Abstract
Thyroid hormones (T3 and T4) regulate bone development, growth, and turnover. Studies have suggested that different skeletal sites respond differently to thyroid hormones. Therefore, we examined the in vitro T3 responsiveness of cells committed to the osteoblast lineage as a function of skeletal location. Bone marrow cells derived from female rat femurs and vertebrae were cultured using conditions that induce osteogenic differentiation. Cells from both sites formed mineralized bone nodules in primary and secondary culture. In femoral cultures, collagen type I (coll I) and osteocalcin (OC) messenger RNA (mRNA) levels increased from the earliest time point examined (day 3) to a maximum on day 12 and thereafter declined to undetectable levels. T3 increased both OC and coll I mRNA, resulting in a continuous expression throughout the culture period. Insulin-like growth factor I (IGF-I) gene expression was detected at very low levels by Northern analysis of femoral total RNA, and T3 only marginally enhanced IGF-I mRNA levels. In vertebral cultures, OC and coll I mRNA levels also increased with time in culture, but remained expressed throughout the culture period. OC and coll I mRNA levels were not markedly altered in response to T3. In contrast to femoral cells, IGF-I gene expression was easily visualized in Northern blots from untreated vertebral cultures and was markedly increased by the addition of T3. The continuous presence of T3 (10−7m) in the medium for 18 days caused a marked decrease in the number of alkaline phosphatase-positive colonies formed in femoral secondary cultures, but only a slight decrease in the number in vertebral cultures. In addition, short term (6 days) exposure to T3 (10−7m) at the beginning of the culture period decreased alkaline phosphatase activity in femoral cultures, but not in vertebral cultures. These findings indicate that there are skeletal site-dependent differences in the in vitro responses of cells of the osteoblastic lineage to thyroid hormone.
THYROID hormone (T3) exerts profound effects on skeletal development and maintenance. Along with other systemic hormones and intrinsic cellular factors, T3 promotes both the formation and breakdown of bone (1, 2). Thyroid hormone deficiency results in abnormalities in skeletal development and bone formation, and thyroid hormone excess increases bone resorption, leading to a loss of bone mass and osteoporosis (3, 4). Thyroid hormone has been reported to act on osteoblasts to indirectly stimulate osteoclastic bone resorption (5). In addition, thyroid hormone-induced bone loss in rats is associated with increased messenger RNA (mRNA) levels of alkaline phosphatase (AP), tartrate-resistant acid phosphatase, and histone H4 in femurs, but not in vertebrae. This suggests a heterogeneous skeletal response to thyroid hormone (6).
The importance of T3 for normal cell function is well established, but the molecular and cellular mechanisms of T3 action in bone are unclear. The molecular effects of T3 are mediated by the binding of T3 to nuclear thyroid hormone receptors (TRs) that bind to response elements on specific genes. A total of four TR isoforms have been identified to date, TRα1, TRα2, TRβ1, and TRβ2, with TRα2 (c-erbAα2) being a nonligand-binding variant (7, 8). The TR isoforms have been identified in rat and human osteosarcoma cell lines (9, 10) as well as in primary osteoblast cultures (11, 12).
In the present study, we describe a reproducible in vitro system in which osteoprogenitors present in bone marrow derived from adult female rat femurs and vertebrae differentiate into osteoblasts capable of mineralizing their extracellular matrix. Specifically in this study, we investigated whether chronic T3 treatment has similar or different effects on osteoblast-related gene expression in cell populations derived from the two skeletal sites. Previously, it was reported that T3 stimulates the production of insulin-like growth factor I (IGF-I) in rat osteoblasts in vitro (13) and increases IGF-I mRNA levels in a mouse osteoblastic clonal cell line (14). This growth factor is highly abundant in bone and acts in an autocrine and paracrine manner to support growth and differentiation of osteoblasts (15). These observations prompted us to examine the effect of T3 on IGF-I mRNA levels in bone marrow cell cultures derived from femurs and vertebrae. We also compared the effect of chronic vs. short term T3 treatment on AP-positive colony formation in femoral and vertebral cultures.
Materials and Methods
The adult rat bone marrow culture system was developed by examining various culture conditions reported to enhance osteoblast differentiation and mineralization (16–18). Culture conditions were optimized to simultaneously study osteoblast differentiation of femoral and vertebral bone marrow cells. The osteoblast phenotype was confirmed by the formation and mineralization of bone nodules in primary and secondary cultures, by the expression of osteoblast-related genes in secondary culture, and by the expression of AP enzyme activity in secondary culture. All reagents were obtained from Sigma Chemical Co. (St. Louis, MO) unless otherwise stated.
Cell culture
The animals were killed by exsanguination after injection with xylazine/ketamine. The experimental protocol was approved by the institutional animal care and use committee at the University of Massachusetts Medical School in Worcester, and the animals were maintained in accordance with NIH Guidelines for the Care and Use of Laboratory Animals.
For each cell culture preparation, the femurs and lumbar vertebrae were aseptically removed from five 100- to 125-g Sprague-Dawley female rats and immediately placed in cold Tyrode’s solution containing 50 ml antibiotic solution (10,000 IU/ml penicillin G, 10 mg/ml streptomycin, and 25 mg/ml amphotericin B; Life Technologies, Grand Island, NY) per liter. The bones were cleaned of all muscle and connective tissue. The marrow was harvested as follows. Femurs were split open down the middle with a scalpel, and all marrow found within the trabecular regions of the proximal and distal femur as well as marrow from the cortical shaft was flushed out using a 21-gauge needle attached to a syringe. Vertebrae were split open down the middle, and all marrow within the trabecular regions was removed. The marrow was collected into two separate tubes (femur and vertebrae) using the following medium: αMEM containing 10 ml antibiotic solution/liter, vitamin K (10−8m), and 20% heat-inactivated FBS (Atlanta Biologicals, Norcross, GA). The cells were pelleted for 10 min at 1000 × g, resuspended in fresh medium, filtered to obtain a single cell suspension, and seeded at 5 × 106 total marrow cells/ml (9 × 105 cells/cm2) onto 100-mm uncoated culture dishes (Corning, Cambridge, MA).
Femoral and vertebral cultures were maintained in a humidified atmosphere (95% air-5% CO2) at 37 C. On day 1, dexamethasone (dex; 10−8m) and freshly prepared ascorbic acid (50 μg/ml) were added to all dishes. Cells were allowed to attach for 4 days, at which time the culture medium was replaced with fresh medium containing αMEM, 20% FBS, ascorbic acid, and dex. Cells were fed every 2 days and were maintained in primary culture to monitor the osteoblast phenotype, as assessed by cell morphology and bone nodule mineralization, or were passaged for gene expression, mineralization, and thyroid hormone response studies. Cell passage was performed on day 8, when femoral and vertebral cells were 60–80% confluent. These cells were detached with 0.25% trypsin in 1 mm EDTA (Life Technologies), resuspended in culture medium, and established in secondary culture by seeding at 2.5 × 104 cells/ml (6.25 × 103 cells/cm2). The culture medium was as described above, except that the dex concentration was increased to 10−7m, and β-glycerophosphate (β-GP; 10 mm) was added. Cells were maintained for up to 18 days in primary or secondary culture. For thyroid hormone response studies, T3 (10−8 and 10−7m) or vehicle (0.0001 n NaOH) was present continuously in the medium beginning on day 0 of secondary culture. Marrow cells were cultured in the continuous presence of dex. In this study, we examined the long term effect of T3 on osteogenic differentiation in culture. Glucocorticoids are essential for promoting osteogenic differentiation in cell cultures derived from adult rat femurs (17, 18) and from adult rat vertebrae (19). Culture conditions are summarized in Table 1.
Conditions for the simultaneous culture and comparison of femoral and vertebral bone marrow-derived osteoblasts
| Rats | Female, 4–6 weeks old, 100–125 g |
| Cell isolation and primary culture medium | αMEM, vitamin K (10−8m), 20% FBS, penicillin G, streptomycin, amphotericin B |
| Cell plating density for primary culture | 5 × 106 total marrow cells/ml (9 × 105 cells/cm2) |
| Primary culture medium additives and time of addition | Ascorbic acid (50 μg/ml) after 24 h on day 1 and thereafter dexamethasone (10−8m) on day 1 |
| Cell feeding and frequency | First medium change on day 4 after plating with fresh medium changes every 2 days thereafter |
| Mineralization of primary cultures | β-GP (10 mm) started on day 8 |
| Cell passage | Day 8 (60–80% confluent) |
| Cell plating density for secondary culture | 2.5 × 104 cells/ml (6.25 × 103 cells/cm2) |
| Secondary culture medium | αMEM, vitamin K (10−8m), 20% FBS, dex (10−7m), ascorbic acid (50 μg/ml), β-GP (10 mm) |
| Rats | Female, 4–6 weeks old, 100–125 g |
| Cell isolation and primary culture medium | αMEM, vitamin K (10−8m), 20% FBS, penicillin G, streptomycin, amphotericin B |
| Cell plating density for primary culture | 5 × 106 total marrow cells/ml (9 × 105 cells/cm2) |
| Primary culture medium additives and time of addition | Ascorbic acid (50 μg/ml) after 24 h on day 1 and thereafter dexamethasone (10−8m) on day 1 |
| Cell feeding and frequency | First medium change on day 4 after plating with fresh medium changes every 2 days thereafter |
| Mineralization of primary cultures | β-GP (10 mm) started on day 8 |
| Cell passage | Day 8 (60–80% confluent) |
| Cell plating density for secondary culture | 2.5 × 104 cells/ml (6.25 × 103 cells/cm2) |
| Secondary culture medium | αMEM, vitamin K (10−8m), 20% FBS, dex (10−7m), ascorbic acid (50 μg/ml), β-GP (10 mm) |
Conditions for the simultaneous culture and comparison of femoral and vertebral bone marrow-derived osteoblasts
| Rats | Female, 4–6 weeks old, 100–125 g |
| Cell isolation and primary culture medium | αMEM, vitamin K (10−8m), 20% FBS, penicillin G, streptomycin, amphotericin B |
| Cell plating density for primary culture | 5 × 106 total marrow cells/ml (9 × 105 cells/cm2) |
| Primary culture medium additives and time of addition | Ascorbic acid (50 μg/ml) after 24 h on day 1 and thereafter dexamethasone (10−8m) on day 1 |
| Cell feeding and frequency | First medium change on day 4 after plating with fresh medium changes every 2 days thereafter |
| Mineralization of primary cultures | β-GP (10 mm) started on day 8 |
| Cell passage | Day 8 (60–80% confluent) |
| Cell plating density for secondary culture | 2.5 × 104 cells/ml (6.25 × 103 cells/cm2) |
| Secondary culture medium | αMEM, vitamin K (10−8m), 20% FBS, dex (10−7m), ascorbic acid (50 μg/ml), β-GP (10 mm) |
| Rats | Female, 4–6 weeks old, 100–125 g |
| Cell isolation and primary culture medium | αMEM, vitamin K (10−8m), 20% FBS, penicillin G, streptomycin, amphotericin B |
| Cell plating density for primary culture | 5 × 106 total marrow cells/ml (9 × 105 cells/cm2) |
| Primary culture medium additives and time of addition | Ascorbic acid (50 μg/ml) after 24 h on day 1 and thereafter dexamethasone (10−8m) on day 1 |
| Cell feeding and frequency | First medium change on day 4 after plating with fresh medium changes every 2 days thereafter |
| Mineralization of primary cultures | β-GP (10 mm) started on day 8 |
| Cell passage | Day 8 (60–80% confluent) |
| Cell plating density for secondary culture | 2.5 × 104 cells/ml (6.25 × 103 cells/cm2) |
| Secondary culture medium | αMEM, vitamin K (10−8m), 20% FBS, dex (10−7m), ascorbic acid (50 μg/ml), β-GP (10 mm) |
Histochemistry
To determine mineralization potential, cells were fixed with cold 2% paraformaldehyde in 0.1 m cacodylic buffer. Calcium phosphate deposition was determined by the von Kossa technique (20). Cell layers were incubated with 3% silver nitrate for 15 min under UV light and viewed macroscopically. To examine the effects of T3 on AP activity, the hormone (10−7m) was included continuously in the medium from day 0 through day 18 of secondary culture or was added for the first 6 days of culture and then withdrawn. On day 18, cells were washed with cold PBS, pH 7.4, and fixed in cold 2% paraformaldehyde in 0.1 m cacodylic buffer. The fixed cultures were incubated in Tris-maleate buffer, pH 8.4, containing naphthol AS-MX phosphate disodium salt (0.5 mg/ml) and Fast Red TR salt (1.0 mg/ml) at 37 C for 30 min (20). The extent of AP-positive staining in control and T3-treated femoral and vertebral cultures was compared and documented by photographing macroscopically. All cultures were examined in triplicate wells of six-well plates.
The establishment of osteoblast culture systems is notoriously serum dependent. To verify that osteoblast differentiation was reproducible by cells derived from both femurs and vertebrae, we compared different commercial serum preparations. Marrow was harvested from five 6-week-old female rats and used for the simultaneous testing of three separate serum lots (4002k, 5006i, and 6008b, Atlanta Biologicals). Primary cultures were examined for mineral deposition on day 18 or were subcultured on day 8, reseeded at 2.5 × 104 cells/ml, and maintained in their starting serum lots. The cells in secondary culture were examined for mineralization on day 10. All plates were examined in duplicate.
Northern analysis
Total RNA from control and T3-treated femoral and vertebral bone marrow cells maintained in secondary culture on six-well plates for various periods of time was isolated by the guanidinium thiocyanate method (21). The RNA (5 μg/lane) was size-fractionated on 1.2% agarose-1.8% formaldehyde gels, transferred onto nitrocellulose membranes (Duralose-UV, Stratagene, La Jolla, CA), and UV cross-linked (Stratagene UV Crosslinker). The blots were hybridized with[ 32P]deoxy-CTP (Amersham, Arlington Heights, IL) probes labeled by the random primer method (22) (DECAprime II kit, Ambion, Austin, TX).
The following complementary DNA (cDNA) probes were used to measure osteocalcin (OC), collagen type I (coll I), IGF-I, and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) mRNA levels, respectively: rat OC (23), rat coll I (cDNA plasmid α IRI) (24), and rat GAPDH (25), provided by Drs. J. B. Lian and G. S. Stein, University of Massachusetts Medical Center (Worcester, MA); and rat IGF-I (26), provided by Dr. Ernesto Canalis, St. Francis Hospital and Medical Center (Hartford, CT).
The blots were exposed to XAR film (Eastman Kodak, Rochester, NY) for detection of OC, coll I, and GAPDH or to BioMax MS film (Kodak) for the more sensitive detection required for IGF-I. For valid comparison, femoral and vertebral Northern blot membranes were always exposed side by side on the same piece of film. Identical film exposure times were used to compare femoral vs. vertebral data and control vs. T3-treated data. The autoradiograms were quantified by scanning laser densitometry (LKB 2400 Gel Scan XL). GAPDH was used as the control, housekeeping gene. The T3 response studies were repeated twice, with similar results obtained.
Results
Growth and morphology of cultured bone marrow cells
Nucleated bone marrow cells from both skeletal sites formed fibroblast-like colonies by day 4 of primary culture. Primary cultures were approximately 80% confluent by day 10 and never reached 100% confluence. Small, tightly packed cuboidal cells were observed within the fibroblastic colonies in both femoral and vertebral primary cultures on days 8–10. There were numerous clusters of various sizes of these cuboidal cells dispersed in the culture dishes. β-GP was added to primary cultures on day 8, and the cuboidal clusters became multilayered to form opaque nodules by day 18. However, although both femoral and vertebral cultures formed nodules, vertebral cells also formed multilayered ridges, indicating a skeletal site difference in bone marrow cells maintained in primary culture under these conditions.
When the cells were subcultured, they attached within 1 day and reached 100% confluence by days 5–6 for vertebral cells and by days 7–8 for femoral cells. Cuboidal cell clusters were evident in both femoral and vertebral secondary cultures by days 5–6 (Fig. 1, A and B). The nodules were much more numerous, more uniformly distributed, and smaller than those formed in primary culture. Femoral and vertebral cells displayed numerous opaque nodules by day 10 in secondary culture. At this time the nodules appeared brown to black in color when viewed with the phase contrast microscope.
Expression of the osteoblast phenotype in secondary (first passage) cultures of femoral and vertebral bone marrow cultures. Colonies of tightly packed cuboidal cells appear in both femoral (A) and vertebral (B) cultures by day 6. These discrete areas multilayer to form nodules where mineralization initiates in both femoral (C) and vertebral (D) cultures. Changes in OC mRNA levels during differentiation of these cultures are shown in E. OC is encoded by a 0.6-kb transcript. The membrane was probed first with 32P-labeled rat OC cDNA and reprobed with 32P-labeled GAPDH cDNA to assess gel loading.
The mineralization ability of femoral and vertebral secondary (day 10) cultures is shown in Fig. 1, C and D. Many clusters of cuboidal cells had mineralized, as shown by positive (black) von Kossa staining. Three separate serum lots were tested for the ability to maintain simultaneous osteogenic differentiation of femoral and vertebral marrow cells. For each serum, duplicate culture dishes showed identical patterns of nodule formation and mineral deposition in primary and secondary cultures (data not shown). Both femoral and vertebral cells in secondary culture showed extensive mineralization with all serum lots. Thus, cell morphology and histochemistry demonstrated bone formation in vitro by both femoral and vertebral osteoblasts in primary and secondary culture.
Sequential osteoblast gene expression and effects of T3
Total cellular RNA from femoral and vertebral marrow cells was analyzed on days 3–12 of secondary culture. A representative Northern blot is shown in Fig. 1E. Cells from these control cultures had undetectable OC mRNA during the earliest time points examined. OC expression was detectable beginning on day 9 in femoral cultures and on day 6 in vertebral cultures, and was abundantly expressed on day 12 for both cultures. Figure 2 confirms that both untreated femoral and vertebral secondary cultures expressed OC mRNA in a pattern consistent with the development of the osteoblast phenotype. A single mRNA band was hybridized at 0.6 kilobase (kb) on Northern blots of both cell populations. OC expression peaked on day 12 in femoral cultures and was not detectable by day 18. In vertebral cultures, the OC transcript increased up to day 15 and continued to be expressed, in contrast to that in femoral cells. The effects of T3 supplementation on OC gene expression in femoral vs. vertebral cells are shown in Fig. 2. T3 differentially affected OC steady state levels in cells derived from the two skeletal sites. Although femoral OC mRNA decreased on day 15 and was undetectable by day 18 in the femoral untreated cells, treatment with T3 (10−8 and 10−7m) maintained OC expression. Vertebral cells did not respond in a similar manner. The effect of T3 on vertebral OC mRNA levels was complex. OC expression on day 12 in T3-treated vertebral cells was increased by 10−8m T3, but decreased by 10−7m T3. On days 15 and 18, T3 caused only a small dose-dependent increase in OC expression.
Time course of OC gene expression in untreated (C) and T3-treated (10−8 and 10−7m) bone marrow cultures. Femoral and vertebral secondary cultures were supplemented with T3 for the indicated number of days. Total RNA was extracted, and 5 μg/lane were used for Northern blot hybridization. The membranes were also probed for GAPDH to assess sample loading. The estimated size of the OC transcript was 0.6 kb. The positions of 28S and 18S ribosomal RNA are indicated.
As shown in Fig. 3, coll I expression in untreated femoral and vertebral cultures followed a pattern similar to OC expression. Coll I expression was highest on day 12, decreased to undetectable expression by day 18 in femoral cultures, and peaked on day 12 with continued expression through day 18 in vertebral cells. In each culture, two mRNA bands (5.8 and 4.7 kb) were detected. Coll I gene expression was dose dependently enhanced by continuous exposure to T3 on days 15 and 18 in femoral cells. Again, the effects of T3 on coll I expression in vertebral cells were complex, with slight stimulation on days 12 and 15 (10−8 and 10−7m) and a decrease on day 18 (10−8m).
Time course of coll I gene expression in untreated (C) and T3-treated (10−8 and 10−7m) bone marrow cultures. The estimated sizes of the coll I transcripts were 5.8 and 4.7 kb. The positions of the 28S and 18S ribosomal RNA are indicated.
IGF-I gene expression and effects of T3
Next, we examined IGF-I steady state mRNA levels as a function of skeletal location in these osteogenic bone marrow cultures. A representative Northern blot obtained using RNA extracted from control cultures is shown in Fig. 4. The temporal expression of the osteoblast-specific gene OC is also shown to compare the appearance of differentiated osteoblasts in culture with the expression of IGF-I. The results clearly indicate that IGF-I expression is differentially regulated in femoral vs. vertebral cultures. IGF-I gene expression was undetectable by Northern analysis of control cultures of femoral cells. In contrast, IGF-I gene expression was evident in vertebral bone marrow cultures, with the highest mRNA levels observed on days 12, and 15, corresponding to peak OC transcript levels.
Time course comparing OC and IGF-I gene expression in untreated femoral and vertebral cultures. Conditions are described in Fig. 2. First, the membrane was probed for OC, then later for GAPDH. The membrane was then stripped and hybridized with cDNA for rat IGF-I. The size of the predominant IGF-I transcript was 6.5 kb.
Figure 5 demonstrates the effects of T3 on IGF-I mRNA isolated from femoral and vertebral control and T3-treated cells. IGF-I expression was again undetectable in femoral untreated cultures, whereas T3 caused a slight increase on day 12. However, in vertebral cultures, T3 treatment markedly increased IGF-I mRNA levels in a dose-dependent manner at each time point examined. Four IGF-I mRNA species were apparent. A 6.5-kb band was predominant, with 4.1-, 1.7-, and 0.9-kb bands also present. T3 caused a proportional increase in these four IGF-I steady state transcripts. The effects of T3 on gene expression in bone marrow cultures is summarized in Table 2.
Time course of the effects of T3 on steady state mRNA levels for IGF-I in femoral and vertebral bone marrow cultures. Conditions are described in Fig. 2. The sizes of the four transcripts detected were 6.5, 4.1, 1.7, and 0.9 kb. The positions of the 28S and 18S ribosomal RNA are indicated by the small arrows.
Relative mRNA levels in rat marrow cultures
| Days in culture | Femoral cultures | Vertebral cultures | |||||
|---|---|---|---|---|---|---|---|
| T3 (m) added | OC | Coll I | IGF-I | OC | Coll I | IGF-I | |
| 9 | None | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 |
| 10−8 | 0.8 ± 0.1 | 0.4 ± 0.1 | 0.8 ± 0.1 | 0.8 ± 0.2 | 0.8 ± 0.1 | 2.3 ± 0.6 | |
| 10−7 | 0.6 ± 0.2 | 0.4 ± 0.2 | 0.6 ± 0.3 | 0.6 ± 0.2 | 0.7 ± 0.1 | 3.6 ± 0.9 | |
| 12 | None | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 |
| 10−8 | 1.1 ± 0.1 | 0.7 ± 0.2 | 1.2 ± 0.1 | 1.3 ± 0.1 | 1.3 ± 0.1 | 2.1 ± 0.5 | |
| 10−7 | 1.0 ± 0.3 | 0.8 ± 0.1 | 2.1 ± 0.3 | 0.6 ± 0.3 | 1.1 ± 0.1 | 4.3 ± 1.1 | |
| 15 | None | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 |
| 10−8 | 2.9 ± 0.3 | 3.1 ± 1.2 | 0.9 ± 0.3 | 1.4 ± 0.2 | 1.1 ± 0.2 | 1.9 ± 0.6 | |
| 10−7 | 3.4 ± 0.5 | 6.0 ± 2.1 | 0.9 ± 0.4 | 1.3 ± 0.1 | 1.3 ± 0.3 | 3.3 ± 1.0 | |
| 18 | None | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 |
| 10−8 | 1.8 ± 0.2 | 1.8 ± 0.3 | 0.9 ± 0.2 | 1.0 ± 0.3 | 0.4 ± 0.2 | 0.9 ± 0.2 | |
| 10−7 | 2.2 ± 0.5 | 2.4 ± 0.6 | 1.1 ± 0.4 | 1.3 ± 0.4 | 1.0 ± 0.1 | 4.1 ± 1.2 | |
| Days in culture | Femoral cultures | Vertebral cultures | |||||
|---|---|---|---|---|---|---|---|
| T3 (m) added | OC | Coll I | IGF-I | OC | Coll I | IGF-I | |
| 9 | None | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 |
| 10−8 | 0.8 ± 0.1 | 0.4 ± 0.1 | 0.8 ± 0.1 | 0.8 ± 0.2 | 0.8 ± 0.1 | 2.3 ± 0.6 | |
| 10−7 | 0.6 ± 0.2 | 0.4 ± 0.2 | 0.6 ± 0.3 | 0.6 ± 0.2 | 0.7 ± 0.1 | 3.6 ± 0.9 | |
| 12 | None | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 |
| 10−8 | 1.1 ± 0.1 | 0.7 ± 0.2 | 1.2 ± 0.1 | 1.3 ± 0.1 | 1.3 ± 0.1 | 2.1 ± 0.5 | |
| 10−7 | 1.0 ± 0.3 | 0.8 ± 0.1 | 2.1 ± 0.3 | 0.6 ± 0.3 | 1.1 ± 0.1 | 4.3 ± 1.1 | |
| 15 | None | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 |
| 10−8 | 2.9 ± 0.3 | 3.1 ± 1.2 | 0.9 ± 0.3 | 1.4 ± 0.2 | 1.1 ± 0.2 | 1.9 ± 0.6 | |
| 10−7 | 3.4 ± 0.5 | 6.0 ± 2.1 | 0.9 ± 0.4 | 1.3 ± 0.1 | 1.3 ± 0.3 | 3.3 ± 1.0 | |
| 18 | None | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 |
| 10−8 | 1.8 ± 0.2 | 1.8 ± 0.3 | 0.9 ± 0.2 | 1.0 ± 0.3 | 0.4 ± 0.2 | 0.9 ± 0.2 | |
| 10−7 | 2.2 ± 0.5 | 2.4 ± 0.6 | 1.1 ± 0.4 | 1.3 ± 0.4 | 1.0 ± 0.1 | 4.1 ± 1.2 | |
Cell cultures were treated with T3 as described in Fig. 3. The levels of OC, coll I, and IGF-I mRNA relative to that of GAPDH mRNA were analyzed by Northern blot hybridization. The values for control (none) were assigned 1.0 for each day in culture analyzed. Values were derived from three different RNA preparations; results are expressed as the mean ± sd. Representative autorads are shown in Figs. 2, 3, and 5.
Relative mRNA levels in rat marrow cultures
| Days in culture | Femoral cultures | Vertebral cultures | |||||
|---|---|---|---|---|---|---|---|
| T3 (m) added | OC | Coll I | IGF-I | OC | Coll I | IGF-I | |
| 9 | None | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 |
| 10−8 | 0.8 ± 0.1 | 0.4 ± 0.1 | 0.8 ± 0.1 | 0.8 ± 0.2 | 0.8 ± 0.1 | 2.3 ± 0.6 | |
| 10−7 | 0.6 ± 0.2 | 0.4 ± 0.2 | 0.6 ± 0.3 | 0.6 ± 0.2 | 0.7 ± 0.1 | 3.6 ± 0.9 | |
| 12 | None | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 |
| 10−8 | 1.1 ± 0.1 | 0.7 ± 0.2 | 1.2 ± 0.1 | 1.3 ± 0.1 | 1.3 ± 0.1 | 2.1 ± 0.5 | |
| 10−7 | 1.0 ± 0.3 | 0.8 ± 0.1 | 2.1 ± 0.3 | 0.6 ± 0.3 | 1.1 ± 0.1 | 4.3 ± 1.1 | |
| 15 | None | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 |
| 10−8 | 2.9 ± 0.3 | 3.1 ± 1.2 | 0.9 ± 0.3 | 1.4 ± 0.2 | 1.1 ± 0.2 | 1.9 ± 0.6 | |
| 10−7 | 3.4 ± 0.5 | 6.0 ± 2.1 | 0.9 ± 0.4 | 1.3 ± 0.1 | 1.3 ± 0.3 | 3.3 ± 1.0 | |
| 18 | None | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 |
| 10−8 | 1.8 ± 0.2 | 1.8 ± 0.3 | 0.9 ± 0.2 | 1.0 ± 0.3 | 0.4 ± 0.2 | 0.9 ± 0.2 | |
| 10−7 | 2.2 ± 0.5 | 2.4 ± 0.6 | 1.1 ± 0.4 | 1.3 ± 0.4 | 1.0 ± 0.1 | 4.1 ± 1.2 | |
| Days in culture | Femoral cultures | Vertebral cultures | |||||
|---|---|---|---|---|---|---|---|
| T3 (m) added | OC | Coll I | IGF-I | OC | Coll I | IGF-I | |
| 9 | None | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 |
| 10−8 | 0.8 ± 0.1 | 0.4 ± 0.1 | 0.8 ± 0.1 | 0.8 ± 0.2 | 0.8 ± 0.1 | 2.3 ± 0.6 | |
| 10−7 | 0.6 ± 0.2 | 0.4 ± 0.2 | 0.6 ± 0.3 | 0.6 ± 0.2 | 0.7 ± 0.1 | 3.6 ± 0.9 | |
| 12 | None | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 |
| 10−8 | 1.1 ± 0.1 | 0.7 ± 0.2 | 1.2 ± 0.1 | 1.3 ± 0.1 | 1.3 ± 0.1 | 2.1 ± 0.5 | |
| 10−7 | 1.0 ± 0.3 | 0.8 ± 0.1 | 2.1 ± 0.3 | 0.6 ± 0.3 | 1.1 ± 0.1 | 4.3 ± 1.1 | |
| 15 | None | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 |
| 10−8 | 2.9 ± 0.3 | 3.1 ± 1.2 | 0.9 ± 0.3 | 1.4 ± 0.2 | 1.1 ± 0.2 | 1.9 ± 0.6 | |
| 10−7 | 3.4 ± 0.5 | 6.0 ± 2.1 | 0.9 ± 0.4 | 1.3 ± 0.1 | 1.3 ± 0.3 | 3.3 ± 1.0 | |
| 18 | None | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 | 1.0 |
| 10−8 | 1.8 ± 0.2 | 1.8 ± 0.3 | 0.9 ± 0.2 | 1.0 ± 0.3 | 0.4 ± 0.2 | 0.9 ± 0.2 | |
| 10−7 | 2.2 ± 0.5 | 2.4 ± 0.6 | 1.1 ± 0.4 | 1.3 ± 0.4 | 1.0 ± 0.1 | 4.1 ± 1.2 | |
Cell cultures were treated with T3 as described in Fig. 3. The levels of OC, coll I, and IGF-I mRNA relative to that of GAPDH mRNA were analyzed by Northern blot hybridization. The values for control (none) were assigned 1.0 for each day in culture analyzed. Values were derived from three different RNA preparations; results are expressed as the mean ± sd. Representative autorads are shown in Figs. 2, 3, and 5.
Effects of T3 on the formation of AP-positive cell colonies
Femoral and vertebral bone marrow secondary cultures both expressed high levels of the enzyme AP. Based upon histochemical staining, AP-positive colonies were first observed on day 3 and steadily increased to a maximum on day 18 (Fig. 6) in cultures derived from both skeletal sites. In two independent experiments, the presence of T3 for the first 6 days of culture decreased the extent of AP-positive staining in femoral bone marrow cultures, but no visible differences were seen in vertebral culture (Fig. 6A). The continuous presence of T3 throughout the 18-day culture period resulted in a further decrease in femur AP-positive colonies and a slight decrease in vertebral AP-positive colonies (Fig. 6B).
Effect of thyroid hormone withdrawal on AP activity of femoral and vertebral bone marrow cultures. Secondary osteogenic bone marrow cultures were treated with vehicle (none) or T3 (10−7m), beginning on day 0. After 6 days of treatment, T3-treated cells were divided into two groups; one was treated with control medium containing no added T3 (A), and the other group continued to receive T3 (B). All cells were harvested on day 18 and stained histochemically for AP. Histochemistry was performed in triplicate for all experimental conditions.
Discussion
These experiments are the first to simultaneously compare osteoblast differentiation in bone marrow cultures isolated from two different skeletal locations and also the first to compare the effects of thyroid hormone on osteoblastic cells in culture as a function of skeletal site. We undertook these studies because of the clinically relevant observation that thyroid hormone differentially affects bone mineral density at the hip and spine (6, 27–29). While the differences in cellular responses of femurs and vertebrae to thyroid hormone are unexplained, it is known that the presence of osteoblasts is required for the hormone to increase bone resorption mediated by osteoclasts (5). To define the molecular basis for the skeletal diversity of responses by osteoblasts to thyroid hormone, it is necessary to culture cells of the osteoblast lineage. Although few studies have compared cells isolated from different parts of the skeleton, one report has confirmed that differences exist between osteoblast populations derived from different bone sites (30).
We have established that osteoprogenitors present in bone marrow cell populations from adult female rat vertebrae and femurs differentiate into osteoblasts in vitro. Vertebral and femoral cells multilayer and form mineralized bone nodules when cultured in the presence of ascorbic acid, β-GP, and dexamethasone. This cellular multilayering and nodule formation are consistent with results previously described for in vitro bone formation by rat femur bone marrow stromal cells (16). Osteoblast differentiation of both cell populations during secondary culture is associated with high AP enzyme activity and the expression of genes for the major bone proteins, coll I and OC. Thus, we have a suitable model to compare osteoblast responses to various factors as a function of skeletal location.
Our results obtained by Northern analysis of secondary cultures support the concept of a skeletal site heterogeneity in response to thyroid hormone. In untreated femoral cultures, OC and coll I steady state mRNA levels increased with time in culture and then decreased to undetectable levels. In the continuous presence of T3, femoral OC and coll I mRNA levels remained expressed throughout the culture period. Vertebral cells did not have the same gene expression pattern; OC and coll I mRNA levels remained elevated throughout the culture period for untreated cells, and T3 had only minimal effects on the expression of these genes. Previously, it was reported that T3 increased the level of OC mRNA in fetal rat calvarial osteoblast cultures (31). Here we have shown that T3 also increased OC mRNA levels in adult femoral osteoblastic cultures. Although multiple TR isoform genes are expressed in both rat femurs and vertebrae (12), thyroid hormone administration in vivo decreases femoral bone mass, but not vertebral bone mass, in rats (6, 28, 29). Our present findings indicate that OC gene expression in vertebral cultures is not dramatically altered in response to T3.
We also found that although cultures derived from both skeletal sites had high AP enzyme activity, T3 markedly depressed AP activity in femoral, but not vertebral, cultures. Furthermore, T3 decreased AP activity in femoral, but not vertebral, cells that had been exposed to the hormone for the first 6 days in culture only, a period during which the osteoblast marker gene OC was not yet expressed. This suggests that thyroid hormone can affect femoral osteoprogenitors at an early stage of differentiation, but its addition does not alter osteoblast differentiation of vertebral osteogenic cells.
To address whether continuous T3 treatment can alter expression of any of the known osteoblast-related genes in vertebral cultures, we studied the effects of T3 on IGF-I mRNA levels. This growth factor has been suggested to mediate some of the effects of thyroid hormone on bone formation (32), and osteoblasts secrete IGF-I, which acts in an autocrine and paracrine manner to regulate osteoblast function (15). In addition, T3 was reported to increase levels of IGF-I mRNA in a murine osteoblastic cell line (MC3T3 cells) (14). The results of our Northern analysis of total RNA showed that IGF-I expression was detectable in vertebral, but not femoral, untreated cultures, and that T3 increased IGF-I mRNA levels dose dependently in vertebral, but not femoral, cells. This skeletal site-dependent difference in the pattern of IGF-I gene expression is in agreement with a recent study indicating that production of the IGF system components by human bone cells is significantly different between the skeletal sites of their origin (33). It is possible that the IGF-I expression correlates with a greater abundance of this growth factor for the vertebral cells compared with the femoral cells, leading to a sustained increase in osteoblast phenotype markers and a prevention of bone loss in the spine. This is consistent with our data showing that chronic exposure to high levels of T3 (10−7m) did not result in a decrease in AP-positive staining in vertebral cultures. This leads to the speculation that if IGF-I were added to femoral cultures, their gene expression and histochemistry profile might resemble those of vertebral cultures.
IGF-I gene expression was detected by Northern analysis in vertebral cultures despite the continuous presence of dexamethasone. The four transcript sizes detected, 6.5, 4.1, 1.7, and 0.9 kb, correspond to those previously reported in rat osteoblast cultures (34). The highest mRNA levels observed correspond to peak OC transcript levels. Glucocorticoids repress transcription of the IGF-I gene (35) and synthesis of IGF-I (36) in rat osteoblasts. The presence of dex was required for both vertebral and femoral cultures to maintain their osteoblastic phenotype with regard to OC and coll I temporal gene expression, mineralization potential, and AP activity, but the activities of the two osteogenic cultures differed in response to added T3; IGF-I gene expression was stimulated dramatically only in vertebral cultures. This suggests the presence of a factor(s) in vertebral cells that overrides gene suppression due to dex.
As the present studies were performed with continuous T3 treatment, it is uncertain whether the increase in IGF-I expression in vertebral cultures was a direct effect on IGF-I gene transcription or whether it was due to an increase in osteoblast differentiation as a consequence of T3 treatment. Also, the marrow culture system has a heterogeneous population of cell types (37), and in addition to osteoblasts, rat marrow stromal cells have been shown to synthesize IGF-I (38). In any case, the effects of T3 on IGF-I gene expression by vertebral cultures far exceeded those on gene expression by femoral cultures. We conclude that vertebral and femoral bone marrow cells populations are differentially regulated by T3in vitro. The simultaneous comparison of osteogenic cultures from different skeletal regions facilitates the examination of mechanisms by which thyroid hormone regulates osteoblast function. The effects of T3 have been studied previously in fetal rat calvaria cells (31, 39). It has been reported that T3 suppresses osteoprogenitor differentiation when cells are maintained in the continuous presence of added T3 (10–9–10−8m) under culture conditions with 15% thyroid hormone-depleted FBS (31). The serum used in the present study was not stripped of hormones. The endogenous concentration of T3 in the 20% FBS culture medium was assayed to be 3.9 × 10−10m. After adding T3 to make the final concentration, 10−8m, the final total T3 concentration in the culture medium with 20% FBS was 1.13 × 10−8m. This value compares well with that found in hyperthyroidism (31) and shows that the T3 responses in gene expression of femoral and vertebral osteoblasts were observed at hormone levels obtainable in vivo.
In conclusion, our results show that differences exist between bone cell populations derived from different skeletal sites. The varied responses of different regions of the skeleton to the hormonal environment is of clinical importance. The culture model presented in this report should prove useful for identifying the regulatory factors responsible for such skeletal site-specific heterogeneity.
Acknowledgments
The authors thank Caroline Kuzia for preparing this manuscript.
References
Author notes
This work was supported by NIH Grant DK-39085.
