The tartrate-resistant purple acid phosphatase of bone osteoclasts—a protein phosphatase with multivalent substrate specificity and regulation

Abstract

Tartate-resistant purple acid phosphatases (TRAP/PAP) are iron-containing, cationic glycoproteins with molecular weights of around 35 kDa and a monomeric peptide structure (Vincent and Averill, 1990; Andersson et al, 1992). When isolated and sufficiently concentrated, these enzymes exhibit a characteristic purple color. The purple acid phosphatases of spleen and uterus have been extensively characterized, in particular with respect to the unique active site containing two iron atoms, which are essential for catalytic activity (Vincent and Averill 1990). This binuclear iron center can exist in two interconvertible states: in the purple enzyme as a diferric pair which is catalytically inactive or in the reduced, catalytically active pink species where the di-iron cluster exist as a mixed-valent Fe(II)-Fe(III) couple (Wang et al. 1992). Low levels of TRAP/PAP enzymes can be detected in most tissues, whereas much higher expression levels are detected in bones of growing animals (Ek-Rylander et al. 1991a). In bones, the tartrate-resistant purple acid phosphatase is abundantly and selectively expressed in actively resorbing osteoclasts (Andersson and Marks, 1989), which is the reason for the wide-spread use of the TRAP/PAP as a histochemical marker for the identification of osteoclasts. Moreover, TRAP/PAP has been tried as an indicator for bone resorption in pathological conditions involving increased bone turnover (Kraenzlin et al. 1990). Despite the fact that the TRAP/PAP enzyme has been recognized as an osteoclast marker enzyme for more than a decade (Minkin, 1982), until recently almost nothing was known concerning the mechanistic role of TRAP/PAP in bone metabolism. Studies employing enzyme histochemical detection of the TRAP/PAP activity as well as immunohisto-chemical detection of the protein at the ultrastructural level have shown a predominant localization of the enzyme to the resorption vacuole (Andersson et al. 1992; Reinholt et al. 1990a), suggesting that the enzyme is secreted by the osteoclast to the site of active resorption. Moreover, inclusion of an antibody against TRAP/PAP blocks resorption in calvarial explants (Zaidi et al. 1989), suggesting that the enzyme has a promotive role in the resorption process. TRAP/PAP enzymes catalyze the hydrolysis of phosphate monoesters (Vincent et al. 1992), and among the known substrates for these enzymes are certain acidic phosphoserine-containing proteins, e.g casein and phosvitin (Andersson et al. 1992). In addition to these extraskeletal phosphoproteins, it was recently demonstrated (Ek-Rylander et al. 1994) that purified bone TRAP/PAP can dephosphorylate the bone matrix phosphoproteins osteopontin (OPN), known as an anchor for the binding of osteoclasts via integrin receptors to bone (Heinegård et al. 1995), and bone sialoprotein. Dephosphorylation of OPN was shown to alter the function of the protein, since the modification abolished the capacity of OPN to promote osteoclast attachment in vitro (Ek-Rylander et al. 1994). This observation suggested that one potential physiological function of TRAP/PAP could be to regulate osteoclast attachment to bone. These initial observations on the substrate specificity of the osteoclast TRAP/PAP enzyme has been extended in the present study, where another abundant phosphorylated protein present in the bone matrix, i.e. osteonectin, as well as a phosphotyrosine-containing gastrin peptide were identified as substrates for the osteoclast TRAP/PAP enzyme. Thus, these studies indicate that the TRAP/PAP enzymes, in contrast to most other protein phosphatases with regulatory functions, exhibit broad specificity both with regard to individual phosphoserine-proteins as well as to preferred phosphoamino acid residues.