Expression and post-translational modification of purple acid phosphatase in insect and mammalian cells

Sammanfattning: Purple acid phosphatase (PAP) is an acidic metallohydrolase that contain a binuclear iron center in its active site. In mammals, this enzyme is also referred to as tartrate-resistant acid phosphatase (TRAPs), uteroferrin or type 5 acid phosphatase. PAP is abundantly expressed in certain macrophages including bone-resorbing osteoclasts and mature dendritic cells as well as in lining epithelia and ganglion cells. The aim of the present study was to investigate the role of different posttranslational modifications on the catalytic properties and distribution of mammalian PAP. PAP contains two potential N-glycosylation sites at Asn97 and Asn128 and is synthesized as a 35-37 kDa monomeric protein that can undergo proteolytic processing of a repressive loop domain to a twosubunit form associated with activation of the enzyme. This study established that rat recombinant PAP expressed from Baculovirusinfected Sf9 insect cells is predominantly secreted as the monomeric form and sitedirected mutagenesis confirmed the presence of two Nlinked oligosaccharides. In contrast, two crystal structures derived from rat PAP expressed by Sf9 cells revealed only one carbohydrate chain linked to Asn97, which highlights possible limitations by protein crystallography to accurately assign oligosaccharides substituted to the protein core. Moreover, the degree of N-glycosylation affected catalytic properties of the enzyme, particularly the monomeric form. It is concluded that the latency of PAP is controlled by an N-glycosylation and depends on the orientation of the repressive loop domain and that differential glycosylation of PAP could provide a physiological mechanism to regulate PAP activity. The presence and effect of N-glycosylation and proteolytic processing on PAP activity and intracellular locations was next assessed in mammalian CHO-K1 cells stably over-expressing rat PAP. PAP from CHO-K1 cells was less extensively N-glycosylated and occurred mainly as the proteolytically processed form. Monomeric PAP containing a single oligosaccharide from CHO-K1 cells exhibited a similar level of basal catalytic activity but could not be activated to the same extent following in vitro digestion with cysteine proteinases compared to the Nglycosylation-deficient monomeric PAP mutants from Sf9 insect cells, suggesting that additional unidentified post-translational modifications in addition to N-glycosylation and proteolytic processing may control PAP activity. Treatment of PAP-transfected CHO-K1 cells expressing cathepsin K with the cysteine proteinase inhibitor E-64 slightly increased the secretion of monomeric PAP, suggesting that cathepsin K is not dominant in the proteolytic processing of PAP and that other proteinases must be implicated in the proteolytic processing of PAP in CHO-K1 cells. Comparison of the PAPs expressed by the two expression systems utilized in this study with native PAPs indicate that Baculovirus-infected Sf9 insect cells are suitable for large-scale production of monomeric PAP for studies on enzyme structure and catalysis, while the CHO-K1 expression system is better suited for investigations related to intracellular transport, proteolytic processing and secretion of the enzyme.