Influence of different decalcifying agents on EGF and EGFR immunostaining

Evaluation of EDTA and nitric acid solutions for decalcification of joints in AG/WT, BALB/c, C57, DBA1/J mice, and in Wistar rats

Decalcification of mineralized samples for microscopic analysis involves competing factors including decalcification time, preservation of tissue integrity and cost. We investigated the utility of different decalcification solutions for studying joints in AG/WT, BALB/c, C57, DBA1/J mice and Wistar rats. The hind paws of the rodents were removed and fixed with 10% buffered formalin. Specimens were divided randomly into three groups for demineralization: 10% nitric acid, 12.5% EDTA at room temperature and 12.5% EDTA at 35 °C with shaking. Sections of joints were stained with hematoxylin and eosin (H & E). We evaluated decalcification time and expense, ease of cutting sections, preservation of nuclear basophilia and intranuclear detail, and intensity of eosin staining. The 10% nitric acid solution produced the most rapid decalcification for the mice, but not the rats. The 12.5% EDTA solution at 35 °C with shaking did not decrease decalcification time. Effects on microtomy were variable as were the effects on H & E staining. The EDTA solution provided the best basophilia and intranuclear detail for the mice. For rats, only 12.5% EDTA at 35 °C with shaking produced good preservation. Preservation of nuclear basophilia and intranuclear detail for rats was best with 10% nitric acid and EDTA 35 °C. For mice, 10% nitric acid failed to preserve nuclear basophilia and intranuclear detail. For intensity of eosin staining, EDTA at room temperature and EDTA 35 °C was best for both mice and rats. Sections also exhibited good H & E staining in most samples decalcified with 10% nitric acid. Although we found considerable variation among groups of animals, we found less variation among the different mouse strains than between mice and Wistar rats.

A comparison of three decalcification agents for assessments of cranial fracture histomorphology

The extraction of mineral calcium from bone by decalcification is a critical step in the preparation of histological samples for light microscopy. This study assessed the time required for complete decalcification and the resultant histomorphological preservation of bone histomorphology by three decalcification agents: 7% hydrochloric acid (HCl), 5% nitric acid, and 10% ethylenediaminetetraacetic acid (EDTA). The goal of this study was to identify which decalcification agent provides the optimal combination of expedient processing and quality histological outcomes of cranial fracture samples. HCl provided the most rapid decalcification ( X ¯  = 3.57 days), nitric acid followed closely ( X ¯  = 10.35 days), while EDTA took significantly longer on average ( X ¯  = 78.97 days) but encompassed a broader range of times. Decalcification agent, sample thickness, sample width, and decedent age are significant predictors of decalcification time. Sample visualization quality, measured for tissues, cells, and nuclei on a five-point Likert scale, was highest for samples decalcified in 10% EDTA, second highest using 5% nitric acid, and lowest for 7% HCl. The quality difference between EDTA and nitric acid was not highly significant for any of the three features. For basic assessments of bone histomorphology, the study results indicate 5% nitric acid is suitable for the decalcification of adult specimens and samples thicker than 3 mm. EDTA is a suitable agent for thin samples of the cranial vault (<3 mm) from infants and young children less than three years old, decalcifying samples in a timeframe comparable to nitric acid while providing the best quality and clarity of samples.

Effect of decalcification protocols on immunohistochemistry and molecular analyses of bone samples

Diagnosis of osteocartilaginous pathologies depends on morphological examination and immunohistochemical and molecular biology analyses. Decalcification is required before tissue processing, but available protocols often lead to altered proteins and nucleic acids, and thus compromise the diagnosis. The objective of this study was to compare the effect of different methods of decalcification on histomolecular analyses required for diagnosis and to recommend an optimal protocol for processing these samples in routine practice. We prospectively submitted 35 tissue samples to different decalcification procedures with hydrochloric acid, formic acid, and EDTA, in short, overnight and long cycles for 1 to >10 cycles. Preservation of protein integrity was examined by immunohistochemistry, and quality of nucleic acids was estimated after extraction (DNA and RNA concentrations, 260/280 ratios, PCR cycle thresholds), analysis of DNA mutations (high-resolution melting) or amplifications (PCR, in situ hybridization), and detection of fusion transcripts (RT-PCR, in situ hybridization). Hydrochloric acid- and long-term formic acid-based decalcification induced false-negative results on immunohistochemistry and molecular analysis. EDTA and short-term formic acid-based decalcification (<5 cycles of 6 h each) did not alter antigenicity and allowed for detection of gene mutations, amplifications or even fusion transcripts. EDTA showed superiority for in situ hybridization techniques. According to these results and our institutional experience, we propose recommendations for decalcification of bone samples, from biopsies to surgical specimens.

Anatomy and function of the vertebral column lymphatic network in mice

Cranial lymphatic vessels (LVs) are involved in the transport of fluids, macromolecules and central nervous system (CNS) immune responses. Little information about spinal LVs is available, because these delicate structures are embedded within vertebral tissues and difficult to visualize using traditional histology. Here we show an extended vertebral column LV network using three-dimensional imaging of decalcified iDISCO+-clarified spine segments. Vertebral LVs connect to peripheral sensory and sympathetic ganglia and form metameric vertebral circuits connecting to lymph nodes and the thoracic duct. They drain the epidural space and the dura mater around the spinal cord and associate with leukocytes. Vertebral LVs remodel extensively after spinal cord injury and VEGF-C-induced vertebral lymphangiogenesis exacerbates the inflammatory responses, T cell infiltration and demyelination following focal spinal cord lesion. Therefore, vertebral LVs add to skull meningeal LVs as gatekeepers of CNS immunity and may be potential targets to improve the maintenance and repair of spinal tissues.

Cellular and collagen reference values of gingival and periodontal ligament tissues in rats: a pilot study

Reference data are lacking on the periodontal ligament and the gingival tissue of the rat model, which would be useful for studies of new medical or biomaterial periodontal treatments. The objective of the current study was to propose cellular and collagen reference values of gingival and periodontal ligament tissues in rat, using a simple and reliable quantitative method after decalcification. Mandibular samples of ten adult Sprague-Dawley rats were used. Mild decalcification was carried out using ethylenediaminetetraacetic acid (EDTA) to preserve the morphology of tissues. Half of the samples were decalcified and the other half were not. The gingiva and the periodontal ligament were analyzed. Descriptive histology and computer-assisted image analysis were performed. The data showed that qualitatively, cellular and extracellular matrix morphologies were well preserved compared to non-decalcified periodontal soft tissue biopsies. Histomorphometrically, constitutive cellularity and the total amount of native collagen, collagen directionality and collagen anisotropy in both experimental conditions did not significantly differ. Taken together, these results suggested that EDTA decalcification did not negatively affect the studied endpoints. Moreover, this mild decalcification method allowed in situ maintenance of the periodontal soft and hard tissue integrity. The structural and compositional computerized assessment performed in the healthy periodontal soft tissue could provide reference values that will be required for future assessment on the effects of pathological, reparative and regenerative processes in rat periodontal soft tissues.

Tissue Morphology and Antigenicity in Mouse and Rat Tibia: Comparing 12 Different Decalcification Conditions

Conventional bone decalcification is a time-consuming process and is therefore unsuitable for clinical applications and time-limited research projects. Consequently, we compared the effect of four different decalcification solutions applied at three different temperatures, and assessed the rate of decalcification and the implications on tissue morphology and antigenicity of mouse and rat tibiae. Bones were decalcified with 10% ethylenediaminetetraacetic acid (EDTA), 10% formic acid, 5% hydrochloric acid, and 5% nitric acid at 4C, 25C, and 37C. Decalcification in both species was fastest in nitric acid at 37C and slowest in EDTA at 4C. Histological and immunohistochemical staining confirmed that the conventional protocols of EDTA at 4C and 25C remain the best option regarding the quality of tissue preservation. Whereas formic acid at 4C is a good alternative saving about 90% of the decalcification time, hydrochloric and nitric acids should be avoided particularly in case of rat tibia. By contrast, due to their smaller size, mouse tibiae had shorter decalcification times and tolerated higher temperatures and exposure to acids much better. In conclusion, this study demonstrated that depending on the specific research question and sample size, alternative decalcification methods could be used to decrease the time of decalcification while maintaining histological accuracy.

Spleen Regulates Hematopoietic Stem/Progenitor Cell Functions Through Regulation of EGF in Cirrhotic Hypersplenism

BACKGROUND

Hematopoietic abnormality is a common cause of cirrhotic hypersplenism (CH) complications and death; it causes serious adverse effects and is associated with bleeding, anemia, infection in CH patients. However, the underlying mechanism is unclear.

AIMS

We aimed to investigate the effects of the spleen on hematopoiesis and hematopoietic stem/progenitor cells (HSPCs) in CH patients.

METHODS

Eleven CH patients were enrolled to assess the effects of the spleen on HSPC functions. Hematopoietic changes were examined by flow cytometry analysis. HSPC functions were detected with colony-forming assays and in vitro cell cultures. Enzyme-linked immunosorbent assay (ELISA) was used to test the concentration of epithelial growth factor (EGF).

RESULTS

The number of HSPCs was decreased in CH patients and was rescued after splenectomy. Serum from CH patients dysregulated HSPCs function, and serum from splenectomy patients restored the dysregulated HSPC function in vitro. The concentration of EGF was decreased in CH patients and was restored to normal level after splenectomy. EGF rescued the dysregulated HSPCs function in vitro.

CONCLUSIONS

The spleen can regulate the functions of HSPCs in CH patients by regulating EGF signaling. EGF may be a therapeutic target for CH treatment.

Immunohistochemical detection of p53, PTEN, Rb, and p16 in canine osteosarcoma using tissue microarray

Although inactivating mutations of tumor suppressor genes are well described in cell lines of canine osteosarcoma (OS), expression of tumor suppressor proteins in spontaneous disease is poorly characterized. We determined the immunohistochemical expression of p53, PTEN, Rb, and p16 in a large cohort of dogs with OS. Formalin-fixed, paraffin-embedded samples of canine OS were analyzed retrospectively. Primary tumor samples from 145 dogs, collected between 2003 and 2008, were evaluated by tissue microarray. Streptavidin-biotin complex immunohistochemistry was performed using monoclonal antibodies for Rb and PTEN and polyclonal antibodies for p16 and p53. The average age of dogs was 7.6 y, and 118 of 145 (81%) were purebred. Most commonly represented purebreds were Greyhound (23%), Rottweiler (11%), and Labrador Retriever (10%). Immunohistochemical detection of p53, PTEN, Rb, and p16 was 81%, 61%, 66%, and 66%, respectively. The staining pattern for p16 was primarily cytoplasmic; the predominant pattern for PTEN, Rb, and p53 was cytoplasmic and nuclear. Exclusively cytoplasmic staining was noted in 19% of samples positive for p53 and 8% of samples positive for Rb. Kaplan-Meier curves showed that protein expression was not associated with significant differences in overall survival ( p > 0.191). We documented heterogeneity in both immunostaining and subcellular localization of tumor suppressor proteins, providing further characterization of canine OS.

Tissue integrity, costs and time associated with different agents for histological bone preparation

The selection of an appropriate demineralizing solution in pathology laboratories depends on several factors such as the preservation of cellularity, urgency of diagnostic and financial costs. The aim of this study was to test different decalcification bone procedures in order to establish the best value of these in formalin-fixed and paraffin-embedded samples. Femurs were removed from 13 adult male Wistar rats to obtain 130 bone disks randomly divided into five groups that were demineralized in different concentrations of nitric acid (Group I); formic acid (Group II); acetic acid (Group III); EDTA, pH7.4 (Group IV) and Morsés solution (Group V). Serial, 3-μm-thick sections were obtained and stained with hematoxylin-eosin to calculate the percentage of osteocyte-occupied lacunae. The sections were also stained with Masson's trichrome in conjunction with picrosirius red under polarized light followed by a semi-quantitative analysis to verify the adjacent muscle-to-bone integrity and preservation of collagen fibres. The highest percentage of osteocyte-occupied lacunae was found with 10% acetic acid solution (95.64 ± 0.95%) and Group I (nitric acid) demanded the shorter time (0.8-5.7days). Of all solutions, 5% nitric acid incurred the lowest cost to achieve complete demineralization compared with other solutions (p < .001). Group IV (EDTA) had the highest integrity of muscle and collagen type I and III (P < 0.01). Demineralization with 10% acetic acid was the most effective at preserving bone tissue, while 5% EDTA was the best at maintaining collagen and adjacent muscle to bone. In conclusion, nitric acid at 5% showed the most efficient result as it balanced both time and cost as a demineralizing solution.

Influence of decalcification procedures on immunohistochemistry and molecular pathology in breast cancer

Distant breast cancer metastases are nowadays routinely biopsied to reassess receptor status and to isolate DNA for sequencing of druggable targets. Bone metastases are the most frequent subgroup. Decalcification procedures may negatively affect antigenicity and DNA quality. We therefore evaluated the effect of several decalcification procedures on receptor status and DNA/RNA quality. In 23 prospectively collected breast tumors, we compared ERα, PR and HER2 status by immunohistochemistry in (non-decalcified) tissue routinely processed for diagnostic purposes and in parallel tissue decalcified in Christensen's buffer with and without microwave, EDTA and Formical-4. Furthermore, HER2 fluorescence in situ hybridization and DNA/RNA quantity and quality were assessed. We found that the percentage of ERα-positive cells were on average lower in EDTA (P=0.049) and Formical-4 (P=0.047) treated cases, compared with controls, and PR expression showed decreased antigenicity after Christensen's buffer treatment (P=0.041). Overall, a good concordance (weighted kappa) was seen for ERα, PR and HER2 immunohistochemistry when comparing the non-decalcified control tissues with the decalcified tissues. For two patients (9%), there was a potential influence on therapeutic decision making with regard to hormonal therapy or HER2-targeted therapy. HER2 fluorescence in situ hybridization interpretation was seriously hampered by Christensen's buffer and Formical-4, and DNA/RNA quantity and quality were decreased after all four decalcification procedures. Validation on paired primary breast tumor specimens and EDTA-treated bone metastases showed that immunohistochemistry and fluorescence in situ hybridization were well assessable and DNA and RNA yield and quality were sufficient. With this, we conclude that common decalcification procedures have only a modest negative influence on hormone and HER2 receptor immunohistochemistry in breast cancer. However, they may seriously affect DNA/RNA-based diagnostic procedures. Overall, EDTA-based decalcification is therefore to be preferred as it best allows fluorescence in situ hybridization and DNA/RNA isolation.

Tooth tissue engineering: tooth decellularization for natural scaffold

AIM

Tissue engineering is a multidisciplinary science that aims to produce replacement organs and biological substitutes. One of the techniques involves decellularizing a biological organ without altering its structure. One challenge is how to demonstrate which method would be better for this process.

METHODOLOGY

Fifty premolar teeth were divided into five groups: G1 (control): solution of 10% formaldehyde; G2: phosphate buffer saline (PBS), 28 g of tetrasodium ethylenediaminetetraacetic (EDTA), sodium hypochlorite 2.5% (SH); G3: PBS, EDTA and 40v hydrogen peroxide (HP); G4: PBS, EDTA, SH, enzymatic detergent (ED); and G5: PBS, EDTA, HP, ED. Each group was analyzed by scanning electron microscopy (SEM), x-ray, measured weights and color and received statistical analysis.

CONCLUSION

This study demonstrated that G5 was the most appropriate method to obtain a natural scaffold.

Mutational profiling of bone metastases from lung adenocarcinoma: results of a prospective study (POUMOS-TEC)

Targeted therapies have improved patient survival in metastatic lung adenocarcinoma. Molecular diagnosis is a key element to identify oncogenic drivers predicting the efficacy of these agents. In stage IV patients, histopathological diagnosis is often performed on bone metastases biopsy, but routine procedure of decalcification may alter DNA quality for subsequent molecular tests. We set up a procedure to perform molecular analyses on bone metastasis and describe the results of mutational profiling. POUMOS-TEC is a prospective study conducted in stage IV lung adenocarcinomas. Bone metastasis specimens from surgery and CT-scan guided biopsies were sent fresh for immediate formalin-fixation. Decalcification was performed, only when necessary, using EDTA. Controls were processed with acid decalcification. DNA extraction was performed after laser microdissection. Mutational profiling of oncogenic drivers was conducted as recommended by the French National Cancer Institute. Diagnosis efficiency of the computed tomography (CT)-scan guided biopsy process was assessed. Among 177 collected bone metastases specimens, 49 came from lung adenocarcinomas. Specimens processed with no decalcification or EDTA (n=45) provided high-quality DNA. Molecular profiling was performed in 44/45 (98%) of cases. The results of the whole panel of oncogenic drivers (EGFR, KRAS, BRAF, PIK3CA, HER2 and ALK) were obtained in 41/45 (91%) of cases. A mutation was observed in 50% of cases including 32% of KRAS and 14% of epidermal growth factor receptor (EGFR) mutations. CT-scan biopsy efficiency rate was 96%. We demonstrated the feasibility to routinely conduct mutational profiling on bone metastases biopsies. We observed a higher rate of EGFR mutations (+42%) in comparison with the average rate of all stage IV lung adenocarcinomas. This procedure is a new step toward the goal of personalized medicine to treat lung cancers and other osteophilic tumors.

Identification of the vitamin D receptor in osteoblasts and chondrocytes but not osteoclasts in mouse bone

Bone is clearly a target of vitamin D and as expected, the vitamin D receptor (VDR) is expressed in osteoblasts. However, the presence of VDR in other cells such as osteocytes, osteoclasts, chondroclasts, and chondrocytes is uncertain. Because of difficulties in obtaining sections of undecalcified adult bone, identification of the site of VDR expression in adult bone tissue has been problematic. In addition, the antibodies to VDR used in previous studies lacked specificity, a property crucial for unambiguous conclusions. In the present study, VDR in the various cells from neonatal and adult mouse bone tissues was identified by a highly specific and sensitive immunohistochemistry method following bone decalcification with EGTA. For accurate evaluation of weak immunosignals, samples from Demay VDR knockout mice were used as negative control. Molecular markers were used to identify cell types. Our results showed that EGTA-decalcification of bone tissue had no detectable effect on the immunoreactivity of VDR. VDR was found in osteoblasts and hypertrophic chondrocytes but not in the multinucleated osteoclasts, chondroclasts, and bone marrow stromal cells. Of interest is the finding that immature osteoblasts contain large amounts of VDR, whereas the levels are low or undetectable in mature osteoblasts including bone lining cells and osteocytes. Proliferating chondrocytes appear devoid of VDR, although low levels were found in the hypertrophic chondrocytes. These data demonstrate that osteoblasts and chondrocytes are major targets of 1α,25-dihydroxyvitamin D, but osteoclasts and chondroclasts are minor targets or not at all. A high level of VDR was found in the immature osteoblasts located in the cancellous bone, indicating that they are major targets of 1α,25-dihydroxyvitamin D. Thus, the immature osteoblasts are perhaps responsible for the vitamin D hormone signaling resulting in calcium mobilization and in osteogenesis.

Tissue preparation and immunostaining of mouse sensory nerve fibers innervating skin and limb bones

Detection and primary processing of physical, chemical and thermal sensory stimuli by peripheral sensory nerve fibers is key to sensory perception in animals and humans. These peripheral sensory nerve fibers express a plethora of receptors and ion channel proteins which detect and initiate specific sensory stimuli. Methods are available to characterize the electrical properties of peripheral sensory nerve fibers innervating the skin, which can also be utilized to identify the functional expression of specific ion channel proteins in these fibers. However, similar electrophysiological methods are not available (and are also difficult to develop) for the detection of the functional expression of receptors and ion channel proteins in peripheral sensory nerve fibers innervating other visceral organs, including the most challenging tissues such as bone. Moreover, such electrophysiological methods cannot be utilized to determine the expression of non-excitable proteins in peripheral sensory nerve fibers. Therefore, immunostaining of peripheral/visceral tissue samples for sensory nerve fivers provides the best possible way to determine the expression of specific proteins of interest in these nerve fibers. So far, most of the protein expression studies in sensory neurons have utilized immunostaining procedures in sensory ganglia, where the information is limited to the expression of specific proteins in the cell body of specific types or subsets of sensory neurons. Here we report detailed methods/protocols for the preparation of peripheral/visceral tissue samples for immunostaining of peripheral sensory nerve fibers. We specifically detail methods for the preparation of skin or plantar punch biopsy and bone (femur) sections from mice for immunostaining of peripheral sensory nerve fibers. These methods are not only key to the qualitative determination of protein expression in peripheral sensory neurons, but also provide a quantitative assay method for determining changes in protein expression levels in specific types or subsets of sensory fibers, as well as for determining the morphological and/or anatomical changes in the number and density of sensory fibers during various pathological states. Further, these methods are not confined to the staining of only sensory nerve fibers, but can also be used for staining any types of nerve fibers in the skin, bones and other visceral tissue.

Gender- and region-specific variations of estrogen receptor α and β expression in the growth plate of spine and limb during development and adulthood

Although estrogen action is indispensable for normal bone growth in both genders, the roles of estrogen receptors (ERs) in mediating bone growth are not fully understood. The effects of ER inactivation on bone growth are sex and age dependent, and may differ between the axial and appendicular regions. In this study, the spatial and temporal expression of ERα and β in the tibial and spinal growth plates of the female and male rats during postnatal development was examined to explore the possible mechanisms. The level of mRNA was examined and compared with quantitative real-time PCR. The spatial location was determined by immunohistochemical analysis. The 1-, 4-, 7-, 12- and 16-week age stages correspond to early life, puberty and early adulthood after puberty, respectively. Gender- and region-specific differences in ERα and β expression were shown in the growth plates. Mainly nuclear staining of ERα and β immunoreactivity was demonstrated in the spinal and tibial growth plate chondrocytes for both genders. Moreover, our study indicated significant effect of gender on temporal ERα and β expression and of region on temporal ERα/ERβ expression ratio. However, spatial differences of region-related ERα and β expression were not observed. Gender-related spatial changes were detected only at 16 weeks of both spine and limb growth plates. ERα and β immunoreactivity was detected in the resting, proliferative and prehypertrophic chondrocytes in the early life stage and during puberty. After puberty, ERα expression was mainly located in the late proliferative and hypertrophic chondrocytes in female, whereas the expression still extended from the resting to hypertrophic chondrocytes in males. Gender- and region-specific expression patterns of ERα and β gene might be one possible reason for differences in sex- and region-related body growth phenotypes. Gender, age and region differences should be taken into consideration when the roles of ERs in the growth plate are investigated.

MT1-MMP expression in the odontogenic region of rat incisors undergoing interrupted eruption

MT1-MMP (membrane type matrix metalloproteinase-1) has been considered an important membrane-type matrix metalloproteinase involved in the remodeling process in tissue and organ development, including the processes of the tooth and root growth and dental eruption. Therefore, the aims of this study were to evaluate MT1-MMP expression in the odontogenic region, as well as the eruption rate and morphology of the lower-left rat incisor, where the eruption process was interrupted for 14 days by a steel wire attached from the center of the incisor labial face and braced to the first molar. In the interrupted eruption group, the eruption rate was significantly reduced, producing drastic morphological alterations in the tooth germ and socket area. The MT1-MMP expression was widespread in the dental follicle, in both groups studied (normal and interrupted eruption groups); however a significant decrease in immunostaining was observed in the interrupted eruption group. Results indicate that MT1-MMP may have an important role in the process of dental eruption.

Age-related variations of leptin receptor expression in the growth plate of spine and limb: gender- and region-specific changes

Leptin is a potent growth-stimulating factor of bone. The effects of leptin on bone growth differ significantly between axial and appendicular regions. Gender differences of leptin function have also been suggested in normal pubertal development. To explore the mechanisms underlying these effects, we investigated the spatial and temporal expressions of the active form of the leptin receptor (Ob-Rb) in the tibial and spinal growth plates of the female and male rats during postnatal development. The 1-, 4-, 7-, 12- and 16-week age stages are representative for early life, puberty and early adulthood after puberty, respectively. Quantitative real-time PCR was used for Ob-Rb mRNA examination and comparison. The spatial location of Ob-Rb was determined by immunohistochemical analysis. There were gender- and region-specific differences in Ob-Rb mRNA expression in the growth plate. Mainly cytoplasm staining of Ob-Rb immunoreactivity was observed in the spinal and tibial growth plate chondrocytes of both genders. Spatial differences of region- and gender-related Ob-Rb expression were not observed. Ob-Rb immunoreactivity was detected in the resting, proliferative and prehypertrophic chondrocytes in early life stage and during puberty. After puberty, staining was mainly located in the late proliferative and hypertrophic chondrocytes. The results of Ob-Rb HSCORE analysis were similar to those obtained from quantitative real-time PCR. Our study indicated direct effects on the chondrocytes of the growth plate in different development stages. The region-specific expression patterns of Ob-Rb gene might be one possible reason for contrasting phenotypes in limb and spine. Different Ob-Rb expression patterns might partly contribute to age- and gender- related differences in trabecular bone mass.