Nature Medicine homepage
Advanced search
 Journal home > Archive > Table of Contents > Article > Full text
Journal home
Advance online publication
Current issue
Archive
Press releases
Supplements
Focuses
Guide to authors
Online submissionOnline submission
For referees
Free online issue
Contact the journal
Subscribe
Advertising
work@npg
Reprints and permissions
About this site
For librarians
 
NPG Resources
Nature
Nature Reviews
Nature Immunology
Nature Cell Biology
Nature Genetics
news@nature.com
Nature Conferences
Dissect Medicine
NPG Subject areas
Biotechnology
Cancer
Chemistry
Clinical Medicine
Dentistry
Development
Drug Discovery
Earth Sciences
Evolution & Ecology
Genetics
Immunology
Materials Science
Medical Research
Microbiology
Molecular Cell Biology
Neuroscience
Pharmacology
Physics
Browse all publications
Article
Nature Medicine  8, 943 - 949 (2002)
Published online: 5 August 2002; | doi:10.1038/nm752

SHIP-deficient mice are severely osteoporotic due to increased numbers of hyper-resorptive osteoclasts

Sunao Takeshita1, 6, Noriyuki Namba1, 2, 6, Jenny J. Zhao3, Yebin Jiang3, Harry K. Genant3, Matthew J. Silva4, Michael D. Brodt4, Cheryl D. Helgason5, Janet Kalesnikoff5, Michael J. Rauh5, R. Keith Humphries5, Gerald Krystal5, Steven L. Teitelbaum1 & F. Patrick Ross1

1 Department of Pathology and Immunology, Washington University School of Medicine, St. Louis, Missouri, USA

2 Department of Pediatrics, Okayama University Graduate School of Medicine and Dentistry, Okayama, Japan

3 Osteoporosis and Arthritis Research Group, Department of Radiology, University of California, San Francisco, USA

4 Department of Orthopedic Surgery, Washington University School of Medicine, St. Louis, Missouri, USA

5 Terry Fox Laboratory, BC Cancer Agency, Vancouver, Canada

6 S.T. and N.N. contributed equally to this study.

Correspondence should be addressed to F. Patrick Ross rossf@medicine.wustl.edu
The hematopoietic-restricted protein Src homology 2−containing inositol-5-phosphatase (SHIP) blunts phosphatidylinositol-3-kinase-initiated signaling by dephosphorylating its major substrate, phosphatidylinositol-3,4,5-trisphosphate. As SHIP-/- mice contain increased numbers of osteoclast precursors, that is, macrophages, we examined bones from these animals and found that osteoclast number is increased two-fold. This increased number is due to the prolonged life span of these cells and to hypersensitivity of precursors to macrophage colony-stimulating factor (M-CSF) and receptor activator of nuclear factor-kappaB ligand (RANKL). Similar to pagetic osteoclasts, SHIP-/- osteoclasts are enlarged, containing upwards of 100 nuclei, and exhibit enhanced resorptive activity. Moreover, as in Paget disease, serum levels of interleukin-6 are markedly increased in SHIP-/- mice. Consistent with accelerated resorptive activity, 3D trabecular volume fraction, trabecular thickness, number and connectivity density of SHIP-/- long bones are reduced, resulting in a 22% loss of bone-mineral density and a 49% decrease in fracture energy. Thus, SHIP negatively regulates osteoclast formation and function and the absence of this enzyme results in severe osteoporosis.
Src homology (SH) 2-containing inositol-5-phosphatase (SHIP) is a protein that becomes tyrosine phosphorylated and associated with Shc (ref. 1) following activation of cell-surface receptors with a range of hematopoietic cytokines2, 3. SHIP has an N-terminal SH2 domain, a central inositol polyphosphate-5-phosphatase catalytic domain, two NPXY sequences that bind phosphotyrosine binding domains when phosphorylated and a proline-rich C terminus2, 3. SHIP specifically recognizes and cleaves the 5'-phosphate group from phosphatidylinositol-3,4,5-trisphosphate (PIP3), the major product of phosphatidylinositol 3-kinase (PI3-K) activity4 and, at least in vitro, from inositol-1,3,4,5-tetrakisphosphate5.

Osteoclasts (OCs) form by fusion of mononuclear progenitors of the monocyte-macrophage lineage6, 7. These multinucleated polykaryons are the principal, if not exclusive, resorptive cells of bone, playing a central role in the formation of the skeleton and regulation of its mass8. Thus, the rate of bone degradation is the product of OC number and the bone-resorptive capacity of the mature OCs.

Two hematopoietic cytokines, macrophage-colony stimulating factor (M-CSF) and receptor activator of nuclear factor-kappaB ligand (RANKL), are necessary and sufficient for differentiation of macrophage precursors into mature OCs. As shown by the osteopetrotic op/op mouse, which lacks M-CSF and is deficient in OCs and macrophages, this cytokine is indispensable for OC precursor proliferation and survival9. RANKL (also known as tumor necrosis factor-related activation induced cytokine10, OC differentiation factor11 and osteoprotegerin ligand12) was initially implicated in the immune response13 and later established as the master cytokine promoting commitment to the OC phenotype. Although macrophage number is normal, RANKL-deficient mice are osteopetrotic due to the complete absence of OCs (ref. 14).

The intracellular signals promoting OC precursor survival and differentiation are emerging and include activation of PI3-K, prompted by either M-CSF or RANKL. Thus, the ability of SHIP to dephosphorylate PIP3 suggests this 5'-phosphatase may also inhibit OC recruitment and function, a hypothesis we tested by examining mice with a homozygous deletion of SHIP. The animals are viable and fertile, but typically fail to thrive past 14 weeks due to a myeloproliferative disorder15. This observation is consistent with SHIP being a negative regulator of cytokine signaling. In this regard, we find that absence of SHIP increases the sensitivity of marrow macrophages to M-CSF and RANKL, resulting in enhanced generation of large, hyper-resorptive OCs. Severe osteoporosis is a consequence of the abundance and hyperactivity of SHIP-/- OCs.

SHIP-/- mice contain numerous enlarged hypernucleated OCs
The tibias of SHIP-/- and wild-type mice, aged 40.9 plusminus 7.8 and 41.3 plusminus 7.0 days, respectively, were decalcified, and histological sections stained for the OC enzyme tartrate-resistant acid phosphatase (TRAP). The bones of mice lacking the 5'-phosphatase exhibited an approximate 2-fold increase in OC number (Fig. 1a and c), establishing that SHIP is a negative regulator of osteoclastogenesis. More striking, however, SHIP-/- OCs were enlarged 2.6 fold (Fig. 1c). Moreover, many mutant OCs contained large numbers of nuclei (Fig. 1b), an appearance mirroring those of patients with Paget disease16.

Figure 1. Absence of SHIP results in increased OCs.
Figure 1 thumbnail

Histological analysis of 6 week-old wild-type and SHIP-/- mice bones. a and b, OCs were identified by TRAP activity reaction product (red). a, Increased number of OCs in SHIP-/- mice (right) compared with wild-type controls (left). Magnification, times40 (upper) and times160 (lower). b, Pagetic, hypernucleated OCs were observed in three separate SHIP-/- mice but not in wild-type (upper left). Magnification, times400. c, Histomorphometric analysis reveals a 2-fold increase in OC number in SHIP1-/- mice () compared with wild-type (). OC size was also increased in SHIP-/- mice. Data are expressed as the mean plusminus s.d. of 25 fields (5 OCs times 5 different animals). *, P < 0.001.



Full FigureFull Figure and legend (93K)
SHIP-/- OC precursors are hypersensitive to M-CSF and RANKL
Because SHIP is a negative regulator of many hematopoietic cytokines, the increased OC numbers in mice lacking this phosphatase may reflect derepression of OC precursor differentiation. To study this, we generated mutant and wild-type OCs, culturing M-CSF-dependent bone-marrow macrophages (BMDMs) in the presence of M-CSF and RANKL. In some experiments, the concentration of M-CSF progressively increased while that of RANKL remained constant. In other cases, RANKL, but not M-CSF was varied. In both circumstances, the level of the invariant cytokine was below that used in our standard osteoclastogenic assays. After six days, TRAP- expressing multinucleated OCs were identified (Fig. 2). At all concentrations of RANKL, osteoclastogenesis was more extensive in SHIP-/- versus wild-type precursors. Although the mutant OC precursors were also more sensitive to M-CSF, this increased sensitivity was most evident at low concentrations of the cytokine. In contrast, differences in OC number disappeared with high M-CSF levels. Thus, SHIP deficiency in OC precursors confers on these cells hypersensitivity to the critical osteoclastogenic cytokines, M-CSF and RANKL.

Figure 2. Osteoclastogenesis by M-CSF and RANKL is increased in SHIP-/-mice.
Figure 2 thumbnail

a and b, MDBMs from wild-type (upper panels) and SHIP-/- (lower panels) mice were cultured in the presence of the indicated concentrations of M-CSF and 12.5 ng/ml of RANKL (a) or in the presence of the indicated concentrations of RANKL and 2.5 ng/ml of M-CSF (b). After 6 d, TRAP staining was performed. Multinucleated OCs stained red are observed at lower concentration of M-CSF (a) and RANKL (b) in SHIP-/-, as compared with wild-type culture.



Full FigureFull Figure and legend (45K)
SHIP-/- OCs are less apoptotic
Having established that the increased number of OCs in SHIP-/- mice reflects, at least in part, stimulated precursor differentiation, we investigated whether prolonged lifespan of the differentiated OCs may also be a contributing factor. Thus, equal numbers of M-CSF-dependent bone-marrow macrophages (MDBMs), derived from SHIP-/- and wild-type mice, were placed in osteoclastogenic conditions containing the same concentrations of M-CSF and RANKL. As expected, osteoclastogenesis in this circumstance was more robust in the cultures containing null precursors (Fig. 3a). Moreover, reflecting the in vivo situation, generated SHIP-/- OCs were considerably larger than their wild-type counterparts. Numerous apoptotic polykaryons were present in day 6 wild-type but not SHIP-/- cultures, further indicating that absence of the phosphatase prolongs OC longevity. To confirm this morphological observation, cell lysates were examined for the presence of histone-associated DNA fragments, a marker of apoptosis. The results demonstrated a greater than three-fold reduction in the death of mutant OCs (Fig. 3b). Thus, the greater number of OCs present in SHIP-/- mice reflects a combination of increased osteoclastogenesis and prolonged OC life span.

Figure 3. SHIP deficiency results in decreased OC apoptosis.
Figure 3 thumbnail

a, MDBMs from wild-type (left panels) and SHIP-/- (right panels) mice were cultured in the presence of M-CSF (10 ng/ml) and RANKL (100 ng/ml). The cells were stained with TRAP on days 3 (top), 4 (middle) and 6 (bottom). Equal numbers of multinucleated OCs (red) are present on day 4. Most OCs in the wild-type culture on day 6, are either apoptotic or dead, as shown with an asterisk. b, In contrast, SHIP-/- OCs () are still all viable versus wild-type (). Cell death on day 6 was detected using a histones/DNA ELISA kit. Data in b are mean plusminus s.d. of 3 samples. *, P < 0.001.



Full FigureFull Figure and legend (90K)
SHIP-/- OCs are hyperresorptive
The OC is the pathogenetic cell in Paget disease and the initial phase of this disorder is characterized by accelerated bone resorption17 and elevated serum levels of interleukin-6 (IL-6) (ref. 18). Given their morphological similarities, we first asked if SHIP-/- OCs, like those of Paget disease, have an increased capacity to resorb mineralized matrix. To this end, we generated TRAP-expressing mononuclear preosteoclasts (pOCs) by culturing wild-type and SHIP-/- MDBMs with RANKL and M-CSF for 2 days on plastic plates, at which time the cells were lifted and placed on whale dentin slices. To assure comparison of the resorptive activity of relatively equal numbers of mutant and wild-type mature OCs, we compensated for the enhanced efficiency of osteoclastogenesis in the absence of SHIP by initially plating fewer mutant pOCs, an approach we validated in preliminary experiments (data not shown). SHIP-/- OCs resorbed more dentin surface and generated deeper resorptive pits than do their wild-type counterparts (Fig. 4a and b). Thus, SHIP blunts the resorptive activity of individual, mature OCs.

Figure 4. SHIP suppresses osteoclastic bone resorption and serum IL-6 levels.
Figure 4 thumbnail

a and b, Wild-type and SHIP-/- pOCs were prepared by culturing MDBMs with RANKL and M-CSF for 2 d, followed by incubation on whale dentin slices for 2 d. After removing cells, resorption pits were stained with Coomassie brilliant blue (a) (Left, wild type; right, SHIP-/-) and resorption pit area was measured (b) (, wild type; , SHIP-/-). Pit depth was determined by confocal microscopy (b). Data are expressed as the mean plusminus s.d. of 10 pits. *, P < 0.001. c, SHIP deficiency results in increased serum IL-6 levels. Serum was obtained by cardiac puncture and IL-6 levels were determined by ELISA. The results for 4 wild-type () and SHIP-/- () mice are shown as pg/mL.



Full FigureFull Figure and legend (70K)
Abundant data show that, IL-6 has important effects on the regulation of the normal OC formation and function19. IL-6 has been implicated as a potentially important mediator of OC function in Paget disease18. Therefore, we next measured serum IL-6 levels in wild-type and SHIP-/- mice. There was a marked increase of serum IL-6 levels in SHIP-/- mice (average 58.8 pg/ml versus undetectable levels in wild-type) (Fig. 4c).

SHIP does not modulate bone formation
Because suppressed osteoblast activity could also contribute to the osteoporosis of SHIP-/- mice, mutant and wild-type animals were administered two doses of the fluorescent bone formation marker, calcein, separated by a 13-day interval. The mineral apposition and bone formation rates of both groups, assessed in non-decalcified histological sections of calvariae, were not significantly different (data not shown). Therefore, the osteoporosis of SHIP-/- mice reflects an accelerated rate of bone resorption in the face of unaltered formation.

SHIP-/- mice are osteoporotic
To determine if the increased number of OCs present in SHIP-/- mice is functionally relevant, trabecular femoral architectural and metaphyseal morphology and biomechanical parameters were examined. Femurs of SHIP-/- mice were slightly (10%) but significantly shorter than normal. On the other hand, all structural parameters of trabecular mass were significantly compromised in the mutant animals. Thus, wild-type metaphysis extended approximately twice as far into the marrow cavity as that from mutant mice (Fig. 5a). Furthermore, the volume of trabecular bone per unit of metaphysis was decreased by more than one third (Fig. 5b). Similarly, trabecular number (Th.N), trabecular thickness (Tb.Th), trabecular separation (Tb.Sp), structure model index (SMI), degree of anisotropy (DA) and connectivity density (CD) were all deficient in SHIP-/- versus wild-type mice. Similarly, cortical thickness was reduced in the mutants (data not shown). Consistent with the decrease in cortical thickness, cross-sectional area and moment of inertia are reduced by 31% and 38%, respectively (data not shown).

Figure 5. SHIP-/- mice are osteoporotic.
Figure 5 thumbnail

a, Representative images of 3D muCT reconstruction of wild-type (left) and SHIP-/- mouse (right) femur. Samples were examined with an isotropic resolution of 5-mum voxel size. Cortical thickness is decreased in SHIP-/- mice and trabecular bone extension into the marrow cavity is reduced by approx50%. Morphometric analysis revealed decreased 3D trabecular bone volume fraction and trabecular number and thinner, more widely-spaced and less connected trabeculae. Moreover, the structural pattern was less plate-like and more rod-like and the cortex was thinner, resulting in lowered resistance to fracture. b, All 3D trabecular structural parameters in the secondary spongiosa of the distal femur, including bone volume fraction (BV/TV), trabecular number (Tb.N), trabecular thickness (Tb.Th), trabecular separation (Tb.Sp), structural model index (SMI) for plate- and rod-like structure, degree of anisotropy (DA), and trabecular connectivity density (CD), were measured from muCT images with isotropic resolution of 5 mum3. They are all statistically significant (*, P < 0.01) between wild-type () and SHIP-/- () using the Wilcoxon test.The specimens were measured blindly. Data are expressed as mean plusminus s.d.



Full FigureFull Figure and legend (60K)
Reflecting these morphological features, areal bone-mineral density (BMD) of SHIP-/- long bones is diminished (Fig. 6). Finally, all biomechanical measurements, including rigidity, ultimate moment and fracture energy, revealed that femoral diaphyseal bone strength of SHIP-/- animals was decreased (Fig. 6). Ultimate and post-yield displacements were not different between groups, indicating that SHIP-/- bones were not abnormally brittle (data not shown). Estimated material properties indicated no differences in Young's modulus (data not shown) and only a 14% decrease in ultimate stress in SHIP-/- bones (Fig. 6). Thus, the stiffness and strength of SHIP-/- bones are significantly reduced due primarily to their reduced cross-sectional size rather than to a material defect.

Figure 6. SHIP-/- bones have reduced areal BMD and whole-bone strength.
Figure 6 thumbnail

Areal BMD measured by DEXA of combined femur and tibia revealed the parameter was decreased by 22% in SHIP-/- mice () versus controls (). Femoral whole-bone mechanical properties were determined using 3-point bending. Rigidity, ultimate moment and fracture energy were dramatically less in SHIP-/- mice, consistent with their decrease in cross-sectional size. Ultimate stress was not significantly reduced in SHIP-/- long bones (P = 0.017). All values are mean plusminus s.d.; *, P < 0.001.



Full FigureFull Figure and legend (23K)
Discussion
Important roles for different members of the phosphoinositide family have been revealed by recent progress in dissecting the role of phospholipids in vesicle trafficking, cytoskeletal organization, cell survival and proliferation, gene transcription and modulation of metabolism. This specificity results from individual phospholipids binding unique protein domains, each of which have one or more distinct functions. Five such domains, designated PH, PX, FYVE, ENTH and TUBBY, engage different phosphoinositides, resulting in their targeting to specific regions of the cell20, 21, 22. Thus, the level of a given PIP can regulate the activity of different subsets of proteins bearing these lipid-binding domains and hence cell function. In turn, the amount of a given phospholipid is determined by the net activity of lipid kinases and phosphatases.

As PIP3 is a major modulator of cell function, enzymes abrogating the signaling capacity of this molecule are of considerable interest. Chief among these are PTEN (phosphatase and tensin homolog deleted on chromosome 10, a tumor suppressor gene that removes the 3' phosphate residue from PIP3)23, the myotubularin family, which also act as 3'-phosphatases24, and two 5'-phosphatases, SHIP and SHIP2. SHIP2 is widely expressed and functions primarily to regulate sensitivity to insulin-mediated signals25. In contrast, SHIP is expressed preferentially by hematopoietic cells, including T- and B- lymphocytes, mast cells and macrophages2. This protein undergoes tyrosine phosphorylation following activation of receptors for various cytokines and growth factors, including M-CSF (refs. 2,3,26, 27, 28, 29, 30).

OCs are ontogenetically related to myeloid cells and as such are also regulated by M CSF. In fact, macrophage precursors require only M-CSF and RANKL to assume the OC phenotype. Given the hyper-proliferation of myeloid cells and their enhanced sensitivity to M-CSF in SHIP-/- mice, we reasoned that mice lacking this phosphatase may also be rich in OCs, which proved to be the case.

Osteoclastogenesis is a multi-step event involving proliferation and survival of early macrophages and their differentiation into mature bone-resorbing polykaryons. Although M-CSF is essential for proliferation and survival of OC precursors, RANKL is the unique OC-differentiating molecule. In fact, the mouse macrophage cell line RAW264.7, which survives without M-CSF, differentiates into polykaryons essentially indistinguishable from actual OCs, when exposed only to RANKL.

Given the complementary, yet distinct properties of M-CSF and RANKL, we assessed the role of each cytokine in the enhanced osteoclastogenesis of SHIP insufficiency. These experiments involved exposing OC precursors, in the form of MDBMs, to increasing concentrations of one cytokine in the face of a constant, relatively low amount of the other. In both circumstances, cells from mice lacking SHIP exhibit increased sensitivity to the varied cytokine, indicating that the lipid phosphatase blunts the osteoclastogenic signals transduced by both c-fms and RANK. Although the molecular mechanisms mediating the responses to the two cytokines are not understood, high-dose M-CSF normalizes other circumstances of arrested osteoclastogenesis, such as that attending deletion of the beta3 integrin in vitro (data not shown).

Modulation of OC longevity, which is enhanced by RANKL and M-CSF, regulates bone resorption, both physiologically and in disorders such as glucocorticoid-induced osteoporosis. We found SHIP-/- OCs are resistant to apoptosis. Although the mechanism by which the lifespan of the mutant cell is prolonged is unknown, it may be related to the role of SHIP as a 5'-phospholipid phosphatase. Specifically, RANKL and M-CSF are each activators of PI3-K, which in turn generates PIP3, an inducer of the anti-apoptotic molecule Akt. Moreover, the attenuation of PIP3 degradation in SHIP deficiency prolongs Akt activation in mast cells, thereby extending mast-cell survival in these mice31.

Increased numbers of OCs may reflect enhanced precursor proliferation, more robust differentiation into mature polykaryons and/or reduced apoptosis. The abundance of OCs in SHIP-/- mice seems to reflect all three events. Compared with their wild-type counterparts, macrophage proliferation is more robust in these mutants in the presence of M-CSF (ref. 15). Moreover, their differentiation into OCs is enhanced by M-CSF, which promotes osteoclastogenesis by SHIP-/- precursors at concentrations of RANKL that fail to do so in wild-type cells. This observation suggests that M-CSF may stimulate not only signals mediating proliferation and/or survival, but in specific circumstances—as is the case for TNF-alpha (ref. 32)—has a permissive and synergistic role in RANKL-induced OC formation.

Increased OC number, size, longevity and, therefore, enhanced resorptive activity, coupled with a failure to increase bone formation, result in loss of bone quantity, quality, density and strength in the mutant mice. Loss of bone structure, mineralization and geometry can explain the decrease in structural biomechanical properties of the tissue, as bone quality or mechanical competence in terms of resistance to fracture is a function of its apparent density and 3D distribution33, 34.

Although OC number in SHIP-/- mice is substantially increased, their appearance is most notable. The mutant polykaryons are enlarged and their cytoplasm essentially filled with nuclei, which, unlike wild-type cells, fail to polarize to the cells' anti-resorptive membrane. Thus, SHIP-/- OCs are morphologically indistinguishable from those of patients with Paget disease and, like these human cells, are exuberant bone resorbers. These similarities may reflect the fact that, like pagetic OC precursors35, those from SHIP-null mice are hyper-responsive to RANKL, resulting in enhanced OC differentiation. Furthermore, activation of PI3-K, a RANKL-mediated event, is essential for organization of the OC cytoskeleton36 and hence the cell's bone degrading activity.

Levels of IL-6 are elevated in the serum of pagetic patients18, and the increased levels of this osteoclastogenic cytokine may be responsible, at least in part, for the pagetic phenotype. Given this hypothesis, we measured serum IL-6 levels in SHIP-/- mice and found a marked increase in this parameter. We also measured IL-6 levels in the medium from cultured resident peritoneal macrophages and found substantially higher levels with macrophages from SHIP-/- mice (data not shown). These observations are consistent with our recent finding that SHIP suppresses IgE-induced IL-6 production by mast cells37.

Although SHIP is an established negative regulator of hematopoietic cell proliferation and survival, this is the first report that it dampens both maturation and function of a particular cell, and does so in the context of a specific differentiation molecule, RANKL. Thus, SHIP may be a 'gatekeeper' of OC number and function via its ultimate impact on the signaling consequences of the two essential cytokines, M-CSF and RANKL. The lack of this 'gatekeeper' activity results in enhanced numbers of super-resorptive OCs.

The genetic defect in a subset of familial Paget disease involves point mutations in RANK, the receptor for RANKL. Mutated RANK is auto-activated, resulting in enhanced signaling and hence increased osteoclastogenesis and function38. The chromosomal locus for SHIP has been established as 2q36 (ref. 39), which is also that of a subset of familial pagetic kindreds40. Given this cosegregation of the SHIP gene and a site of loss of heterozygosity (LOH) for Paget disease, plus the fact we demonstrate here that SHIP is downstream of RANK with respect to OC differentiation and function, the possibility exists that SHIP is mechanistically important in some patients suffering from familial Paget disease. Clearly, analysis of the putative causative gene will be necessary in such patients.

 Top
Methods
Animals.
The generation of SHIP-/- mice has been described15. All mice used in these experiments were 4−10-wk-old, bred and maintained with sterilized food, water and bedding at the Animal Facility of the Washington University School of Medicine. All experiments were approved by the Animal Ethics Committee of Washington University

Assessment of 3D trabecular microstructure.
The distal femurs of 5 wild-type and 6 SHIP-/- mice were scanned using a microcomputed tomography scanner (muCT 40, Scanco, Switzerland), with matrix size 2,048 times 2,048 and isotropic resolution of 5 mum3. 3D trabecular structural parameters in the secondary spongiosa were directly measured, which has advantages over static bone histomorphometry in that it does not make the stereological model assumption of an underlying fixed plate or rod model of trabecular structure41. The specimens were blindly measured, without knowledge of the group code. Comparisons between groups were performed using the Wilcoxon test.

Morphological and biomechanical assessment of femora from SHIP-/- mice.
Whole-bone mechanical properties of left femurs and cross-sectional morphology of contra-lateral right femurs were assessed as described42. Right femurs were embedded in plastic and sectioned transversely at their midpoint. Cross-sectional images were obtained to determine bone area and bone moment of inertia about the medial-lateral axis. Left femurs were loaded to failure in three-point bending, with the loading point at midpoint, with support points spaced at a distance of 6 mm. Force and displacement data were recorded and converted to moment and normalized displacement values. We determined the following whole-bone structural properties: rigidity (a measure of resistance to bending), ultimate moment (a measure of strength), ultimate and post-yield displacements (measures of brittleness) and fracture energy (a measure of resistance to fracture). We estimated two material properties, Young's modulus (a measure of resistance to elongation) and ultimate stress (a measure of strength) using standard equations. Differences between groups were assessed using unpaired t-tests (Statview 5.0, SAS, Cary, North Carolina).

Bone mass measurements by dual energy x-ray absorptiometry.
Bone-mineral density (BMD) was measured using dual-energy X-ray absorptiometry (DEXA) on femurs and tibias and analyzed by a Lunar MDL model (LUNAR Corporation, Madison, Wisconsin).

Histology and histomorphometric analyses.
Mouse femurs and tibias were excised, cleaned of soft tissue and decalcified in EDTA. Histological sections were stained with H&E or for tartrate-resistant acid phosphatase (TRAP) activity using a commercial kit (Sigma). The double-labeling technique has been described43: calcein (25 mg/kg; Sigma) was injected twice, with a 13-day interval, and animals were killed 2 days later. Static and dynamic histomorphometric analyses were performed according to standard protocol44 using the Osteomeasure Analysis System (OsteoMetrics, Incorporated, Decatur, Georgia). Statistical differences between groups (n = 11−13) were assessed by Students' t-test.

OC generation.
MDBMs were prepared from whole bone marrow of 6- to 10-wk-old mice and cultured as described45. In some experiments, MDBMs were maintained with RANKL (100 ng/mL) and M-CSF (10 ng/mL), which induces these cells to undergo OC differentiation45. Glutathione-S-transferase (GST)-RANKL was expressed and purified as described32. Recombinant mouse M-CSF was purchased from R&D Biosystems Incorporated (Minneapolis, Minnesota). OCs were identified by TRAP staining and apoptotic cells were assessed using the cell death detection ELISAPLUS kit (Roche Molecular Biochemicals, Mannheim, Germany).

Resorption pit assay.
TRAP-positive mononuclear pOCs were generated by culturing MDBMs with RANKL and M-CSF for 2 days and then were cultured on whale dentin slices. Dentin slices were harvested at day 2. Cells were removed from the dentin slices with 0.25 M ammonium hydroxide and mechanical agitation. Maximum resorption lacunae depth was measured using a confocal microscope (Microradiance; Bio-Rad Laboratories Incorporated, Hercules, California). For evaluation of resorption pit area, dentin slices were stained with Coomassie brilliant blue (CBB) and the stained resorption area was analyzed with light microscopy and Osteomeasure software (OsteoMetrics).

 Top
Received 24 April 2002; Accepted 9 July 2002; Published online: 5 August 2002.

REFERENCES
  1. Liu, L. et al. The Src homology 2 (SH2) domain of SH2-containing inositol phosphatase (SHIP) is essential for tyrosine phosphorylation of SHIP, its association with Shc, and its induction of apoptosis. J. Biol. Chem. 272, 8983−8998 (1997). | Article | PubMed  | ISI | ChemPort |
  2. Krystal, G. Lipid phosphatases in the immune system. Semin. Immunol. 12, 397−403 (2000). | Article | PubMed  | ISI | ChemPort |
  3. Rohrschneider, L.R., Fuller, J.F., Wolf, I., Liu, Y. & Lucas, D.M. Structure, function, and biology of SHIP proteins. Genes Dev. 14, 505−520 (2000). | PubMed  | ISI | ChemPort |
  4. Huber, M. et al. The src homology 2-containing inositol phosphatase (SHIP) is the gatekeeper of mast cell degranulation. Proc. Natl. Acad. Sci. USA 95, 11330−11335 (1998). | Article | PubMed  | ChemPort |
  5. Damen, J.E. et al. The 145-kDa protein induced to associate with Shc by multiple cytokines is an inositol tetraphosphate and phosphatidylinositol 3,4,5-triphosphate 5-phosphatase. Proc. Natl. Acad. Sci. USA 93, 1689−1693 (1996). | Article | PubMed  | ChemPort |
  6. Ibbotson, K.J., Roodman, G.D., McManus, L.M. & Mundy, G.R. Identification and characterization of osteoclast-like cells and their progenitors in cultures of feline marrow mononuclear cells. J. Cell Biol. 99, 471−480 (1984). | Article | PubMed  | ISI | ChemPort |
  7. Tondravi, M.M. et al. Targeted deletion of PU.1 induces osteopetrosis rescuable by bone marrow transplantation. Nature 386, 81−84 (1997). | Article | PubMed  | ISI | ChemPort |
  8. Teitelbaum, S.L. Bone resorption by osteoclasts. Science 289, 1504−1508 (2000). | Article | PubMed  | ISI | ChemPort |
  9. Kodama, H., Nose, M., Niida, S. & Yamasaki, A. Essential role of macrophage colony-stimulating factor in the osteoclast differentiation supported by stromal cells. J. Exp. Med. 173, 1291−1294 (1991). | Article | PubMed  | ISI | ChemPort |
  10. Wong, B.R. et al. TRANCE is a novel ligand of the tumor necrosis factor receptor family that activates c-Jun N-terminal kinase in T cells. J. Biol. Chem. 272, 25190−25194 (1997). | Article | PubMed  | ISI | ChemPort |
  11. Yasuda, H. et al. Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/RANKL. Proc. Natl. Acad. Sci. USA 95, 3597−3602 (1998). | Article | PubMed  | ChemPort |
  12. Lacey, D.L. et al. Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 93, 165−176 (1998). | Article | PubMed  | ISI | ChemPort |
  13. Anderson, D.M. et al. A homologue of the TNF receptor and its ligand enhance T-cell growth and dendritic-cell function. Nature 390, 175−179 (1997). | Article | PubMed  | ISI | ChemPort |
  14. Kong, Y.Y. et al. OPGL is a key regulator of osteoclastogenesis, lymphocyte development and lymph-node organogenesis. Nature 397, 315−323 (1999). | Article | PubMed  | ISI | ChemPort |
  15. Helgason, C.D. et al. Targeted disruption of SHIP leads to hemopoietic perturbations, lung pathology, and a shortened life span. Genes Dev. 12, 1610−1620 (1998). | PubMed  | ISI | ChemPort |
  16. Teitelbaum, S.L. Metabolic and other non-tumorous disorders of bone. in Pathology, 7th edn (eds. Anderson, W.A.D. & Kissane, J.M.) 1905−1977 (C.V. Mosby Company, St. Louis, 1977).
  17. Hosking, D.J. Paget's disease of bone. Br. Med. J (Clin. Res. Ed.) 283, 686−688 (1981). | PubMed  | ChemPort |
  18. Roodman, G.D. Paget's disease and osteoclast biology. Bone 19, 209−212 (1996). | Article | PubMed  | ISI | ChemPort |
  19. Horowitz, M.C. & Lorenzo, J.A. Local regulators of bone: IL-1, TNF, lymphotoxin, interferon-gamma, IL-8, IL-10, IL-4, the LIF/IL-6 family and additional cytokines. in Principles of Bone Biology, 2nd edn (eds. Bilezikian, J.P., Raisz, L.G. & Rodan, G.A.) 961−977 (Academic Press, San Diego, 2002). | ChemPort |
  20. Cantley, L.C. Translocating tubby. Science 292, 2019−2021 (2001). | Article | PubMed  | ISI | ChemPort |
  21. Hurley, J.H. & Meyer, T. Subcellular targeting by membrane lipids. Curr. Opin. Cell Biol. 13, 146−152 (2001). | Article | PubMed  | ISI | ChemPort |
  22. Xu, Y., Seet, L.F., Hanson, B. & Hong, W. The Phox homology (PX) domain, a new player in phosphoinositide signalling. Biochem. J. 360, 513−530 (2001). | Article | PubMed  | ISI | ChemPort |
  23. Simpson, L. & Parsons, R. PTEN: Life as a tumor suppressor. Exp. Cell Res. 264, 29−41 (2001). | Article | PubMed  | ISI | ChemPort |
  24. Nandurkar, H.H. et al. Characterization of an adapter subunit to a phosphatidylinositol (3)P 3-phosphatase: Identification of a myotubularin-related protein lacking catalytic activity. Proc. Natl. Acad. Sci. USA 98, 9499−9504 (2001). | Article | PubMed  | ChemPort |
  25. Clement, S. et al. The lipid phosphatase SHIP2 controls insulin sensitivity. Nature 409, 92−97 (2001). | Article | PubMed  | ISI | ChemPort |
  26. Kavanaugh, W.M. & Williams, L.T. An alternative to SH2 domains for binding tyrosine-phosphorylated proteins. Science 266, 1862−1865 (1994). | PubMed  | ISI | ChemPort |
  27. Liu, L., Damen, J.E., Cutler, R.L. & Krystal, G. Multiple cytokines stimulate the binding of a common 145-kilodalton protein to Shc at the Grb2 recognition site of Shc. Mol. Cell. Biol. 14, 6926−6935 (1994). | PubMed  | ChemPort |
  28. Matsuguchi, T. et al. Shc phosphorylation in myeloid cells is regulated by granulocyte macrophage colony-stimulating factor, interleukin-3, and steel factor and is constitutively increased by p210BCR/ABL. J. Biol. Chem. 269, 5016−5021 (1994). | PubMed  | ISI | ChemPort |
  29. Saxton, T.M., van Oostveen, I., Bowtell, D., Aebersold, R. & Gold, M.R. B cell antigen receptor cross-linking induces phosphorylation of the p21ras oncoprotein activators SHC and mSOS1 as well as assembly of complexes containing SHC, GRB-2, mSOS1, and a 145-kDa tyrosine-phosphorylated protein. J. Immunol. 153, 623−636 (1994). | PubMed  | ISI | ChemPort |
  30. Crowley, M.T., Harmer, S.L. & DeFranco, A.L. Activation-induced association of a 145-kDa tyrosine-phosphorylated protein with Shc and Syk in B lymphocytes and macrophages. J. Biol. Chem. 271, 1145−1152 (1996). | Article | PubMed  | ISI | ChemPort |
  31. Liu, Q. et al. SHIP is a negative regulator of growth factor receptor-mediated PKB/Akt activation and myeloid cell survival. Genes Dev. 13, 786−791 (1999). | PubMed  | ISI | ChemPort |
  32. Lam, J. et al. TNF-alpha induces osteoclastogenesis by direct stimulation of macrophages exposed to permissive levels of RANK ligand. J. Clin. Invest. 106, 1481−1488 (2000). | PubMed  | ISI | ChemPort |
  33. Jiang, Y. et al. Trabecular bone mineral and calculated structure of human bone specimens scanned by peripheral quantitative computed tomography: Relation to biomechanical properties. J. Bone Miner. Res. 13, 1783−1790 (1998). | PubMed  | ISI | ChemPort |
  34. Jiang, Y., Zhao, J., Rosen, C., Geusens, P. & Genant, H.K. Perspectives on bone mechanical properties and adaptive response to mechanical challenge. J. Clin. Densitometry 2, 423−433 (1999). | ISI | ChemPort |
  35. Neale, S.D., Smith, R., Wass, J.A. & Athanasou, N.A. Osteoclast differentiation from circulating mononuclear precursors in Paget's disease is hypersensitive to 1,25-dihydroxyvitamin D(3) and RANKL. Bone 27, 409−416 (2000). | Article | PubMed  | ISI | ChemPort |
  36. Nakamura, I. et al. Wortmannin, a specific inhibitor of phosphatidylinositol-3 kinase, blocks osteoclastic bone resorption. FEBS Letters 361, 79−84 (1995). | Article | PubMed  | ISI | ChemPort |
  37. Kalesnikoff, J. et al. SHIP negatively regulates IgE + antigen-induced IL-6 production in mast cells by inhibiting NF-kappaB activity. J. Immunol. 168, 4737−4746 (2002). | PubMed  | ISI | ChemPort |
  38. Hughes, A.E. et al. Mutations in TNFRSF11A, affecting the signal peptide of RANK, cause familial expansile osteolysis. Nature Genet. 24, 45−48 (2000). | Article | PubMed  | ISI | ChemPort |
  39. Ware, M.D. et al. Cloning and characterization of human SHIP, the 145-kD inositol 5- phosphatase that associates with SHC after cytokine stimulation. Blood 88, 2833−2840 (1996). | PubMed  | ISI | ChemPort |
  40. Hocking, L.J. et al. Genomewide search in familial Paget disease of bone shows evidence of genetic heterogeneity with candidate loci on chromosomes 2q36, 10p13, and 5q35. Am. J. Hum. Genet. 69, 1055−1061 (2001). | Article | PubMed  | ISI | ChemPort |
  41. Jiang, Y., Zhao, J. & Genant, H.K. Macro and Micro Imaging of Bone Architecture. in Principles of Bone Biology, 2nd edition (eds. Bilezikian, J.P., Raisz, L.G. & Rodan, G.A.) 1599−1623 (Academic Press, San Diego, 2002).
  42. Brodt, M.D., Ellis, C.B. & Silva, M.J. Growing C57Bl/6 mice increase whole bone mechanical properties by increasing geometric and material properties. J. Bone Miner. Res. 14, 2159−2166 (1999). | PubMed  | ISI | ChemPort |
  43. Vignery, A. & Baron, R. Dynamic histomorphometry of alveolar bone remodeling in the adult rat. Anat. Rec. 196, 191−200 (1980). | PubMed  | ISI | ChemPort |
  44. Parfitt, A.M. et al. Bone histomorphometry: Standardization of nomenclature, symbols, and units. Report of the ASBMR histomorphometry nomenclature committee. J. Bone Miner. Res. 2, 595−610 (1987). | PubMed  | ISI | ChemPort |
  45. Takeshita, S., Kaji, K. & Kudo, A. Identification and characterization of the new osteoclast progenitor with macrophage phenotypes being able to differentiate into mature osteoclasts. J. Bone Miner. Res. 15, 1477−1488 (2000). | PubMed  | ISI | ChemPort |