_________________________________________________________________________http://www.paper.edu.cn BBRC Biochemical and Biophysical Research Communications 297 (2002) 38–45 www.academicpress.com Inhibition of MEPE cleavage by Phex Rong Guo,a Peter S.N. Rowe,b Shiguang Liu,a Leigh G. Simpson,a Zhou-Sheng Xiao,a and L. Darryl Quarlesa,* a Department of Medicine, The Center for Bone and Mineral Disorders, Duke University Medical Center, Box 3036, Durham, NC 27710, USA b The University of Texas Health Science Center at San Antonio, Institute for Drug Development, Molecular Therapies, Endocrinology, Texas Research Park, 14960 Omicron Drive, 78245 San Antonio, TX, USA Received 7 August 2002 Abstract X-linked hypophosphatemia (XLH) and the Hyp-mouse disease homolog are caused by inactivating mutations of Phex which results in the local accumulation of an unknown autocrine/paracrine factor in bone that inhibits mineralization of extracellular matrix. In these studies, we evaluated whether the matrix phosphoglycoprotein MEPE, which is increased in calvaria from Hyp mice, is a substrate for Phex. Using recombinant full-length Phex (rPhexWT) produced in Sf9 cells, we failed to observe Phex- dependent hydrolysis of recombinant human MEPE (rMEPE). Rather, we found that rPhex-WT inhibited cleavage of rMEPE by endogenous cathepsin-like enzyme activity present in Sf9 membrane. Sf9 membranes as well as purified cathepsin B cleaved MEPE into two major fragments of 50 and 42 kDa. rPhexWT protein in Sf9 membrane fractions, co-incubation of rPhexWT and cathepsin B, and pre-treatment of Sf9 membranes with leupeptin prevented the hydrolysis of MEPE in vitro. The C-terminal domain of Phex was required for inhibition of MEPE cleavage, since the C-terminal deletion mutant rPhex (1–433) [rPhex30M] failed to inhibit Sf9-dependent metabolism of MEPE. Phex-dependent inhibition of MEPE degradation, however, did not require Phex enzymatic activity, since EDTA, an inhibitor of rPhex, failed to block rPhexWT inhibition of MEPE cleavage by Sf9 membranes. Since we were unable to identify interactions of Phex with MEPE or actions of Phex to metabolize cathepsin B, Phex may be acting to interfere with the actions of other enzymes that degrade extracellular matrix proteins. Although the molecular mechanism and biological relevance of non-enzymatic actions of Phex need to be established, these findings indicate that MEPE may be involved in the pathogenesis defective mineralization due to Phex deficiency in XLH and the Hyp-mouse. Ó 2002 Elsevier Science (USA). All rights reserved. Keywords: M13 endopeptidase; Mineralization; Bone; XLH; Hyp Phex is a zinc-dependent type II cell-surface mem- Related endopeptidases have multiple substrates that brane metalloprotease that is involved in regulating are often co-expressed with the enzyme in different tis- phosphate and mineral homeostasis [1–4]. Inactivating sues [10]. mutations of Phex lead to defective calcification of bone Parabiosis experiments [11] and kidney cross-trans- and cartilage and renal phosphate wasting in the clinical plantations [12] in the Hyp-mouse homolog of XLH disorder X-linked hypophosphatemic rickets (XLH) indicated that defective Phex function leads to the ac- [3,5,6] and the Hyp and Gy mice disease homologs [7–9]. cumulation of a circulating phosphaturic factor, called The pathogenesis of phosphate wasting and impaired phosphatonin [6]. Recent investigations have identified mineralization in these disorders remain poorly under- FGF23, a phosphaturic factor that is expressed by tu- stood, but may represent the accumulation of distinct mors causing oncogenic hypophosphatemic osteomala- Phex substrates or the accumulation of factors as an cia (OHO) [13,14], and mutated in autosomal dominant indirect consequence of inactivating Phex mutations. hypophosphatemic rickets (ADH) [15], as a possible Phex substrate and candidate for phosphatonin [16,17]. Increased circulating FGF23 levels, however, do not * Corresponding author. Fax: 1-919-684-4476. fully explain the impaired mineralization associated with E-mail address: [email protected] (L. Darryl Quarles). inactivation of Phex. First, inhibition of mineralization 0006-291X/02/$ - see front matter Ó 2002 Elsevier Science (USA). All rights reserved. PII: S0006-291X(02)02125-3 __________________________________________________________________________中国科技论文在线 http:www.paper.edu.cn R. Guo et al. / Biochemical and Biophysical Research Communications 297 (2002) 38–45 39 is observed ex vivo in Phex deficient osteoblast cultures modulin binding peptide (CBP), N-terminal tagged human MEPE [18], which lack FGF23 [19]. Moreover, the mineral- protein comprising amino acids 95–525, and anti-human MEPE an- ization defect in Hyp-derived osteoblasts is not totally tisera were generated as previously described [24]. Cathepsins B and D were obtained from CalBiochem (San Diego, CA, USA). Leupeptin, dependent on hypophosphatemia, since partial correc- aprotinin, and pefabloc, a protease inhibitor cocktail containing all tion of defective mineralization but not hypophosphat- three of these inhibitors were obtained from Roche Molecular Bio- emia can be achieved by the targeted expression of Phex chemicals (Indianapolis, IN, USA). The anti-V5 monoclonal antibody in bone of Hyp mice [19]. Rather, the impaired miner- was obtained from Invitrogen (Carlsbad, CA, USA). Z-Ala-Ala-Leu- alization appears to result from the autocrine/paracrine p-nitroanilide (ZAAL-pNA) was purchased from Bachem Biosciences (King of Prussia, PA, USA). Leucine aminopeptidase was purchased secretion by osteoblasts of a mineralization inhibitory from Sigma (St. Louis, MO, USA). factor [8,20–22]. Recently, several peptides have been Assessing Phex enzyme activity. We assessed Phex activity by identified that are cleaved by Phex in vitro, including modifications of previously described methods [18]. We used mem- ZAAL-pNA, leu[enkephalin], and human parathyroid brane fractions from the Sf9 cells (50 lg) expressing vector or the V5- hormone related peptide [18,23], but none of these His epitope tagged Phex constructs. Protein samples were incubated with 50 lM ZAAL-pNA in 100 ll of 100 mM MES (pH 6.5) for 1 h at substrates account for the mineralization defects asso- 37 °C. After completion of the initial incubation, the reaction mixture ciated with inactivating mutations of Phex. was further incubated with 0.4 mU leucine aminopeptidase for 20 min Another candidate for the putative bone derived at 37 °C. The reaction was stopped by the addition of 100 mM EDTA Phex substrate is the matrix extracellular phosphogly- and the absorbency was measured at 405 nm after centrifugation. In coprotein (called MEPE or OF45) [24,25]. MEPE, which some studies, membrane fractions were preincubated with 100 mM EDTA for 30 min before the addition of ZAAL-pNA. was originally isolated and cloned from mesenchymal Assessing MEPE cleavage. Purified recombinant human MEPE tumors derived from OHO patients, is expressed in bone (100 gg) was incubated with various enzymes preparations, including and has molecular similarities with several extracellular 30 lg solubilized membranes derived from Sf9 cell infected with either matrix RGD-containing phosphoglycoproteins involved wild-type baculovirus, rPhexWT or rPhex30M. In addition, rMEPE in mineralization, including dentin sialophosphoprotein was incubated with 5 lg cathepsin B or 0.5 U cathepsin D. For all studies, the reaction mixture contained 50 mM NaCl, 20 mM Tris pH (DSPP), osteopontin (SPP1), and dentin matrix protein- 7.4 for 1 h at 37 °C. The reactions were stopped by adding protein 1 (DMP1). Preliminary studies in MEPE deficient mice loading buffer (Invitrogen Carlsbad, CA, USA). In some studies, either indicate that the absence of MEPE is associated with leupeptin (1 lM), aprotinin (0.3 lM), and pefabloc (4 lM), a protein- accelerated mineralization [26]. Conversely, MEPE ase inhibitor cocktail containing all three of these inhibitors, or transcripts are increased in poorly mineralizing bone 200 mM EDTA pre-incubated with 100 gg CBP-tagged MEPE for 20 min at 37 °C, followed by addition of the various enzymes. All derived from Hyp-mice [27]. These associations suggest the reactions were analyzed by Western blot using anti-MEPE that the Phex gene product might control bone matrix antisera. mineralization by directly or indirectly regulating the SDS–PAGE and Western blot analyses. Immunoblot analysis was metabolism of MEPE. To date, no studies have exam- carried out by modifications of previously described methods [28]. The ined whether MEPE is a substrate for Phex. specified amount of membrane proteins was dissolved in SDS–gel loading buffer (Invitrogen Carlsbad, CA, USA). Separated proteins In the present investigation, we investigated whether were transferred to a nitrocellulose membrane (0.45 lm, Bio-Rad, MEPE is a substrate for Phex. Using recombinant epi- Chicago, IL USA) over a 30 min period at 2.5 mA/cm2 at room tem- tope tagged MEPE and Phex proteins, we were unable perature using a Semi-dry blotting system (Millipore, Chicago, IL to show that MEPE is a substrate for Phex; however, we USA). Immunoblotting was performed using anti-MEPE antisera di- found that Phex inhibited the proteolytic cleavage of luted 1:2000 in TBST and 1% bovine albumin (BSA) or using antiV5 monoclonal antibody (Invitrogen Carlsbad, CA, USA) for 1 h at room MEPE by cathepsin B-like activity present in Sf9 cell temperature. The blots were washed with TBST for 60 min and incu- membranes. These findings support