Cell Biology International 1998, Vol. 23, No. 1, 41–49 Article No. cbir.1998.0316, available online at http://www.idealibrary.com on
TISSUE TRANSGLUTAMINASE EXPRESSION IN QUAIL EPIPHYSEAL CHONDROCYTES ELISA GIONTI1,3*, MASSIMO SANCHEZ2, ANTONIETTA ARCELLA3, GIANFRANCO PONTARELLI3, SIMONA TAVASSI4, VITTORIO GENTILE4, ANNA COZZOLINO4 and RAFFAELE PORTA5 Dipartimento di Medicina Clinica e Sperimentale, Via T. Campanella, 88100 Catanzaro, Italy; 2Istituto Superiore di Sanita`, viale Regina Elena, 299–00161 Roma, Italy; 3Dipartimento di Biochimica e Biotecnologie Mediche, Universita` di Napoli ‘Federico II’, Via S. Pansini, 5–80131 Napoli, Italy; 4Dipartimento di Biochimica e Biofisica, II Universita` di Napoli, Via Costantinopoli, 16–80138 Napoli, Italy; 5CIRPEB Centro Interdipartimentale di Ricerca sui peptidi bioattivi, Universita` di Napoli ‘‘Federico II’’, via Mezzocannone, 4, 80138 Napoli, Italy
1
Received 31 March 1998; accepted 25 September 1998
Tissue transglutaminase (tTGase) is a GTP-binding Ca2+ -dependent enzyme which catalyses the post-translational modification via å(ã-glutamyl)lysine bridges. The physiological role of tTGase is not fully understood. It has been shown that in cartilage the expression of tTGase correlates with terminal differentiation of chondrocytes. Recent evidence suggests that the GTP-binding activity of tTGase may play a role in the control of cell cycle progression thus explaining some of the suggested roles for the enzyme. tTGase activity is present in primary cultures of epiphyseal chondrocytes and increases transiently upon retinoic acid (RA) treatment. Increase in enzyme activity occurs upon RA addition and is accompanied by a parallel increase in protein and mRNA levels. Stimulation of tTGase expression by RA correlates with suppression of cell growth and occurs independently of cell adhesion and cell differentiation. tTGase expression is not observed in MC2, a permanent chondrocyte cell line derived from retrovirus infected chondrocytes. RA treatment fails to activate tTGase expression in MC2 cells and to completely suppress cell proliferation. Our findings lend support to the idea that tTGase might play a role in non-dividing cultured 1999 Academic Press chondrocytes. K: retinoic acid; tissue transglutaminase; cell adhesion; cell growth; chondrocyte phenotype.
INTRODUCTION Transglutaminases (TGases, EC 2.3.2.13) are a family of calcium-dependent enzymes which catalyse post-translational modifications of proteins by the introduction of an isopeptide bond between the ã-carboxamide group of proteinbound glutamines and the å-amino group of protein-bound lysines (for reviews see Folk, 1980; Lorand and Conrad, 1984; Greenberg et al., 1991; Aeschlimann and Paulsson, 1994). Various transglutaminases have been characterized both in *To whom correspondence should be addressed: Dipartimento di Biochimica e Biotecnologie Mediche, Via S. Pansini, 5-80131 Naples, Italy; E-mail:
[email protected] 1065–6995/99/010041+09 $30.00/0
intracellular and extracellular compartments (Aeschlimann and Paulsson, 1994). Tissue transglutaminase (tTGase) is a predominantly cytosolic enzyme whose expression is regulated by all transretinoic acid (RA) in vivo and in vitro. Induction of tTGase occurs soon after the addition of RA; the mediators of RA-induced tTGase expression are the nuclear receptors for RA, RAR-á, RAR-â and RAR-ã, as well as the nuclear receptors for 9-cisretinoic acid, RXR’s (Moore et al., 1984; Chiocca et al., 1988; Nara et al., 1989; Piacentini et al., 1992; Nakanishi et al., 1991; Zhang et al., 1995; Nagy et al., 1995). The physiological function of tissue transglutaminase has been associated with the activation of the programmed cell death 1999 Academic Press
42
process (apoptosis) (Fesus et al., 1991; Nagy et al., 1995). The enzyme, however, can also be associated with the cell membrane to exert a receptor-coupled signaling function (Nakaoka et al., 1994; Mian et al., 1995; Chen et al., 1996) or expressed at the extracellular surface of many cells (Barsigian et al., 1991; Martinez et al., 1994). Several extracellular matrix components are tTGase substrates (for review see: Aeschlimann and Paulsson, 1994). In fibroblasts and in endothelial cells, the enzyme plays a role in the extracellular matrix stabilization (Martinez et al., 1994), cell adhesion and cell morphology (Gentile et al., 1992; Jones et al., 1997). The expression of tTGase in cartilage tissues correlates with the terminal differentiation of chondrocytes (Aeschlimann et al., 1993). Cartilage development proceeds through distinct stages starting from mesenchymal cell condensation and resulting in the differentiation of resting chondrocytes. During endochondral ossification, resting chondrocytes undergo a further process of differentiation/maturation involving proliferation, hypertrophy, calcification, degradation and replacement of cartilage by bone and marrow (for reviews see Poole, 1991; Sandberg, 1991; Hunziker, 1994; Erlebacher et al., 1995). In developing long bones tTGase expression correlates with chondrocyte maturation (Aeschlimann et al., 1993). The enzyme is externalized from hypertrophic chondrocytes and might be involved in matrix cross-linking before cartilage undergoes calcification. Furthermore, osteonectin and collagen II are target proteins in hypertrophic chondrocytes (Aeschlimann et al., 1993; Hohenaldl et al., 1995). Extracellular matrix of cartilage contains a variety of proteins including cartilage specific proteoglycan aggrecan, collagen II, tenascin and several minor collagens (Hascall, 1988; van der Rest and Garrone, 1991; Mayne and Brewton, 1993; Mackie et al., 1987). The expression of genes for extracellular matrix components varies during cartilage differentiation/maturation process (for reviews see Poole, 1991; Sandberg, 1991; Hunziker, 1994; Erlebacher et al., 1995). Mesenchymal cells produce collagens I and III; chondrocytes synthesize aggrecan and collagens II, IX, XI and X. The phenotype of cartilage cells undergoes extensive modulation in cell culture as well and the mechanisms regulating the differential expression of extracellular matrix genes are largely unknown (for reviews see Ramirez and Di Liberto, 1990; Petit et al., 1992). Retinoids are a well established means of studying the regulation of extracellular matrix genes in cultured chondrocytes (Benya and Padilla,
Cell Biology International, Vol. 23, No. 1, 1999
1986; Horton et al., 1987; Oettinger and Pacifici, 1990). Primary cultures of quail epiphyseal chondrocytes synthesize the cartilage-specific collagens type II and type IX. Treatment of suspension chondrocytes with retinoic acid inhibits the expression of cartilage collagens and induces cell adhesion along with fibronectin expression (Sanchez et al., 1991, 1993). RA-induced inhibition of the chondrocyte phenotype correlates with cell adhesion. In fact, RA-chondrocytes undergo their phenotypic modulation as they begin to attach. Further, prevention of cell adhesion blocks many of the effects on the chondrocyte phenotype induced by RA (Sanchez et al., 1996). Here we used primary cultures and a permanent cell line of chondrocytes to study tissue transglutaminase expression and its relationships with cell proliferation, cell differentiation and cell adhesion.
MATERIALS AND METHODS Cell culture Primary chondrocytes were isolated from day 10 quail embryo tibiae as previously described (Gionti et al., 1985) with the following modifications. Tibiae were incubated for 1 h in a saline solution containing trypsin/collagenase and the cells released after this incubation are discarded as they consisted mainly of perichondrial cells. Epiphysis were dissected, minced and incubated at 37C in fresh enzyme mixture; cells released by sequential enzymatic digestion from about 50 embryos were pooled and seeded at low initial cell density (2000 cells/cm2). Floating cells were resuspended in Coon’s modified F12 medium (AmbesiImpiombato et al., 1980) supplemented with 10% foetal calf serum. Expression of the differentiated phenotype was assessed by SDS-PAGE of metabolically labeled proteins. all trans-RA (generously provided by Hoffmann-La Roche) was dissolved in 95% ethanol and stored at 80C in the dark. On day 0, this solution was diluted with growth medium and control cultures received an equivalent amount of ethanol. After addition of 0.5 ì RA, cells were plated at 1105 cells/ml. No toxic effect was ever detected with this concentration of RA in treated cultures. All the experiments described in this report were carried out with ordinary foetal calf serum. Parallel experiments in which delipidised serum was used in either control or RA treated cultures did not show significant differences
Cell Biology International, Vol. 23, No. 1, 1999
in tissue transglutaminase activities. To prevent RA-induced cell adhesion, culture dishes were coated with a film of agarose. The continuous cell line MC2 was derived from quail epiphyseal chondrocytes infected with avian myelocitomatosis virus/Rous associated virus-1, subgroup A -MC29(RAV-1)-, a retrovirus carrying the v-myc oncogene (Gionti et al., 1985). Primary epiphyseal chondrocytes can grow either in suspension or in monolayer. Because of their higher sensitivity to viral infection, chondrocytes grown in monolayer were originally used in infection experiments; here, monolayer cells are used as the normal counterpart of the infected cell line. DNA probes The tTGase probe was a full-length clone harbouring a 3200-bp cDNA insert isolated from a cDNA library made from chicken heart (kindly provided by Drs P. Davies and V. Thomazy, Houston, TX, U.S.A.). The coding and the 3 untranslated region shows high homology to the cDNA clone isolated by Weraarchakul-Boonmark et al. (1992). The RAR-â2 probe used was g RAR-â, harbouring a 1600-bp cDNA insert including the entire coding region and the 3 untranslated region of the chicken gene (Smith and Eichele, 1991). The GAPDH probe was a full-length cDNA clone (1200 bp) encoding rat glyceraldehyde-3-phosphate dehydrogenase (Fort et al., 1985). The inserts of these plasmids were used as probes. High specific activity random primed probes were prepared with the Promega kit as specified by the supplier. RNA extraction and Northern blot hybridization Total cellular RNA (2107 cells/sample) extracted as previously described (Chomczynski and Sacchi, 1987), was denatured, fractionated on formaldehyde/agarose gels (Lehrach et al., 1977) and blotted onto an Amersham Hybond N nylon membrane. Pre-hybridization and hybridization were performed in 500 m NaH2PO4 pH 7.2, 7% SDS, 1 m EDTA for 30 min and 16 h, respectively; the hybridization temperature was 65C. Filters were washed three times at 65C in 50 m NaH2PO4 pH 7.2, 1% SDS. Washed filters were dried and exposed to Fuji films with intensifying screens at 80C. Densitometric scanning analysis was performed using software from NIH-Image. Transglutaminase assay Cells (3106 cells/sample) were mechanically removed from dishes, rinsed twice with PBS, resus-
43
pended and homogenized in Tris–HCl 20 m (pH 7.4) containing 2 m EGTA, pepstatin (0.1 ìg/ 100 ìl), leupeptin (0.1 ìg/100 ìl), PMSF 0.1 m, bestatin (0.1 ìg/100 ìl) and sonicated at 4C for 20 s. Transglutaminase activity was measured by detecting the incorporation of [3H]spermidine into N,N -dimethylcasein as previously reported (Lorand, 1972). The incubation mixture contained 125 m Tris–HCl (pH 8.0), 2.5 m CaCl2, 10 m dithiothreitol (DTT), 200 ìg N,N -dimethylcasein, 12.5 ì spermidine containing 0.15 ìCi [3H]spermidine and 5 to 25 ìg of protein from total cell homogenate in a final volume of 0.1 ml. After 30 min incubation at 37C, the samples were precipitated with 10% TCA. Free [3H]spermidine was eliminated by washing with large volumes of cold 10% TCA containing 1 m spermidine. The pellet was resuspended and bound [3H]spermidine was counted with 5 ml of Pico Fluor 30 scintillation cocktail (United Technologies/Packard). Chondrocytes enzyme activity was inhibited by adding EGTA thus suggesting that it is calciumdependent. Western blot analysis Determination of tissue transglutaminase protein was carried out with condrocytes lysates by Western blot analysis. tTGase-positive bands were revealed by using as a primary antibody an affinity purified IgG raised in goat against purified guinea pig liver tTGase proven to be not cross-reactive with other forms of transglutaminase (Chiocca et al., 1988). Statistical analysis Statistical significance of tTGase mRNA levels between control and RA-treated chondrocytes were calculated using Student’s t-test. RESULTS Retinoic acid stimulates tTGase expression in primary chondrocytes Like chondrocytes from many species, quail epiphyseal chondrocytes (QEC) can grow either in monolayer as polygonal cells or in suspension. Since suspension quail chondrocytes convert poorly into attached cells, they are a suitable system in which to study the alterations in cellmatrix interactions that occur during RA-induced changes in cell shape (Sanchez et al., 1991).
44
Cell Biology International, Vol. 23, No. 1, 1999
Fig. 1. Morphology of control and RA treated primary chondrocytes. A: control untreated chondrocytes (the picture was taken 3 days after plating); B, C and D: chondrocytes treated with 0.5 ì RA for 1, 2 and 3 days, respectively. Cells were photographed under phase contrast: magnification 75.
Exposure of suspension QEC to RA for three days blocks cell growth, inhibits the expression of cartilage-specific collagens type II and IX, turns on fibronectin expression and induces cell adhesion (Sanchez et al., 1991, 1996). One day after RA addition about 50% of the cells adhere to the substrate; by 2 days of treatment adhesion is almost complete (Fig. 1). By 3 days many cells display a dedifferentiated flattened morphology consistent with the results obtained by Hassel et al. (1979). Transglutaminase activity was found in chondrocyte lysates (Table 1). Enzyme activity increases upon RA treatment and reaches its maximum after around 16 h of treatment then gradually declines to slightly lower values than the base-line activity detected in control chondrocytes. Stimulation of enzyme activity is accompanied by parallel and equivalent increases of a 77-kDa band as detected by Western blot analysis in cell lysates from control and RA treated chondrocytes (not shown). A single 3.2-kb band was detected by hybridization of total RNA from suspension QEC with a cDNA clone encoding chicken tissue-transglutaminase (Fig. 2). In the experiment shown in Figure 2, stimulation of tTGase expression occurs between 4 and 8 h after RA addition and remains unchanged over the subsequent 16 h. Thereafter, tTGase mRNA levels decline until at 72 h they reach the basal levels detected in untreated control chondrocytes (the results are not shown but have been made available
to the referees). RAR-â2 expression is readily induced upon RA addition (Fig. 2), consistently with the results obtained in chicken sternal chondrocytes by Iwamoto et al. (1993a). RA-treated chondrocytes display a 2–4-fold increase of tTGase mRNA over the levels detected in untreated control chondrocytes as evaluated in different experiments by densitometric scanning analysis and normalization to GAPDH expression (Table 2). Cell adhesion does not correlate with tTGase expression in untreated chondrocytes: similar levels of tTGase expression were observed in Table 1. Transglutaminase activity in chondrocyte homogenates from primary cultures treated with 0.5 ì RA for the indicated times Treatment (h)
Control 8 16 24 40 72
Transglutaminase activity* (pmoles of [3H]spermidine incorporated/mg protein) 26.00.49 45.50.88 66.91.37 52.31.21 43.80.71 15.61.4
*Values represent the mean value of three different experiments carried out by triplicate determinations.. Further experimental details are reported in the Materials and Methods.
Cell Biology International, Vol. 23, No. 1, 1999
45
Fig. 2. Time course of RA-induced stimulation of tissue transglutaminase and RAR-â2 genes in primary cultures of quail epiphyseal chondrocytes. Steady-state levels of tTGase, RAR-â2 and GAPDH mRNAs in control chondrocytes (lane 1) and in chondrocytes treated with 0.5 ì RA for 4, 8 and 24 h (lanes 2, 3 and 4, respectively).
untreated cultures growing either in suspension or in monolayer as judged by Northern blot (Fig. 2, lane 1; Fig. 3, lane 1), by Western blot analysis and by enzyme activity assays (not shown). To investigate whether the stimulation of tTGase expression correlates with RA-induced cell Table 2. Relative levels of tTGase mRNA (fold of stimulation) in RA-treated chondrocytes
Exp. 1 Exp. 2 Exp. 3 Exp. 4 Exp. 5 Exp. 6 Exp. 7 Means ..
Monolayer
Suspension
3.5 1.6 3.45 4.4 — — 3 3.19** 1.023719
2.75 — — 2.25 4.8 2.75 — 3.1375* 1.133119
Primary chondrocytes were cultured either in permissive conditions for RA-induced cell adhesion (Monolayer) or in non-permissive conditions for RA-induced cell adhesion (Suspension). Total RNAs extracted from control and RA-treated chondrocytes were isolated and analysed by Northern blot for tTGase expression. tTGase and GAPDH bands were scanned and the scan units were converted into -fold stimulation over the basal levels detected in control chondrocytes. In the experiments 1, 3, 6, the reported values refer to stimulations obtained after 6 h of RA-treatment; in the other experiments (2, 4, 5, 7) the reported values refer to shorter treatments (4 h). Significant differences from controls are designated as *P
