Immunohistological analysis of transglutaminase factor XIIIA expression in mouse embryonic growth plate

ELSEWER

Journal of Orlliopaedic Research 20 (2002) 575-578

Journal of Ort hopaedic Research www.eIsevier.com/locatelort hres

Immunohistological analysis of transglutaminase factor XIIIA expression in mouse embryonic growth plate

Abstract Previously we demonstrated the expression of Factor XIIIA (FXIIIA), a coagulatioli traasglutaminase, in avian embryonic growth plate. To explore whether FXIIIA is also expressed by chondrocytes of the iiiamliialian cartilage anlagen of bones, we analyzed tlie iiiouse embryonic growth plate by iiiiinuiiostainilig using anti-FXIIIA antibodies developed against human zind chicken proteins. We revealed the expression of FXIIIA in the epipliyseal growth plate, where FXIIIA appears first intracellularly in the zone of proliferation/iiiaturatioii, and remains intra- and extracellularly throughout the hypertrophic zone. Externalizatioii of FXIIIA O C C L I ~ Sbefore mineralization. Transglutaminase activity was assayed in organ cultures using rliodamine-labeled synthetic substrate Pro-Val-Lys-Gly. Enzyiiiatic activity shows a restricted distribution in cartilage and correlates with FXIIIA expression pattern, suggesting that cartilagenous transglutaminase activity is due, at least partially, to the FXIIIA isoform. We conclude, that coagulation factor FXIIIA is expressed by chondrocytes of embryonic mouse long bone cartilages in a strictly regulated pattern. which correlates with chondrocyte differentiation and matrix mineralization. 0 2002 Orthopaedic Research Society. Published by Elsevier Science Ltd. All rights reserved. K q ~ i ~ ~ o Mouse r ~ l . ~ ;embryo; Cartilage; Transgltitaiiiinase FXIIIA

Transglutaiiiinases (TGases) form a fainily of eiizynies, which catalyze the reaction of g-glytaniyl-e-lysine cross-link forination (reviewed in [ 1,5]). In maininals at least seven isoforins, with a wide variety of tissue distribution and proposed physiological functions, have been described. Transglutaminase activity in cartilage and bone is thought to be correlated with matrix mineralization. I n normal development, the tissue forin of transglutaminase (TGase 11) is up-regulated in the hypertrophic zone of the growth plate of epiphyseal bones [2,16]. I n aging articular cartilage TGase I 1 is involved in pathologic iiiineralization that results in calcium pyrophosphate dehydrate crystal formation [14]. 0 t h pathologic processes in articular cartilage, e.g., osteoarthritis, may also result in abnormal mineralization, which might involve transplutaiiiinase activity [6,8]. The tissue form of transglutaminase is expressed in many cells and tissues and has been ascribed a role i n many physiological cellular events. Recently, the mice ‘Corresponding author. Tel.: +1-617-636-6652; fax: +I-617-6366536. E-rr iiril (rt/ihiw; ilia I-ia.notiriiii nsk aia@t ti ft s.ed ti ( M , V , N ti riii in skaya).

lacking the TGase I1 gene have been generated by two groups. Surprisingly, in both cases the aiiiinals are vital, develop and reproduce normally, and have no noticeable abnormalities in skeleton development [3,10]. Since many tissues express inore than one isoform of transglutaminase with very similar substrate specificity, the observed phenotype in the knockout mice can be explained by redundancy in transglutaminase action. Therefore, we sought to analyze whether chondrocytes of the mouse growth plate express, in addition to tissue transglutaiiiiiiase, other isoforins of the enzyme. Previously, we reported the expression of a coagulation trangslutami~iase,Factor XIIIA, in the hypertrophic choiidrocytes of avian embryonic growth plate, along with the tissue isoform [I I]. This study was undertaken to analyze the expression of FXIIIA i n niaminaliaii growth plate, using mouse embryo as a model. To analyze the distribution of the FXIIlA protein in the developing growth plate, we employed immunofluorescence on sections of tlie growth plate of 18-20 day mouse embryos. For this we used the following polycloiial antibodies: one antibody against human FXIIIA (commercially available from Calbiochem, catalog nuinber 233502, used a t I300 dilution), and two-against

0736-0166/02/$ - see l‘ront matter 0 2002 Ortholxiedic Research Society. Published by Elsevicr Science Ltd. All rights reserved PII: S 0 7 3 6 - 0 2 6 6 ( 0 1 ) 0 0 1 4 6 - 2

chicken FXIIIA antibody Acl6 which detects the activation peptide and therefore is specific for thc zyniogen foi-in, and antibody A3, which detects the C-terniinal portion of the molecule (kindly provided by Dr. Nimpf, used at I :I00 dilution as described elsewhere [ 131). None of these antibodies cross-reacts with the purified cominercially available guinea pig liver transglutaniinase (TGase 11), while all three show reactivity with mouse plasma samples, as dcterniiiicd by Western blot analysis (data not shown). Moreover, as the tissue form of transglutaminase has no activation peptide or any sequcnce homologous to it, tlie use of Ac16 antibody, generated against the activation polypeptide of FXIIIA, confirms thc specificity of imniuiioliistoclieiiiical analysis. In the embryonic mouse growth plate the FXIIIA (Fig. 1, panel A) is not present within tlie proliferative zone, appearing first within tlie cells of tlie zone of maturation, as revealed by cross-reactivity with tlie antihunian FXIllA antibody. Its expression progressively increases throughout this region, reaching a maximum signal in the prehypertrophic zone. This level is retained throughout the zone of hypertrophy, as shown at higher magnification in Fig. 1(C). Externalizatioii of enzyme occurs before mineralization, and FXIIIA protein is detected in tlie zone of hypertrophy both intra- and extracellularly (Fig. l(C)j. Since no antibody exists which distinguishes the active form of the enzyme from tlie zyniogen form, we chose to localize tratisglutaiiiinase activity itself by binding of tlie rliodaniiiie-labeled synthetic substrate peptide Pro-Val-Lys-Gly (SY2011). This peptide has previously been shown to be a good traiisglutaiiiinase substrate, a s determined by its ability to serve a s competitor in a transglutaminase activity assay involving the inco rpora tion/cross- I i n king of C 4 - pt resci ~ ~ ne into casein, while others of a similar size but lacking the putative traiisglutaiiiinase substrate sequence were not cross-linked by transglutaiiiinases ([9]; our unpublished results). Organ cullures of 18-20 day mouse embryonic long bones were incubated for 3-1 2 11 in medium containing the rliodaiiiiiie-coiijugated SY201 1 peptide. Then the cultures were fed with fresh medium without tlie peptide, and following an additional incubation for 12-24 ti wcrc processed for cryostat sectioning. Sections of such cultures showed incorpoi-ation of the peptide in both the matrix and tlie cells of tlie hypertrophic zone, but reactivity was absent from non-hypertrophic cells such 21s those of tlie proliferative zone (Fig. l(C)).The overall distribution of tlie peptide incorporated into cartilagenous protcins corresponds to that observed for FXIIIA. In addition, the absence of activity 111 the nonhypertrophic cartilage confirms that the retention of the pcptide within thc hypertrophic region does reflect cioss-linking, and not, for example. just artifactual Irnpping of tlie probe.

'

Our previous studies on embryonic chicken growth plate idcntified FXJIIA as a gene up-regulated during cliondrocytc differentiation, with the increase in mRNA level at least fivefold in hypertrophic versus non-hypertrophic cartilagcs [I 1 ,I 21. Recently, we have confirmed these results by iinniuiiohistocheniistry (in press), using two antibodies against chicken FXIIIA protein [13]. The data obtained in this study confirm that FXIIIA expression is characteristic for differentiating chondrocytes i n mammalian, as well as avian species. To confirm tlie nature of observed imniunostaining, we employed antibodies against whole human FXIIIA molecule along with antibodies against chicken FXllIA peptides corresponding to the activation peptide and to tlie carboxyl-terminus. All antibodies revealed the same pattern of FXI IIA localization in developing growth plate of long bones, showing high enrichment of FXIIIA protein in the hypertrophic chondrocytes, with subsequent release into extracelluaIr matrix. We conclude that transglutaminases, in particular the FXIIIA form of the molecule, are likely to be involved in the formation of long bones through activities in tlie growth plate. Tlie main source of plasma FXIIIA is the bone marrow; however, the studies by others suggested that tlie bone marrow is not tlie only source for FXIIIA synthesizing cells. Tlie mRNA for FXIIIA has been detected at various levels in many other tissues, including placenta, kidney, lung, skeletal muscle and heart [7]. The protein expression has been shown by iminunostaining in several cell types in embryonic human skin [4], and several other embryonic connective tissues, including tendosynovial anlagen and tlie stroma of chorionic vilIi [ 151. Our data identify differentiating chondrocytes as a new cell type producing transglutaminase FXIIIA. It is noteworthy, that Silverinan and Knapik [I 51 reported tlie appearance of FXIIIA protein in tlie cliondroblasts of hunian embryonic bone anlagen synchronously with primary matrix mineralization prior to invasion of the lacunae by the perichondrial capillary network. More recently, Johnson et al. [6] also reported expression of coagulation transglutaminase FXIIIA in tlie human fetus growth plate at 160 days gestation. Therefore, expression of FXIIIA in embryonic long bones is common for all higher inammatian species. The presence of at least two transglutaminase isoforms in the embryonic mouse cartilage, e.g., FXIIIA and TGaseII, results in redundancy of traiisglutaiiiiiiase activity. The extent to which either traiisglutaiiiinase contributes to cartilage maturation is yet to be seen. Since m i t rix cross-1i n ki ng by t ransglu taminase correlates with tissue mineralization, not only in normal development, but also i n pathological processes such as artherosclerosis [S], our data on identification of FXIIIA cxpressioii i t i d ifferent i at ing chond rocy t es is import a n t for further iinalysis of its role in this disease. Recent

577

Fig. I , FXlllA protein localization i n the mouse embryonic growth plate. (A) Foriiialdeliyde-fixed cryostat 6 mm sections 01' IS day iiioiise embryo hyaluronidase (Sigma) treatment (at 37 "C for 30 miti), and immtinoliistocliemical labeling with antibody ;against long bones were subjected to 0. 15'%~ huiiinn FXI IIA (Calbiochem). followed by Ruorescent detection. HSrE staining in the right panel allows the identification o f ditkrent zones in the growth plate, as indicated on the side (M-zone of maturation; Hyp-zone of IiyperIrophy). (B) 11iiiiiunoliistocliemic;ll labeling 01' serial tissue sections with anti-chicken FXIllA antibodies. with Acl6 antibody detecting the ;activation peptide. and A3 antibody detecting the C-terminus o1'thc molecule. (C) Co-localization of tr~iiigslut~iiiiiiiase activity With FXlllA protein. Rliod;iiiiine-l;ibeled 1r;insglutaminase substrate SY20 I I was incubated at 5 pM with 20 day niouse embryo long bones for 1.5 It. and then kept in Tresh tinlabeled medium for 3 11. Dissection and immllnost~iininpwith anti-hiiman FXIIIA were perl'ormed :as described above. Some or the cells positive for substrate incorporation and FXlllA protein ure aiinrketl with (*). Nuclear staining by Hoechst dye is shown for tissue orientatioii. and to show that cells i n pre-hypertrophic zone are negative for both the FXllIA staining and transgltitaiiiinase substrate incorporation. (A, B) l o x magnification; (C) 20x magnilic~ation.

studies on FXIIIA up-regulation in human osteoarthritic tissues and its possible correlation with pathological mineralization support this suggestion [6]. The data described i n this report will allow thc dctailed analysis of this correlation using iiiouse model of osteoarthritis.

Acknowledgements

The authors are grateful to Dr. Niinpf for providing antibodies. Dr. Nuriiiinskaya is the recipient or Charles A. King Trust fellowship and of CHarIron Faculty Research Award. The study \YX also supported

by National Institutes of Health, Grant number HD023681-12.

References [I] Aeschliniann D, Thomazy V. Protein cross-liiiking in assembly and reniodelling of extracellular matrices: the role of transglutamines. Connect Tissue Res 2000;41:1-27. [2] Aescliliniann D, Wetterwald A, Fleisch H, Pausson M. Expression of tlie tissue transglutaininase in skeletal tissues correlated with events of terminal differentiation of chondrocytes. J Cell Biol 1993:12O: 1461-70. [3] De Laurenzi V, Melino G. Gene disruption of tissue transglutaniinase. Mol Cell Biol 2O01;21:148-55. [4] Gibraii NS, Nickoloff BJ, Holbrool KA. Ontogeny and cliaracterization of factor X l l l a + cells in developing human skin. Anat Embryo1 1996;193:3541, [5] Greenberg CS, Bii-ckbicliler PJ. Rice RH. Traiisglutaminases: iiiultiftinctional cross-linking enzymes that stabilize tissues. FASEB J 1991:5:3071-7. [GI Johnson K. Hashimoto S. Lotz M. Pritzker K, Terkeltaub R. Interleukin-l induces pro-mineralizing activity of cartilage tissue traiisglutaiiiinase and Factor XIIIa. Am J Pathol2001;159:149-63. [7] Kaczmarek E, Liu Y, Berse B, Chen CS. McDonagh J. Biosynthesis of plasma factor XIII: evidence for transcription and tl.anslation in hepatoma cells. Biochim Biophys Acta 1995; 1247: 121-34.

[S] Karpouzas GA, Terkeltaub RA. New developmeiits in the pathogenesis of articular cartilage calcification. Curr Rheumatol Rep 1999;I :12 1-7. [9] Kim SY, Park WM, Jung SW, Lee J . Novel transg1tit;iminase inhibitors reduce tlie cornified cell envelope formation. Biocliem Biophys Res Coniniun 1997;233:3944. [lo] Nanda N, Iismaa EE, Owens WA, Husain A, Mackay F, GI.nliam A. Targeted inactivation of Ghltissue traii~gl~itaminase 11. J Biol Chem 2001;276:20673-5. [ I I ] Nurminskaya M, Liiiseiiiiiayer TF. Analysis of up-regulated genes during chondrocyte hypertrophy. Ann N Y Acad Sci 1996; 785:309-10. [ 121 Nurniinskaya M, Magee C, Nurminsky D. Linseniiiaper TF. Plasma tl-ansglutaiiiinase in hypertrophic chondrocytes: expression and cell-specific intracellular activation produce cell death and externalization. J Cell Biol 199S;142:1 13544. [I31 Recheis B, Osiunger A, Haubenwallner S. Schneider WJ, Niinpf J. Chicken coagulation factor XIIIA is produced by the theca exlerna and stabilizes the ovarian follicular wall. J Biol Cheiii 2000;275:35320-7. [ 141 Rosenrbal AK, Derfus BA, Henry LA. Transglutaiiiiiiase activity in aging articular chondrocytes and articular cartilage vesicles. Arthritis Rheum 1997:40:966-70. [I51 Silverman JS, Knapik M. Fibroma of tendon sheath and thendosynovial giant cell tumors are rich in factor XIIIa+ dendrophages. J Histoteclinol 1996;I9:45-53. [ 161 Thomazy VA, Davies PJ. Expression of tissue tr~~iisgl~~ta~iiinase in the developing chicken linib is associated both with apoptosis and endochondral ossification. Cell Death Differ 1999:6:146-54.