4; Fig. 3). In addition, bFGF-treated animals had an MLT that was significantly higher than all other animal groups at day 120. Osteoclast surface increased by more than 2-fold at day 60 after OVX and remained increased at day 120. At day 120, all animals treated with hPTH(1–34) and bFGF had similar osteoclast surface compared with the sham + vehicle animals (Table 4; Fig. 3).

Mechanical properties measurements

The elastic modulus (E) and hardness (H) were obtained by indentation along a line crossing the edge of the samples with an interval of 2 μm covering a length of at least 30 μm to attempt evaluating new bones formed by bFGF and PTH treatments. Figure 4 shows the typical variation of elastic modulus and hardness for one set of sham, OVX, PTH, and bFGF samples. The figure shows topographic AFM images covering the indentations length. The fleshes indicate the directions of the indentations, and “0” designates the starting points of indents. The bone marrow zones are labeled as “a” in these images.

Figure FIG. 4..

Elastic modulus and hardness variation close to edges of trabeculae bones for a group of sham + vehicle-, OVX + vehicle-, hPTH-, and bFGF-treated samples. The pictures represent topographic AFM images obtained after the indentations. The origins and the directions of indentation are indicated by white arrows. The areas labeled as “a” are bone marrow zones.

The mean and SD of elastic modulus and hardness obtained on each group are shown in Fig. 5. The results clearly show no significant deviation of mechanical properties of newly grown bones treated by PTH and bFGF compared with sham and OVX + vehicle-treated animals.

Figure FIG. 5..

The mean value and SD of elastic moduli and hardness deduced from three different samples, each different location was sectioned in both longitudinal and proximal orientations.

Occasionally, however, areas with lower elastic modulus and hardness were observed on PTH- and bFGF-treated samples compared with sham and OVX + vehicle-treated animals. Figure 6 shows E-mapping of a “normal” bFGF-treated bone versus assumed “osteoid” bone on the outer edge of the trabeculae. The quantitative elastic moduli values along the black lines drawn on images are plotted as open and closed circles for comparison. Significant reduction in the elastic modulus is apparent on the osteoid bone in compare to normally observed bFGF samples. This change in elastic modulus at the edge of the trabeculae corresponds with the osteoid surface observed by histomorphometry for both the PTH- and bFGF-treated animals.

Figure FIG. 6..

Elastic modulus mapping of “normal” and assumed osteoid bones (insets). The values along the black lines on the images are plotted as open and closed circles.


This study shows that systemic treatment with either hPTH(1–34) or bFGF produced anabolic responses and restored cancellous bone mass to severely osteopenic OVX rats. However, treatment of OVX rats with hPTH(1–34) increased bone mass mainly by thickening trabeculae; PTH was not able to create new trabecular elements and thereby increase trabecular connectivity. In contrast, treatment of OVX rats with bFGF increased cancellous bone mass by forming new trabecular connections and therefore increased trabecular connectivity.

bFGF has been found to stimulate bone marrow stromal cells to differentiate into osteoblasts and form osteoid.(18–22) Liang et al.(18) reported that when osteopenic OVX rats were treated with 200 μg/kg bFGF for 14 days, osteoblast surface and osteoid volume were markedly increased. It seems that bFGF may form new trabeculae de novo and that the newly formed osteoid or trabeculae may serve as “bridges” to connect the pre-existing trabeculae together.(18) Dunstan et al.(36) treated OVX osteopenic rats with FGF-1 for 28 days with a dose of 200 μg/kg and found this peptide also increased bone formation and restored trabecular bone volume to nearly the level of sham control animals. Results from our own research group also found that bFGF treatment for 14 days created new trabecular elements, but with withdrawal of bFGF treatment, the new trabeculae were lost. Treatment with bFGF followed by estrogen or PTH seemed to maintain the newly formed trabeculae.(23) Although these findings suggest that bFGF may be especially useful for the treatment of severe osteoporosis, they did not answer questions such as whether the unmineralized bone (osteoid) induced by short-term bFGF treatment can be mineralized during prolonged bFGF treatment period and whether long-term bFGF administration would restore or increase trabecular connectivity. Our current study demonstrated that, by prolonging bFGF treatment period to 60 days with a subcutaneous dose of 1 mg/kg, it increased both unmineralized bone (osteoid) but also increased mineralized tissue as evidenced by increasing bone volume measured by μCT and bone histomorphometry. In addition, bFGF differed from PTH in that it increased bone volume by increasing new trabecular spicules, trabecular number, and connectivity whereas PTH primarily increased trabecular bone volume by thickening existing trabeculae.

The mechanical properties of bone are influenced by “bone mass” as well as bone size, bone quality, and cancellous bone architecture.(37–40) It is known that when an equivalent amount of bone is distributed as well connected, numerous, and thin trabeculae, it is biomechanically more competent than when arranged as disconnected, widely separated, and thick trabeculae.(39–42) Trabecular perforation, which results in shifting plate-like trabeculae to rod-like trabeculae, is a well-established characteristic of cancellous bone osteoporosis.(16,17,29,), (30,40,43,), (44,45) PTH administration may improve trabecular architecture somewhat as it increased trabecular thickness. However, PTH does not seem to increase connectivity, because there was no significant change in either trabecular number or spacing. On the other hand, bFGF administration restored trabecular number, partially restored trabecular connectivity, and reduced trabecular separation. Because some studies have reported that restoration of trabecular connectivity may be important for strength recovery,(39,46,47) it would be of great interest in future studies to find out which agent would have a greater effect in increasing cancellous bone strength and whole bone strength, because both hPTH(1–34) and bFGF increased trabecular bone mass to a similar level, but through different architectural routes.

bFGF treatment generated new trabecular bone spicules and increased trabecular connectivity, whereas PTH added new bone to existing bone surfaces. To determine if the newly formed bone was mechanically competent, we performed AFM nanoindentation on the individual trabeculae in all treatment groups and found that elastic modulus and hardness of the new bone was the same as the pre-existing bone. While we did not test whole bone strength in this study, other investigators have reported both bFGF and PTH increase whole bone strength of the femur and tibia in other estrogen deplete animal models.(17,21) In a large phase III clinical osteoporosis trial, recombinant hPTH(1–34) [rhPTH(1–34)] was found to reduce new vertebral fracture risk,(48) showing that rhPTH(1–34) both builds new bone and improves bone strength. Because our study only evaluated the mechanical properties of individual trabeculae, whole bone strength will need to be tested because other data suggest it may not reflect the changes we have observed.

The AFM indentation technique provides a powerful tool for mapping mechanical properties of biomaterials. In particular, because hard tissues such as bone and teeth(49,50) are anisotropic and heterogeneous, the local variations in mechanical properties detectable with the AFM technique often play an important role in controlling the mechanical behavior of these materials at macroscopic length scales. In the present work, AFM determined the mechanical properties of the newly formed bone that resulted from the different treatments. By establishing the equivalence of the elastic properties of the new bone tissue with that of the pre-existing tissue, we have facilitated future use of the finite-element method for simulating the effects of bFGF treatment on trabecular bone strength.

Bone resorption, as determined from deoxypyridinoline cross-links, as well as osteoclast surface and osteoclast number, was not affected by treatment with bFGF. In previous studies in which OVX rats were treated with bFGF for 14 days, increases in osteoblast surface, osteoid surface, and osteoid volume indicated that new bone was being formed.(16,19,20) However, bone mineralization was found to be impaired during the treatment period, as indicated by the lack of fluorochrome labeling despite the increased osteoblast surface. In the current study, by prolonging the injection period to 60 days, we found that, apart from increasing undermineralized bone tissue (osteoid), bFGF also increased mineralized bone as well. This was confirmed by increased bone volume measured by μCT (Fig. 2) and bone histomorphometry using von Kossa staining that showed mineralized tissue stained in black (Fig. 3). In addition, OSC level, double fluorochrome labels, and MARs, which reflect the osteoblastic function, were increased with bFGF treatment (Fig. 3). As a result, bone formation rate was increased to a similar level to that of hPTH(1–34) treatment, resulting in augmentation of overall cancellous bone mass.

As bFGF is a broad spectrum mitogen that affects the proliferation and differentiation of numerous cell types and tissue,(51,52) serious precautions should be considered before it is administered systemically. While many investigators have administrated bFGF intravenously,(17–19,36) we chose instead to inject bFGF subcutaneously, but with a five times higher dose (1 mg/kg sc versus 200 μg/kg iv).(53) The rats treated with this bFGF regimen seemed to tolerate the injection reasonably well, and they did not lose body weight during the injection period compared with the OVX rats treated with vehicle. However, hematocrit levels were about 50% lower in the bFGF-treated animals than those of vehicle-treated OVX rats and were similar to intravenous injection at the dose of 200 μg/kg for 14 days.(18,21) Because the side effects caused by bFGF treatment would most likely disappear with the termination of bFGF,(18,52) our current results suggest that systemic administration of bFGF could become a beneficial anabolic therapy for osteoporosis, given that it can restore the lost trabecular connectivity in the severely osteoporotic skeleton.

In summary, we demonstrated that treatment of OVX osteopenic rats with bFGF resulted in new trabecular bone formation and improved trabecular microarchitecture by increasing trabecular connectivity. In addition, we have demonstrated that the new trabecular elements were mechanically competent in that their elastic properties were equivalent to those of the pre-existing bone tissue. The anabolic effects of bFGF differ from those of hPTH(1–34), in that hPTH(1–34) mainly increased bone formation on pre-existing bone surfaces but failed to improve trabecular connectivity. The biomechanical implications of increased connectivity require further study.


This work was supported by National Institutes of Health Grant 1R01AR43052 and grants from the Rosalind Russell Arthritis Research Center.