Comparative transcriptomic analysis identifies distinct molecular signatures and regulatory networks of chondroclasts and osteoclasts.

BACKGROUND: Chondroclasts and osteoclasts have been previously identified as the cells capable of resorbing mineralized cartilage and bone matrices, respectively. While both cell types appear morphologically similar, contain comparable ultrastructural features, and express tartrate-resistant acid phosphatase (TRAP), however, no information is available about the genomic similarities and differences between osteoclasts and chondroclasts. METHODS: To address this question, we laser captured homogeneous populations of TRAP-positive cells that interact with bone (osteoclasts) and TRAP-positive cells that interact with mineralized cartilage (chondroclasts) on the same plane from murine femoral fracture callus sections. We then performed a global transcriptome profiling of chondroclasts and osteoclasts by utilizing a mouse genome Agilent GE 4X44K V2 microarray platform. Multiple computational approaches and interaction networks were used to analyze the transcriptomic landscape of osteoclasts and chondroclasts. RESULTS: Our systematic and comprehensive analyses using hierarchical clustering and principal component analysis (PCA) demonstrate that chondroclasts and osteoclasts are transcriptionally distinct cell populations and exhibit discrete transcriptomic signatures as revealed by multivariate analysis involving scatter plot, volcano plot, and heatmap analysis. TaqMan qPCR was used to validate the microarray results. Intriguingly, the functional enrichment and integrated network analyses revealed distinct Gene Ontology terms and molecular pathways specific to chondroclasts and osteoclasts and further suggest that subsets of metabolic genes were specific to chondroclasts. Protein-protein interaction (PPI) network analysis showed an abundance of structured networks of metabolic pathways, ATP synthesis, and proteasome pathways in chondroclasts. The regulatory network analysis using transcription factor-target gene network predicted a pool of genes including ETV6, SIRT1, and ATF1 as chondroclast-specific gene signature. CONCLUSIONS: Our study provides an important genetic resource for further exploration of chondroclast function in vivo. To our knowledge, this is the first demonstration of genetic landscape of osteoclasts from chondroclasts identifying unique molecular signatures, functional clustering, and interaction network.

Deletion of Wnt5a in osteoclasts results in bone loss through decreased bone formation.

Bone remodeling is achieved through the coupled activities of osteoclasts and osteoblasts that are controlled by many locally generated secreted factors, including WNT5A. While previous studies have demonstrated that osteoblast-derived WNT5A promotes osteoclastogenesis, the function of osteoclast-derived WNT5A on bone remodeling has remained unexplored. We examined the effects of osteoclast-derived WNT5A on bone homeostasis by utilizing the Cathepsin K-Cre (Ctsk-Cre) mouse to conditionally delete Wnt5a in mature osteoclasts. These mice exhibited reduced trabecular and cortical bone. The low bone-mass phenotype was driven by decreased bone formation, not osteoclast-mediated bone resorption, as osteoclast number and serum CTX marker were unchanged. Furthermore, molecular analysis of osteoclast- and osteoblast-derived WNT5A identified a serine-phosphorylated WNT5A that is unique to RANKL-treated macrophages mimicking osteoclasts. This study suggests a new paradigm in which WNT5A has opposing effects on bone remodeling that are dependent on the cell of origin, an effect that may result from cell type-specific differential posttranslational modifications of WNT5A.

Articular cartilage protection in Ctsk-/- mice is associated with cellular and molecular changes in subchondral bone and cartilage matrix.

Osteoarthritis (OA) is a degenerative disease and a major cause of chronic disability in aging individuals. Cathepsin K (CatK), encoded by the Ctsk gene, has been implicated in the pathogenesis of pycnodysostosis and osteoporosis. The use of a selective inhibitor of CatK was recently shown to delay OA progression in rabbits. However, the cellular mechanisms underlying these protective effects remain unexplored. We examined articular cartilage maintenance and joint bone remodeling using Ctsk null mice (Ctsk-/- ) which underwent destabilization of the medial meniscus (DMM). We found that Ctsk-/- mice displayed delayed remodeling of subchondral and calcified cartilage by osteoclasts and chodroclasts respectively in DMM-induced osteoarthritis. While WT mice displayed a more severe OA phenotype than Ctsk-/- mice at 16 weeks, higher subchondral bone volume and lower trabecular spacing were also observed in surgically-induced OA joints of Ctsk-/- mice. However, no differences were seen in non-surgical controls. During OA progression, TRAP+ osteoclast numbers were increased in both WT and Ctsk-/- mice. However, Ctsk-/- mice had fewer physis-derived chondroclasts than WT when OA was present. These data suggest that CatK may differentially regulate chondroclastogenesis in the growth plate. Targeted PCR arrays of RNA harvested from laser captured osteoclasts in the subchondral bone and chondroclasts in the growth plate demonstrated differential expression of Atp6v0d2, Tnfrsf11a, Ca2, Calcr, Ccr1, Gpr68, Itgb3, Nfatc1, and Syk genes between WT and Ctsk-/- mice at 8- and 16-weeks post-DMM. Our data provide insight into the cellular mechanisms by which cathepsin K deletion delays OA progression in mice.

Role of Cbl-PI3K Interaction during Skeletal Remodeling in a Murine Model of Bone Repair.

Mice in which Cbl is unable to bind PI3K (YF mice) display increased bone volume due to enhanced bone formation and repressed bone resorption during normal bone homeostasis. We investigated the effects of disrupted Cbl-PI3K interaction on fracture healing to determine whether this interaction has an effect on bone repair. Mid-diaphyseal femoral fractures induced in wild type (WT) and YF mice were temporally evaluated via micro-computed tomography scans, biomechanical testing, histological and histomorphometric analyses. Imaging analyses revealed no change in soft callus formation, increased bony callus formation, and delayed callus remodeling in YF mice compared to WT mice. Histomorphometric analyses showed significantly increased osteoblast surface per bone surface and osteoclast numbers in the calluses of YF fractured mice, as well as increased incorporation of dynamic bone labels. Furthermore, using laser capture micro-dissection of the fracture callus we found that cells lacking Cbl-PI3K interaction have higher expression of Osterix, TRAP, and Cathepsin K. We also found increased expression of genes involved in propagating PI3K signaling in cells isolated from the YF fracture callus, suggesting that the lack of Cbl-PI3K interaction perhaps results in enhanced PI3K signaling, leading to increased bone formation, but delayed remodeling in the healing femora.

Regulation of Osteoblast Migration Involving Receptor Activator of Nuclear Factor-kappa B (RANK) Signaling.

Bone remodeling requires osteoclast activation, resorption, and reversal, prior to osteoblast migration into the bone pit. The Receptor Activator of NF-κB (RANK) signaling pathway plays an important role in bone remodeling. Two components of the RANK signaling pathway, RANK Ligand (RANKL) and the decoy receptor Osteoprotegerin (OPG), are expressed predominantly on the surface of osteoblasts, while RANK is principally expressed on the surface of osteoclasts. However, RANK has also been reported to be expressed on the surface of osteoblasts and osteosarcoma tumor cells. Treatment with soluble RANKL (sRANKL) of both normal osteoblasts and osteosarcoma tumor cells activated phosphorylation of ERK, p38(MAPK) , Akt, and p65(NF-κB). However, modified Boyden chamber assays and wound repair assays showed differential response to sRANKL-induced chemotactic migration in normal osteoblasts and osteosarcoma tumor cells. In contrast to previously published results, both normal osteoblasts and osteosarcoma tumor cells responded to sRANKL-induced chemotactic migration but the normal osteoblasts did so only in the presence of an ERK pathway inhibitor. For both normal and tumor cells, the chemotactic response could be blocked by inhibiting the PI3K/Akt or p65(NF-κB) pathway. Response to sRANKL in normal and tumor cells suggests a role for RANK/ERK-mediated signaling in normal osteoblasts chemotactic migration during bone remodeling that is altered or lost during osteosarcoma tumorigenesis.

Somatic mutations in SQSTM1 detected in affected tissues from patients with sporadic Paget's disease of bone.

Paget's disease of bone (PDB) is a focal disorder of bone remodeling that leads to overgrowth of affected bone, with rare progression to osteosarcoma. Extensive studies of familial PDB showed that a majority of cases harbor germline mutations in the Sequestosome1 gene (SQSTM1). In contrast, little is known about the mutational status of SQSTM1 in sporadic PDB. We hypothesized that somatic SQSTM1 mutations might occur in the affected tissues of sporadic PDB and pagetic osteosarcoma. We used laser capture microdissection to capture homogeneous populations of cells from the affected bone or tumor of patients with sporadic PDB or pagetic osteosarcoma, respectively. DNA from these samples and appropriate controls was used for sequence analysis and allelic discrimination analysis. Two of five patients with sporadic PDB had SQSTM1(C1215T) mutations detected in their affected bone but not in their blood samples, indicating a somatic origin of the mutations. Samples from three of five sporadic pagetic osteosarcoma patients had the SQSTM1(C1215T) mutation, whereas the normal adjacent tissue from two of these tumors clearly lacked the mutation, again indicating an occurrence of somatic events. No SQSTM1 mutations were found in primary adolescent osteosarcomas. The discovery of somatic SQSTM1 mutations in sporadic PDB and pagetic osteosarcoma shows a role for SQSTM1 in both sporadic and inherited PDB. The discovery of somatically acquired mutations in both the diseased bone and tumor samples suggests a paradigm shift in our understanding of this disease.

Biomarkers in Osteosarcoma.

Osteosarcoma is the most common primary tumor of bone and accounts for approximately 19% of all malignant tumors of bone. It is the third most common malignant tumor in teenagers. More than twenty years ago, the advent of a multidisciplinary approach that combined multi-agent chemotherapy and limb-sparing surgery greatly improved the survival rate of patients with osteosarcoma. Unfortunately, since that time, survival rates have not dramatically improved. To date, the most powerful predictors of outcome have remained the ability to detect metastatic disease at diagnosis and the histopathologic response of the tumor to preoperative chemotherapy. Presently, 80% of patients who do not have distant metastases at initial diagnosis will become long-term survivors. Unfortunately, this means that approximately 20% of patients who do not present with metastases at diagnosis will not survive. This group of patients appears to be resistant to current treatment as attempts to intensify therapy after surgery for patients with a poor histopathologic response has not significantly improved survival rates. It is these patients that are in the greatest need of additional clinically relevant markers for prognosis and who can be most helped by molecular analysis. While steady progress has been made in the identification of genetic alterations in osteosarcoma, no individual molecular marker has thus far been demonstrated to have a better prognostic significance in the treatment of osteosarcomas than the current clinical markers. Thus there is clearly a need to employ new comprehensive analysis technologies to develop significantly more informative classification systems and to identify new therapeutic targets.

Osteosarcoma in Paget's disease of bone.

Paget's disease of bone (PDB) is a focal disorder of bone metabolism first described by Sir James Paget in 1876. It is presumed benign in nature and mediated by abnormal osteoclast function. The incidence of osteosarcomas complicating PDB is estimated at <1%. These cancers occur mostly in persons with long-standing, polyostotic disease and affect patients in their seventh decade or when osteosarcoma is remarkably rare in the general population. Epidemiological studies suggest that this late peak of osteosarcomas is absent in regions where Paget's is infrequently reported. Whereas PDB has a predilection for the axial skeleton, skull, femurs, and tibias, pagetic osteosarcoma tend to spare the spine, and are reported more commonly in the pelvis, femur, humerus, and skull. A molecular basis for the association of osteosarcoma with Paget's disease is unclear. These osteosarcomas are osteogenic in origin, consistently arise in sites of pagetic bone, and may present as metachronous, multifocal lesions. On histopathology, the lesions are usually osteoblastic, and the tumor phenotype is sometimes characterized as an exaggerated, chaotic form of the accelerated bone remodeling that characterizes PDB. New insights from the biology of adolescent osteosarcomas, VCP and SQSTM1 mutations now defined in patients with Paget's disease, and emerging evidence that stromal lesions are present in patients with Paget's disease are changing the way we think about the pathogenesis of PDB and the rare complication of pagetic osteosarcomas.

Expression of receptor activator of nuclear factor-kappaB is inversely correlated with metastatic phenotype in breast carcinoma.

During normal bone remodeling, the receptor activator of nuclear factor-kappaB (RANK) interacts with its ligand RANKL, which is present on pre-osteoclasts, resulting in bone resorption and initiation of new bone formation. When breast cancer metastasizes to bone, normal bone remodeling is disturbed by invasion of tumor cells, resulting in osteolytic lesions. We have studied the expression of both RANK and RANKL in 10 nonneoplastic breast samples, 58 infiltrating ductal carcinoma (IDC), and 43 breast cancer bony metastases (BTM). RANK seemed to be present in all samples tested. However, whereas RANKL expression was observed in 90% of nonneoplastic breast, RANKL expression was only observed in 62% of nonmetastatic IDC, 31% of metastatic IDC, and 2% of osteolytic BTM lesions. This decreased or absent expression of RANKL in the tumor cells may allow RANK, which is normally expressed as a receptor on the cell surface, to target RANKL present on the cell surface of normal osteoblasts and stromal cells of the bone. Stimulation of the normal osteoblasts and stromal cells by the tumor cells may then lead to secondary osteoclastogenesis, resulting in the osteolytic phenotype common to breast metastases.

Three novel mutations in SQSTM1 identified in familial Paget's disease of bone.

UNLABELLED: Mutations in Sequestosome 1 (SQSTM1) have been shown to segregate with familial Paget's disease of bone (PDB). We examined the coding sequence of SQSTM1 in five PDB pedigrees and found three novel mutations clustered around the C-terminal ubiquitin associated domain. Disruptions of the C-terminal domain of SQSTM1 seem to be a leading cause of familial PDB. INTRODUCTION: The characteristic features of Paget's disease of bone (PDB) are caused by focal areas of excessive and uncoordinated bone remodeling. A total of seven genetic loci (PDB1-PDB7) have been reported to be associated with the disease. The gene for Sequestosome 1 (p62; SQSTM1) has been identified as the causative gene for PDB3 in numerous French-Canadian families and families predominantly of British descent. To date, a total of three mutations, all affecting the ubiquitin-associated domain of SQSTM1, have been identified: a single 1215 C to T (P392L) transversion in exon 8, a T insertion in exon 8 (E396X), and a G to A mutation at the splice junction of exon 7 (IVS7 + 1). MATERIALS AND METHODS: DNA was isolated from blood collected from the members of five U.S. PDB pedigrees. Mutation analysis of the coding sequence of the SQSTM1 gene was performed on the proband and other key individuals in the pedigrees. RESULTS: Four of the five families had SQSTM1 mutations. Three of these mutations were novel: a single base deletion in exon 8 at position 1210 (1210delT) resulting in a premature stop codon at amino acid 394, a single C deletion in exon 8 at position 1215 (1215delC) also resulting in a premature stop codon at amino acid 394, and a single 1200 C to T (P387L) transversion in exon 7. CONCLUSION: Noteworthy is the fact that these three SQSTM1 mutations, in addition to the three previously described mutations, are clustered near the C-terminal of the protein. These mutations may be acting in a dominant-negative fashion to disrupt the ubiquitin-binding function, which could result in abnormal activation of the NF-kappaB pathway and the subsequent activation of the osteoclasts. These findings imply that SQSTM1 mutations may play a role in the majority of familial PDB in the United States.

Analysis of environmental factors in familial versus sporadic Paget's disease of bone--the New England Registry for Paget's Disease of Bone.

UNLABELLED: A registry for Paget's Disease (PD) was created to study the environmental and familial features of this disorder. This study examines the first 202 people enrolled. Those with a family history of PD (20%) suffered earlier deformity and fracture, and tended to have grandparents born abroad. These findings suggest heritage is important in PD; the role of environment remains unclear. INTRODUCTION: The New England Registry for Paget's Disease of Bone is a database that was created to explore the distribution and determinants of disease frequency. METHODS: Using a case-series design, we explored the association of environmental factors in expression of the disease, comparing those patients with familial Paget's disease to those with sporadic Paget's disease (PD). RESULTS AND CONCLUSIONS: Analysis of the first 202 patients enrolled in the registry revealed a positive family history (FH) in 41 (20%). Significant findings in this cohort included an earlier age of onset (51 years FH+ versus 59 years FH-, p < 0.05), a trend to a higher incidence of bone deformity (49% versus 33%, p < 0.1), and an increased fracture rate (27% versus 11%, p < 0.05). Persons with a FH of PD were less likely to record the United States or Canada as their grandparents' birthplace (p < 0.01), and sibships tended to be larger (p < 0.05). A history of measles infection, childhood exposure to pets, milk ingestion, year of immigration to the United States, birth order, level of education, and functional status did not distinguish the two groups. The database confirms the high prevalence of familial PD and supports the theory that heritable factors are important in the pathogenesis of this focal disorder of bone metabolism.

An adaptive strategy for single- and multi-cluster gene assignment.

Strict assignment of genes to one class, dimensionality reduction, a priori specification of the number of classes, the need for a training set, nonunique solution, and complex learning mechanisms are some of the inadequacies of current clustering algorithms. Existing algorithms cluster genes on the basis of high positive correlations between their expression patterns. However, genes with strong negative correlations can also have similar functions and are most likely to have a role in the same pathways. To address some of these issues, we propose the adaptive centroid algorithm (ACA), which employs an analysis of variance (ANOVA)-based performance criterion. The ACA also uses Euclidian distances, the center-of-mass principle for heterogeneously distributed mass elements, and the given data set to give unique solutions. The proposed approach involves three stages. In the first stage a two-way ANOVA of the gene expression matrix is performed. The two factors in the ANOVA are gene expression and experimental condition. The residual mean squared error (MSE) from the ANOVA is used as a performance criterion in the ACA. Finally, correlated clusters are found based on the Pearson correlation coefficients. To validate the proposed approach, a two-way ANOVA is again performed on the discovered clusters. The results from this last step indicate that MSEs of the clusters are significantly lower compared to that of the fibroblast-serum gene expression matrix. The ACA is employed in this study for single- as well as multi-cluster gene assignments.

Determination of a minimal region of loss of heterozygosity on chromosome 18q21.33 in osteosarcoma.

Previous analysis of tumor-specific constitutional LOH had identified a putative tumor-suppressor gene (LOH18CR) active in osteosarcoma tumorigenesis, which mapped to a subregion of chromosome 18q linked to both familial Paget's disease and FEO. Using 9 new polymorphic loci within the previous minimal region of LOH, we have reduced the minimal region of LOH in osteosarcoma tumors to localize the LOH18CR locus to the distal end of chromosome 18q21.33. This new region is approximately 500 kb and contains at least 7 known genes; however, it excludes 2 previous candidate genes: TNFRSF11A (RANK) and BCL2.

Identification of a novel tandem duplication in exon 1 of the TNFRSF11A gene in two unrelated patients with familial expansile osteolysis.

Familial expansile osteolysis (FEO) is a rare autosomal dominant disorder characterized by striking focal expansile osteolytic bone lesions and generalized osteopenia, often accompanied by characteristic early hearing loss and dental disease. The TNFRSF11A gene encodes the receptor activator of nuclear factor-kappaB (RANK), which has been demonstrated to be essential in bone remodeling and osteoclast differentiation. Identical insertional mutations in the first exon of RANK have been identified in all published FEO kindreds. The mutation is an 18 base pair tandem duplication in the sequence coding for the signal peptide of RANK, which causes an increase in NF-kappaB signaling. We report the identification and mutational analysis of two unrelated FEO patients. One had no family history of FEO, but presented with bilateral hearing loss at an early age, deterioration of teeth, and severe pain and swelling in the distal tibia before the age of 20. The second patient had a family history of FEO and exhibited an extensive expansile tibial lesion and lesions in one humerus and a phalanx. She also had early hearing loss and dental disease. Mutational analysis of the TNFRSF11A gene in our patients demonstrated an 18 base pair tandem duplication, one base proximal to the duplications previously reported. This novel mutation results in addition of the same six amino acids to the RANK signal peptide that has been observed previously. Further analysis of the exon 1 sequence demonstrated that it has the ability to form a stable secondary structure that may facilitate the generation of tandem duplications.