1. Top of page
  2. Abstract
  7. Acknowledgements

We have used homologous recombination in embryonic stem cells to generate mice with a targeted disruption of the osteopontin (Opn, or Spp1, for secreted phosphoprotein 1) gene. Mice homozygous for this disruption fail to express osteopontin (OPN) as assessed at both the mRNA and protein level, although an N-terminal fragment of OPN is detectable at extremely low levels in the bones of −/− animals. The Opn−/− mice are fertile, their litter size is normal, and they develop normally. The bones and teeth of animals not expressing OPN are morphologically normal at the level of light and electron microscopy, and the skeletal structure of young animals is normal as assessed by radiography. Ultrastructurally, proteinaceous structures normally rich in OPN, such as cement lines, persist in the bones of the Opn−/− animals. Osteoclastogenesis was assessed in vitro in cocultures with a feeder layer of calvarial osteoblast cells from wild-type mice. Spleen cells from Opn−/− mice cells formed osteoclasts 3- to 13-fold more frequently than did control Opn+/+ cells, while the extent of osteoclast development from Opn−/− bone marrow cells was about 2- to 4-fold more than from the corresponding wild-type cells. Osteoclast development occurred when Opn−/− spleen cells were differentiated in the presence of Opn−/− osteoblasts, indicating that endogenous OPN is not required for this process. These results suggest that OPN is not essential for normal mouse development and osteogenesis, but can modulate osteoclast differentiation.


  1. Top of page
  2. Abstract
  7. Acknowledgements

OSTEOPONTIN (OPN) is a secreted phosphoprotein found in the collagenous extracellular matrix of mineralized tissues and in many body fluids, notably plasma, urine, bile, and milk.1–3 The protein has a GRGDS integrin-binding sequence that interacts with integrins of the αv class, and it can facilitate attachment of cells to various surfaces, for example during the attachment of osteoclasts to bone.4,5 Sequence motifs in OPN that have been well conserved among avian and mammalian species include the RGD sequence just N-terminal to a thrombin cleavage site, an Asp-rich sequence with possible importance in binding to calcified tissues, a C-terminal heparin-binding domain, and multiple serine residues in contexts appropriate for phosphorylation by casein kinase II or mammary gland kinase.6 The synthesis of OPN is induced when T cells are activated,7 when JB6 epidermal cells are treated with 12-O-tetradecanoyl-phorbol-13-acetate,8 and when Ras becomes activated and cells acquire a metastatic phenotype.9 Indeed, various experiments strongly support the hypothesis that OPN promotes the metastatic process.10–12

In addition to a cell attachment capability, OPN has properties of a cytokine.7 For example, it can activate c-src and stimulate phosphoinositide 3-kinase activity in target cells.13,14 OPN can inhibit the induction by lipopolysaccharide and γ-interferon of inducible nitric oxide synthase (iNOS, type II nitric oxide synthase).15 This inhibition of iNOS transcription correlates with the ability of OPN to protect tumor cells from being killed by activated macrophages,16,17 suggesting that perhaps this is how OPN contributes to the metastatic phenotype.18 OPN is produced at high levels by the macrophages found in granulomas of diverse etiology, including those induced by Mycobacterium tuberculosis,19,20 consistent with its having an anti-inflammatory role. An anti-infectious role has long been suspected because of its association with resistance to certain infectious agents.7 OPN also induces cellular chemotaxis and haptotaxis,21,22 and it stimulates the infiltration of monocytes and macrophages to sites of subcutaneous OPN injection,23 possibly through a mechanism involving CD44.24 There is a strong association between enhanced OPN expression and monocyte/macrophage infiltration at sites of focal injury in the kidney.25–27

Despite the variety of activities attributed to OPN, and its prominence in many normal and pathological tissues, its significance to the vertebrate organism remains to be elucidated. It is frequently found in pathological calcifications such as atherosclerotic plaques,2 sclerotic glomeruli,28 and kidney stones.29,30 OPN's high expression in osteogenic cells and its accumulation in the calcified extracellular matrices of bone and teeth have been well established, seemingly implicating it in the development and remodeling processes of mineralized tissues.3 Its presence at mineralized tissue surfaces and interfaces31 and its facilitation of phagocytosis of OPN-coated particulates are consistent with a role in promoting cell attachment and removal of foreign bodies.32 Its prominent distribution throughout bone, and in particular its concentration at cement lines, has prompted the suggestion that OPN participates in hard tissue cohesion and may promote interfacial adhesion between apposing substrata.31,33 Other in vitro studies have identified OPN as a potent inhibitor of hydroxyapatite (calcium phosphate) crystal formation and growth.34–36

As a means to define the role that OPN plays in mammalian systems, we have generated mice that cannot make OPN because of a targeted mutational disruption of the Opn gene. (Although spp1 [secreted phosphoprotein 1] is the officially recognized name for the mouse OPN gene, we prefer to use OPN because it has become the name commonly used in the literature.) We report here that these mice develop normally and are fertile. Although no histologically detectable phenotype is apparent in the bones and teeth of mice lacking OPN, the frequency with which spleen and bone marrow cells from Opn−/− mice form osteoclasts in in vitro cocultures is elevated in comparison with cells from Opn+/+ mice.


  1. Top of page
  2. Abstract
  7. Acknowledgements

Generation of Opn−/− mice

OPN genomic clones were obtained from a mouse strain 129 genomic library (a generous gift from F. Alt) by screening with a fragment of the BALB/c Opn gene.37 Positive clones were mapped and a 4.8 kb BamHI–HindIII fragment subcloned into pBluescript. The targeting construct was made by inserting the neo cassette from pMC1 neo38 into this plasmid at the unique EagI site in exon 6, in the reverse orientation relative to OPN transcription. A thymidine kinase cassette from pMC1TK139 was inserted just 3′ of the Opn sequences, in the reverse transcriptional orientation. This construct was linearized with BamHI and 100 μg of purified DNA electroporated into 4 × 108 AB2.1 cells.40 Transfected cells were plated onto mitomycin-C treated SNL-767 fibroblasts, and drug-resistant cells were selected in G418 plus gancyclovir. Surviving clones were placed into 96-well plates and expanded. Correctly targeted clones were identified by polymerase chain reaction and confirmed by Southern blotting (Fig. 1). Cells from two clones that had undergone the desired recombination event were injected into C57Bl/6 blastocysts, which were then implanted into pseudopregnant CD-1 female mice. One of the two clones gave germline transmission of the ES cell phenotype. Genomic DNA from cells or mouse tail fragments was isolated by proteinase K digestion, extracted with phenol, and precipitated with ethanol. Chimeric males were mated to C57Bl/6 females, and the subsequent heterozygous F1 animals were crossed to generate Opn+/+ and Opn−/− lines. All animal studies were conducted using protocols approved by the Rutgers Institutional Review Board for the Use and Care of Animals.

thumbnail image

Figure FIG. 1.. Targeted disruption of the Opn locus. (A) Map of the Opn locus, below, and the targeting construct, above. Open boxes are exons, stippled box is the promoter element in the neo cassette, and the open box labeled oen is the neomycin phosphotransferase gene. Dashed lines indicate where the ends of the targeting construct fall in the Opn gene. Selected restriction sites are indicated: H, HincII; E, EcoRI; B, BamHI; Bx, BstXI; Ea, EagI; H3, Hind III. The sizes of the expected HincII fragments are indicated. (B) Southern analysis of DNA from a targeted cell line and from two mice. Genomic DNA was prepared from cells or tail DNA and digested with HincII. The fragments were separated and hybridized to the probe indicated in part A (hatched box labeled A) which hybridizes to a region of the Opn gene that is outside the region of homology between the Opn gene and the targeting allele. The positions of the wild-type (WT) and disrupted (DIS) alleles are indicated. Lane 1 is DNA from the parental, wild-type AB2.1 cell line; lane 2 is DNA from the targeted 9B cell line; lane 3 is DNA from a mouse heterozygous for the Opn disruption; and lane 4 is DNA from a mouse homozygous for the Opn disruption.

Download figure to PowerPoint

Analysis of OPN mRNA and protein

RNA was prepared by using TriReagent (GIBCO BRL, Gaithersburg, MD, U.S.A.). Total cellular RNA was fractionated on 1% agarose gels in the presence of formaldehyde and transferred to Gene Screen Plus (DuPont NEN, Boston, MA, U.S.A.). These blots were hybridized at 42°C overnight in the presence of 50% formamide. Western blotting was used to detect OPN in various tissues and body fluids. Serum-free Dulbecco's minimal essential medium, conditioned by mouse embryo fibroblasts for 16 h, was concentrated about 50-fold prior to analysis. Urine was not concentrated. Protein was extracted from bones as described.41 Briefly, bones were flash frozen in liquid N2, pulverized, and extracted with 4 M guanidine-HCl in 50 mM Tris-HCl, pH 7.3. This extract was discarded, and the residue was further extracted with 4 M guanidine-HCl in 50 mM Tris-HCl, pH 7.3, containing 0.5 M Na2EDTA, twice for 24 h each time. The EDTA extracts were combined and the buffer was changed to 6 M urea in 50 mM Tris-HCl, pH 7.3. Proteins were extracted from kidney and lactating mammary glands in RIPA buffer as previously described.42 The protein concentration was determined by using the bicinchoninic acid assay (Pierce Chemical, Rockford, IL, U.S.A.). Proteins were separated on 12% sodium dodecyl sulfate (SDS)-polyacrylamide gels and transferred to Immobilon-P membranes (Millipore, Bedford, MA, U.S.A.). These blots were blocked with 1% nonfat dry milk and reacted with the indicated antibody preparations. Antibody reactivity was visualized with enhanced chemiluminescence (Amersham, Chicago, IL, U.S.A.).


Goat anti-rat OPN antiserum 19921 was kindly provided by Dr. Cecilia Giachelli, and was used in Westerns at a dilution of 1:1500 and in immunocytochemistry at a dilution of 1:10. Antiserum 732 is a mouse anti-mouse OPN polyclonal serum developed in our laboratory in the Opn−/− mice (Kowalski et al. unpublished data) and was used in Westerns at a dilution of 1:1500 or less. Antiserum to bone sialoprotein (BSP) was LF-6, kindly provided by Dr. Larry Fisher.43

Bone histology and immunocytochemistry

Mandibles, tibiae, and calvariae from 2- to 4-month-old mice were fixed in 0.1 M sodium cacodylate-buffered 4% paraformaldehyde/1% glutaraldehyde and analyzed as described.31 Briefly, bones were left undecalcified or were decalcified for 2 weeks in 4% disodium EDTA, dehydrated, and embedded in Epon or LR White acrylic resin. One-micrometer-thick sections were cut and stained with von Kossa reagent or with toluidine blue for light microscopy; 80–100 nm sections on nickel grids were used for ultrastructural analyses by transmission electron microscopy and for colloidal-gold immunocytochemistry. Post-embedding immunolabeling for OPN44 was performed using the antibody OP-19921 and for BSP using the antibody LF-6 followed by protein A-gold (10–14 nm diameter gold particles) and conventional staining with uranyl acetate and lead citrate. Incubation of sections with preimmune serum, irrelevant polyclonal antibody, or protein A–gold alone served as controls. Morphological observations and immunocytochemical labeling patterns were recorded using a Zeiss Axiophot light microscope and a JEOL TEM 2000 FX II electron microscope operated at 80 kV.

Osteoclast formation in vitro

Osteoblast cultures were prepared from calvariae of neonatal mice of the indicated strain by sequential collagenase digestion as described45 and maintained in alpha minimal essential medium (α-MEM) with 10% fetal calf serum (Gibco BRL, Grand Island, NY, U.S.A.). Bone marrow cells were obtained by flushing the cells from the medullary cavity of femurs with α-MEM. The dispersed cells were washed, counted, and 2.5 × 105 cells/cm2 plated on 1 × 104 osteoblasts in 24-well plates. Similarly, 1 × 105 spleen cells, obtained as described,46 were plated on osteoblasts in 24-well plates. These cultures were maintained in α-MEM in 10% fetal calf serum in the presence of 10−8 M 1α,25-dihydroxyvitamin D3 for 7 days. Osteoclasts were identified by staining for tartrate-resistant acid phosphatase (TRAP) and classified according to the number of nuclei.47


  1. Top of page
  2. Abstract
  7. Acknowledgements

Derivation of Opn−/− mice

The targeting construct used to disrupt the Opn gene comprised 4.8 kb of Opn sequence from 129 strain genomic DNA containing a neo cassette inserted into the EagI site in exon 6 (Fig. 1A). This EagI site lies immediately 5′ of the RGD sequence, so that any truncated protein made from the 5′ end of the gene would lack this integrin-binding sequence. A thymidine kinase-coding sequence in the targeting vector just 3′ of the Opn sequence, and in the opposite transcriptional orientation to that of the Opn gene, allowed for enrichment of targeted clones by negative selection.39 The linearized construct was introduced into AB2.1 embryonic stem cells by electroporation, and clones that had undergone the desired homologous recombination event were identified by polymerase chain reaction. The genotype was subsequently confirmed by Southern analysis. Correctly targeted clones, grown in the absence of G418, were injected into C57Bl/6 blastocysts. One cell line, 9B, gave rise to male chimeras that were able to transmit the disrupted Opn allele to their offspring. The resulting heterozygous F1 animals were mated to generate animals homozygous for the targeted disruption of the Opn gene, which were obtained in the expected Mendelian ratio. Southern analysis of DNA from the targeted 9B cell line and two mice containing the disrupted Opn allele confirmed that these animals were homozygous for the disrupted Opn allele (Fig. 1B).

Assays for OPN expression in mice homozygous for the disrupted Opn gene

To verify that OPN expression was indeed extinguished in the Opn−/− animals, we analyzed Opn mRNA and protein levels in a variety of different tissues and cell preparations (Fig. 2). The probe used in the experiment of Fig. 2A was a fragment of the Opn cDNA extending from the 5′ end of the mRNA to the EagI site in exon 6, the site of insertion of the neo cassette in the targeting construct. This probe will hybridize to any truncated mRNA fragments that might be transcribed from the endogenous promoter in the disrupted Opn gene. No normal-sized or truncated Opn transcripts were detectable in RNA derived from kidneys of the Opn−/− mice. A higher molecular weight RNA species hybridizing with this probe was seen when large amounts of RNA from Opn−/− kidneys were analyzed (Fig. 2B, lane 6). This transcript hybridizes with both 5′ and 3′ probes and is seen in RNA preparations from mice of both genotypes. Its identity is at present unknown.

thumbnail image

Figure FIG. 2.. Absence of OPN expression in Opn−/− mice. (A) Northern analysis of kidney RNA prepared from mice with different Opn alleles. Total RNA was prepared from kidneys of mice of different genotypes and fractionated on an agarose gel. The resulting blot was probed with a fragment of the Opn cDNA extending from the 5′ end of the RNA to the EagI site in exon 6. Lane 1: +/+, 5 μg; lane 2: +/−, 5 μg; lane 3: +/−, 0.5 μg; lane 4: +/−, 0.1 μg; lane 5: −/−, 5 μg; lane 6: −/−, 20 μg. Identical results were obtained with a probe representing Opn sequences 3′ of the EagI site. (B) Western blot of OPN protein in various tissues. Protein samples were separated on 12% SDS polyacrylamide gels and transferred to Immobilon-P membranes. These blots were incubated with goat anti-rat OPN IgG (OP-199, lanes labeled 199) or with control IgG (lanes labeled nIgG), and visualized by enhanced chemiluminescence. Lane 1: 4 μl of medium conditioned by RAW264.7 cells; lanes 2–5, CM—concentrated medium conditioned by primary mouse embryo fibroblast cells, 20 μg protein/lane; lanes 6–9: 10 μl of undiluted mouse urine; and lanes 10–13: 5 μg of bone extract protein. OPN from bone migrates more rapidly on this gel than do the other forms of OPN, possibly because of a lower phosphate content. Smearing at the top of the bone +/+ lane (10) incubated with anti-OPN probably represents high molecular weight aggregates of OPN.67 (C) Presence and relative concentration of cross-reacting fragment in −/− bones. Protein extracts from +/+ and −/− bones were fractionated on a 12% SDS-polyacrylamide gel. (Left panel) Lane 1: +/+ bone extract, 0.5 μg; lane 2: +/+ bone extract, 0.05 μg; lane 3: +/+ bone extract, 0.01 μg; lane 4: −/− bone extract, 5 μg. This blot was reacted with antiserum 199 as described above. (Right panel) Lane 1: +/+ bone extract, 0.5 μg; lane 2: +/+ bone extract, 0.05 μg; lane 3: −/−bone extract, 5 μg; reaction was with antiserum 732. Positions of molecular weight markers (in kDa) are shown, and the position of wt OPN is indicated (OPN). The arrows indicate the position of the cross-reacting 35 kDa species. Antiserum 732 to mouse OPN was made in the Opn−/− mice (Kowalski et al., unpublished data) so that the secondary antibodies used also detect endogenous mouse IgG; the position of these bands in the right panel is indicated by dots.

Download figure to PowerPoint

Western blotting of a variety of tissues, fluids, and cells from these mice with the anti-OPN antiserum OP-199 confirmed that OPN protein was not detectable in the Opn−/− animals (Fig. 2B). Samples assayed included medium conditioned by mouse embryo fibroblasts (lanes 2–5), urine (lanes 6–9), and an extract of bone (lanes 10–13). In many cases, these results were difficult to interpret because cross-reactivity of OP-199 and other antibodies was seen with several unidentified proteins, particularly in the tissue extracts. For this reason, comparisons of identical samples incubated with immune and control immunoglobulin G (IgG) are shown Fig. 2B. For example, in lanes 3 and 4, showing conditioned medium from embryo fibroblast cultures, several proteins migrating more rapidly than OPN in the Opn−/− sample exhibited reactivity with the 199 antiserum; however, this reactivity was also seen with the control IgG in lane 4.

In bone extracts from the Opn−/− animals, antisera OP-199 and 732, both specific for OPN, detect a protein migrating with an apparent molecular weight of ∼35 kDa in long exposures (Fig. 2C). It is likely that this protein represents a truncated form of OPN. In principle, a transcript could be generated from the endogenous OPN promoter and be completely processed to generate a 2.8 kb mRNA containing the neo sequences in exon 6. If this transcript were translated, it would give rise to an amino terminal fragment of OPN, which would contain sequences represented in exons 2–5 and part of exon 6. Such a protein would not contain the RGD sequence, or the C-terminal half of the protein. We have estimated that the 35 kDa protein is present at a level 100- to 200-fold lower than that of wildtype OPN, and we have been unable to detect it in any body fluids or tissues other than bone. This fragment of OPN would be unlikely to have any effect on the phenotype of the animals. First, it is predicted to lack the RGD sequence that has been shown to be important for OPN function in several systems. Second, while this fragment would be expected to retain the poly Asp sequence, which might allow it to function in mineral binding, its extremely low concentration (Fig. 2C) renders it unlikely that this fragment can have any effect on the bone phenotype. Independent support for this idea comes from observations of animals with a different disruption of the OPN gene in which exons 4–7 are deleted.48 These animals lack the immunoreactive 35 kDa OPN fragment, yet their bone morphology is indistinguishable from that described here (Figs. 3 and 4; McKee, Rittling, and Liaw, unpublished data).

thumbnail image

Figure FIG. 3.. Histology of the proximal tibial growth plate in Opn−/− and Opn−/− mice. Light microscopic features of both wildtype (A, Opn+/+) and mutant (B, Opn−/−) tissues are similar in that the growth plates (GP) subjacent to epiphyseal bone (EB) contain columns of chondrocytes that typically proceed through proliferative and hypertrophic stages. In mice of both genotypes, bone (B) is deposited by osteoblasts onto spicules of calcified cartilage (C) to form the primary spongiosa (PS). These are epoxy resin (Epon) sections obtained from decalcified specimens and stained with toluidine blue.

Download figure to PowerPoint

Characteristics of the OPN-deficient mice

Mice homozygous for the targeted disruption appear phenotypically normal. They are fertile and can lactate, and their litter size is normal. Weights of the animals of the different genotypes between 25 and 52 days of age do not differ significantly (data not shown). Histologic examination of liver, spleen, kidney, pancreas, and lung revealed no obvious abnormalities in the Opn−/− animals (data not shown).

Bone morphology in the absence of OPN

OPN was originally isolated from bone,49 and its name reflects its presumed importance in this tissue, in which it is especially abundant.3 We have extensively compared the bones of Opn+/+ and Opn−/− animals using radiography, light and electron microscopy, and ultrastructural immunocytochemistry. The skeletal structure of the Opn−/− animals appeared radiographically normal (data not shown). Morphologically, the cells and extracellular matrix organization and composition of the bones and teeth in the Opn−/−mice were indistinguishable from those of wild-type animals (Fig. 3 and data not shown). In bone, osteogenic cell types were readily identifiable and were present with their expected frequency and distribution. Identical results have been obtained with an independently derived strain of Opn−/− mice (McKee and Liaw, unpublished data). These results lend support to the idea that the cross-reacting 35 kDa protein seen on Western blots, if it is an OPN fragment, is not responsible for the lack of a phenotype in the bones of the Opn−/− mice. The disruption in the OPN gene in the mutant mice generated by Liaw and coworkers was achieved by a strategy that would not be expected to generate a similar 35 kDa fragment.48

Ultrastructurally, extracellular matrix organization of bone tissue in the mutant mice was unchanged, and prominent organic structures within the bone, such as collagen fibrils, cement lines, and laminae limitantes, were all readily discernable. Calcification of the matrix appeared unaffected by the absence of OPN. Osteoclasts with well-developed ruffled borders and otherwise normal histology were present, and numerous crenated cement (reversal) lines, indicative of bone resorption activity by these cells, were distributed throughout the bone matrix. Colloidal gold immunocytochemistry for OPN in wild-type mice revealed intense immunolabeling of mineralized matrix in bone, tooth cementum, laminae limitantes at bone surfaces, and cement lines at sites of bone remodeling. However, in the Opn−/− mice, while normal hard tissue architecture and organization were retained (Fig. 4A), cement lines and other structural elements normally known to contain OPN (Fig. 4B) showed a complete absence of immunolabeling for this protein (Fig. 4C). Other noncollagenous extracellular matrix proteins abundant in bone, such as BSP (Fig. 4D) and osteocalcin (data not shown), exhibited essentially normal immunolabeling patterns in the OPN-deficient mice.

thumbnail image

Figure FIG. 4.. Bone ultrastructure and immunocytochemistry in Opn+/− (B) and Opn−/− (A, C, D) animals. (A) As observed here by transmission electron microscopy of undecalcified samples of tibia from mutant mice, and as similarly noted for wild-type animals, bone-forming osteoblasts (Ob) secrete a layer of generally unmineralized osteoid matrix (OS) that subsequently calcifies to become the mineralized matrix (MM) proper of bone. As for normal bone, calcification commences as small foci within the osteoid (arrows), with mineral confluence being achieved at the interface between the osteoid and the mineralized matrix—the so called mineralization front. Osteoblast lineage cells become trapped in the matrix and are identified as osteocytes (Oc). (B) Post-embedding, colloidal-gold immunocytochemistry for OPN in heterozygous (illustrated here) and wild-type mice reveals immunolabeling throughout the bone matrix, particularly in cement lines (CL). (C) Immunocytochemistry performed as in (B) on sections of bone form OPN−/− mice. The absence of colloidal-gold particles over cement lines confirms the lack of OPN in these structures. (D) Bone matrix immunolabeled for BSP in Opn−/− mice showed an otherwise normal distribution of gold particles throughout the bone and also at cement lines (CL). (A) Epon section of undecalcified tibia stained with uranyl acetate and lead citrate. (B–D), LR White sections of decalcified alveolar bone from the mandible immunolabeled for OPN or BSP and counterstained with uranyl and lead.

Download figure to PowerPoint

Altered osteoclastogenesis in vitro

OPN has been implicated in osteoclast function,4 (reviewed in Ref. 3) so the consequences of a lack of this protein on osteoclast differentiation from monocyte precursors were assessed in vitro. When in contact with osteoblasts, and in the presence of 1α,25-dihydroxyvitamin D3, osteoclast precursor cells from bone marrow and spleen can be induced to differentiate into osteoclast-like cells.45,46 In these coculture systems, cells derived from bone marrow and spleen differentiate in vitro over 7 days into multinucleated cells with the characteristics of osteoclasts: they stain for TRAP, resorb bone, and bind calcitonin.50 Spleen cells from Opn−/− animals in such cocultures gave rise to markedly more TRAP positive (TRAP+) cells than did spleen cells from Opn+/+ mice (Table 1, Fig. 5). Spleen cells from Opn+/− animals gave an intermediate result. While the absolute number of osteoclasts formed varied among individual animals (as has been previously shown to occur51), on average, about 7-fold more multinucleated cells stained for TRAP after 7 days in culture in the Opn−/− cultures as compared with the Opn+/+ cultures (Table 1, and data not shown). Cells derived from the Opn+/− animals were on average 3-fold more efficient at forming osteoclasts than were wild-type cells (Table 1, and data not shown). These TRAP+ cells derived from Opn−/− spleens were confirmed as osteoclast-like in that they were able to form resorption pits in bone slices (data not shown), and the morphology of these pits was similar for both the Opn+/+ and Opn−/− osteoclasts. When bone marrow cells from Opn+/+ and Opn−/− mice were placed in such cocultures with primary osteoblasts derived from wild-type mice (either 129xC57Bl/6 or ddy, Table 1), a similar increase in the numbers of TRAP+ cells developing in 7 days was observed, although the magnitude of the difference, 2- to 4-fold increased numbers of TRAP+ cells in the Opn−/− cultures, was not as great as for the spleen cells.

thumbnail image

Figure FIG. 5.. Osteoclast formation in vitro. TRAP staining of osteoclasts (red) developing in cocultures with ddy osteoblasts,(47) as described in the Materials and Methods. A, +/+; B, +/−; C, −/−; D, +/+; E, −/−. (A–C) Osteoclasts developed from spleen precursors; (D–E) osteoclasts from bone marrow precursors. Original magnification ×40.

Download figure to PowerPoint

Thumbnail image of

The osteoblast cells used in this coculture system produce OPN at readily detectable levels (Fig. 6), so that the osteoclasts from the Opn−/− spleens were exposed to OPN during the culture period. This observation implies that the observed difference in osteoclast formation is due to differences in the spleen cells themselves, or that OPN plays an autocrine role in this system, such that the osteoclast precursors can distinguish endogenously synthesized from exogenously supplied OPN. To distinguish between these possibilities, spleen cells from Opn−/− mice were differentiated on osteoblasts derived from Opn−/− calvariae (Table 1). The results were similar to those obtained with wild-type osteoblasts, indicating that OPN is not required for this process in excess of the amount provided in the FBS.

thumbnail image

Figure FIG. 6.. OPN expression in ddy osteoblast cultures. Osteoblasts were prepared from ddy calvaria, and cultured for 8 days. At the end of the culture period, the cells were incubated in serum-free medium for an additional 1 day (1d CM), 2 days (2d CM), or 3 days (3d CM) as indicated above the lanes, and this conditioned medium was collected. Fifteen microliters of these conditioned media were fractionated directly on an SDS polyacrylamide gel, transferred to Immobilon-P, and reacted with OP-199 IgG as described in the legend to Fig. 2B and the Materials and Methods.

Download figure to PowerPoint


  1. Top of page
  2. Abstract
  7. Acknowledgements

Significance of normal development in Opn−/− mice

The OPN protein sequence is highly conserved among species,52 and the protein is expressed by cells in a wide variety of tissues throughout the body.53 OPN is found in most if not all body fluids, is very abundant in mineralized tissues, and has long been implicated in bone formation and remodeling.3,54 For these reasons, the apparently normal phenotype of mice lacking OPN was unexpected. Opn mRNA is expressed at high levels in kidney, for example, yet the kidneys of the mice that do not express OPN are morphologically normal. We have been unable to detect OPN protein in normal (+/+) kidneys by Western blotting (data not shown), which implies that under nonpathological conditions, there is little OPN in soft tissues. Thus, while OPN is a ubiquitous component of body fluids, perhaps acting to prevent mineral precipitation from these solutions,55,56 it does not appear to play an essential role in the normal processes of soft tissue differentiation or homeostasis. It follows that a lack of OPN in these soft tissues has little consequence to the healthy, unstressed organism. Interestingly, mice with disruptions in genes coding for vitronectin and tenascin, which are also RGD-containing proteins,57,58 or for both OPN and vitronectin48 similarly develop and grow normally.

Role of OPN in bone morphology and mineralization

OPN is abundant in the mineralized tissues; its ability to bind to calcified matrices is due to its overall acidity, including a poly Asp stretch, and a high degree of phosphorylation.59 The accumulation of OPN in cement lines demarcating the reversal site of bone remodeling by osteoclasts, and at bone surfaces—laminae limitantes—where osteocytes, osteoblasts, bone lining cells and osteoclasts routinely interface directly with the extracellular matrix, has led to speculation that OPN regulates cell adhesion and dynamics at bone surfaces.4,5,32 It has also been proposed that OPN present at cement lines (resting, or reversal, lines) and elsewhere in bone mediates hard tissue integrity by binding various extracellular matrix components as well as mineral, thus linking organic and inorganic phases to provide tissue adhesion/cohesion (reviewed in Ref. 33).

In the present study, we have documented that morphologically defined structures known to be rich in OPN persist in the bones and teeth of Opn−/− mice, and that a lack of OPN apparently has no effect on either the structure or the distribution of cells within these tissues. While no histologically detectable phenotype is apparent in the mineralized tissues of mice lacking OPN, biochemical and crystallographic studies are in progress to test for differences in bone strength and mineral organization in these animals. Since OPN is a member of a family of RGD-containing proteins, some of which, such as BSP, are abundant in bone, it may be that some of these other proteins, or perhaps heretofore unidentified proteins, can subserve the putative function of OPN in its absence.

With regard to extracellular matrix mineralization in bones and teeth, our data suggest either that OPN is not normally involved in the calcification of these tissues or that such hard tissues can utilize alternative calcification strategies not involving OPN. A variety of anionic proteins have been identified as regulators of calcification in vertebrate and invertebrate mineralizing systems.35,60,61 In light of the vital importance of the vertebrate skeleton in maintaining form and locomotion capability, in defining internal cavities and protecting organs and tissues, and in acting as an ion reservoir for calcium homeostasis, it is reasonable that redundant strategies exist for developing and maintaining hard tissue extracellular matrices such as found in bone.

Function of OPN in osteoclastogenesis

Although there is no obvious alteration in the morphology or ultrastructure of bone cells and extracellular matrix in the Opn−/− animals, the formation of osteoclast-like cells is enhanced up to 13-fold in cocultures with calvarial osteoblasts when the cells are prepared from the spleen or bone marrow of the Opn−/− animals compared with those from the Opn+/+ animals. This result suggests two possibilities: that OPN inhibits the differentiation of osteoclast precursors into osteoclasts in cell culture, or that OPN affects the formation or accumulation of osteoclast precursors in the spleen and in the bone marrow. Our observation that osteoclasts are formed with similar efficiencies on wild-type and Opn−/− osteoblasts implies, however, that OPN expression is not required for this differentiation process in vitro, and that the difference observed in vitro reflects differences in the cellular composition of the spleen and bone marrow.

Yamate et al.62 demonstrated that in cultures of bone marrow cells, a specific antiserum to OPN inhibited the formation of TRAP+ cells, as did RGD-containing peptides, suggesting that the binding of OPN to cell surface integrins is important in the development of osteoclasts in the in vitro system. Our results differ from these observations in that we describe an inhibitory effect of OPN on the process of osteoclast differentiation. The major difference between our experiments and those of Yamate and coworkers is in the culture conditions; our experiments were performed on calvarial osteoblasts while those of Yamate et al. utilized cells from the bone marrow cultures themselves as stromal cells. One possible explanation for these divergent results is that there are multiple differentiation pathways leading to osteoclastogenesis, and the pathway used depends on the specific cellular and molecular composition of the culture system used. We hypothesize, then, that OPN plays different roles in the different pathways, stimulating differentiation along one pathway, inhibiting it along another. Indeed, our results demonstrate that OPN is dispensable for the differentiation process in vitro altogether, in that osteoclast formation occurs when Opn−/− spleen cells are cocultured with Opn−/− osteoblasts.

In any case, the alteration in osteoclast precursors that we detect in this assay does not appear to affect osteoclast differentiation in vivo under nonpathological conditions. An expected result of increased osteoclast development in vivo might be an osteoporotic/osteopenic phenotype in the Opn−/− animals, yet this has not been detected. Thus, mechanisms to compensate for a lack of OPN appear to exist in the whole animal, but possibly not in the isolated cell cultures. Additionally, if different pathways of osteoclast differentiation exist in vivo, it may be that the pathway used for osteoclastogenesis during normal bone development does not depend on OPN, while a different pathway is used in pathological situations, in which OPN may have a function.

Function of OPN in pathological settings

OPN expression in a variety of tissues is elevated in certain pathologies, and the protein is thought to function in several important aspects of immune cell function. For example, OPN expression is known to be increased in the kidney in association with the interstitial fibrosis occurring with glomerulonephritis, with cyclosporinee nephropathy, with angiotensin II–induced tubulointerstitial nephritis, and with hydronephrosis.2,26,27,63 In each case, OPN was hypothesized to play a role in the recruitment of macrophages to these sites of tissue injury. OPN interacts with murine macrophages,23 attenuates their response to specific stimuli,17 and stimulates IgG and IgM production in mixed cultures of macrophages and B cells.64 The protein is important in macrophage infiltration in vivo,65 and is implicated in macrophage adhesion and perhaps functions in bone wound healing.66 Taken together, these observations implicate OPN expression as a cellular response to tissue injury of various sorts.66 Indeed, Liaw et al.48 have presented evidence that OPN does have a role in soft tissue remodeling, i.e., wound healing. Since the mice in our colony, housed under specific pathogen-free conditions, are not subject to such pathologies, the effect of an absence of OPN in these animals is minimal. However, experiments are in progress to test whether Opn−/− mice subjected to various physiological stresses and insults will respond differently than their wild-type counterparts.


  1. Top of page
  2. Abstract
  7. Acknowledgements

We thank Drs. Mario Capecchi and members of his laboratory for making various reagents and constructs available for us, Allan Bradley for the gift of ES cell lines, Fred Alt and Dennis Willerford for the gift of the 129 library, and Cecilia Giachelli for the gift of the OP-199 antibody. The advice and assistance of Dr. Kiran Chada and colleagues are gratefully acknowledged. We thank Dr. John Hoyer for assistance in evaluating kidney histology. M.D.M. and A.N. wish to thank I. Turgeon for technical assistance and S. Zalzal for preparation of the colloidal-gold complexes This work was supported by National Institute of Health grant DC01295 (D.T.D.), by a Johnson & Johnson Discovery Research Award (D.T.D.), by New Jersey Commission on Cancer Research Grants #795–035 and #796–031 (S.R.R.), and by MRC of Canada grants #MT11360 and #MTB674 (M.D.M.).


  1. Top of page
  2. Abstract
  7. Acknowledgements
  • 1
    Denhardt DT, Guo X 1993 Osteopontin: A protein with diverse functions FASEB J 7:14751482.
  • 2
    Giachelli CM, Schwartz SM, Liaw L 1995 Molecular and cellular biology of osteopontin: Potential role in cardiovascular disease Trends Cardiovasc Med 5:8895.
  • 3
    Butler WT, Ridall AL, McKee MD 1996 Osteopontin. In: Bilezekian,JP, Raisz,LG, Rodan,GA (eds.) Principles of Bone Biology. Academic Press, San Diego, CA, U.S.A., pp. 167181.
  • 4
    Reinholt FP, Hultenby K, Oldberg C, Heinegård D 1990 Osteopontin—A possible anchor of osteoclasts to bone Proc Natl Acad Sci USA 8:44734475.
  • 5
    McKee MD, Nanci A 1995 Osteopontin and the bone remodeling sequence: Colloidal-gold immunocytochemistry of an interfacial extracellular matrix protein Ann NY Acad Sci 760:177189.
  • 6
    Sørensen ES, Højrup P, Petersen TE 1995 Post-translational modification of bovine osteopontin: Identification of twenty-eight phosphorylation and three O-glycosylation sites Protein Sci 4:20402049.
  • 7
    Patarca R, Saavedra RA, Cantor H 1993 Molecular and cellular basis of genetic resistance to bacterial infection: The role of the early T-lymphocyte activation-1/osteopontin gene Crit Rev Immunol 13:225246.
  • 8
    Smith JH, Denhardt DT 1987 Molecular cloning of a tumor promoter-inducible mRNA found in JB6 mouse epidermal cells: Induction is stable at high, but not at low, cell densities J Cell Biochem 34:1322.
  • 9
    Craig AM, Bowden GT, Chambers AF, Spearman MA, Greenberg AH, Wright JA, McLeod M, Denhardt DT 1990 Secreted phosphoprotein mRNA is induced during multi-stage carcinogenesis in mouse skin and correlates with the metastatic potential of murine fibroblasts Int J Cancer 46:133137.
  • 10
    Gardner HAR, Berse B, Senger DF 1994 Specific reduction in osteopontin synthesis by antisense RNA inhibits the tumorigenicity of transformed Rat1 fibroblasts Oncogene 9:23212326.
  • 11
    Feng B, Rollo EE, Denhardt DT 1995 Osteopontin (OPN) may facilitate metastasis by protecting cells from macrophage NO-mediated cytotoxicity: Evidence from cell lines down-regulated for OPN expression by a targeted ribozyme Clin Exp Metastasis 13:453462.
  • 12
    Oates AJ, Barraclough R, Rudland PS 1996 The identification of osteopontin as a metastasis-related gene product in a rodent mammary tumour model Oncogene 13:97104.
  • 13
    Chellaiah M, Fitzgerald C, Filardo EJ, Cheresh DA, Hruska KA 1996 Osteopontin activation of c-src in human melanoma cells requires the cytoplasmic domain of the integrin αv-subunit Endocrinology 137:24322440.
  • 14
    Chellaiah M, Hruska K 1996 Osteopontin stimulates gelsolin-associated phosphoinositide levels and phosphatidylinositol triphosphate-hydroxyl kinase Mol Biol Cell 7:743753.
  • 15
    Hwang S-M, Lopez CA, Heck DE, Gardner CR, Laskin DL, Laskin JD, Denhardt DT 1994 Osteopontin inhibits induction of nitric oxide synthase activity by inflammatory mediators in mouse kidney epithelial cells J Biol Chem 269:711715.
  • 16
    Rollo EE, Denhardt DT 1996 Differential effects of osteopontin on the cytotoxic activity of macrophages from young and old mice Immunology 88:642647.
  • 17
    Rollo EE, Laskin DL, Denhardt DT 1996 Osteopontin inhibits nitric oxide production and cytotoxicity by activated RAW264.7 macrophages J Leukoc Biol 60:397404.
  • 18
    Denhardt DT, Chambers AF 1994 Overcoming obstacles to metastasis—Defenses against host defenses: Osteopontin (OPN) as a shield against attack by cytotoxic host cells J Cell Biochem 56:4851.
  • 19
    Carlson I, Tognazzi K, Manseau EJ, Dvorak HF, Brown LF 1997 Osteopontin is strongly expressed by histiocytes in granulomas of diverse etiology Lab Invest 77:103108.
  • 20
    Nau GJ, Guilfoile P, Chupp GF, Berman JS, Kim SJ, Kornfeld H, Young RA 1997 A chemoattractant cytokine associated with granulomas in tuberculosis and silicosis Proc Natl Acad Sci USA 94:64146419.
  • 21
    Liaw L, Almeida M, Hart CE, Schwartz SM, Giachelli CM 1994 Osteopontin promotes vascular cell adhesion and spreading and is chemotactic for smooth muscle cells in vitro Circ Res 74:214224.
  • 22
    Senger DR, Ledbetter SR, Claffey KP, Papadopoulos-Sergiou A, Perruzzi CA, Detmar M 1996 Stimulation of endothelial cell migration by vascular permeability factor/vascular endothelial growth factor through cooperative mechanisms involving the αvβ3 integrin, osteopontin, and thrombin Am J Pathol 149:293305.
  • 23
    Singh RP, Patarca R, Schwartz J, Singh P, Cantor H 1990 Definition of a specific interaction between the early T lymphocyte activation 1 (ETA-1) protein and murine macrophages in vitro and its effect upon macrophages in vivo J Exp Med 171:19311942.
  • 24
    Weber GF, Ashkar S, Glimcher MJ, Cantor H 1996 Receptor–ligand interaction between CD-44 and osteopontin (ETA-1) Science 271:509512.
  • 25
    Giachelli CM, Pichler R, Lombardi D, Denhardt DT, Alpers CE, Schwartz SM, Johnson RJ 1994 Osteopontin expression in angiotensin II-induced tubulointerstitial nephritis Kidney Int 45:515524.
  • 26
    Pichler R, Giachelli CM, Lombardi D, Pippin J, Gordon K, Alpers CE, Schwartz SM, Johnson RJ 1994 Tubulointerstitial disease in glomerulonephritis: Potential role of osteopontin (uropontin) Am J Path 144:915926.
  • 27
    Diamond JR, Kees-Folts D, Ricardo SD, Pruznak A, Eufemio M 1995 Early and persistent up-regulated expression of renal cortical osteopontin in experimental hydronephrosis Am J Pathol 146:14551466.
  • 28
    Lopez CA, Hoyer JR, Wilson PD, Waterhouse P, Denhardt DT 1993 Heterogeneity of osteopontin expression among nephrons in mouse kidney and enhanced expression in sclerotic glomeruli Lab Invest 69:355363.
  • 29
    Kohri K, Nomura S, Kitamura Y, Nagata T, Yoshioka K, Iguchi T, Yamate T, Umekawa Y, Suzuki H, Sinohara H, Kurita T 1993 Structure and expression of the mRNA encoding urinary stone protein (osteopontin) J Biol Chem 258:1518015184.
  • 30
    McKee MD, Nanci A, Khan SR 1995 Ultrastructural immunodetection of osteopontin and osteocalcin as major matrix components of renal calculi J Bone Miner Res 10:19131929.
  • 31
    McKee MD, Nanci A 1996 Osteopontin at mineralized tissue interfaces in bone, teeth, and osseointegrated implants: Ultrastructural distribution and implications for mineralized tissue formation, turnover, and repair Microsc Res Tech 33:141164.
  • 32
    McKee MD, Nanci A 1996 Secretion of osteopontin by macrophages and its accumulation at tissue surfaces during wound healing in mineralized tissues: A potential requirement for macrophage adhesion and phagocytosis Anat Rec 245:394409.
  • 33
    McKee MD, Nanci A 1996 Osteopontin: An interfacial extracellular matrix protein in mineralized tissues Connect Tissue Res 35:197205.
  • 34
    Boskey AL, Maresca M, Ullrich W, Doty SB, Butler WT, Prince CW 1993 Osteopontin–hydroxyapatite interactions in vitro: Inhibition of hydroxyapatite formation and growth in a gelatin-gel Bone Miner 22:147159.
  • 35
    Hunter GK 1996 Interfacial aspects of biomineralization Curr Opin Solid State Mater Sci 1:430435.
  • 36
    Hunter GK, Hauschka PV, Poole AR, Rosenberg LC, Goldberg HA 1996 Nucleation and inhibition of hydroxyapatite formation by mineralized tissue proteins Biochem J 317:5964.
  • 37
    Craig AM, Denhardt DT 1991 The murine gene encoding secreted phosphoprotein 1 (osteopontin): Promoter structure, activity, and induction in vivo by estrogen and progesterone Gene 100:163171.
  • 38
    Thomas KR, Capecchi MR 1987 Site-directed mutagenesis by gene targeting in mouse embryo-derived stem cells Cell 51:503512.
  • 39
    Mansour SL, Thomas KR, Capecchi MR 1988 Disruption of the proto-oncogene int-2 in mouse embryo-derived cells: A general strategy for targeting mutations to non-selectable genes Nature (Lond) 336:348352.
  • 40
    Soriano P, Montgomery C, Geske R, Bradley A 1991 Targeted disruption of the c-src proto-oncogene leads to osteopetrosis in mice Cell 64:693702.
  • 41
    Prince CW, Oosawa T, Butler WT, Tomana M, Bhown AS, Bhown M, Schrohenloher RE 1987 Isolation, characterization, and biosynthesis of a phosphorylated glycoprotein from rat bone J Biol Chem 262:29002907.
  • 42
    Rittling SR, Novick KE 1997 Osteopontin expression in mammary gland development and tumorigenesis Cell Growth Differ 8:10611069.
  • 43
    Fisher LW, Hawkins GR, Tuross N, Termine JD 1987 Purification and partial characterization of small proteoglycans I and II, bone sialoproteins I and II, and osteonectin from the mineral compartment of developing human bone J Biol Chem 262:97029708.
  • 44
    Bendayan M 1995 Colloidal gold post-embedding immunocytochemistry Prog Histochem Cytochem 29:1159.
  • 45
    Takahashi N, Akatsu A, Udagawa N, Sasaki T, Yamaguchi A, Moseley J, Martin TJ, Suda T 1988 Osteoblastic cells are involved in osteoclast formation Endocrinology 123:26002602.
  • 46
    Udagawa N, Takahashi N, Akatsu T, Sasaki T, Yamaguchi A, Kodama H, Martin TJ, Suda T 1989 The bone marrow-derived stromal cell lines MC3T3-G2/PA6 and ST2 support osteoclast-like cell differentiation in cocultures with spleen cells Endocrinology 125:18051813.
  • 47
    Matsumoto HN, Tamura M, Denhardt DT, Obinata M, Noda M 1995 Establishment and characterization of bone marrow stromal cell lines that support osteoclastogenesis Endocrinology 136:40844091.
  • 48
    Liaw L, Birk DE, Ballas CB, Whitsitt JS, Davidson JS, Hogan BLM 1998 Altered wound healing in mice lacking a functional osteopontin gene (spp1) J Clin Invest 101:14681478.
  • 49
    Franzen A, Heinegård D 1985 Isolation and characterization of two sialoproteins present only in bone calcified matrix Biochem J 232:715724.
  • 50
    Akatsu T, Tamura T, Takahashi N, Udagawa N, Tanaka S, Sasaki T, Yagamushi A, Nagata N, Suda T 1992 Preparation and characterization of a mouse osteoclast-like multinunucleated cell population J Bone Miner Res 7:12971306.
  • 51
    Grigoriadis AE, Wang ZQ, Cecchini MG, Hofstetter W, Felix R, Fleisch HA, Wagner EF 1994 c-fos: A key regulator of osteoclast-macrophage lineage determination and bone remodeling Science 266:443448.
  • 52
    Butler WT 1995 Structural and functional domains of osteopontin Ann NY Acad Sci 760:611.
  • 53
    Brown LF, Van Berse B, de Water LI, Papadopoulos-Sergiou A, Peruzzi CA, Manseau EJ, Dvorak HR, Senger DR 1992 Expression and distribution of osteopontin in human tissues; widespread association with luminal epithelial surfaces Mol Biol Cell 3:11691180.
  • 54
    Chen J, Singh K, Mukherjee BB, Sodek J 1993 Developmental expression of osteopontin (OPN) mRNA in rat tissues: Evidence for a role for OPN in bone formation and resorption Matrix 13:113120.
  • 55
    Shiraga H, Min W, VanDusen WJ, Clayman MD, Miner D, Terrell CH, Sherbotie JR, Foreman JW, Przysiecki C, Neilson EG, Hoyer JR 1992 Inhibition of calcium oxalate crystal growth in vitro by uropontin, a new member of the aspartic-acid rich protein superfamily Proc Natl Acad Sci USA 89:426430.
  • 56
    Worcester EM, Blumenthal SS, Beshensky AM, Lewand DL 1992 The calcium oxalate crystal growth inhibitor protein produced by mouse kidney cortical cells in culture is osteopontin J Bone Miner Res 7:10291036.
  • 57
    Zheng X, Saunders TL, Camper SA, Samuelson LC, Ginsburg D 1995 Vitronectin is not essential for normal mammalian development and fertility Proc Natl Acad Sci USA 92:1242612430.
  • 58
    Saga Y, Yagi T, Ikawa Y, Sakakura T, Aizawa S 1992 Mice develop normally without tenascin Genes Dev 6:18211831.
  • 59
    Goldberg HA, Hunter GK 1995 The inhibitory activity of osteopontin on hydroxyapatite formation in vitro Ann NY Acad Sci 760:305308.
  • 60
    Moradian-Oldak J, Frolow F, Addadi L, Weiner S 1992 Interactions between acidic matrix macromolecules and calcium phosphate ester crystals: relevance to carbonate apatite formation in biomineralization Proc R Soc Lond (Biol) 24:4755.
  • 61
    Aizenberg J, Hanson J, Ilan M, Leiserowitz L, Koetzle TF, Addadi L, Weiner S 1995 Morphogenesis of calcitic sponge spicules: A role for specialized proteins interacting with growing crystals FASEB J 9:262268.
  • 62
    Yamate T, Mocharla H, Taguchi Y, Igietseme JU, Manolagas SC, Abe E 1997 Osteopontin expression by osteoclast and osteoblast progenitors in the murine bone marrow: Demonstration of its requirement for osteoclastogenesis and its increase after ovariectomy Endocrinology 138:30473055.
  • 63
    Young BA, Burdmann EA, Johnson RJ, Alpers CE, Giachelli CM, Eng E, Andoh T, Bennett WM, Couser WG 1995 Cellular proliferation and macrophage influx precede interstitial fibrosis in cyclosporine nephrotoxicity Kidney Int 48:439438.
  • 64
    Lampe MA, Patarca R, Iregui MV, Cantor H 1991 Polyclonal B-cell activation by the Eta-1 cytokine and the development of autoimmune disease J Immunol 147:29022906.
  • 65
    Giachelli CM, Lombardi D, Johnson RJ, Murry CE, Almeida M 1998 Evidence for a role of osteopontin in macrophage infiltration in response to pathological stimuli in vivo Am J Pathol 152:353358.
  • 66
    Rodan G 1995 Osteopontin overview Ann NY Acad Sci 760:15.
  • 67
    Sørensen ES, Rasmussen LK, Møller L, Jensen PH, Højrup P, Petersen TE 1994 Localization of transglutaminase-reactive glutamine residues in bovine osteopontin Biochem J 304:1316.