OUP user menu

Osteopontin—a molecule for all seasons

M. Mazzali, T. Kipari, V. Ophascharoensuk, J.A. Wesson, R. Johnson, J. Hughes
DOI: http://dx.doi.org/10.1093/qjmed/95.1.3 3-13 First published online: 1 January 2002


Osteopontin (OPN) is a multifunctional protein, and although highly expressed in bone, it is also expressed by various cell types including macrophages, endothelial cells, smooth muscle cells and epithelial cells.1,,2 OPN is involved in diverse biological processes and the aim of this review is to give a broad outline of some important aspects of the biology of OPN. Although this review is written primarily from a renal perspective, it is hoped that the reader will appreciate that OPN is involved in both physiological and pathological processes in multiple organs and tissues including biomineralization, inflammation, leukocyte recruitment and cell survival.

OPN structure

OPN is a negatively‐charged acidic hydrophilic protein of approximately 300 amino acid residues, and is secreted into all body fluids. The OPN cDNA from various mammalian species exhibits a high degree of sequence homology. There is evidence of alternative splicing, although the functional significance of this is unclear at present. The molecule undergoes considerable post‐translational modification, and is phosphorylated and glycosylated. OPN has an arginine‐glycine‐aspartic acid (RGD) cell binding sequence, a calcium binding site and two heparin binding domains.

Cells may bind OPN via multiple integrin receptors including the vitronectin receptor (αvβ3) as well as various β1 and β5 integrins. Integrin binding may be RGD‐dependent or ‐independent (e.g. via the motif SVVYGLR). OPN does not bind the standard form of CD44 (hyaluronic acid receptor) but does bind various isoforms of CD44. These CD44 isoforms bind OPN via multiple sites. OPN may be cleaved by thrombin, resulting in the exposure of additional cryptic binding sites as well as the production of functional chemotactic fragments.3

Regulation of OPN expression

The regulation of OPN expression is incompletely understood at present and may differ between various cell types. Studies indicate that the OPN promoter contains various motifs including a purine‐rich sequence, an ets‐like sequence, glucocorticoid and vitamin D response elements and interferon‐inducible elements.4–,6 Pro‐inflammatory cytokines stimulate OPN gene transcription and expression. For example, activation of macrophages with lipopolysaccharide (LPS) and nitric oxide (NO) induces OPN gene expression and protein secretion.7 Classical mediators of acute inflammation such as tumour necrosis factor α (TNFα) and interleukin‐1β (IL‐1β) strongly induce OPN expression4,5,,8 while other mediators that can induce OPN upregulation include angiotensin II, transforming growth factor β(TGFβ), hyperglycaemia and hypoxia.9–,14

OPN and pathological mineralization

OPN is intimately involved in the regulation of both physiological and pathological mineralization. In normal bone tissue, OPN is expressed by both osteoclasts and osteoblasts which are the cells responsible for bone remodelling. During normal bone mineralization, osteoclast‐derived OPN inhibits the formation of hydroxyapatite.15,,16 Despite the fact that OPN knockout mice exhibit a normal bone structure,17 there is evidence indicating that OPN is involved in bone resorption. For example, OPN knockout mice are markedly protected from bone loss following ovariectomy.18 In addition, bone resorption mediated by parathyroid hormone is also OPN‐dependent.19

Prominent OPN expression is found in regions of dystrophic calcification such as that associated with degenerative and atheromatous vascular disease.20–,24 Experiments using an in vitro model of smooth muscle cell calcification indicate that OPN exerts an inhibitory effect upon this process,25 thereby suggesting that the upregulation of OPN evident in areas of dystrophic calcification represents an attempt to either prevent or limit vascular calcification.

OPN is also an important facet of the urinary tract's defence against the formation of renal stones, the majority of which are comprised of calcium oxalate. As a consequence of the essential homeostatic conservation of water, normal urine is supersaturated with respect to crystalline components, and this empirically suggests the existence of physiological mechanisms that actively inhibit urinary crystallization of calcium salts. Various inhibitory macromolecules including urinary OPN have now been isolated and identified from both normal urine and kidney stones, e.g. nephrocalcin, crystal matrix protein, bikunin and Tamm‐Horsfall protein.26–,30

OPN is synthesized within the kidney and secreted into the urine by epithelial cells, including the loop of Henle, distal convoluted tubule and papillary epithelium.31,,32 OPN can inhibit the nucleation, growth and aggregation of calcium oxalate crystals in vitro26,,33 and directly inhibits the binding of calcium oxalate crystals to cultured renal epithelial cells.34 In addition, OPN directs calcium oxalate crystallization to the calcium oxalate dihydrate phase, which is significantly less adherent to renal tubular epithelial cells than the calcium oxalate monohydrate phase.35,,36

Are the inhibitory properties of OPN regarding calcium oxalate crystal formation and adhesion to tubular epithelial cells important in vivo? Experiments involving hyperoxaluric rats that develop renal calcium oxalate deposition do suggest an important in vivo role, since the upregulation of OPN mRNA and protein expression specifically colocalized with areas of renal calcium oxalate deposition.37,,38 However, definitive data regarding the protective role of OPN were lacking until recently. We have used OPN knockout mice to test the hypothesis that OPN is required for the effective prevention of renal calcium oxalate deposition in vivo. Previous studies have indicated that non‐manipulated OPN knockout mice exhibit normal kidney architecture and renal function.39–,41 We induced comparable hyperoxaluria in OPN knockout and control OPN wild‐type mice by the administration of 1% ethylene glycol, an oxalate precursor, in the drinking water for 4 weeks. Hyperoxaluric OPN knockout mice developed significant intratubular deposition of calcium oxalate whilst OPN wild‐type mice were completely unaffected (Figure 1).42 Crystal inclusions were comprised exclusively of calcium oxalate monohydrate, and typically affected the distal nephron and collecting duct, sites where the concentration of oxalate would be predicted to be highest. Interestingly, the absence of calcium oxalate deposition in hyperoxaluric OPN wild‐type mice was associated with significant upregulation of renal OPN expression by immunocytochemistry, thereby lending further support for a renoprotective role for OPN (Figure 2). The fact that OPN knockout mice do not spontaneously develop renal calcium oxalate deposition indicates that other urinary inhibitors can adequately compensate for the absence of OPN. However, these studies do indicate that urinary OPN is essential in adverse conditions predisposing to calcium oxalate deposition. OPN‐mediated protection is likely to result from effects at various sites along the pathway of crystal formation, growth and retention, as indicated schematically in Figure 3. Currently, the exact molecular mechanism whereby OPN exerts its various inhibitory effects is unclear, although it is known that phosphorylation of the OPN molecule is critically important.43,,44

We believe that these research findings will be of relevance to clinic patients with renal stone disease. Recent data indicates that defective inhibition of urinary crystallization is present in the majority of stone‐forming patients and may well be of critical importance in determining individual susceptibility to stone formation.45 It is therefore noteworthy that OPN is present in normal urine at levels that can effectively inhibit crystallization.46,,47 However, clinical studies to date are inconclusive regarding the relationship between OPN and renal stone disease. Some investigators have found reduced concentrations of OPN in urine from stone formers compared to normal individuals,48 while others have not.49 One study has described a single base mutation in the OPN gene which has a significantly higher incidence in patients with recurrent stone formation or familial nephrolithiasis.50 Therefore, while these results suggest that OPN may play an important role in protecting the kidney from stone formation, a cause and effect relationship in humans has not been established. However, this may not be of critical importance, since the future determination of the exact molecular mechanism of action of OPN may allow the development of novel water‐soluble drugs that will significantly increase the capacity of urine to inhibit pathological crystal formation and deposition. Such a therapeutic strategy may clearly be of great benefit to patients with recurrent renal stone disease.

Figure 1.

Hyperoxaluric OPN knockout mice exhibit marked intratubular deposition of calcium oxalate. Photomicrographs of renal tissue sections from an OPN knockout mouse administered 1% ethylene glycol for 4 weeks. Intratubular deposition of calcium oxalate (examples arrowed) is evident on both von Kossa (a) staining (which specifically detects calcium deposition) as well as staining with haematoxylin and eosin (b).

Figure 2.

Hyperoxaluric OPN wild‐type mice exhibit marked upregulation of tubular OPN expression. Representative photomicrographs of renal tissue sections immunostained for OPN expression. (a) Control untreated OPN wild‐type mouse. (b) OPN wild‐type mouse following 4 weeks administration of 1% ethylene glycol in the drinking water.

Figure 3.

Proposed schema indicating the potential sites of OPN‐mediated modulation (inhibition [−] or stimulation [+]) of calcium oxalate monohydrate (COM) or calcium oxalate dihydrate (COD) crystal formation, growth, aggregation and subsequent deposition within the kidney.

OPN and inflammation

OPN plays an important role during both acute and chronic inflammation where it may be expressed by resident epithelial, endothelial and smooth muscle cells as well as infiltrating macrophages and T cells. OPN may exert both pro‐inflammatory and anti‐inflammatory actions, with the net effect of OPN depending upon the nature of the biological scenario.

Leucocyte recruitment and function

OPN is involved in the recruitment and retention of macrophages and T cells to inflamed sites. Dermal injection of purified OPN induces an inflammatory macrophage infiltrate, while macrophage recruitment in response to N‐formyl‐met‐leu‐phe, a potent macrophage chemotactic peptide, is significantly attenuated by administration of neutralizing anti‐OPN antibodies.51 OPN may not only be expressed by T cells but also induces T‐cell chemotaxis and costimulates T‐cell proliferation.3 In addition, OPN fragments generated by thrombin cleavage, and which are likely to be present at sites of inflammation, are more effective supporters of T‐cell adhesion than is the native molecule.3 Furthermore, the administration of neutralizing anti‐OPN antibodies to rats with glomerulonephritis significantly inhibited T‐cell accumulation, T‐cell activation and the delayed type hypersensitivity response.52

OPN expression correlates with macrophage infiltration in various animal models of acute and chronic renal injury such as glomerulonephritis, hypertensive glomerulosclerosis and cyclosporine nephropathy,53–,55 with similar findings being reported in human crescentic glomerulonephritis.56 In addition, upregulation of OPN expression precedes the macrophage infiltrate present in injured kidneys.57 An important functional role for OPN in renal disease is supported by studies involving the administration of function blocking anti‐OPN antibodies in rat models of acute glomerulonephritis. The administration of neutralizing anti‐OPN antibody markedly reduced glomerular and tubulointerstitial macrophage infiltration and T‐cell accumulation, as well as various histological and functional parameters of renal injury.52,,53 Similarly, the administration of intravenous OPN antisense oligodeoxynucleotide to rats with glomerulonephritis resulted in decreased OPN expression by tubular epithelial cells and diminished interstitial macrophage accumulation.58 Studies involving OPN knockout mice have produced somewhat conflicting results, since the absence of OPN resulted in diminished macrophage infiltration following ureteric obstruction but had no effect upon macrophage infiltration in a model of glomerulonephritis.40,,59

OPN expression is pronounced in granulomatous lesions, whether secondary to infections such as tuberculosis or conditions such as sarcoidosis or silicosis.60 OPN knockout mice are more susceptible to infection with mycobacteria.61 Similarly, OPN expression in humans also contributes to resistance to mycobacterial infection and disease.62 It is now established that OPN is critically involved in the initiation of cell‐mediated immune responses, and stimulates Th1 cytokine expression and inhibits Th2 cytokine expression.63,,64 OPN knockout mice exhibit a defective Th1 response to infection with either Herpes simplex virus or Listeria monocytogenes; this defective response is associated with diminished macrophage production of interleukin‐12 (IL‐12) and interferon‐γ and increased production of interleukin‐10 (IL‐10).65 Furthermore, recombinant OPN directly modulates macrophage production of these cytokines in vitro, with exogenous OPN stimulating macrophage IL‐12 production and inhibiting IL‐10 production.65 An additional level of complexity is that the differential effect of OPN upon macrophage cytokine production is mediated by ligation of different cell surface receptors by OPN, since stimulation of IL‐12 and inhibition of IL‐10 result from β3 integrin and CD44 ligation, respectively.65

OPN modulates activation of T cells as well as cytokine production,66,,67 while gene expression profiling studies indicate that OPN is selectively upregulated in Th1 cells.68 Lastly, the elevation in serum OPN levels evident in models of autoimmune disease coupled with the fact that OPN may induce polyclonal activation of B cells and associated immunoglobulin production implicates OPN in the pathogenesis of autoimmunity.67,69–,71

OPN and cell survival

OPN is a cell survival factor and may protect cells from undergoing apoptosis. For example, treatment of serum‐starved NRK52E renal epithelial cells with neutralizing anti‐OPN antibody markedly enhances the level of apoptosis.40 Furthermore, despite diminished macrophage infiltration and interstitial fibrosis, the obstructed kidneys of OPN knockout mice exhibit increased levels of tubular cell apoptosis compared to wild‐type mice, suggesting that OPN is capable of providing survival signals to tubular epithelial cells in vivo.40 The binding of OPN to the αvβ3 integrin of endothelial cells activates the pro‐survival transcription factor NFκB and protects endothelial cells from undergoing apoptosis.72 There are also reports of OPN exerting pro‐survival effects in other cell types, including vascular smooth muscle cells and haemopoietic cells.73,,74.

OPN and the regulation of inducible nitric oxide synthase

Recent data indicate that other actions of OPN may also exert important protective effects in tissues. OPN inhibits the expression of inducible nitric oxide synthase (iNOS) in both macrophages and primary renal tubular epithelial cells.75,,76 Since iNOS‐derived nitric oxide (NO) is a well‐documented mediator of tissue injury, the OPN‐mediated inhibition of iNOS would be predicted to exert beneficial effects upon surrounding tissues. Indeed, OPN knockout mice fare poorly following renal ischaemia‐reperfusion injury, a model documented to involve NO.39 The absence of OPN is associated with significantly worse renal dysfunction, increased tubular expression of iNOS and nitrotyrosine residues and more severe histological injury.39 Interestingly, since the capacity of cultured human tubular epithelial cells to upregulate OPN expression diminishes with increasing age of the kidney from which they were derived, one can speculate that this may be involved in the increased susceptibility of aged kidneys to ischaemia‐reperfusion injury.77

NO is an important facet of macrophage cytotoxicity and is involved in macrophage killing of tumour cells78 and glomerular mesangial cells in vitro.79 Recently, we have investigated the mechanism underlying the induction of apoptosis in MDCK renal tubular epithelial cells during coculture with macrophages activated with lipopolysaccharide and interferon‐γ. Interestingly, MDCK cell death is also dependent upon macrophage generation of NO, since macrophage‐mediated MDCK cell apoptosis is significantly blocked by the iNOS inhibitor L‐NAME. In addition, activated macrophages from iNOS wild‐type mice induce a 3.5‐fold higher level of MDCK cell apoptosis compared to activated macrophages from iNOS knockout mice (unpublished data, Figure 4). Since inflammatory macrophages are implicated in the induction of tubular epithelial cell death in vivo,80 it can be appreciated that OPN‐mediated inhibition of iNOS may protect resident renal cells from macrophage cytotoxicity during renal inflammation. Interestingly, NO induces OPN production by macrophages,7 thereby providing a negative feedback loop predicted to limit NO‐mediated tissue injury. This raises the possibility that the absence of OPN‐mediated iNOS inhibition may be at least partly responsible for the increased level of tubular epithelial cell death evident in the obstructed kidneys of OPN knockout mice.40

Figure 4.

Primary murine bone marrow derived macrophages activated with lipopolysaccharide and interferon‐γ induce apoptosis in MDCK renal tubular epithelial cells. (a) Immunofluorescent microscopy of a coculture of unlabelled activated macrophages and MDCK cells specifically labelled with a fluorescent green dye. As a consequence, macrophages are not visualized, while apoptotic MDCK cells appear rounded (examples arrowed) and exhibit bright cytoplasmic condensation. (b) High power of a green condensed apoptotic MDCK cell in the coculture. (c) Hoechst staining demonstrates the pyknotic nucleus of the apoptotic MDCK cell (green arrow) together with the normal nucleus of an adjacent macrophage (red arrow). (d) The merged image demonstrates the unlabelled macrophage in close contact to the apoptotic target MDCK cell.

Other biological functions of OPN

OPN is important in tissue repair. OPN knockout mice exhibit defective repair of incisional skin wounds with abnormal collagen fibrillogenesis,81 although the mechanism underlying this defect is unclear at present. OPN is also associated with cellular regeneration. For example, OPN expression was localized to PCNA‐positive regenerating tubules in a rat model of acute renal failure induced by the administration of the nephrotoxic drug gentamicin.82 In addition, OPN is involved in the proliferation of vascular smooth muscle cells and glomerular mesangial cells induced by hypoxia.83,,84

OPN expression is also associated with tissue scarring and fibrosis, with OPN upregulation evident in models of renal disease exhibiting extensive fibrotic changes such as progressive glomerulonephritis and the remnant kidney model.85,,86 The correlation of OPN with tissue fibrosis is most likely to be secondary to the concomitant macrophage infiltrate and associated production of TGFβ.40 However, recent data suggest that OPN may play a more direct role in the development of tissue fibrosis, since OPN is chemotactic for fibroblasts, potentiates growth factor mediated fibroblast proliferation and modulates fibroblast secretion of metalloproteinases.87–,89

OPN is important in tumour cell biology. Indeed, OPN was originally isolated from epithelial and fibroblastic cells that had undergone malignant transformation.90 Treatment of neoplastic cell lines with OPN antisense oligodeoxynucleotide results in diminished malignant potential in vitro and in vivo.91 In a study of patients with breast carcinoma, increased expression of OPN by tumour tissue was associated with a significantly reduced disease‐free survival and overall survival.92 It may well be significant that the binding of OPN to cells may stimulate metalloproteinase production and thereby facilitate cell migration through extracellular matrix,93,,94 although it should be noted that the opposite effect may occur in some cell types.95


OPN is an impressively multifunctional molecule (Figure 5). Although non‐manipulated OPN knockout mice exhibit a normal phenotype, they exhibit markedly abnormal responses to a variety of injurious stimuli, revealing much about the biological role of OPN in vivo. Further work is required to determine the full role of OPN in human disease. However, it is envisaged that the elucidation of the molecular mechanisms underlying the myriad actions of OPN will allow the development of novel therapeutic agents designed to modulate the action of OPN for patient benefit in a broad range of diseases, including renal stone disease, inflammation and cancer.

Figure 5.

Simplified schema indicating various important biological functions of OPN.


JH was a Wellcome Trust Advanced Fellow and is now in receipt of a Wellcome Trust Senior Research Fellowship in Clinical Science. RJJ is supported by NIH grants DK‐43422, DK‐47659 and DK‐52121. JAW was primarily supported through a Research Career Development Grant from the Department of Veterans Affairs (RCD9305) and received additional support from the Medical College of Wisconsin.


  • Address correspondence to Dr J. Hughes, Phagocyte Laboratory, MRC Centre for Inflammation Research, University of Edinburgh Medical School, Teviot Place, Edinburgh, UK. e‐mail: jeremy.hughesed.ac.uk


View Abstract