Disruption of structural and functional integrity of a 2 -macroglobulin by cathepsin E Mitsue Shibata 1 , Hideaki Sakai 1 , Eiko Sakai 1 , Kuniaki Okamoto 1 , Kazuhisa Nishishita 1 , Yoshiyuki Yasuda 2 , Yuzo Kato 1 and Kenji Yamamoto 2 1 Department of Pharmacology, Nagasaki University School of Dentistry, Japan; 2 Department of Pharmacology, Graduate School of Dental Science, Kyushu University, Fukuoka, Japan a 2 -Macroglobulin (a2M) is an abundant glycoprotein with the intrinsic capacity for capturing diverse proteins for rapid delivery into cells. After internalization by the receptor- mediated endocytosis, a2M-protein complexes were rapidly degraded in the endolysosome system. Although this is an important pathway for clearance of both a2M and biological targets, little is known about the nature of a2M degradation in the endolysosome system. To investigate the possible involvement of intracellular aspartic proteinases in the dis- ruption of structural and functional integrity of a2M in the endolysosome system, we examined the capacity of a2M for interacting with cathepsin E and cathepsin D under acidic conditions and the nature of its degradation. a2M was effi- ciently associated with cathepsin E under acidic conditions to form noncovalent complexes and rapidly degraded through the generation of three major proteins with apparent molecular masses of 90, 85 and 30 kDa. Parallel with this reaction, a2M resulted in the rapid loss of its antiproteolytic activity. Analysis of the N-terminal amino-acid sequences of these proteins revealed that a2M was selectively cleaved at the Phe811-Leu812 bond in about 100mer downstream of the bait region. In contrast, little change was observed for a2M treated by cathepsin D under the same conditions. Together, the synthetic SPAFLA peptide corresponding to the Ser808– Ala813 sequence of human a2M, which contains the cath- epsin E-cleavage site, was selectively cleaved by cathepsin E, but not cathepsin D. These results suggest the possible involvement of cathepsin Ein disruption of the structural and functional integrity of a2M in the endolysosome system. Keywords: a 2 -macroglobulin; aspartic proteinase; cathepsin D; cathepsin E; endolysosome system. 1 a 2 -Macroglobulin (a2M) is an abundant plasma glycopro- tein composed of four identical subunits of M r  185 kDa [1]. a2M inhibits the activity of all classes of endopeptidases from both endogenous and foreign sources. The proteinases cleave an accessible region of the polypeptide chain of a2M, the bait region, thereby leading to the activation of internal thiol esters and the subsequent conformational change that entraps the responsible proteinase [1]. Then, the a2M- proteinase complexes are recognized by the low-density lipoprotein receptor-related protein (LRP)/CD91 on the surface of different cell types such as hepatocytes [2], fibroblast-like cells [3], and monocytes/macrophages [4] and become destined to rapid clearance and degradation in the endolysosome system [3,5]. Recent studies have also dem- onstrated that a2M has other important intrinsic capacity for capturing diverse molecules, including cytokines [6–9], growth factors [10–13], hormones [14], and soluble b-amy- loid peptide [15], for rapid delivery into cells and degrada- tion. The association of these molecules with a2M induces neither cleavage of the a2M peptide bond [16,17] nor the a2M conformational change [18,19]. The nature of this association is therefore distinct from the trapping mechan- ism for proteinases. However, these molecules bound to a2M are similarly targeted to cells expressing the a2M signaling receptor and become destined to rapid degrada- tion in the endolysosome system. Therefore, a2M also plays an important part in the clearance of these molecules or regulates their biological activity. Meanwhile, a2M has also been shown to mediate immune responses through the delivery of foreign antigens to macrophages [20]. It is thus considered that a2M is involved in a wide range of physiological processes with the intrinsic capacity for capturing diverse target proteins for rapid delivery into cells and efficient degradation in the endolysosome system. However, the nature of the endocytosed a2M degradation in the endolysosome system is poorly understood. A phy- siological inactivator of a2M has not yet been identified. Cathepsins E and D are analogous endolysosomal aspartic proteinases in mammalian cells [21]. Cathepsin E represents a major portion of the proteolytic activity in the endosomal compartment in certain cell types such as macrophages and microglia [22–24], gastric cells [25], and antigen-presenting B cell lymphoblasts [26]. Cathepsin E is also associated with the plasma membrane of various cell types such as erythrocytes [27,28], osteoclasts [29], gastric parietal cells [30], renal proximal tubule cells [30], and hepatic cells [30]. On the other hand, cathepsin D is widely distributed in almost all the mammalian cells as the most abundant Correspondence to K. Yamamoto, Department of Pharmacology, Graduate School of Dental Science, Kyushu University, Fukuoka 812–8582, Japan. Fax: + 81 92 6426342, Tel.: + 81 92 6426337, E-mail: kyama@dent.kyushu-u.ac.jp Abbreviations: a2M, a 2 -macroglobulin; Hb, hemoglobin; LRP, low-density lipoprotein receptor-related protein. (Received 30 November 2002, accepted 24 January 2003) Eur. J. Biochem. 270, 1189–1198 (2003) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03479.x endolysosomal proteinase. More recently, it has been demonstrated that cathepsin E has a potential for foreign antigen processing for presentation by class II major histocompatibility complex [24] and possible regulation of activities of substance P and related tachykinins [31], whereas cathepsin D is indispensable for protection of the onset and development of a certain type of neuronal ceroid lipofucinosis [32] and proteolysis of proteins regulating cell growth and tissue homeostasis [33]. However, there is no unequivocal evidence for the participation of these two proteinases in the degradation of a2M. The inherent problem is that both cathepsins E and D were essentially inactive at around neutral pH where a2M is very stable, whereas they are most active at around pH 4.0 where a2M is unstable. Although previous work has suggested that cathepsin E is inhibited by a2M at pH 6.2 [34] and 5.5 [35], the ability of a2M to interact with cathepsins E and D below about pH 5.0 is still uncertain. In this report, we demon- strate that a2M is selectively associated with cathepsin E below pH 5.0 and rapidly cleaved it at a specific site distinct from the bait region, thereby losing its structural and functional integrity. Materials and methods Materials Trypsin (Type XIII) and human a2M were purchased from Sigma-Aldrich. Bovine liver cDNA was purchased from Clontech. Cathepsin E was purified from rat spleen [36] and human erythrocytes [27] as previously described. Cathepsin D was purified from rat [37] and bovine spleen [38] as described. The fluorogenic decapeptide substrate MOCAc- Gly-Lys-Pro-Ile-Ile-Phe-Phe-Arg-Leu-Lys(Dnp)- D -Arg-NH 2 was synthesized as described previously [39]. Antibodies specific for rat cathepsin E and cathepsin D were raised in rabbits and purified by affinity chromatography as des- cribed previously [28]. Antiserum against bovine a2M was purchased from Yagai Research Center (Yamagata, Japan). All other chemicals were of reagent grade and were purchased from various commercial sources. Purification of a2M from bovine serum a2M was purified from bovine serum as described previ- ously [40], with a slight modification. Briefly, the serum was applied to a Ni/nitrilotriacetic acid column (QIAGEN, 1.5 cm · 3.5 cm) and eluted with 50 m M sodium acetate buffer, pH 5.0, containing 50 m M NaCl. The eluate was applied to a Mono Q column equilibrated with 33 m M sodium phosphate buffer, pH 6.0. The column was washed with the same buffer and eluted with a linear gradient of NaCl (50–500 m M ) in the buffer. Fractions containing a2M were determined by assaying the inhibitory activity against rat spleen extract at pH 3.8 using hemoglobin as a substrate. Fractions containing a2M, whose activity was determined by inhibition of the hemoglobin (Hb)-hydrolyzing activity of rat spleen extracts at pH 3.8, were pooled and concen- trated and then subjected to gel filtration on Superose 6. The a2M fractions were pooled and subjected to the second Mono Q anion exchange chromatography. The pooled a2M fractions were concentrated and dialyzed against 20 m M Hepes buffer, pH 7.2, containing 140 m M NaCl. Assays The proteinases activity of cathepsins E and D were measured in 0.1 M sodium acetate buffer, pH 3.8, using 1.5% acid-denatured Hb [27] or MOCAc-Gly-Lys-Pro-Ile- Ile-Phe-Phe-Arg-Leu-Lys(Dnp)- D -Arg-NH 2 [39] as des- cribed previously. Sequencing of bovine a2M cDNA The degenerate nucleotide sequences of the primers for polymerase chain reaction (PCR) were designed based on the conserved regions of the mammalian a2M. Bovine liver cDNA was used as the template for PCR. The PCR products were isolated and subcloned into pBluescript II SK vector (Stratagene) and sequenced by an ABI automatic DNA sequencer model 310 (Perkin Elmer-Applied Bio- systems). Twenty-five clones were highly homologous to a2M. Gel electrophoresis and N-terminal sequence analysis SDS/PAGE and immunoblotting were carried out follow- ing the procedure as described previously [28]. For the N-terminal amino-acid sequencing, the purified bovine a2M and the cathepsin E-digested protein were separated by SDS/PAGE (8% gel) under reducing conditions and then transferred onto poly(vinylidene difluoride) membranes and stained with Coomassie blue R-250. The stained bands were excised and the adsorbed proteins were subjected to an automatic protein/peptide sequencer (Applied Biosystems Model 477A). Interaction of a2M with cathepsin E Purified bovine a2M (140 pmol) was incubated with or without cathepsin E (70 pmol) at pH 3.8 at appropriate time intervals. After neutralization, the incubation mixtures were applied to gel filtration on Superose 6 equilibrated with 10 m M sodium phosphate buffer, pH 6.0, containing 150 m M NaCl. Fractions were analyzed by SDS/PAGE, immunoblotting and cathepsin E activity. Analysis of synthetic peptides treated with cathepsin E and cathepsin D Synthetic peptides corresponding to the cleavage site region of a2M by cathepsin E were designed and custom-synthes- ized at the Peptide Institute (Osaka). Each peptide (20 n M ) was incubated with or without either cathepsin E (20 p M )or cathepsin D (20 p M )in0.1 M sodium acetate buffer, pH 3.8, and then subjected to reversed phase high-performance liquid chromatography (RP-HPLC) 2 with a lBondapak C18 column (3.9 mm · 300 mm) (Waters). The column was eluted with a gradient of acetonitorile (0–60% in 30min) in 0.1% trifluoroacetic acid at a flow rate of 1.0 mLÆmin )1 . Each peak fraction was pooled and the amino-acid sequence was analyzed by an Applied Biosystems automated derivatizer-analyzer (model 477A/120). 1190 M. Shibata et al. (Eur. J. Biochem. 270) Ó FEBS 2003 Results Effect of a2M on the hemoglobin hydrolyzing activity of cathepsin E and cathepsin D To assess the inhibitory capacity of a2M from various sources for cathepsin D and cathepsin E, we first purified bovine a2M. The final preparation gave a single protein band with an apparent molecular mass of 170 kDa when analyzed by SDS/PAGE under reducing conditions (data not shown). The N-terminal amino-acid sequence of this protein was found to be AVDGKEPQYM, which was identical to the N-terminal sequence of the intact a2M as reported previously [41]. Addition of bovine a2M,aswellas human a2M, to cathepsin E purified from human and rat sources (0.01 ng each) at pH 3.8 resulted in the significant decrease of Hb-hydrolyzing activity in a dose-dependent manner though the rate and extent of inhibition by bovine a2M were slightly but significantly higher than those by human a2M (Fig. 1). In contrast, little change was observed for the Hb-hydrolyzing activity of cathepsin D purified from rat and bovine spleen when a2M was added under the same conditions. This selective inhibition was further substan- tiated by experiments using the cell extract of rat spleen, in which cathepsin E and cathepsin D comprise 55 and 45% of the total Hb-hydrolyzing activity [25]. The cell extract treated with discriminative antibodies specific for cathepsin D to remove this protein showed a significant decrease in the Hb-hydrolyzing activity by addition of either human or bovine a2M, whereas the extract devoid of cathepsin E by immunoprecipitation with specific antibodies to cathepsin E showed no significant change in the Hb-hydrolyzing activity by each a2M (not shown). The results indicate that the selective reduction of cathepsin E activity by a2M occurs through the specific interaction between a2M and cathepsin E. Effect of pH on the interaction of cathepsin E with a2M Previous studies have shown that cathepsin E binds a2M at pH 6.2 and that the enzymatic activity of the complex toward the synthetic substrate Pro-Pro-Thr-Ile-Phe-Phe- (4-NO 2 )-Arg-Leu is not significantly affected [34]. Strong inhibition of the cathepsin E activity toward the protein substrate ribonuclease A by a2M was also observed at pH 5.5, where the a2M was cleaved by cathepsin E at the Phe-Tyr bond in the bait region [35]. Similarly, complete inhibition of the cathepsin D activity toward hemoglobin by a2M was observed at pH 6.2 [34]. These observations suggest that the inhibition of these enzymes by a2M at mild acidic pH values is similar to that observed with other classes of proteinases at neutral pH values. To assess whether the pH is crucial for the action of a2M on these aspartic proteinases, the effect of lowering the pH below 5.5 on the association of a2M with cathepsins E and D and the structural change in a2M upon complex formation with these enzymes were analyzed. After incubation of a2M at 37 °C for 1 h with or without cathepsin E at the indicated pH values, the reaction mixtures were analyzed by non- denaturing PAGE (Fig. 2). a2M treated with cathepsin E at pH 5.5 migrated faster than the native a2M, indicating that a2M became bound to cathepsin E and thereby underwent conformational change into the more compact form. a2M treated with cathepsin E at pH 4.5 was diffusely stained and its mobility was faster than that observed at pH 5.5. a2M treated with cathepsin E at pH 3.8 was seen near the dye front. Fig. 1. Effect of bovine a2M on the proteolytic activities of cathepsin E and cathepsin D. The Hb-hydrolyzing activity of cathepsin E purified from human erythrocytes and rat spleen and cathepsin D purified from bovine spleen and rat spleen (0.01 ng each) was measured at pH 3.8 in the absence or presence of increasing amounts of a2M. Ó FEBS 2003 Degradation of a 2 -macroglobulin by cathepsin E (Eur. J. Biochem. 270) 1191 To assess the mechanism of action of cathepsin E on a2M, we next analyzed whether the mobility changes of cathepsin E-treated a2M was due to cleavage of the bait region or degradation of bovine a2M. The cathepsin E-treated a2M molecule was rapidly degraded to generate two major protein bands with apparent molecular masses of 90 and 30 kDa within a 15-min incubation at pH 3.8 and 37 °C (Fig. 3A). The N-terminal amino-acid sequences of the 90- and 30-kDa peptides were found to be AVDGKEP and LAIPVE, respectively. Within a 30-min incubation, an additional 85-kDa peptide was generated and its N-terminal amino-acid sequence was identical to that of the 90-kDa peptide. All these peptides were further degraded by prolonged incubation. Similar results were obtained with human a2M (Fig. 3B), although the N-terminal amino-acid sequence of human 85-kDa peptide was not identical to that of the 90-kDa peptide from bovine a2M. The N-terminal amino-acid sequence of the 85-kDa peptide derived from human a2M corresponded to the sequence of the N-termi- nus of the original a2M, and the N-terminal amino-acid sequence of the 90-kDa was identical to the sequence starting with 685th Tyr (YESDVM). Although human a2M treated with cathepsin E generated the 30-kDa peptide, its N-terminal amino-acid sequence could not be determined by overlapping of additional peptides in the vicinity of 30 kDa. These three peptides were also generated from cathepsin E-treated human a2M at pH 4.5 and 5.5 (Fig. 3C). As no detectable accumulation of the other protein bands was observed, cleavage at the other sites probably much more rapid than at the one that was slow enough to allow the cleaved fragments to build up. This may explain some discrepancy in molecular sizes between the intact 170-kDa polypeptide and the generated 90- and 30-kDa fragments. In contrast, no degraded protein bands were observed for the cathepsin D-treated bovine a2M under the same conditions (Fig. 3A). Figure 4 shows Fig. 2. Association of cathepsin E with a2M under acidic conditions. Cathepsin E was incubated with bovine a2M at a molar ratio of 2 : 1 at 37 °C for 60 min at the indicated pH values. Then the reaction mixture was subjected to native PAGE at pH 8.9. As a control, a2M treated with trypsin (an a2M/enzyme ratio, 1 : 1.5) at pH 7.5 and 37 °C for 10 min was run on the same gel. Fig. 3. SDS/PAGE of a2M treated with cathepsin E or cathepsin D under acidic conditions. a2M from bovine (A) and human sources (B, C) was incubated with cathepsin E or cathepsin D at molar ration of 1 : 1 at 37 °C at the indicated pH values for various times, and then analyzed by SDS/ PAGE under reducing conditions. A, bovine a2M at pH 3.8; B, human a2M at pH 3.8; C, human a2M at pH 4.5 and 5.5. A, 8% gel; B and C, 15% gel. Fig. 4. Effects of substrate/enzyme ratios on generation of the cathepsin E-cleavage peptides of a2M. Bovine a2M was incubated with cathepsin E at various a2M/enzyme ratios at pH 3.8 and 37 °C for 30 min, and then the reaction products were analyzed by SDS/PAGE (6%) under reducing conditions. 1192 M. Shibata et al. (Eur. J. Biochem. 270) Ó FEBS 2003 SDS/PAGE profiles of bovine a2M treated with cathepsin E at different molar ratios at 37 °Cand30minatpH3.8. The a2M-cathepsin E complex gave the 90- and 30-kDa peptides at a a2M/cathepsin E molar ratio of below 10. More than 90% of the original a2M disappeared at the molar ratio of approx. 2 : 1, where the generation of the 90- and 30-kDa peptides reached the maximal value. N-terminal sequencing of bovine a2M and sequencing its cDNA Because the inhibition capacity of bovine a2M for cathepsin E was stronger than that of human a2M, and because there is no report on the sequence of the subunit of bovine a2M so far, we determined the partial amino-acid sequence of Fig. 5. Comparison of the partial amino-acid sequence of bovine a2M with those of human, rat, and mouse a2M species. The partial amino-acid sequence of bovine a2M, including the bait region to the thiol ester bond site, was shown aligned with those of human, rat and mouse a2M species. This region corresponds to the residues 538–954 of human a2M. Shading indicates identity relative to the bovine a2M sequence, and numbering is relative to the human a2M. The bait region is underlined and the thiol ester bond site is double-underlined. The arrow and the box indicate the cathepsin E cleavage site and the sequence used for peptide synthesis, respectively. Human, rat and mouse a2M sequences are from refs [45], [46] and [47], respectively. Ó FEBS 2003 Degradation of a 2 -macroglobulin by cathepsin E (Eur. J. Biochem. 270) 1193 bovine a2M predicted from its cDNA sequence (corres- ponding to the residues 538–954 of human a2M), which contained the bait region. Comparison of the amino-acid sequence of bovine a2M determined with those of human, rat, and mouse origins revealed that the sequence was strongly related to those of these species (Fig. 5). Although the overall sequence of this region of bovine a2M was significantly similar to those of human, mouse, and rat sources (59%, 62%, and 55% identities, respectively), the bait region (corresponding to the residues 666–706 of human a2M) was dissimilar to those of other species and of different length. The N-terminal amino-acid sequences of the 90- (AVDGKEP) and 30-kDa peptides (LAIPVE) generated from cathepsin E-treated bovine a2M correspon- ded to the sequences of N-terminus and starting with 812th Leu of the intact a2M, respectively. It is worth emphasizing that the amino-acid sequences of cathepsin E-cleavage sites (shown by the box) were highly conserved among mammalian species. Taken together, these results indicate that cathepsin E specifically cleaved at the Phe811-Leu812 bond present in about 100mer downstream of the bait region of both bovine and human a2M. Analysis of the specific cleavage of a2M by cathepsin E with various synthetic peptides To further confirm the specific cleavage of a2M by cathepsin E, synthetic peptides of a2M including the cathepsin E cleavage site region were designed and synthesized. These synthetic peptides were incubated with cathepsin E or cathepsin D at pH 3.8 and 37 °C for various time intervals and then the reaction products were analyzed by HPLC C18 column. The hexapeptide SPAFLA corresponding to the cleavage site of human a2M was efficiently cleaved by cathepsin E at the Phe-Leu bond, whereas it was not cleaved by cathepsin D (Table 1). The hepta peptide SSAFLAF corresponding to the cleavage site of bovine a2M was also cleaved by cathepsin E. However, differing from the peptide SPAFLA, this peptide was efficiently cleaved by cathepsin D. On the other hand, the pentapeptides SPAFL and SSAFL were not cleaved either cathepsin E or cathepsin D, indicating that the presence of Ala in the P¢2siteiscrucialfor the selective cleavage of a2M by cathepsin E and that the addition of Phe to the P¢3 site causes a loss of this selective action of cathepsin E. In agreement with these results, the cathepsin E-induced a2M degradation was significantly inhibited by these three peptides, most strongly by SPAFLA (Fig. 6). Structural changes in a2M upon complex formation with cathepsin E as determined by gel filtration analysis a2M was first incubated with cathepsin E (a molar ratio of 14 : 1) at pH values between 5.5 and 3.8 at various time intervals. The reaction mixtures were adjusted to pH 6.0 and then subjected to gel filtration on Superose 6. At every pH value, more than 80% of cathepsin E rapidly disappeared from the original position corresponding to a molecularmassashighas80kDaandappearedatthe position where a2M was eluted. However, the a2M- cathepsin E complex was rapidly dissociated at pH 3.8 within a 2-min incubation (Fig. 7B). Parallel to this change, a significant amount of the original a2M disappeared and a Table 1. The ability of cathepsin E and cathepsin D to cleave synthetic peptides. The synthetic peptides (20 n M ) containing the cathepsin E-cleavage site of a2M were incubated with or without either cathepsin E or cathepsin D (20 p M each)atpH 3.8and37 °C for the indicated time. The samples were then subjected to a HPLC C18 column chromatography and the resultant peak fractions were subjected to the amino-acid sequence analyzer. The values are expressed as a percentage of the initial intact peptide. The asterisk indicates the values for SPAF and SSAF generated by cleavage of SPAFLA and SSAFLAF, respectively. ND, not determined. Peptides Cathepsin E Cathepsin D 0 0.5 4 24 0 0.5 24 SSAFL 100 ND ND 99 100 100 100 SPAFL 100 ND ND 98 100 100 100 SPAFLA 100 93 63 21 100 100 99 *SPAF 0 7 37 79 0 0 1 SSAFLAF 100 2 ND 0 100 33 1 *SSAF 0 99 ND 100 0 68 99 Fig. 6. Effects of synthetic peptides on degradation of a2M by cathepsin E. Bovine a2M was incubated with cathepsin E at pH 3.8 and 37 °C for 30 min in the absence or presence of various synthetic peptides (500 l M ), and then the reaction products were analyzed by SDS/ PAGE under reducing conditions. Peptides used are Ser-Ser-Ala-Phe- Leu (1), Ser-Pro-Ala-Phe-Leu (2), and Ser-Pro-Ala-Phe-Leu-Ala (3). 1194 M. Shibata et al. (Eur. J. Biochem. 270) Ó FEBS 2003 few protein peaks were produced. After the 1-h incubation at pH 3.8, the original a2M completely disappeared and generated two major protein peaks at the positions corresponding to molecular masses as high as 90 and 30 kDa. In agreement with this change cathepsin E reappeared at the original position with no loss of activity. On the other hand, the dissociation of cathepsin E from the a2M complex was relatively slow at pH 4.5 and 5.5 (Fig. 7A). Although the formation of a2M–cathepsin E complex was rapid at pH 5.5 and 4.5, the rate of dissociation of this complex was pH-dependent. On the other hand, the antiproteolytic activity of a2M for trypsin was lost by incubation with cathepsin E in a dose- dependent manner (Fig. 8). Discussion a2M has a wide range of physiological activities via its interaction with various target proteins, such as the control of the activity of proteinases, the regulation of the activities of numerous cytokines and growth factors, and the enhancement of antigen presentation. Once taken up inside the cell the a2M–protein complexes are rapidly degraded in the endolysosome system. However, the fate of a2M in the Fig. 7. Gel filtration on Superose 6 of a2M treated with cathepsin E at various pH values. Bovine a2M was incubated with cathepsin E at 37 °Catthe indicated pH values. The reaction mixtures were adjusted to pH 6.0 and then run on a Superose 6 column equilibrated with 10 m M sodium phosphate buffer, pH 6.0, containing 150 m M NaCl. Fractions were analyzed by the cathepsin E activity at pH 3.8 with the synthetic substrate MOCAc-Gly-Lys-Pro-Ile-Ile-Phe-Phe-Arg-Leu-Lys(Dnp)- D -Arg-NH 2 , SDS/PAGE and immunoblotting with antibodies to cathepsin E. (a) The elution profiles of a2M and cathepsin E when each protein was run on the column independently. The a2M was treated with cathepsin E at pH 4.5 for 1 h (b), at pH 5.5 for 1 h (c), at pH 3.8 for 45 s (d), at pH 3.8 for 2 min (e), and at pH 3.8 for 1 h (f). The solid and dotted lines illustrate the cathepsin E activity and the absorbance at 280 nm, respectively. Ó FEBS 2003 Degradation of a 2 -macroglobulin by cathepsin E (Eur. J. Biochem. 270) 1195 complexes and the nature of its degradation are not clear. The present findings on the biochemical nature of the interaction of cathepsin E with a2M are unique and unexpected. This is the first report of the structural and functional disruption of a2M integrity by cathepsin E. It is surprising that cathepsin E can be associated with a2M at very low pH values and rapidly cleaves it at the specific site distinct from the bait region. Under mild acidic conditions, a2M appears to interact with cathepsin E [35], as well as other proteinases, to be cleaved a peptide bond in the bait region and thereby undergoes a conformational change similar to that occurring at neutral pH values [42], where cathepsin E is essentially inactive [27,36]. However, at pH values below 5.0, a2M is unstable and is likely to lose the proteinase-binding activity [43], where cathepsin E is more active. The endolysosomal compartment is the major site of endogenous protein degradation. Degrada- tion of a2M, like other endocytosed proteins, is known to occur rapidly inside the endolysosomes, where the pH is maintained below 5.0. The intravacuolar pH is important because most of the endolysosomal hydrolases including cathepsins E and D require a pH of 3.5–5.0 for maximal activity. Similarly, microbicidal systems such as peroxi- dase-hydrogen peroxide need such an acid pH for optimal activity [44]. At pH values between 3.8 and 5.5, cathepsin E selectively bound a2M and cleaved it at the Phe811-Leu812 bond at a distance from the bait region. Therefore, the cathepsin E–a2M interaction below pH 5.0 appears to be unique and is different from that occurring at mild acidic and neutral pH values. Namely, cathepsin E is efficiently associated with a2M at pH values below 5.0 without loss of the proteolytic activity and rapidly cleaves it at the Phe811-Leu812 bond distinct from the bait region and then leaves the associated site. Considering that a very low pH treatment results in dissociation of a2M into the dimers, which do not reassociate normally but tend to aggregate [43], it is more likely that cathepsin E interact with the distinct region from thebaitregionofthea2M. Under these conditions, however, the analogous aspartic proteinase cathepsin D neither interacts with a2M nor cleaves it. As previous work has demonstrated that the action of cathepsin D, as well as cathepsin E, toward protein substrates was blocked by a2M at pH 6.2 [34], a slight decrease in pH below 5.5 is very likely to cause an additional conformational change of a2M and thereby abolish the ability of cathepsin D to bind a2M. This study also described for the first time the cloning and sequencing of partial cDNA for bovine a2M, as its primary structure is likely to provide significant informa- tion regarding the finding that bovine a2M is more sensitive to cathepsin E digestion than that from other species (data not shown). Analysis of the amino-acid sequence of bovine a2M deduced from isolated cDNA clones revealed that its gross structure was homologous to those of human, rat, and mouse a2M, although the bait region was dissimilar to that of either a2M from other species. In particular, the sequence around the cathepsin E cleavage site (Phe811–Leu812 bond) was highly conserved in all of the other species. To further confirm the selective cleavage of the Phe811–Leu812 bond in a2M by cathepsin E, we synthesized some peptides corresponding to part of the cleavage site and analyzed the susceptibility of these peptides to cleavage by cathepsin E. The hexapeptide SPAFLA corresponding to part of the cathepsin E cleavage site (the Ser808–Ala813 sequence of human a2M) was selectively cleaved at the Phe-Leu bond by cathepsin E, but not cathepsin D. The peptides SPAFL and SSAFL corresponding to the sequence Ser808–Leu812 of human and bovine a2M, respectively, were not cleaved by either cathepsin E or cathepsin D. In agreement with the results, the peptide SPAFLA significantly inhibited the degradation of a2M by cathepsin E. These results indicate that the presence of Ala in the P¢2 is essential for selective cleavage the synthetic peptides by cathepsin E. The presence of Pro in the P3 site in a2M, however, is unlikely to be crucial for its selective cleavage, as the 809th residue in bovine a2M is Ser in place of Pro found in other species. Acknowledgements We thank Dr Haruki Uemura (Department of Protozoology, Institute of Tropical Medicine, Nagasaki University) for helpful discussion and technical advice and Drs Tsutomu Iwamoto (1st Department of Oral and Maxillofacial Surgery, Nagasaki University School of Dentistry), Masayo Okaji, Kazuhiro Kanaoka, Fumio Hashimoto and Yasuhiro Kobayashi (Department Orthodontics, Nagasaki University School of Dentistry) for fruitful discussions. References 1. Barrett, A.J. & Starkey, P.M. (1973) The interaction of alpha 2-macroglobulin with proteinases. Characteristics and specificity of the reaction, and a hypothesis concerning its molecular mechanism. Biochem. J. 133, 709–724. 2. Gliemann, J., Larsen, T.R. & Sottrup-Jensen, L. (1983) Cell association and degradation of alpha 2-macroglobulin-trypsin complexes in hepatocytes and adipocytes. Biochim. Biophys. Acta 756, 230–237. Fig. 8. Effect of cathepsin E-treated a2M on the antiproteolytic activity for trypsin. Bovine a2M was incubated with cathepsin E at various enzyme/a2M ratios at pH 3.8 and 37 °C for 30 min, and then the reaction products were neutralized and analyzed for the antiproteinase activity on trypsin using 1% casein as a substrate. The values are expressed as percentages of the trypsin activity in the absence of a2M. 1196 M. Shibata et al. (Eur. J. Biochem. 270) Ó FEBS 2003 3. Van Leuven, F., Cassiman, J.J. & Van Den Berghe, H. (1979) Demonstration of an alpha 2-macroglobulin receptor in human fibroblasts, absent in tumor-derived cell lines. J. Biol. Chem. 254, 5155–5160. 4. Debanne, M.T., Bell, R. & Dolovich, J. (1975) Uptake of protei- nase-alpha-macroglobulin complexes by macrophages. Biochim. Biophys. Acta 411, 295–304. 5. Van Leuven, F., Cassiman, J.J. & Van den Berghe, H. (1978) Uptake and degradation of alpha 2-macroglobulin-protease complexes in human cells in culture. Exp. Cell Res. 117, 273–282. 6. Borth, W. & Luger, T.A. (1989) Identification of alpha 2-macro- globulin as a cytokine binding plasma protein. Binding of inter- leukin-1 beta to ÔFÕ alpha 2-macroglobulin. J. Biol. Chem. 264, 5818–5825. 7. Teodorescu,M.,McAfee,M.,Skosey,J.L.,Wallman,J.,Shaw,A. & Hanly, W.C. (1991) Covalent disulfide binding of human IL-1 beta to alpha 2-macroglobulin: inhibition by D -penicillamine. Mol. Immunol. 28, 323–331. 8. Matsuda, T., Hirano, T., Nagasawa, S. & Kishimoto, T. (1989) Identification of alpha 2-macroglobulin as a carrier protein for IL-6. J. Immunol. 142, 148–152. 9. James, K., van den Haan, J., Lens, S. & Farmer, K. (1992) Pre- liminary studies on the interaction of TNF alpha and IFN gamma with alpha 2-macroglobulin. Immunol. Lett. 32, 49–57. 10. O’Connor-McCourt, M.D. & Wakefield, L.M. (1987) Latent transforming growth factor-beta in serum. A specific complex with alpha 2-macroglobulin. J. Biol. Chem. 262, 14090–14099. 11. Dennis, P.A., Saksela, O., Harpel, P. & Rifkin, D.B. (1989) Alpha 2-macroglobulin is a binding protein for basic fibroblast growth factor. J. Biol. Chem. 264, 7210–7216. 12. Huang, J.S., Huang, S.S. & Deuel, T.F. (1984) Specific covalent binding of platelet-derived growth factor to human plasma alpha 2-macroglobulin. Proc.Natl.Acad.Sci.U.S.A.81, 342–346. 13. Ronne, H., Anundi, H., Rask, L. & Peterson, P.A. (1979) Nerve growth factor binds to serum alpha-2-macroglobulin. Biochem. Biophys. Res. Commun. 87, 330–336. 14. Chu, C.T., Rubenstein, D.S., Enghild, J.J. & Pizzo, S.V. (1991) Mechanism of insulin incorporation into alpha 2-macroglobulin: implications for the study of peptide and growth factor binding. Biochemistry 30, 1551–1560. 15. Mettenburg, J.M., Webb, D.J. & Gonias, S.L. (2002) Distinct binding sites in the structure of alpha 2-macroglobulin mediate the interaction with beta-amyloid peptide and growth factors. J. Biol. Chem. 277, 13338–13345. 16. Gonias, S.L. (1992) Alpha 2-macroglobulin: a protein at the interface of fibrinolysis and cellular growth regulation. Exp. Hematol. 20, 302–311. 17. Borth, W. (1992) Alpha 2-macroglobulin, a multifunctional bind- ing protein with targeting characteristics. FASEB J. 6, 3345–3353. 18. Crookston,K.P.,Webb,D.J.,Lamarre,J.&Gonias,S.L.(1993) Binding of platelet-derived growth factor-BB and transforming growth factor-beta 1 to alpha 2-macroglobulin in vitro and in vivo: comparison of receptor-recognized and non-recognized alpha 2-macroglobulin conformations. Biochem. J. 293, 443–450. 19. Philip, A. & O’Connor-McCourt, M.D. (1991) Interaction of transforming growth factor-beta 1 with alpha 2-macroglobulin. Role in transforming growth factor-beta 1 clearance. J. Biol. Chem. 266, 22290–22296. 20. Chu, C.T. & Pizzo, S.V. (1993) Receptor-mediated antigen deli- very into macrophages. Complexing antigen to alpha 2-macro- globulin enhances presentation to T cells. J. Immunol. 150, 48–58. 21. Yamamoto, K. (1999) Cathepsin E & cathepsin D. In Proteases: New Perspectives (Turk, V., ed.), pp. 59–71. Birkhauser-Verlag, Basel, Switzerland. 22. Amano, T., Nakanishi, H., Oka, M. & Yamamoto, K. (1995) Increased expression of cathepsins E and D in reactive microglial cells associated with spongiform degeneration in the brain stem of senescence-accelerated mouse. Exp. Neurol. 136, 171–182. 23. Sastradipura, D.F., Nakanishi, H., Tsukuba, T., Nishishita, K., Sakai, H., Kato, Y., Gotow, T., Uchiyama, Y. & Yamamoto, K. (1998) Identification of cellular compartments involved in pro- cessing of cathepsin E in primary cultures of rat microglia. J. Neurochem. 70, 2045–2056. 24. Nishioku, T., Hashimoto, K., Yamashita, K., Liou, S.Y., Kagamiishi, Y., Maegawa, H., Katsube, N., Peters, C., von Figura,K.,Saftig,P.,Katunuma,N.,Yamamoto,K.& Nakanishi, H. (2002) Involvement of cathepsin E in exogenous antigen processing in primary cultured murine microglia. J. Biol. Chem. 277, 4816–4822. 25. Sakai, H., Saku, T., Kato, Y. & Yamamoto, K. (1989) Quantita- tion and immunohistochemical localization of cathepsins E and D in rat tissues and blood cells. Biochim. Biophys. Acta 991, 367–375. 26. Bennett, K., Levine, T., Ellis, J.S., Peanasky, R.J., Samloff, I.M., Kay, J. & Chain, B.M. (1992) Antigen processing for presentation by class II major histocompatibility complex requires cleavage by cathepsin E. Eur. J. Immunol. 22, 1519–1524. 27. Yamamoto, K. & Marchesi, V.T. (1984) Purification and char- acterization of acid proteinase from human erythrocyte mem- branes. Biochim. Biophys. Acta. 790, 208–218. 28. Takeda,M.,Ueno,E.,Kato,Y.&Yamamoto,K.(1986)Isola- tion, and catalytic and immunochemical properties of cathepsin D -like acid proteinase from rat erythrocytes. J. Biochem. 100, 1269–1277. 29. Yoshimine, Y., Tsukuba, T., Isobe, R., Sumi, M., Akamine, A., Maeda, K. & Yamamoto, K. (1995) Specific immunocytochemical localization of cathepsin E at the ruffled border membrane of active osteoclasts. Cell Tissue Res. 281, 85–91. 30. Saku, T., Sakai, H., Shibata, Y., Kato, Y. & Yamamoto, K. (1991) An immunocytochemical study on distinct intracellular localiza- tion of cathepsin E and cathepsin D in human gastric cells and various rat cells. J. Biochem. 110, 956–964. 31. Kageyama, T., Ichinose, M. & Yonezawa, S. (1995) Processing of the precursors to neurotensin and other bioactive peptides by cathepsin E. J. Biol. Chem. 270, 19135–19140. 32. Nakanishi, H., Zhang, J., Koike, M., Nishioku, T., Okamoto, Y., Kominami, E., von Figura, K., Peters, C., Yamamoto, K., Saftig, P. & Uchiyama, Y. (2001) Involvement of nitric oxide released from microglia-macrophages in pathological changes of cathepsin D-deficient mice. J. Neurosci. 21, 7526–7533. 33.Saftig,P.,Hetman,M.,Schmahl,W.,Weber,K.,Heine,L., Mossmann, H., Koster, A., Hess, B., Evers, M., von Figura, K. & Peters, C. (1995) Mice deficient for the lysosomal proteinase cathepsin D exhibit progressive atrophy of the intestinal mucosa and profound destruction of lymphoid cells. EMBO J. 14, 3599– 3608. 34. Thomas, D.J., Richards, A.D. & Kay, J. (1989) Inhibition of aspartic proteinases by alpha 2-macroglobulin. Biochem. J. 259, 905–907. 35. Athauda, S.B., Arakawa, H., Nishigai, M., Takahashi, T., Ikai, A. & Takahashi, K. (1993) Inhibition of cathepsin E by alpha 2-macroglobulin and the resulting structural changes in the inhibitor. J. Biochem. 113, 526–530. 36. Yamamoto, K., Katsuda, N. & Kato, K. (1978) Affinity puri- fication and properties of cathepsin-E-like acid proteinase from rat spleen. Eur. J. Biochem. 92, 499–508. 37. Yamamoto, K., Katsuda, N., Himeno, M. & Kato, K. (1979) Cathepsin D of rat spleen. Affinity purification and properties of two types of cathepsin D. Eur. J. Biochem. 95, 459–467. 38. Shibata, M. (1994) Activation and stabilization of endoplasmic aspartic proteinases by ATP. Jpn. J. Oral Biol. 36, 289–298. 39. Yasuda, Y., Kageyama, T., Akamine, A., Shibata, M., Komi- nami, E., Uchiyama, Y. & Yamamoto, K. (1999) Characterization Ó FEBS 2003 Degradation of a 2 -macroglobulin by cathepsin E (Eur. J. Biochem. 270) 1197 of new fluorogenic substrates for the rapid and sensitive assay of cathepsin E and cathepsin D. J. Biochem. 125, 1137–1143. 40. Feldman, S.R., Gonias, S.L., Ney, K.A., Pratt, C.W. & Pizzo, S.V. (1984) Identification of ÔembryoninÕ as bovine alpha 2-macro- globulin. J. Biol. Chem. 259, 4458–4462. 41. Warburton, M.J., Coles, B., Dundas, S.R., Gusterson, B.A. & O’Hare, M.J. (1993) Hydrocortisone induces the synthesis of alpha 2-macroglobulin by rat mammary myoepithelial cells. Eur. J. Biochem. 214, 803–809. 42. Pochon, F., Barray, M. & Delain, E. (1989) Dissociation of alpha 2-macroglobulin into functional half-molecules by mild acid treatment. Biochim. Biophys. Acta 996, 132–138. 43. Barrett, A.J., Brown, M.A. & Sayers, C.A. (1979) The electro- phoretically ÔslowÕ and ÔfastÕ forms of the alpha 2-macroglobulin molecule. Biochem. J. 181, 401–418. 44. Klebanoff, S.J. (1975) Antimicrobial mechanisms in neutrophilic polymorphonuclear leukocytes. Semin. Hematol. 12, 117–142. 45. Kan,C.C.,Solomon,E.,Belt,K.T.,Chain,A.C.,Hiorns,L.R.& Fey, G. (1985) Nucleotide sequence of cDNA encoding human alpha2-macroglobulin and assignment of the chromosomal locus. Proc.NatlAcad.Sci.USA82, 2282–2286. 46. Gehring, M.R., Shiels, B.R., Northemann, W., de Bruijn, M.H.L., Kan, C C., Chain, A.C., Noonan, D.J. & Fey, G.H. (1987) Sequence of rat liver alpha-2-macroglobulin and acute phase control of its messenger RNA. J. Biol. Chem. 262, 446–454. 47. Umans, L., Serneels, L., Hilliker, C., Stas, L., Overbergh, L., De Strooper,B.,VanLeuven,F.&VandenBerghe,H.(1994) Molecular cloning of the mouse gene coding for alpha2-macro- globulin and targeting of the gene in embryonic stem cells. Genomics 22, 519–529. 1198 M. Shibata et al. (Eur. J. Biochem. 270) Ó FEBS 2003 . the presence of Ala in the P¢2 is essential for selective cleavage the synthetic peptides by cathepsin E. The presence of Pro in the P3 site in a2M, however,. further confirm the specific cleavage of a2M by cathepsin E, synthetic peptides of a2M including the cathepsin E cleavage site region were designed and synthesized.