Isobel Braidman

Localization of estrogen receptor beta protein expression in adult human bone.

Evidence suggests that the newly described estrogen receptor β (ER‐β) may be important for estrogen (17β‐estradiol) action on the skeleton, but its cellular localization in adult human bone requires clarification. We addressed this by using indirect immunoperoxidase with a novel affinity purified polyclonal antibody to human ER‐β, raised to hinge domain (D) sequences from the human receptor. Bone was demineralized in 20% EDTA and all biopsy specimens were formalin‐fixed and wax‐embedded. Vigorous retrieval was essential for ER‐β detection. In sections (5 μm) of benign prostate hyperplasia, used as positive control, clear nuclear immunoreactivity was seen in glandular epithelial cells, with a 1:500 dilution of ER‐β40. For bone sections, optimal antibody dilutions were 1:100–1:250. We found that in normal bone (from graft operations), in fracture callus from both men and women (>25 years old), pagetic bone, osteophytes, and secondary hyperparathyroid bone, all from older patients, ER‐β was expressed clearly in osteoclast nuclei, with little cytoplasmic immunoreactivity. Nuclear immunoreactivity was still prominent in osteoclasts, with antibody diluted 1:500, although it faded in other cells. Osteoblasts, in areas of active bone formation or bone remodeling, also expressed ER‐β, as did some osteocytes. However, hypertrophic chondrocytes were negative, unlike mesenchymal cells, adjacent to the osteogenesis. Megakaryocytes and some capillary blood vessels cells were receptor positive. All ER‐β expression was blocked totally by preincubation of antibody with antigen. We conclude that ER‐β is expressed in cells of osteoblast lineage and in osteoclasts. The latter appear relatively abundant in this receptor and this might provide a means for direct action of estrogen on osteoclasts.

Demonstration of estrogen receptor mRNA in bone using in situ reverse-transcriptase polymerase chain reaction.

Falling estrogen levels affect the female skeleton profoundly. Following menopause, estrogen lack is a major cause of osteoporosis. The site of estrogen action in human bone, however, is unclear, but responsive cells must express the estrogen receptor (ER). One obstacle to localizing these cells is that mRNA for ER is expressed in low copy number. Hence, conventional molecular techniques are either too insensitive to detect receptor transcripts (in situ hybridization) or necessitate amplification of RNA extracted from tissue [Northern analysis and polymerase chain reaction (PCR)], thus failing to identify the specific target cells within the mixed-cell population of bone. In situ PCR (IS-PCR) is a technique that combines the sensitivity of PCR with the localization of conventional in situ hybridization. The technique has previously been used primarily to detect single-copy genes and viral DNA within cells. More recently, incorporation of a reverse-transcriptase reaction (IS-RT-PCR) has allowed the technique to be used to identify rare mRNAs within tissues. We have therefore applied the technique of IS-RT-PCR to localize ER mRNA first in human breast tumors, a known positive tissue, and then in bone. Using conventional riboprobe in situ hybridization, ER transcripts were not detectable in any bone cells within sections taken from normal bone and several actively remodeling bone tissues, namely, Pagets disease, renal hyperparathyroidism, and healing fracture callus. The technique of IS-RT-PCR, however, allowed amplification of transcripts to a detectable level. Following two cycles of amplification, hybridization signal was observed in osteoblasts and to a lower level in osteoclasts and occasional osteocytes. This positive signal was more obvious after five cycles, particularly in osteoclasts and osteocytes. After ten cycles, although signal was increased in osteoclasts and osteocytes, it appeared to be decreased in osteoblasts, suggesting that overamplification leads to loss of target complex from these cells. We conclude that several cell types in human bone express ER mRNA in vivo.

Evidence for cell-specific changes with age in expression of oestrogen receptor (ER) alpha and beta in bone fractures from men and women.

Oestrogen is recognized as important for maintaining bone mass in men and women. Oestrogen receptor (ER) α and the recently described ER‐β are both expressed in bone cells, but have different affinities for oestrogen agonists and plant oestrogens, which could be important in developing treatments for bone loss in both men and women. It is unclear, however, which isoform predominates in bone; cell type and age may influence their relative expression. The present study has compared ER‐α and ER‐β expression in serial sections of human fracture callus from males (n = 19, age range 5–72 years) and females (n = 15, age range 3–86 years) by indirect immunoperoxidase. Fracture callus was used as it can be readily obtained from individuals over a wide age range and contains a variety of bone cells. Antibody specificity was confirmed by western blotting and comparison of immunoreactivity in sections of breast tumour and benign prostate hyperplasia. No gender difference in ER expression was found in bone from individuals less than 40 years old. Proliferative chondrocytes were positive for both isoforms, but few larger hypertrophic cells were immunoreactive. ER‐α and ER‐β were co‐expressed in osteoclasts, suggesting that oestrogen may act directly on these cells. Osteoblasts, osteocytes, and mesenchymal cells also expressed both isoforms. In women over 40 years of age, however, relatively fewer biopsies contained osteocytes positive for ER‐α and ER‐β. Likewise, the proportions of osteoblasts and mesenchymal cells expressing ER‐β were reduced but ER‐α remained unaffected. In contrast, in men over 40 years, only the proportion of biopsies containing ER‐β‐positive mesenchymal cells was lower. In these older men and women, ER‐α and ER‐β expression was retained by the small proliferative chondrocytes. These results demonstrate that gender, age, and cell type are important determinants of ER isoform expression in skeletal cells. Copyright

Preliminary evidence for impaired estrogen receptor-α protein expression in osteoblasts and osteocytes from men with idiopathic osteoporosis

Although osteoporosis is usually associated with women, 1 in 12 men in the UK have the disease, and a third of these cases are idiopathic. Estrogen is now known to be associated with bone loss in older men, but we found, previously, that levels of this hormone were normal in younger cases of male idiopathic osteoporosis (MIO) in the age range 33-61 years. We therefore hypothesized that their estrogen responses in bone might be defective, through impaired estrogen receptor-alpha (ER)-alpha expression. Consequently, in the present study, we compared expression of ER-alpha by indirect immunofluorescence, semiquantitative image analysis, and in situ reverse transcription-polymerase chain reaction in bone sections from MIO patients (33-56 years) (N = 7); age-matched control men (N = 7); and, for reference, ovarian steroid (OS)-replete (N = 7) and OS-deficient women (N = 6). In the control men, 23 +/- 6% (mean +/- SEM) of osteoblasts and 14 +/- 2% of osteocytes expressed ER-alpha protein, similar to OS-replete women. Although receptor expression decreased in OS-deficient women, the loss of ER-alpha protein in MIO patients was more severe (1 +/- 0.5% osteocytes, 2 +/- 1% osteoblasts expressed receptor); however, ER-alpha messenger RNA (mRNA) was still expressed in controls and MIO patients. Bone loss in these patients may be due to deficient ER-alpha protein expression.

Demonstration of vitamin D receptor transcripts in actively resorbing osteoclasts in bone sections

The effects of the active metabolite of vitamin D, 1,25 dihydroxyvitamin D3 (1,25D), are mediated via the vitamin D receptor (VDR). 1,25D is known to have profound effects on bone resorption, but proof that the human osteoclast expresses VDR in vivo is absent. Receptors have been demonstrated in osteoblasts, and it has been generally accepted that the effects of 1,25D on formed osteoclasts are mediated via osteoblasts. Using conventional riboprobe in situ hybridization, VDR transcripts were readily detectable in osteoblasts within sections taken from normal bone and several actively remodelling bone tissues, namely, Pagets disease, renal hyperparathyroidism, and healing fracture callus. However, VDR transcripts also appeared to be present at low levels within osteoclasts from two pagetic samples and two hyperparathyroid samples. To examine this latter finding further, we have used the novel technique of in situ-reverse transcriptase-polymerase chain reaction (IS-RT-PCR) for specific amplification and detection of VDR mRNA within sections taken from the same conditions described above, and also from osteoclastoma samples. As expected, VDR transcripts were amplified and detected in osteoblasts and marrow cells, but were also prominently found in osteoclasts at approximately 50% of the level detected in osteoblasts in normal bone and at 60% in the active bone tissues. This suggests that in addition to effects on osteoclast precursors and those mediated via osteoblasts, 1,25D could exert direct effects on the active bone resorbing cells in vivo.

Immunofluorescent localization of estrogen receptor-α in growth plates of rabbits, but not in rats, at sexual maturity

Estrogens are considered essential to the mechanism for closure of epiphyses in both males and females. The mechanism for this, however, is still unclear. It is likely that estrogen acts directly on growth plate chondrocytes, but the localization of the cells expressing the estrogen receptor (ER) has yet to be ascertained. Moreover, in rodents, growth plates remain open well into adult life. Whether the distribution of estrogen target cells in rodent epiphyses differs from that in other species, is also unclear. We therefore compared localization of estrogen target cells (denoted by ER-alpha protein expression) in species in which growth plates fuse, with that in rodents. Thus, we have investigated ER-alpha protein expression in femoral growth plates from male and female rabbits, just at sexual maturity (6 months), when growth plate fusion was just commencing, and in rats of equivalent developmental stage (9 weeks). ER-alpha was detected in undecalcified cryosections by immunofluorescence with 1D5 monoclonal antibody, raised to human ER-alpha; uterine sections were positive controls. ER-alpha-positive cells were localized to the proliferative/early hypertrophic zone of male and female rabbits. By contrast, cells in the similar region of the mature rat growth plate were ER-alpha-negative in both genders, although receptor could be readily detected in uteri of mature female rats. In growth plates of immature male and female rats (6 weeks), however, ER-alpha was clearly expressed by cells of the proliferative/early hypertrophic zone, but was barely detectable in uteri from immature females. Our findings support the view that estrogen may act directly on the growth plate and, in species in which there is epiphyseal fusion, may thus have a role in this process. If ER-alpha expression is lost at sexual maturity, as in rodents, growth plates may remain open into adulthood. Our results also show the changes in ER-alpha expression in growth plates of maturing rats may be opposite to that in the uterus and raise the possibility that receptor expression may be controlled differently in reproductive and skeletal tissues.

Effect of ovarian steroid deficiency on oestrogen receptor alpha expression in bone.

The mechanism by which oestrogen and hormone replacement therapy (HRT) maintain bone mass in women is still unclear. It has previously been shown that cells of osteoblast lineage in vivo, particularly osteocytes, express oestrogen receptor α (ERα). Nevertheless, it is still debatable whether oestrogen and the ovarian steroids have a direct affect on osteocytes. If they could regulate osteocyte ERα expression, this would be strong evidence for the involvement of these cells in the hormonal regulation of bone mass. This study therefore aimed to compare bone biopsies from women who were replete with ovarian steroids (pre‐ovariectomy or post‐HRT) with those from the same women when hormone‐deficient (post‐ovariectomy or pre‐HRT) for cellular localization of ERα protein or mRNA expression by indirect immunofluorescence, or by in situ hybridization combined with reverse transcriptase‐polymerase chain reaction (IS‐RT‐PCR) respectively. Image analysis showed that proportions of osteocytes positive for immunodetectable ERα were higher in hormone‐replete than in hormone‐deficient women (25 ± SEM 3 per cent, 12 ± SEM 4 per cent, respectively; n = 5), with similar but non‐statistically significant changes in osteoblasts. This was observed even when HRT was commenced 18 years after menopause. In contrast, grain volume/unit cell area of osteoblast mRNA signal was markedly higher when hormone‐deficient (0·055 ± 0·01) than when hormone‐replete (0·016 ± 0·004), with similar but non‐significant differences in osteocytes. This preliminary study indicates up‐regulation of osteocyte ERα protein by ovarian steroids in these patients, which is accompanied by decreased osteoblast ERα mRNA expression, providing further evidence for the involvement of osteocytes in the regulation of skeletal structure by ovarian steroids. Copyright

Effects of two novel bisphosphonates on bone cells in vitro.

Bisphosphonates are now widely used in the treatment of bone diseases, particularly where there is uncontrolled bone resorption, as they are known to be potent inhibitors of osteoclasis. It is still unclear whether the bisphosphonates act by inhibiting osteoclast maturation or by blocking the mechanism of bone resorption, and little is known of their effects on osteoblasts. Recent studies with 3-amino-1, hydroxypropylidene-1,1-bisphosphonic acid (APD) in the treatment of osteolytic metastases in breast cancer have suggested that APD may affect osteoblasts directly. We have now investigated the effects of two novel bisphosphonates, CGP 47072 and CGP 42446A on osteoclastogenesis in fetal rat calvariae cultured on collagen gels and on human osteoblasts (hOB) cultured as explants from bone taken from patients at surgery. We also compared the action of these new bisphosphonates with that of APD, which at concentrations of 2.5 x 10(-6) M to 2.5 x 10(-10) M inhibited osteoclast recruitment, even when this was stimulated by conditioned medium from MCF7 breast cancer cells. This bisphosphonate was particularly potent if cultured with calvaria taken at 19 days gestation, when more immature osteoclast precursors are present. If calvariae from 20 days gestation were used, which contain more mature cells, it produced less inhibition. In contrast, CGP 42446A (2.5 x 10(-6) M to 2.5 x 10(-8) M) was more effective in inhibiting osteoclast maturation in calvariae from 20 days gestation than in those from 19 days. CGP 47072 had a similar pattern of effects but was less potent than either of the other two compounds. APD or CGP 42446A at 10(-5) M significantly inhibited hOB numbers and DNA synthesis, but lower concentrations had little effect. CGP 47072 did not inhibit human osteoblast replication. It is unlikely that these effects are due to calcium chelation, as none of these compounds mimicked results obtained with EDTA, which was effective only at 2.5 x 10(-6) M in reducing osteoclast size and 10(-4) M in human osteoblast cultures. These results demonstrate that all three bisphosphonates are able to inhibit osteoclast formation at low concentrations. APD may be able to influence less mature osteoclast precursors and CGP 42446A and CGP 47072 may exert their effects on the fusion of more mature precursor cells on the bone surface. At these concentrations, however, there is little or no effect on osteoblasts.

EXTRA‐ENDOCRINE FUNCTIONS OF VITAMIN D

I t is well established that 1.25 dihydroxyvitamin D3 (1,25 (OH)2D& the active metabolite of vitamin D3, has a major role in the endocrine control of circulating calcium levels through its actions on the upper small intestine. This picture has become confused of late by extensive reports that a wide variety of tissues and cells both have receptors for the sterol and respond to i t by changes in their biological activity. These other target tissues are, in most cases, not obviously part of the process that regulates circulating calcium levels, and may have functions outside this endocrine system.

Interleukin-6 does not mediate the stimulation by prostaglandin E2, parathyroid hormone, or 1,25 dihydroxyvitamin D3 of osteoclast differentiation and bone resorption in neonatal mouse parietal bones

The cytokine interleukin-6 (IL-6) was produced by neonatal mouse parietal bones during a 6- or 48-hour culture period in response to prostaglandin E2 (PGE2) and bovine parathyroid hormone (PTH) 1-34 fragment but not 1,25-dihydroxyvitamin D3 [1,25(OH)2D3]. At the same time there was an increase in tartrate-resistant, acid phosphatasepositive osteoclasts (TRAP+OC) with all three osteotropic effectors over 6 hours, and an increase in 45Ca release over 48 hours. TRAP+OC numbers on PGE2-stimulated bones were positively correlated with IL-6 concentration. Our aim was to determine if IL-6 mediated this response. Recombinant human IL-6 (rhIL-6) was added to parietal bones in culture at concentrations within the range that PGE2 or PTH would produce during incubation. However, over 6 or 48 hours, rhIL-6 did not stimulate TRAP+OC to increase in number nor did it cause an increase in calcium release over 48 hours. Adding an antibody against mouse IL-6 to bone cultures stimulated with PTH or PGE2 neutralized the resulting IL-6 bioactivity by up to 92% but did not inhibit TRAP+OC formation. We conclude that although IL-6 is produced in response to two important stimulators of bone resorption, it does not mediate osteoclast differentiation or bone resorption in this model.