| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ß+) Have Bone Phenotypes
Department of Oral & Dental Sciences (S.J.Y., K.W., N.L.B., J.R.S., J.P.M.), Division of Child Dental Health, University of Bristol Dental School, Bristol BS1 2LY, United Kingdom; Department of Orthopaedics (M.J.P., T.J.C.), University of Bristol, Bristol BS2 8EJ, United Kingdom; and Department of Physiology (T.L., S.B.R., F.-P.Z., I.H.), University of Turku, FIN-20520 Turku, Finland
Address all correspondence and requests for reprints to: Dr. Jason P. Mansell, Division Child Dental Health, University of Bristol Dental School, Lower Maudlin Street, Bristol BS1 2LY, United Kingdom. E-mail: j.p.mansell{at}bris.ac.uk.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
ß+); and 3) ovariectomized (OVX) hCG ß+ model. Male LuRKO mice showed a decrease in BMD after 5 months, apparently secondary to suppressed gonadal steroid production. Similarly, 9- to 10-wk-old female LuRKO mice exhibited decreases in histomorphometric parameters tested. The data indicate that loss of LH signaling results in a reduction in bone formation or an increase in bone resorption. By contrast, there were significant increases in BMD and histomorphometric indices for female, but not male, hCG ß+ mice, indicating that chronic exposure to hCG results in bone formation or a decrease in bone resorption. However, OVX of the hCG ß+ mice resulted in a significant reduction in BMD comparable to OVX WT controls. Although gonadotropin levels are tightly linked to sex steroid titers, it appears that their effects on the skeleton are indirect. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Both LH and hCG bind to the same seven-transmembrane domain G protein-coupled receptor (8), which is now thought to have a role in the metabolism of a number of tissues. The identification of receptors for LH outside of the ovarian-pituitary axis is relatively recent (9). There appears to be an association between LH receptor expression and the sensitivity of the same site to estrogen. Skin, mammary gland, uterus, and urinary bladder express LH receptors and respond to changes in estrogen status. Moreover, estrogen can influence LH receptor expression and therefore the sensitivity of target tissues to this gonadotropin as observed, for example, in the epithelial tissue of the pig oviduct (10). Collectively, these findings imply that LH has a role in a variety of cell and tissue responses and is not restricted to the regulation of gonadal function.
We sought to determine whether bone tissue and bone-forming osteoblasts might also be a target for LH/CG. LH and hCG are heterodimers made up of noncovalently associated subunits, the -subunit, which is common to all the gonadotropins, and the hormone-specific ß-subunit. LH and hCG belong to the cystine knot superfamily (11), sharing several topological features with other family members known to influence bone tissue metabolism, for example activins (12), TGFß, and bone morphogenetic proteins (13). In light of the changes that take place in the skeleton during periods of elevated LH/hCG levels combined with the reported effects of estrogen on bone (14) and the strong association between LH/hCG and estrogen, we hypothesized that bone tissue may also be a target for LH/hCG. We used primary human osteoblasts (hOBs) and the osteoblast-like cell lines, mC3T3-E1, MG63, and SAOS2, to determine whether these cells expressed LH receptors. We investigated changes in bone mineral density (BMD) in vivo, with an LH receptor null mutant (LuRKO) mouse model and a murine transgenic model overexpressing both hCG subunits (hCG ß+).
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
At confluence, cells were passaged into the wells (104 cells in 0.5 ml of media/well) of eight-well chamber slides (Lab-Tek Chamber slide systems, Nalge Nunc International, Roskilde, Denmark). Cells were left under conventional cell culture conditions for 16 h, the media and cell chambers removed, and the adherent cells probed for the LH receptor. Briefly, slides were fixed by immersion in 2% (wt/vol) paraformaldehyde (pH 7.4) for 30 min at room temperature, followed by incubation for 10 min at -20 C in methanol. Slides were then rinsed in PBS before goat serum blocking. Goat serum was diluted five times in PBS supplemented with 0.1% (wt/vol) BSA. Blocking solution (300 µl/well) was dispensed into each well and left at room temperature for 1 h. This was discarded and 50 µl of the LH receptor antibody (gift from Dr. Patrick Roche, Mayo Clinic, Rochester, MN) [1:10 dilution with 1% (wt/vol) BSA] added to half of the wells. The remaining (secondary control) wells received the antibody diluent solution alone. For these experiments, we used a rabbit polyclonal antibody, which recognized a specific extracellular sequence, residues 15–38 (15), of the LH receptor. Following antibody application, slides were left at room temperature for 1 h and then rinsed in PBS. A goat antirabbit fluorescein isothiocyanate (FITC) conjugate (Sigma, Poole, UK) was diluted 75-fold in the antibody diluent solution, and 50 µl dispensed into each well and left for 1 h at room temperature. Slides were rinsed briefly in PBS, and each well was then treated with 25 µl of a propidium iodide mountant (Vectashield, H-1300, Vector Laboratories, Peterborough, UK) for nuclear staining and a cover slip placed on top. LH receptors were visualized using a Leica (Buckinghamshire, UK) DMLB fluorescence microscope.
Preparation of tissue and cells for LH receptor identification by Western blotting
Freshly isolated ovaries from a rat were homogenized using a glass tissue homogenizer (3 ml Jencons, Leighton Buzzard, UK) and 1 ml of 5 mM MgSO4, 40 mM Tris (pH 7.4) containing 5 mM N-ethylmaleimide, 200 µM Pefabloc SC (Roche Molecular Biochemicals, Mannheim, Germany), 1 µM BB-3103 (metalloproteinase inhibitor, British Biotechnology, Oxford, UK), 0.1% wt/vol Triton X-100, and 20% vol/vol glycerol. The resultant homogenate was then extracted over a period of 1 h under refrigerated conditions, the sample was centrifuged for 10 min at 2000 x g, and the supernatant removed and stored at -20 C until required.
hOBs, the human osteoblast-like cell line MG63, murine mC3T3-E1 osteoblast-like cells, and murine Leydig tumor cells (mLTC-1; ATCC, Manassas, VA) were grown under conventional culturing conditions in media recommended for each cell type. When cell monolayers, or islands in the case of mLTC-1 cells, covered approximately 80% of the 75-cm2 culture flask, the media were removed and the cells rinsed with sterile PBS. Washed cells were then treated with the extraction buffer as described for the ovarian tissue homogenate. Before electrophoresis, ovarian tissue and cultured cell extracts were assayed for their total protein content by the Bradford method using detergent compatible protein assay reagents (Bio-Rad, Hertfordshire, UK). For electrophoresis, samples were diluted twice in sample buffer as described by Laemmli (16) but containing twice the concentration of sodium dodecyl sulfate (2% wt/vol) and 5% vol/vol 2-mercaptoethanol. Samples were boiled for 10 min and then centrifuged at 9000 rpm for 2 min. Samples were loaded at the same protein concentration (15 µg per lane) onto 1-mm-thick, 10% polyacrylamide gels, and electrophoresed using the mini-protean II electrophoresis apparatus (Bio-Rad). Proteins were subsequently transferred onto nitrocellulose (Hybond, Amersham, Buckinghamshire, UK) for 90 min at 250 mA in the presence of 20% vol/vol methanol. Following transfer, the membrane was blocked overnight at 4 C in milk-blocking buffer (MBB: 5% wt/vol fat-free milk powder; 0.5% vol/vol Tween 20; 20 mM Tris-HCl; and 150 mM NaCl, pH 7.4).
The membrane was probed using a rabbit polyclonal antibody that identifies residues 1–11 (15) of the LH receptor extracellular domain (gift from Dr. P. Roche, Mayo Clinic, Rochester, MN). The antibody was diluted 1:500 in MBB, combined to the blot and left for 2 h, at room temperature. Following a wash in MBB, a secondary goat antirabbit antibody, conjugated to alkaline phosphatase (Sigma) was diluted 1:5000 in the same solution, applied to the blot, and left for 3 h at room temperature. The blot was rinsed free of antibody using the same buffer but without milk powder, and developed using nitroblue tetrazolium (1.5 mM) and 3-bromo-4-chloro-5-indolylphosphate (0.6 mM) in 0.5 M MgCl2, 0.5 M Tris-HCl (pH 9.7).
Identification of LH receptor mRNA in primary human osteoblasts using nested RT-PCR and Southern blotting
Total RNA was extracted from human osteoblasts once they had reached approximately 80% confluence in 75-cm2 flasks. The application of QIAshredder columns and RNeasy minicolumns (QIAGEN, Hilden, Germany), as per the manufacturer’s instructions, enabled total RNA extraction of lysed cell monolayers. RT-PCR was used to identify the presence of LH/hCG mRNA in human osteoblasts. The omission of template RNA served as a negative control and total RNA from prostate cells and human placenta were used as positive controls (gifts from Dr. K. Whittington and Dr. L. Armstrong, Bristol, UK). RT-PCR was performed using the Titanium one-step RT-PCR kit (CLONTECH, Palo Alto, CA) as instructed by the manufacturer. Briefly, a master mix was prepared that contained a single optimized buffer with a single enzyme mixture to enable the RT-PCR to be performed in a single tube. One microgram of total RNA (measured by spectrophotometry) together with 1 µl of each primer (45 µM, initial concentration) from set one (Table 1
, primer sequences provided by Prof. C. V. Rao, University of Louisville, Louisville, KY) were combined to the master mix to give a final reaction volume of 50 µl. Thermal cycling consisted of 50 C for 1 h to enable first strand cDNA synthesis, followed by denaturation at 95 C for 5 min and 30 cycles of denaturation at 94 C for 45 sec, annealing at 58 C for 1 min, and extension at 72 C for 1 min. The final extension lasted 7 min, after which 5 µl of the reaction products were combined to fresh master mix and 1 µl each of each primer from set two (Table 1) to perform the second, nested, PCR. The nested PCR primers are designed within the region amplified by the initial RT-PCR. Thermal cycling was carried out as described except that the initial step in cDNA synthesis was not necessary. Amplified products were resolved within 2% agarose gels containing ethidium bromide. The identities of the bands corresponding to the LH receptor were corroborated by Southern blotting by hybridization with a 32P-labeled, full-length, human LH/hCG receptor cDNA.
|
-spectrometer (1260 Multigamma II, Wallac, Turku, Finland). Nonspecific binding was determined in the presence of 50 IU Pregnyl, and all data were corrected for nonspecific binding. The ability of hCG binding to LH receptors in was determined by Scatchard analysis.
Assessment of LH receptor functionality in osteoblasts by quantification of cAMP after stimulation with hCG
hOBs, mC3T3-E1, and mLTC-1 cells were cultured as previously described. At confluence cells were passaged and subcultured into 30-mm culture dishes at a density of 15 x 104 cells/dish until they reached 80% confluency. A 100-µM final concentration of 3-isobutyl-1-methylxanthine (Sigma) was added to each dish to inhibit endogenous phosphodiesterase. Five minutes later, the test ligands were added: either vehicle, 1 x 105 IU/liter hCG (Pregnyl, Organon), or 25 µM prostaglandin E2 (Sigma). Test ligands were added for 10 min, 2 h, or 8 h. Transferring dishes to ice and adding 1 ml of ice-cold acidified ethanol (0.2 M HCl in absolute ethanol) quenched reactions. Dishes were stored overnight at -20 C to extract cAMP. Medium from each dish was then transferred to a universal tube (Merck, Hertfordshire, UK), stored at -70 C and subsequently subjected to freeze drying overnight.
Total cAMP in each sample was measured using a cAMP 3H assay system kit (Amersham) as per the manufacturer’s instructions. Each sample was combined with 4 ml of scintillation cocktail (Optiphase HiSafe 2, Fisher Chemicals, Loughborough, UK) and counted using a liquid scintillation counter (1217 Rack beat, LKB Wallac). Data were expressed as the mean nanomolar concentration of cAMP.
Assessment of LH receptor functionality in osteoblasts by detection of phosphorylated ERK1/2 after stimulation with hCG
MG63 and mC3T3-E1 cells were cultured as previously described. At confluence, cells were passaged and subcultured into 60-mm culture dishes at a density of 2 x 104 cells/dish. Cells were cultured for a further 72 h and then transferred to serum free medium overnight. Cells were stimulated with either 1 x 104 IU/liter recombinant hCG (Ovitrelle, Serono, London, UK), 10 ng/ml fibroblast growth factor (FGF)-1 (positive control) or vehicle for 15 min, 2 h, or 4 h. The medium was aspirated and the reaction terminated by the addition of 100 µl of hot (70 C) lysis buffer (3% sodium dodecyl sulfate; 7% sucrose; 45 mM Tris, pH 6.8; 0.01% bromophenol blue; 35 mM dithiothreitol). Cells were scraped and the cell lysates boiled for 5 min. Cell lysates were subjected to SDS-PAGE followed by transfer onto nitrocellulose as described previously for the identification of the LH receptor. The membrane was probed using a rabbit polyclonal antiactive MAPK antibody (Promega, Southampton, UK) diluted 1:1000 and the secondary antibody used was an antirabbit peroxidase conjugate. Bands were detected using chemiluminescent reagents (ECL, Amersham). Equal loading of protein was confirmed by reprobing the membrane using a mouse monoclonal anti--tubulin antibody (Sigma).
Experimental animals and treatments
Care of animals.
The animals were housed in a specific pathogen-free environment, under controlled temperature and light conditions, and were provided tap water and commercial mouse chow ad libitum. All mice were handled in accordance to the institutional animal care policies of the University of Turku. Using tissues obtained from a previous study, which had been approved by the institutional ethical committee, reduced the number of animals required.
Transgenic hCG overexpressors
Transgenic mice were generated to overexpress both the hCG and ß subunit cDNA under the control of the ubiquitin C promoter as described by Rulli et al. (18). Briefly, transgenic mice carrying the hCG- or the hCGß-subunit were generated by pronuclear microinjection, as described by Rulli et al. (18). Double transgenic hCGß+ mice producing elevated levels of hCG (about 1 x 104 IU/liter in serum) were obtained by crossbreeding independent lines of hCG and hCGß mice. PCR analyses of genomic DNA from tail biopsies were used to identify transgenic animals. The genetic background of the hCGß+, hCGß+ OVX and WT mice was FVB/N. Adult hCGß+ females showed increased serum levels of progesterone (60-fold), testosterone (6-fold), and prolactin (80-fold), and a transient increase in estradiol action at peripuberty, as indicated by the presence of cornified vaginal mucosa and enlarged, fluid-filled uteri. Adult hCGß+ males presented with elevated levels of testicular testosterone (10-fold), and enlarged accessory sex organs.1 Animals were ovariectomized at 3 wk of age and killed at 2 months of age by cervical dislocation. The controls for ovariectomized mice were sham operated.
LuRKO
LuRKO mice were generated by inactivating, through homologous recombination, exon 11 on the LHR gene identically as described by Zhang et al. (19). Mice were genotyped by PCR analysis of DNA from tail biopsies. WT (+/+) and LuRKO (-/-) mice were obtained from the same colony with the same genetic background (129/SvEv/C57BL) at different ages. Animals were treated with avertin anesthesia, and whole body was stored at -70 C until further analysis.
Dual x-ray absorptiometry (DXA) analysis
BMD was measured by DXA using the PIXImus scanner (Lunar, Madison, WI) that is specifically designed for use with small animals. Rear limbs were removed from each mouse and excess flesh trimmed away. The limb was separated into femur and tibia and each bone cleaned using gauze. Bones were orientated in the same position for each scan.
Static histomorphometry
Clean bones were fixed in 70% ethanol for 48 h then dehydrated through sequentially increasing concentrations of ethanol: 80% ethanol, 90% ethanol, and finally three changes of 100% ethanol for 24 h each. Tibiae were then immersed in chloroform for 24 h and in 100% ethanol for a further 24 h. Processed tibiae were cut using a small circular saw a few millimeters distally from the proximal tibial-fibula anastomosis. Tibiae were embedded without decalcification in hard grade acrylic white resin (The London Resin Co., Reading, UK) and baked for 24 h at 60 C. Longitudinal sections of the proximal and distal metaphysis were prepared using a Reichert-Jung 2050 microtome with a "d" profile tungsten carbide blade at a width of 7 µm and stained using toluidine blue [0.25% toluidine blue in 0.01 M citrate phosphate buffer (pH 3.7)].
Histomorphometric analysis was performed using transmitted microscopy linked to a computer assisted image analyzer (Osteomeasure, Osteometrics, Atlanta, GA). Two nonconsecutive sections were analyzed blindly per animal for each parameter.
The sample site used was a defined area of 0.364 mm2, the proximal border of which was situated 0.25 mm below the growth plate to exclude primary spongiosa. The parameters measured for each section included: cancellous bone volume, which was expressed as a percentage of total tissue volume (BV/TV) (percentage), trabecular width (TbWi) (micromolar concentration), and number (TbN) (per millimeter).
Statistical analyses
The Student’s t test was used to look for significant differences in the DXA and histomorphometric data in which two experimental animal groups were compared. Welch’s correction was applied where the data failed to meet the criterion of equal variance. One-way ANOVA was used to examine for significant differences in the data where the four experimental animal groups were compared. When P < 0.05 was found, a Tukey multiple comparisons posttest was performed between all groups. In some instances, natural log transformations were applied to the data to ensure that variance between groups showed less than a 5-fold difference.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
shows the presence of the LH receptor protein in extracts of both human (hOB and MG63) and murine (mC3T3-E1) osteoblasts. The mass of the detected component, which comigrated with the ovary standard under reducing conditions, is approximately 68 kDa.
|
shows cells exposed to both primary and secondary antibodies and reveals the presence of immunolabeled LH receptors. Figure 2B shows the control when cells were treated with the FITC-conjugated secondary antibody alone. The staining of cell nuclei was achieved using a propidium iodide mountant.
|
) and Southern blotting (Fig. 3B) revealed the presence of LH receptor mRNA in hOBs, MG63, and SAOS-2 cells.
|
). Scatchard analysis of the data showed that mLTC-1 cells, under the identical conditions as those used for MG63 cells, possess high-affinity LH receptors. HEK293 cells served as a negative control and did not bind hCG.
|
shows that there were no significant differences in mean cAMP levels when cells were treated with hCG compared with vehicle control for each time point measured. In additional experiments, hCG stimulation of mC3T3-E1 cells showed no increase in cAMP levels (data not shown). These results indicate that osteoblasts are unable to generate cAMP upon receipt of hCG. As a control, hOBs were treated with 25 µM prostaglandin E2 to demonstrate that these cells have the capacity to generate detectable levels of cAMP. Furthermore, mLTC-1 cells were used as a positive cell line to ensure that the hCG was bioactive.
|
shows that ERK1/2 is increased when the mC3T3-E1 cells were stimulated with FGF-1 (positive control), but when cells were treated with hCG there was no evidence for ERK1/2 phosphorylation. Figure 6B confirms similar total protein loadings for the immunoblot after probing for -tubulin. Results obtained with MG63 cells (data not shown) were similar to those obtained for mC3T3-E1 cells.
|
ß+ transgenic mice have markedly raised BMDß+ mice compared with WT controls. Figure 7A depicts marked changes in both tibial (P < 0.001) and femoral (P < 0.0001) BMD observed in the hCG ß+ compared with WT controls. In contrast, Fig. 7B shows that male transgenics have comparable BMD in both femur and tibia compared with WT.
|
ß+ transgenic miceß+ transgenic mice was either due to the direct effect of raised serum hCG, or an indirect effect of the ovary, a group of female hCG ß + and WT mice were bilaterally OVX at 3 wk of age. After 8 wk of age, the bones from each group were analyzed using DXA and static histomorphometry (Fig. 8 and Table 2). hCG ß+ transgenic mice had raised BMD of both the femur (31% increase) (hCG ß+, 77 ± 1.5; WT, 59 ± 1.7 mg/cm2; P < 0.001) and tibia (27% increase) (hCG ß+: 66 ± 2.4; WT: 52 ± 2.3 mg/cm2; P < 0.001) in addition to an increase in histomorphometric indices measured compared with WT animals. For example, hCG ß+ transgenic mice had a 7-fold increase in proximal cancellous bone volume compared with WT controls (hCG ß+, 31 ± 3.3; WT, 4.6 ± 1.6 BV/TV %; P < 0.05).
|
|
ß+ transgenic mice resulted in a significant reduction of femoral (36% decrease) (hCG ß+, 77 ± 1.5; hCG ß+ OVX, 49 ± 1.3 mg/cm2; P < 0.001) and tibial (33% decrease) (hCG ß+, 66 ± 2.4; hCG ß+ OVX, 44 ± 3.0 mg/cm2; P < 0.001) BMD and a reduction in all histomorphometric indices compared with their non-OVX counterparts. The data for the hCG ß+ OVX were comparable to age-matched WT OVX controls. The findings clearly support a role of the ovary in precipitating the increase in bone volume observed in the hCG-overexpressing mice.
LuRKO mice have a decrease in BMD
DXA and static histomorphometry was used to analyze long bones from LuRKO mice. Figure 9A shows no significant decreases in femoral (LuRKO, 47.2 ± 0.9; WT, 51.3 ± 2.4 mg/cm2; P = 0.14) and tibial (LuRKO, 43.6 ± 1.1; WT, 45.2 ± 1.3 mg/cm2; P = 0.34) BMD of 9- to 10-wk-old female LuRKO mice compared with age- and sex-matched WT controls. This declining trend in BMD in the 9- and 10-wk-old female LuRKO mice was corroborated with static histomorphometry; the BV/TV, TbWi, and TbN all decreased compared with WT (Table 3). The differences in BMD of 5-month-old male LuRKO mice vs. controls reached statistical significance with marked decreases in femoral (LuRKO, 53 ± 1.6; WT, 76 ± 1.6 mg/cm2; P < 0.0001) and tibial (LuRKO, 48 ± 4.2; WT, 54 ± 1.9 mg/cm2; P < 0.003) BMD (Fig. 6B
).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Our initial studies sought to address whether osteoblasts, which are known targets for estrogen (20, 21), expressed the LH receptor. Immunolocalization of hOBs using a well-characterized antibody that recognizes a region (residues 15–38) of the extracellular domain of the LH receptor (15) indicated that primary osteoblasts expressed the LH receptor. To confirm these findings, cell monolayers and appropriate controls were processed for protein extraction, subsequent SDS-PAGE, and immunoblotting using a different antibody, recognizing residues 1–11, to that applied for immunolocalization studies. The data indicated that osteoblasts expressed LH receptors, and these findings were reinforced through the identification of LH receptor mRNA by nested PCR and subsequent Southern blotting.
We were unable to demonstrate any increase in cAMP or ERK1/2 phosphorylation in hOBs or mC3T3-E1 cells treated with hCG. Cells were assessed for their ability to phosphorylate ERK1/2, to identify if MAPKs might be implicated; in a recent review of LH/hCG actions (9), it was reported that MAPK may be an important signaling pathway in nongonadal sites expressing LH receptors. Our data indicate that stimulation of osteoblast LH receptors does not increase cAMP or ERK phosphorylation, suggesting that there is either expression of a low number of LH receptors or that the receptor expressed is nonfunctional. Nonfunctional LH receptors have been previously reported in the turkey where three different, alternatively spliced, partial LH receptor cDNA isoforms have been identified. In the turkey, the alternatively spliced isoforms are differentially expressed in a tissue-specific manner with a relatively high expression of the receptor isoforms in peripheral nongonadal tissue. It is thought that the alternative splicing of the LH receptor has been evolutionarily conserved as similar isoforms have also been detected in chicken and swine (22).
To explore the possible role of LH/hCG on the skeleton in vivo, several murine models were used. These consisted of LuRKO (19), a transgenic model overexpressing hCG (hCG ß+) (18),1 and an hCG ß+ that had also been OVX. The generation of the null mutant was by inactivating exon 11 on the LHR gene. The transgenic was developed by overexpression of the hCG - and ß-subunit cDNA in two individual mouse lines that were subsequently crossed to yield double-transgenic mice producing very high levels of bioactive hCG (about 1 x 104 IU/liter) (18).1
BMD of the murine models was measured by DXA using the PIXImus scanner (Lunar), which is specifically designed for use with small animals. The accuracy of the PIXImus scanner in measuring calcium content was found to be highly accurate in previous studies in which highly significant correlations between femoral total BMD and ash weight (r = 0.86, P < 0.0001) were found. The coefficient of variation for femoral BMD, obtained after scanning 30 mouse femurs 5 times each with repositioning between scans, was found to be 2.7% (23). The analysis of bone mass by DXA is very useful at determining the amount of bone at a given site; however, it is important to note that DXA is unable to identify the cellular mechanisms responsible for any changes observed.
Ablation of the LH receptor resulted in a significant reduction in BMD by 5 months of age, as assessed by DXA, the reduction in femoral BMD was approximately 43% compared with WT controls. A similar pattern was observed for the younger (9–10 wk) animals, although the data obtained did not reach statistical significance. Furthermore static histomorphometry of these bones revealed reductions in cancellous bone volume, trabecular width and number. This reduction occurred in the face of more than 90% reduction of gonadal sex steroid production. The decline in bone mass for the LuRKO mice could be a consequence of either heightened bone resorption and/or reduced osteoblast activity.
In marked contrast was the profound increase in both BMD and histomorphometric parameters in the female hCG ß+ mice. The "sclerotic-like" presentation of the hCG ß+ could occur through reduced bone resorption and/or increased osteoblastic activity resulting in the apposition of new bone. The increase in BMD in these mice was approximately 1.3-fold, a change that has only been reported for mice when treated with supra-pharmacological levels of estrogen (24). Interestingly, the female hCG ß+ mice have a transient increase in estradiol action at peripuberty, as indicated by the presence of cornified vaginal mucosa and enlarged, fluid-filled uteri.1
The increase in the BMD of hCG ß+ mice may be evidence for a synergistic effect of moderately raised estrogen with the large increase of hCG. Alternatively, increases in the other serum hormones in these animals, most notably progesterone (60-fold) testosterone (6-fold) and prolactin (80-fold), may account for rise in BMD.1 Progesterone is known to stimulate bone formation and calcification in the presence of estrogen (25) and pseudopregnant rats, characterized by low estrogen but elevated progesterone, have a far higher rate of periosteal bone formation than that seen in OVX or intact animals (26). There has also been interest in the role of testosterone on the skeleton. The recent development of the androgen receptor knockout mouse (27) clearly supports a role for testosterone in male skeletal development because androgen receptor knockout mice have a reduced BMD. Bone also responds to changes in prolactin secretion. Raised prolactin levels are often accompanied by increases in bone loss (osteopenia). Furthermore, human osteoblasts express the mRNA for prolactin receptor (28) and generation of prolactin receptor null mutant mice results in a decrease in BMD (29).
It is also possible that additional ovarian factors are responsible for the increase in BMD in the female hCG ß+ mice because the circulating levels of estrogen in this model are not known to raise BMD to this extent. Indeed, daily estradiol injections of 4 mg/kg body weight, which result in a serum estradiol concentration of 100 nM, i.e. an increase in serum estrogen of approximately 500-fold, are required to obtain a similar rise in BMD and the histomorphometric data presented in this report (24). This implies that an agent, working in conjunction with estrogen is responsible for the changes in BMD. However, the presence of raised estrogen from pre-puberty, as would occur in the female hCG ß+ mice might still be sufficient to precipitate the changes.
Because estrogen levels were higher in the young female hCG ß+ mice, it was important to examine the role of the ovary in this model. A group of hCG ß+ mice were subjected to bilateral OVX and their bones analyzed as described. The data obtained clearly indicated that the ovary was responsible, directly or otherwise, for eliciting the changes observed because the BMD and histomorphometric analyses were comparable to WT OVX controls. The effect of OVX resulted in a reduction of BMD by 36%. Given that the female hCG ß+ mice start off with more bone (as determined by histomorphometry) than their WT littermates, the result suggests that OVX may produce a more rapid rate of bone loss in the presence of raised hCG.
Age-matched male hCG ß+ transgenics had comparable BMD to WT animals. This may be because the bone density in male mice reaches a plateau response at physiological sex steroid concentrations in WT mice; thus, a further increase in sex steroid concentrations as seen in the transgenic mice1 has no additional effect on the bone.
Overall, these in vivo and in vitro investigations support an indirect effect of gonadotropins on the skeleton. In view of the immunolocalization, immunoblotting and nested RT-PCR, our findings suggest that osteoblasts may express either low receptor numbers and/or the presence of nonfunctional receptors. This latter possibility is supported by the presence of different LH receptor isoforms that have been reported for nongonadal sites (22). The differential regulation of alternatively spliced LH receptor transcripts may be a physiologically conserved mechanism, but what is not known is whether these isoforms actually function.
The increase in bone observed in the hCG ß+ mice may be due to either heightened bone formation or a reduction in bone resorption. The identification of a potentially novel mechanism of bone formation may provide fundamental information for the development of anabolic bone therapies. Furthermore, if the increase in BMD in these mice is due to exposure of raised prepubertal estrogen, this may help develop predictions of bone diseases, such as osteoporosis. Although LH receptors have been described for many nongonadal estrogen-sensitive sites (9), bone does not appear to be a direct target for LH/hCG. Our findings strengthen evidence for the significant role by the gonads in regulating skeletal tissue development and turnover. The specific contribution of the different hormones on bone metabolism could be determined by crossing the hCG ß+ mice individually with knockout mice for estrogen, progesterone, and prolactin receptors. Potentially these studies may have implications for a new approach to therapeutic strategies.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: BMD, Bone mineral density; BV/TV, cancellous bone volume/tissue volume; D-PBS, Dulbecco’s PBS containing 0.1% BSA; DXA, dual energy x-ray absorptiometry; FGF, fibroblast growth factor; FITC, fluorescein isothiocyanate; hCG, human chorionic gonadotropin; hOB, human osteoblast; LuRKO, LH receptor knockout; MBB, milk-blocking buffer; OVX, ovariectomized; TbN, trabecular number; TbWi, trabecular width; WT, wild-type.
Rulli, S. B., P. Ahtiainen, S. Mdkeld, J. Toppari, M. Poutanen, and I. Huhtaniemi, manuscript submitted.
Received January 8, 2003.
Accepted for publication April 30, 2003.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  E. Dos Santos, M.-N. Dieudonne, M.-C. Leneveu, R. Pecquery, V. Serazin, and Y. Giudicelli In vitro effects of chorionic gonadotropin hormone on human adipose development J. Endocrinol., August 1, 2007; 194(2): 313 - 325. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T R. Kumar What have we learned about gonadotropin function from gonadotropin subunit and receptor knockout mice? Reproduction, September 1, 2005; 130(3): 293 - 302. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. L. Bowen, G. Verdile, T. Liu, A. F. Parlow, G. Perry, M. A. Smith, R. N. Martins, and C. S. Atwood Luteinizing Hormone, a Reproductive Regulator That Modulates the Processing of Amyloid-{beta} Precursor Protein and Amyloid-{beta} Deposition J. Biol. Chem., May 7, 2004; 279(19): 20539 - 20545. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP< |
), mLTC-1 (
), and HEK293 (
) cells. The mLTC-1 cells, which served as a positive control, possess high-affinity LH receptors. Both MG63 and HEK293 (negative control) show no hCG binding. These experiments were repeated three times, and data from a representative experiment are presented.
) compared with WT controls (
). Differences between groups were assessed using an unpaired Student’s t test. B, In contrast, age-matched male transgenic hCG 
