Acta Histochem. Cytochem. Vol. 28 No. 3 281-286 (1995)

Aminopeptidases and Dipeptidyl Peptidases Activities in Chicken Bone Tissue

Osamu Fukushima and Hiroshi Yamashita

Department of Anatomy (I), The Jikei University, School of Medicine, Minato-ku, Tokyo 105

Received for publication April 14, 1995 and in revised form July 4, 1995

There have been very few papers reporting ed. These activities were sensitive to on the localizations of aminopeptidases and 10 mM of ethylenediaminetetraacetic acid dipeptidyl peptidases activities in bone (EDTA). A DPP-I was seen in osteoclasts, tissue, although these peptidases seem to osteoblasts and osteocytes in the presence have important roles in bone resorption and of mercaptoethylamine (MEA), however Gly- osteoid degradation. The present study Arg-MNA hydrolyzing activity in the absence demonstrates these peptidases in fixed and of MEA was restricted to only osteoclasts. decalcified tibial metaphyses of 3-week-old DPP-II activity was present in osteoblasts chickens using the azo-dye methods at light and osteocytes, using Lys-Pro-MNA as a microscopic level. As substrates, amino substrate. There was no Lys-Ala-MNA acid derivatives of 4-methoxy-2-naphthyl- hydrolyzing activity in bone cells. The pre- amine (MNA) were used: Leu- or Ala-MNA sent study failed to demonstrate DPP-IV for aminopeptidase-M (AP-M, EC 3.4.11.2), activity in either 4% paraformaldehyde- or Glu-MNA for aminopeptidase-A (AP-A, EC 2.5% glutaraldehyde-fixed samples, 3.4.11.7), Gly-Arg-MNA for dipeptidyl pep- although it was observed in the glomerulus tidase-I (DPP-I, EC 3.4.14.1), Lys-Ala- or and proximal convoluted tubule cells of the Lys-Pro-MNA for dipeptidyl peptidase-II rat kidney fixed with glutaraldehyde. The (DPP-II, EC 3.4.14.2), and Gly-Pro-MNA for present study suggests that AP-M and DPP-I dipeptidyl peptidase-IV (DPP-IV, EC in osteoclasts, and AP-A, DPP-I and DPP-II in 3.4.14.5). An AP-M in osteoclasts, and AP- osteoblasts seem to function in the bone A in osteoblasts and osteocytes were observ- remodeling process.

Key words: Bone, Aminopeptidase, Dipeptidyl peptidase, Enzyme histochemistry

I. Introduction 2-naphthylamide substrate caused significant diffusion of the reaction product due to the slow rate of coupling of The bone matrix consists mainly of hydroxyapatite, 2-naphthylamide to a diazonium salt and considerable collagen, non-collagenous proteins and proteoglycans, aqueous solubility of the reaction product. They recom- and is constantly remodeled. Therefore, proteases have mended 4-methoxy-2-naphthylamine (MNA) substrates important roles in matrix resorption. However, there for investigating precise localizations of proteases with en- have been very few papers addressing histochemical zyme histochemical techniques (azo-dye methods). San- localizations of proteases, with the exception of cathepsins nes and colleagues [15] employed MNA substrates for B, L and D [5, 6, 9, 13, 15, 20]. Lipp [10] first reported demonstration of dipeptidyl peptidases I and II (DPP-I, that aminopeptidase-M (AP-M) activity was present in DPP-II), and showed DPP-I in osteoblasts, osteocytes, osteoclasts, osteoblasts and endothelial cells at the and chondrocytes in proliferating and hypertrophic vascular invasion in growth plate using Leu-2-naphthyl- regions of the growth plate, and DPP-II in osteoblasts, amide as a substrate. Smith and van Frank [18] showed osteocytes, and chondrocytes in the resting zone. In the same experiment, they failed to demonstrate cathepsin B This work was supported by a Grand-in Aid for Scientific Research, activity in the bone, although it was strongly positive in No. 05670030, from the Ministry of Education, Science and Culture, osteoclasts with other enzyme histochemical [5, 20] and im- Japan. munohistochemical studies [6, 9, 13, 16]. Further, there Correspondence to: Dr. Osamu Fukushima, Department of Anatomy has been no trial to demonstrate AP-A and DPP-IV (I), The Jikei University, School of Medicine, 25-8 Nishishinbashi-3- activities in the bone. chome, Minato-ku, Tokyo 105, Japan. The present study, using azo-dye methods with MNA

281 282 Fukushima and Yamashita substrates, showed AP-M and DPP-I in osteoclasts, and B. DPP-IV medium: 1.0 mM Gly-Pro-MNA, 0.1 M AP-A, DPP-I and DPP-II in osteoblasts and osteocytes. cacodylate buffer (pH 7.4), 2.0 mM Fast Blue B. All substrate was purchased from Enzyme System Products II. Materials and Methods (CA, U.S.A.). As controls, substrate-omitted media (all media), 10 mM EDTA containing media (AP-M and AP-A For the present study, 9 domestic chickens of 3 weeks media), and MEA-omitted medium (DPP-I medium) were of age (body weight 120-140g), were purchased from also prepared. After incubation, sections were rinsed Saitamajikkendoubutsu-kyokyusho (Saitama, Japan). with distilled water, and immersed in 2% CuSO4 for The chickens were anesthetized with ether, then tibial 5 min. Sections were rinsed in distilled water again, then metaphyses were dissected out quickly and cut into small mounted with glycerin jelly, and observed by conventional blocks. The blocks were divided into three groups, and light microscopy. each group was immersed in one of the following three As positive controls, kidneys (cortex) of Wistar male kinds of fixative for 60 min at 0-4•Ž: 4% parafor- rats (8-weeks old) fixed with 1% GLA-4% PFA were also maldehyde (PFA) in 0.1 M cacodylate buffer (pH 7.2) con- prepared. taining 8% sucrose, 1% glutaraldehyde (GLA)-4% PFA in 0.1 M cacodylate buffer (pH 7.2), or 2.5% GLA in 0.1 M III. Results cacodylate buffer (pH 7.2) containing 8% sucrose. The blocks were rinsed in 0.1 M cacodylate buffer (pH 7.2) 1. AP-M activity containing 8% sucrose (cacodylate-buffered sucrose), AP-M activity was seen in osteoclasts which were then decalcified with 5% ethylenediaminetetraacetic acid attached to the bone matrix, using either Leu-MNA or

(EDTA) in 30 mM N-(2-hydroxyethyp-piperazine-N'-2- Ala-MNA as a substrate. There was no difference in ethanesulfonic acid (HEPES) buffer (pH 7.2) containing localization and intensity of the activity between the 8% sucrose [4] at 4•Ž for 3 days. The blocks were treated two substrates. No activity was observed in osteoblasts, sequentially with 10%, 15%, 20% and 25% sucrose, then osteocytes and endothelial cells (Fig. 1). There was no embedded in OCT compound (Tissue Tek, Miles Inc. IN, reaction product in the bone matrix. The activity was U.S.A.), and frozen in dry-ice acetone. Frozen blocks markedly inhibited by 10 mM EDTA (Fig. 2). The were stored at -80•Ž until use. Cryostat sections, 8 ƒÊm substrate-omitted medium yielded no activity (not in thickness, were placed on poly-L-lysine (Sigma, P-1524) shown). The activity was observed in 4% PFA-, 1% coated slide glasses, and air dried. Sections were in- GLA-4% PFA-, and 2.5% GLA-fixed samples. cubated in the following media for 30 min at 37•Ž, accor- In the rat kidney, the activity was restricted to the ding to Sannes [14]. AP-M medium: 1.6 mM Leu-MNA brush border of proximal convoluted tubule cells (not or Ala-MNA, 0.1 M phosphate buffer (pH 6.7), 2.0 mM shown). Fast Blue B (Sigma, D-3502). AP-A medium: 1.7 mM Glu-MNA, 0.1 M phosphate buffer (pH 6.7), 2.0 mM Fast 2. AP-A activity Blue B. DPP-I medium: 1.0 mM Gly-Arg-MNA, 0.1 M AP-A activity was present in osteoblasts and phosphate buffer (pH 6.0), 10 mM EDTA, 1.0 mM mercap- osteocytes. Osteoclasts did not possess the activity. The toethylamine (MEA, Sigma, M-6500), 0.2 mM Fast Blue activity was seen only in 4% PFA-fixed samples (Fig. 3), B. DPP-II medium: 1.0 mM Lys-Pro-MNA or Lys-Ala- with no reaction product in 1% GLA-4% PFA-, and 2.5% MNA, 0.05 M cacodylate buffer (pH 5.5), 2 mM Fast Blue GLA-fixed bone tissues (not shown). EDTA had a strong

Fig. 1. Aminopeptidase-M (EC 3.4.11.2) activity using Leu-MNA as a substrate in 2.5% GLA-fixed chicken tibial metaphyses. The activity was present in osteoclasts which were attached to the bone matrix. There was no activity in osteoblasts, osteocytes, pre-osteoclasts and detach- ed osteoclasts. Arrows indicate attached osteoclasts. •~480. Fig. 2. Aminopeptidase-M activity in the presence of 10 mM EDTA in 2.5% GLA-fixed bone tissues. As a substrate, Leu-MNA was employed. The activity was significantly inhibited by EDTA. Arrow indicates an attached osteoclast. •~480. Fig. 3. Aminopeptidase-A (EC 3.4.11.7) activity in 4% PFA-fixed bone tissues. The activity was seen in osteoblasts and osteocytes. Osteoclasts (arrow) did not possess the enzyme. •~240. Fig. 4. Aminopeptidase-A activity in the influence of 10 mM EDTA in 4% PFA-fixed bone tissues. The activity was significantly inhibited by EDTA. Arrow indicates an osteoclast. •~240. Fig. 5. Dipeptidyl peptidase-I (EC 3.4.14.1) activity in 2.5% GLA-fixed bone tissues. In the presence of MEA, a SH-reagent, the activity was

present in osteoblasts, osteocytes and osteoclasts (arrows). •~240. Fig. 6. A Gly-Arg-MNA hydrolyzing activity in the absence of MEA in 2.5% GLA-fixed bone tissues. The activity was restricted to only osteoclasts. There was no reaction product in osteoblasts and osteocytes. Arrow indicates an osteoclast. •~480. Fig. 7. Dipeptidyl peptidase-II (EC 3.4.14.2) activity with Lys-Pro-MNA as a substrate in 2.5% GLA-fixed bone tissues. The activity was

positive in osteoblasts and osteocytes. No significant activity was seen in osteoclasts (arrow). •~480. Fig. 8. Lys-ALa-MNA hydrolyzing activity at pH 5.5 in 2.5% GLA-fixed bone tissues. There was no reaction product in bone sections. Arrow indicates an osteoclast. •~240. Peptidases Activities in the Bone 283

Figs. 1-8. 284 Fukushima and Yamashita inhibitory effect on AP-A activity (Fig. 4). In the rat tions at the light microscopic level. kidney fixed with 1% GLA-4% PFA, the activity was AP-M, DPP-I and Gly-Arg-MNA hydrolyzing en- strongly positive on the brush border of proximal con- zyme were restricted to osteoclasts, suggesting that these voluted tubule cells (not shown). enzymes have roles in bone resorption. AP-A, DPP-I and DPP-II activities were positive in osteoblasts and 3. DPP-I activity osteocytes, indicating that these enzymes relate to DPP-I activity was observed in osteoblasts, osteocytes osteoblastic functions, for example, bone matrix calcifica- and osteoclasts under the presence of MEA, a SH-reagent tion, osteoid degradation, and hydrolysis of local polypep- (Fig. 5). A substrate-omitted medium yielded no reaction tide growth factors. product (not shown). Interestingly, in the absence of MEA, the activity was restricted to only osteoclasts (Fig. 1. AP-M (EC 3.4.11.2, also called aminopeptidase-N, 6). Both DPP-I activity and Gly-Arg-MNA hydrolyzing microsomal aminopeptidase or CD 13) activity in the absence of MEA were detectable in 4% The present study indicated that AP-M was positive in PFA-, 1% GLA-4% PFA-, and 2.5% GLA-fixed bone active osteoclasts which were attached to the bone matrix, tissues. not in detached osteoclasts [4] or pre-osteoclasts. However, the present study failed to demonstrate AP-M is a membrane bounded exopeptidase, and DPP-I activity in the rat kidney. metalloglycoprotein containing two atoms of Zn per molecule [11]. Therefore, the activity is sensitive to 4. DPP-II activity chelating agents such as EDTA (Fig. 2). The major DPP-II activity with Lys-Pro-MNA as a substrate was physiological role of AP-M in the brush border of prox- demonstrated in osteoblasts and osteocytes (Fig. 7). imal tubule cells in the kidney and in microvilli of the in- There was no activity in osteoclasts. The reaction product testinal absorption epithelial cells is its participation in the was observed in 4% PFA-, 1% GLA-4% PFA- and 2.5% terminal digestion of proteins [11]. AP-M in osteoclasts GLA-fixed bone tissues. No reaction product was present might participate in terminal digestion of denatured col- in sections using Lys-Ala-MNA as a substrate (Fig. 8). A lagen and non-collagenous proteins which have been substrate-omitted medium yielded no activity (not shown). degraded by cysteine endoproteases in resorption laculae. In the rat kidney, the activity was observed in the On the other hand, recent biochemical papers suggested cytosol of proximal and distal convoluted tubule cells, enkepalines [11], interleukin (IL)-lbeta, IL-2, tumor with Lys-Ala-MNA and Lys-Pro-MNA as substrates (not necrosis factor (TNF)-beta and IL-6 [7] as natural shown). substrates. It is well known that ILs, especially IL-1 and 6, are potent stimulators for osteoclastic bone resorption. 5. DPP-IV activity AP-M in osteoclasts might also inactivate these peptides to DPP-IV activity was not detected in the bone tissues regulate resorbing activity. employed in the present study (not shown). In the rat kidney fixed with 1% GLA-4% PFA, the 2. AP-A (EC 3.4.11.7, also called glutamyl aminopep- activity was clearly localized in the glomerulus and on the tidase, angiotensinase-A or BP-1/6C3) brush border of proximal convoluted tubule cells (not The present study indicated AP-A activity was localiz- shown). ed in osteoblasts and osteocytes. The activity is sensitive to the chelating agent, EDTA (Fig. 4), and glutaraldehyde- fixation. Although AP-A activity was seen on the brush IV. Discussion border of rat kidney fixed with 1% GLA-4% PFA, it was Proteases are involved in bone matrix turnover [19]. not detected in GLA-fixed samples of bone tissue. In In osteoclastic bone resorption, cysteine endoproteases many tissues, capillary endothelial cells possess AP-A ac- which can degrade collagen in the resorption lacunae tivity which participate in regulating the blood pressure (acidic compartment), have been investigated intensively. [11], but no significant activity was present in endothelial In remodeling bone, osteoblasts secrete neutral col- cells in the bone marrow in the present study. lagenase and degrade the osteoid [17]. Furthermore, it is AP-A is a membrane-bound exopeptidase which acts known that proteases in osteoblasts modify proteoglycans specifically on peptides with unsubstituted N-terminal L- for matrix calcification [3], and have an effect on the level Glu and Asp, removing them. It is also a Zn-protein as is of polypeptide growth factors in bone matrix [8]. AP-M. There is 34% amino acid homology between However, very few studies have been done on aminopep- human AP-A and AP-M [21]. Angiotensin-II, chole- tidases and dipeptidyl peptidases in the bone tissues, cystokinin (CCK-8) [11], IL-7 and IL-7 receptor [22] are although the established azo-dye methods [11, 14, 18] have known natural substrates. One of the possible roles of been available. One of the reasons seems to be that the AP-A in osteoblast plasma membrane could be degrada- natural substrates of these peptidases in the bone tissues tion of the osteoid which should be removed by osteoblasts are unknown. To investigate the roles of these peptidases, before osteoclasts attack the bone matrix. the present study attempted to demonstrate their localiza- Peptidases Activities in the Bone 285

3. DPP-I (EC 3.4.14.1, also called cathepsin C, cathep- fibrin, IL-2, promellitine and collagen [7, 11]. Further sin J, glucagon-degrading enzyme, or lysosomal cys- technical improvements are needed to investigate the role teine exopeptidase) in the bone tissues. Gly-Arg-MNA hydrolyzing activity in the presence of MEA, a SH-reagent, was localized in osteoblasts, V. Acknowledgment osteocytes and osteoclasts, but in the absence of MEA was restricted to only osteoclasts (Figs. 5, 6). The authors wish to thank Nancy L. Kief for her ex- DPP-I, also known as cathepsin C, is a cysteine cellent editorial assistance with this article. exopeptidase and localized in lysosomes. The recent biochemical investigation by Nikawa and colleagues [12] indicated that cathepsin C was identical to cathepsin J in VI. References rat liver. The enzyme can degrade benzyloxycarbonyl-L- 1. Baron, R., Neff, L., Courtony, P.J., Louvard, D. and phenylalanyl-L-arginine-MNA (Z-Phe-Arg-MNA), as well Farquhar, M.G.: Polarized secretion of lysosomal enzymes: as Gly-Arg-MNA. Our previous study [5] showed a Co-distribution of cation-independent mannose-6-phosphate MEA-dependent Z-Phe-Arg-MNA hydrolyzing activity receptors and lysosomal enzymes along the osteoclast exo- was observed only in osteoclasts. There was no MEA- cytic pathway. J. Cell Biol. 106; 1863-1872, 1988. independent Z-Phe-Arg-MNA hydrolyzing activity in 2. Chilosi, M., Lestani, M., Menestrina, F. and Fiore-Donati, L.: osteoclasts. On the other hand, MEA-dependent Gly- Immunohistochemical characterization of osteoclasts and Arg-MNA hydrolyzing activity was present in osteoblasts osteoclast-like cells with monoclonal antibody MB1 on paraffin- embedded tissues. J. Pathol. 156; 251-254, 1988. and osteocytes, as well as in osteoclasts (Fig. 5). In the 3. Einhorn, T.A. and Majeska, R.: Neutral proteases in chicken bone, MEA-dependent Gly-Arg-MNA hydrolyz- regenerating bone. Clin. Orthop. Rel. Res. 262; 286-297, ing enzyme (DPP-I) is not the same as MEA-dependent 1991. Z-Phe-Arg-MNA hydrolyzing enzyme. Sannes and 4. Fukushima, O., Bekker, P.J. and Gay, C.V.: Characteriza- colleagues [15] failed to demonstrate it in osteoclasts, tion of the functional stages of osteoclasts by enzyme although Baron and colleagues [1] found it in lysosomes of histochemistry and electron microscopy. Anat. Rec. 231; 298- 315, 1991. all bone cells (osteoblasts, osteocytes and osteoclasts) with 5. Fukushima, O., Awatake, T., Kishimoto, K., Hua, Q. and an electron microscopic immunohistochemical technique. Yamashita, H.: An enzyme histochemical investigation of cys- The present results are identical with Baron's study. teine endoprotease activity in chicken bone tissue. Acta Gly-Arg-MNA hydrolyzing activity in the absence of Histochem. Cytochem. 27; 245-249, 1994. MEA was restricted to only attached osteoclasts. It is 6. Goto, T., Kiyoshima, T., Tsukuba, T., Nishimura, Y., Himeno, M., Yamamoto, K. and Tanaka, T.: Localization of different from DPP-I, because of its MEA-independence cathepsins B, D, and L in the rat osteoclast by immuno-light and localization. and -electron microscopy. Histochemistry 101; 33-40, 1994. 7. Hoffmann, T., Faust, J., Neubert, K. and Ansorge, S.: Dipep- 4. DPP-II (EC 3.4.14.2, also called dipeptidyl aminopep- tidyl peptidase IV (CD 26) and aminopeptidase N (CD 13) tidase-II) catalyzed hydrolysis of cytokines and peptides with N-terminal As previously reported by Sannes and colleagues [15], cytokine sequences. FEBS 336; 61-64, 1993. 8. Howell, S., Caswell, A.M., Kenny, A.J. and Turner, J.: Mem- Lys-Pro-MNA hydrolyzing activity was positive in brane peptidases on human osteoblast-like cells in culture: osteoblasts and osteocytes (Fig. 7). There was no Lys- hydrolysis of calcitonin and hormonal regulation of endopep- Ala-MNA hydrolyzing activity in the bone tissues. The tidase-24.11. Biochem. J. 290; 159-164, 1993. present results are identical with Sannes' report. 9. Karhukorpi, E.-K., Vihko, P. and Vaananen, K.: A difference The enzyme is known as a serine exopeptidase occurr- in the enzyme contents of resorption lacunae and secondary ing in lysosomes, and can degrade collagen fragments lysosomes of osteoclasts. Acta Histochem. 92; 1-11, 1992. 10. Lipp, W.: Aminopeptidase in bone cells. J. Histochem. [11]. One of the osteoblastic functions is a removal of Cytochem. 7; 205, 1959. osteoid before osteoclasts attack the bone matrix. 11. Lojda, Z., Gossrau, R. and Stoward, P.J.: Proteases. In Osteoblasts possess a neutral collagenase. This might lead "Histochemistry . Theoretical and Applied, vol. 3, Enzyme us to believe that DPP-II may have a role in intralysosomal Histochemistry, 4th ed." ed. by P.J. Stoward and A.G.E. digestion of collagen fragments which have been split by Pearse, Churchill Livingstone, London, 1991, pp. 281-335. 12. Nikawa, T., Towatari, T. and Katunuma, N.: Purification and osteoblastic neutral collagenase. characterization of cathepsin J from rat liver. Eur. J. Biochem. 204; 381-393, 1992. 5. DPP-IV (EC 3.4.14.5, also called CD 26) 13. Ohsawa, Y., Nitatori, T., Higuchi, S., Kominami, E. and The present study using decalcified chicken bone Uchiyama, Y.: Lysosomal cysteine and aspartic proteinases, tissues failed to demonstrate it. Sannes and colleagues acid phosphatase, and an endogenous cysteine protease inhibitor, cystatin-beta, in rat osteoclasts. J. Histochem. [15] also did not observe it. However, an immunohisto- Cytochem. 41; 1075-1083, 1993. chemical study showed DPP-IV in osteoclasts [2]. The 14. Sannes, P.L.: The histochemical and cytochemical localiza- activity might be inhibited by an EDTA-treatment. tion of proteases. Prog. Histochem. Cytochem. 18; 1-48, 1988. The enzyme is a serine exopeptidase, and degrades 15. Sannes, P.L., Schofield, B.H. and McDonald, D.F.: substance P, casomorphin 5, kentsin, alpha-chain of Histochemical localization of cathepsin B, dipeptidyl peptidase 286 Fukushima and Yamashita

I, and dipeptidyl peptidase II in rat bone. J. Histochem. pp. 193-249. Cytochem. 34; 983-988, 1986. 19. Vaes, G.: Cellular biology and biochemical mechanism of bone 16. Sasaki, T. and Ueno-Matsuda, E.: Cystein-proteinase localiza- resorption. Clin. Orthop. Rel. Res. 231; 239-271, 1988. tion in osteoclasts: An immunocytochemical study. Cell Tissue 20. Van Noorden, C.J.F., Vogels, I.M.C. and Smith, R.E.: Res. 271; 177-179, 1993. Localization and cytophotometric of cathepsin B activity in 17. Simmons, D.J. and Grynpas, M.D.: Mechanisms of bone for- unfixed and undecalcified cryostat sections of whole rat knee mation in vivo. In "Bone, vol. 1, The Osteoblast and joints. J. Histochem. Cytochem. 37; 617-624, 1989. Osteocyte", ed. by B.K. Hall, The Telford Press, Caldwell, 21. Wang, L.L.J. and Cooper, M.D.: cDNA cloning and expres- 1990, pp. 193-302. sion of human glutamyl aminopeptidase (aminopeptidase A). 18. Smith, R.E. and Van Frank, R.M.: The use of amino acid Genomics 17; 657-664, 1993. derivatives of 4-methoxy-beta-naphthylamine for the assay and 22. Wu, L.L.Q., Wang, J., Bucy, R.P. and Cooper, M.D.: subcellular localization of tissue proteinases. In "Lysosomes in Widespread tissue distribution of aminopeptidase A, an evolu- Biology and Pathology, vol. 4.", ed. by J.T. Dingle and R.T. tionarily conserved ectoenzyme recognized the BP-1 antibody. Dean, North Holland Publishing Company, Amsterdam, 1975, Tissue Antigens 42; 488-496, 1993.