ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS

Vol. 288, No. 2, August 1, pp. 488-494, 1991

Collagenase Synthesis by Cloned Rabbit Pulp Cells- Molecular and Immunological Identity of Pulp Cell and Fibroblast Enzyme’

Laura Goldberg, Masako Sakamoto, Yves Bolender,’ and Seizaburo Sakamoto3 Laboratory for the Study of Connective Tissue Metabolism, Harvard School of Dental Medicine, Boston, Massachusetts 02115

Received January 2, 1991, and in revised form March 15, 1991

Cultures of cloned rabbit pulp (RP) cells without stim- ulation produced collagenase of a concentration as high as reference rabbit skin fibroblast cultures which were stimulated with phorbol myristate acetate (PMA, 100 ng/ ml). The RP cell collagenase was compared with refer- ence fibroblast collagenase in Western blot analysis using monoclonal antibodies prepared against RP cell colla- genase and a polyclonal antibody prepared against rabbit fibroblast collagenase. Both enzyme preparations re- vealed, with either antibody, identical bands of approx- imate molecular masses 57,000, 52,500, and 45,000. These antibody preparations variously inhibited RP cell collagenase activity. Intracellular collagenase in RP cells in culture was demonstrated by the indirect immunoflu- orescence antibody technique using polyclonal anti-fi- broblast collagenase antibody. RNA samples from RP cells hybridized with rabbit fibroblast collagenase cDNA (clone H9) and showed a distinct band at 2.7 kb. Both control and PMA-stimulated RP cells and PMA-stimu- lated reference skin fibroblasts demonstrated strong cy- toplasmic hybridization between H9 and collagenase mRNA. The results indicate that RP cell collagenase is identical to rabbit fibroblast collagenase, and that the RP cell line provides a useful in vitro reference system for the study of collagenolysis in the rabbit model. o lesl

Academic Press, Inc.

A group of metalloproteinases with neutral optimal pH, including collagenase, stromelysin, and gelatinase, are

’ This research was supported in part by a grant from the National Institutes of Health (DE 05255).

‘Present address: Louis Pasteur University, 1 Place de L’Hopital 67000 Strasbourg, France.

3 To whom correspondence and reprint requests should be addressed at present address: Department of Preventive Dentistry, Tohoku Uni- versity School of Dentistry 4-1, Seiryo-machi, Sendai 980 Japan.

488

thought to be important for the degradation of connective tissue proteins. The concerted activity of these neutral metalloproteinases is required for extracellular degrada- tion at neutral pH. The best characterized of them is the enzyme collagenase [for a review see (1)] .

Various types of cells produce collagenase under certain physiological and pathological conditions. In particular, studies of fibroblast collagenase of human and rabbit or- igin have provided an insight into the molecular biology of this enzyme (2, 3). It is believed that even among the same type of cells, e.g., fibroblasts, collagenase synthesis and its activity level are regulated differently in various tissues under certain physiological or pathological con- ditions (l-3).

The presence of collagenase has been demonstrated in cultures of isolated fibroblast-like dental pulp cells (4) and cloned rabbit pulp (RP)* cells (5). Unlike ordinary fibroblasts, RP cells were found to produce relatively high levels of collagenase without a particular stimulant, such as phorbol myristate acetate (PMA) (3). The RP cells used in the above study were initially cloned from the odontogenic cells of a rabbit incisor before dentine for- mation (6). The study of procollagen synthesis by RP cells, which are diploid, revealed that these cells exhibit a unique predominance of type III collagen in culture, a distinction from the character of collagen in uiuo (7).

Although further characterization of the RP cell line remained to be studied, it appeared that this cell line would give rise to an ideal model for the study of colla- genase in dental pulp tissue and in connective tissue in general (5). Collagenase secreted by RP cells in culture

’ Abbreviation used: RP, rabbit pulp; PMA, phorbol myristate acetate; DMEM, Dulbecco’s modified Eagle’s medium; FBS, fetal bovine serum; RF, rabbit fibroblast; Tris, tris(hydroxymethyl)aminomethane; p-APMA, p-aminophenymercuric acetate; SDS-PAGE, sodium dodecyl sulfate- polyacrylamide gel electrophoresis; GIT, guanidine isothiocyanate; SSC, standard saline citrate; PBS, phosphate-buffered saline; DTT, dithio- threitol; fi-GP, &glycerophosphate.

0003-9861/91 $3.00 Copyright 0 1991 by Academic Press, Inc.

All rights of reproduction in any form reserved.

COLLAGENASE SYNTHESIS BY CLONED RABBIT PULP CELLS

was highly purified and shown to comprise three bands of approximate molecular masses 45, 52.5, and 57 kDa (5). Recently, a set of monoclonal antibodies against this enzyme was prepared and the above three bands were identified as derivatives of the same enzyme molecule by Western blot analysis using the monoclonal antibodies (8). This work extends the study of RP cell collagenase synthesis to its nuclear signal, compares RP cell colla- genase with fibroblast collagenase, and presents evidence suggesting that RP cell collagenase is identical to rabbit skin fibroblast collagenase.

EXPERIMENTAL PROCEDURES

Rabbit pulp cell culture. RP cells (kindly provided by Dr. A. Sato of Tokyo Medical and Dental University, Tokyo, Japan) were inoculated at 5 X lo4 cells/ml in T-25 flasks (unless otherwise specified), and grown in 5 ml of Dulbecco’s modified Eagle’s medium (DMEM), high glucose (GIBCO Laboratories, Grand Island, NY) supplemented with 10% fetal bovine serum (FBS) (Irvine Scientific, Santa Ana, CA), 100 unit/ml of penicillin, and 100 @g/ml of streptomycin, at 37”C, in a fully humidified atmosphere of 5% COz/air. The medium was changed at 2- or S-day intervals. Conditioned media were centrifuged to remove free cells, then pooled and stored at -20°C for subsequent collagenase analysis. Cultures prepared in different containers were not treated or treated with PMA at 10 or 100 rig/ml as indicated at appropriate sections.

Rabbit fibrobhst (RF) culture. Fibroblasts were isolated from skin pieces of a 2-day-old New Zealand white rabbit by a conventional method (9) using DMEM/lO% FBS culture medium. Fibroblasts outgrown from the skin pieces were collected by trypsinization, seeded, and cultured into T-25 flasks. Conditioned media were collected and processed as RP cell cultures as described above.

Isolation and assay of collagenuse. Conditioned medium samples from both RP cell and RF cultures were passed at room temperature through multiple columns (gel bed volume: 1 ml) of heparin-Sepharose 4B gel, which had been equilibrated with 0.05 M Tris-HCl buffer, pH 7.6, con- taining 5 mM CaCl,, to remove most of the serum proteins, gelatinase as well as stromelysin. The bound collagenase was eluted with 2.5 ml of Tris buffer containing 0.5 M NaCl and 0.5 M Ca(CH$O&, and the eluates were promptly desalted by gel filtration using PD-10 columns (Pharmacia Fine Chemicals Co., Piscataway, NJ) as described (5).

Since the isolated collagenase was virtually all latent enzyme, its ac- tivity was assayed after activation with 1 mM ofp-aminophenymercuric acetate (p-APMA), using reconstituted “C-labeled rat skin collagen fi- brils containing approximately 3000 cpm per assay as the substrate. The unit of collagenase activity was defined as described (5). The activity level was expressed as units/lo6 cells/24 h of the cultures.

Sodium dodecyl sulfate-polyacryhmide gel ekctrophoresis (SDS- PAGE) and Western blot analysis. Samples of both RP cell and RF collagenase were subjected after reduction or without reduction to SDS- PAGE according to the method of Laemmli (10). The protein bands were transferred electrophoretically onto a nitrocellulose membrane after the method of Towbin (11). and antigen bands were detected using the Immuno-Blot system from Bio-Rad Labs. (Richmond, CA).

Preparation of immunoglobulins (ZgGs). Monoclonal antibodies against RP cell collagenase were prepared by a standard method as described (8). Sheep anti-rabbit fibroblast collagenase antiserum (12) was kindly provided to us by Dr. C. E. Brinckerhoff (Dartmouth Uni- versity Medical School, Hanover, NH). IgGs were prepared from various monoclonal hybridoma culture media and sheep antiserum using Affi-Gel protein A MAPS II columns (Bio-Rad Labs.) as previously de- scribed (8).

Colkgeme inhibition test. Antibody IgGs were tested for their ability to inhibit collagenase. Partially purified RP cell collagenase (0.5 units

per assay) was mixed with various quantities of monoclonal or polyclonal IgGs, as well as control, nonimmune IgGs, and incubated for 2 h in the assay system described above. The concentrations of IgGs, at which approximately 50% inhibition of the added activity was observed, were chosen for the computation of inhibition activity of antibody.

Immunocytochemical demonstration of collagenase in RP cells. RP cells were seeded at 5 X 10” cells/ml in dishes (35 mm diameter, Falcon No. 1008) containing glass coverslips. After reaching confluency at 6 days of culture, RP cells were incubated further in fresh medium with or without PMA (100 rig/ml) for 6 h. The cells on glass coverslips were washed, air-dried, and fixed with 10% buffered formalin at 4°C for 15 min. The specimens were subjected to indirect immunofluorescence an- tibody staining for collagenase as described (13) using sheep anti-rabbit fibroblast collagenase antiserum as the first antibody and fluorescein- conjugated second antibody against sheep IgG (Miles Labs., Elk- hart, IN).

Preparation of mRNA from RP cells. RP cell mRNA was prepared by the method of Davis et al. (14). A brief description follows: larger cultures (80 cm2 flasks, NUNC) were grown to superconfluency, rinsed three times in PBS, then suspended and lysed in 4 M guanidine iso- thiocyanate (GIT) buffer containing 25 mM sodium acetate and 0.8% (V/V) 2-mercaptoethanol. Approximately 2.5 X lo7 cells were layered into each Beckman SW41 tube above a cushion of 4 ml of 5.7 M cesium chloride buffer containing 25 mM sodium acetate. The tubes were then filled with GIT buffer and ultracentrifuged overnight at 32,000 rpm at 20°C. The RNA pellets were resuspended in 0.3 M sodium acetate then precipitated and rinsed in 80% ethanol.

The poly(A)+RNA was isolated from above RNA samples by oligo(dT) cellulose (Type 3, Collaborative Res. Inc., Bedford, MA) chromatography, and its quantity was estimated by optical density at 260 nm.

Preparation of collagenase cDNA probe. The cDNA clone H9, which contains a 650.bp insert complementary to rabbit synovial fibroblast collagenase mRNA (15), was kindly provided to us by Dr. C. E. Brinck- erhoff. Clone H9 was amplified through large scale bacterial culture, then isolated and purified by a method of alkaline lysis. Digestion of the clone by the restriction enzyme PstI (Bethesda Research Labs., Gaithersburg, MD) released the collagenase insert from the vector plas- mid pBR322 in two sections: a 575bp fragment and a 75.bp fragment. The 575-bp fragment alone was recovered by electroelution and used as our probe. For Northern blot analysis, the probe was labeled with [32P]dTTP (New England Nuclear, Boston, MA) using a random primed labeling kit (Boehringer Mannheim, Indianapolis, IN). For hybridization with collagenase mRNA within RP cells, the probe was labeled with “S-dATP (NEN) also using the above kit. As a control probe, bacte- riophage X DNA provided with the kit was labeled similarly.

Northern blot analysis of isolated RP cell mRNA. The RNA was elec- trophoresed through a 1% agarose/formaldehyde gel and passively transferred overnight to a Zetabind membrane (Curio Laboratory Prod- ucts, Meriden, CT). The blot was washed, then baked for 2 h at 80°C under vacuum. Prehybridization and hybridization were each carried out overnight at 42°C in a shaking water bath. The composition of the prehybridization and hybridization solutions, as well as the sequence of posthybridization washes, was according to the method of Davis et al. (14). The final wash performed was a low stringency wash of 1X SSC + 0.1% SDS at 42°C. For autoradiography, the blot was contacted with Kodak XAR film using a DuPont Cronex Lightening-Plus intensifying screen and exposed for 3 h at room temperature.

Hybridization of collagenase mRNA in RP cells. For the preparation of monolayer specimens, RP cells were cultured in four-chamber-slide flasks (Lab-Tek). Cultures of RF stimulated with PMA (10 rig/ml) were also prepared as references.

The cells were fixed in 4% paraformaldehyde in PBS for 20 min, rinsed, dehydrated, and temporarily stored at 4°C. The cell slides were pretreated by being immersed in 0.2 N HCl for 20 min, incubated in 2X SSC at 70°C for 15 min, and postfixed in 4% paraformaldehyde in PBS for 5 min. The samples were acetylated, blocked, and dehydrated. Dried slides were stored overnight at -20°C with a dessicant. The hybridization

490 GOLDBERG ET AL.

0 10 100

FIG. 1. Collagenase synthesis in rabbit pulp (RP) cell and rabbit fi- broblast (RF) cultures and the effects of phorbol myristate acetate (PMA) on their collagenase synthesis. Confluent cultures (25-cm’ flasks) of RP cell and RF were prepared as described in the section and incubated for three days with or without PMA (10 or 100 rig/ml, indicated on abscissa). Hatched bars, RP cell cultures; open bars, RF cultures. The values are averages of duplicate cultures. Two other experiments of the same pro- tocol gave rise to similar results.

solution contained the following: 50% deionized fromamide, 0.3 M NaCl, 10 mM Tris, 1 mM EDTA, 1X Denhardt’s solution, 500 &g/ml yeast tRNA, 80 pg/ml salmon spleen DNA, 10% dextran sulfate, and 100 nM DTT.

After a 2-h prehybridization with this solution, each sample was cov- ered with 20 al of hybridization buffer containing either the collagenase probe or the control probe, sealed with a coverslip, and hybridized over- night in a humidified chamber at 42°C. The next day the slides were subjected to a series of increasing stringent washes and dehydrated. The slides were dipped in Ilford k5 Nuclear Research Emulsion diluted 1:l with deionized water. Slides were dried, incubated 3-7 days, developed, and stained with hematoxylin (16).

RESULTS

In both size plastic culture flasks, the RP cells had a spindle-shaped appearance at confluency at 6 days under the present culture conditions. Thereafter, the cells con- tinued to grow slowly to a thick and extremely tight monolayer. However, in the glass-bottom slide chambers, the cells often overlapped each other without forming a smooth monolayer. Several sites of multiple cell layers resembled those of osteoblast cultures, which undergo mineralization in the presence of /I-glycerophosphate (/3- GP) (17). The addition of P-GP to the RP cell cultures, however, did not give rise to mineralization (our unpub- lished results). Despite the difference in appearance of the cells in plastic and glass containers, both cultures produced comparable amounts of collagenase.

RP cells without stimulation synthesized and released fairly high levels of latent collagenase into the serum- containing culture medium. In contrast, skin fibroblast cultures released little or often undetectable amounts of collagenase under the same culture conditions. Under stimulation with PMA, however, the fibroblasts released significantly increased amounts of collagenase; these re- sults were consistent with previous studies (1,3). RP cells also released greater amounts of collagenase than the control when they were stimulated with PMA (Fig. 1).

Collagenase samples from both RP cell and RF, par- tially purified by heparin-Sepharose affinity chromatog- raphy, were examined by Western blot analysis using monoclonal and polyclonal antibodies. Both enzyme preparations revealed identical bands: 57, 52.5, and 45 kDa with different antibody preparations, indicating that these bands are derived from the same enzyme molecule and that RP cell collagenase and RF collagenase are iden- tical (Fig. 2).

Since a difference in antibody activity of various hy- bridoma clones against RP cell collagenase was observed in our previous study (8), these monoclonal antibodies were compared with a polyclonal antibody for their en- zyme inhibition and Western blot reaction. Although the monoclonal antibodies variously inhibited RP cell colla- genase activity, their inhibition activity was far less than that of the polyclonal antibody (Table I). Three mono- clonal clones (2A6,2A2,2E2) revealed Western blot bands with RP cell collagenase only after its reduction, not without reduction, and one monoclonal clone (F9) did not reveal a Western blot band at all. The polyclonal an- tibody revealed Western blot bands regardless of whether the antigen was reduced or not prior to electrophoresis (Table I and Fig. 2).

Intracellular collagenase in RP cells in culture was demonstrated by indirect immunofluorescence antibody

92.5 Kd -

66.2 Kd --

45.0 Kd -

31.0 Kd -

21.5 Kd --

4 4

4

FIG. 2. Western blots of RP cell collagenase and reference RF col- lagenase with either monoclonal antibody or polyclonal antibody. Lanes 1 and 4, RF collagenase from control culture without stimulation; lane 2, RF collagenase from PMA (100 rig/ml)-stimulated culture; and lanes 3 and 5, RP cell collagenase from RP cell culture without stimulation. Lanes l-3, reacted with polyclonal antibody; lanes 4 and 5, reacted with monoclonal antibody (2E2). The scale ofprotein molecular weight stan- dards (Bio-Rad Labs., Richmond, CA) is presented on the left. The bands of approximate molecular masses of 57,52.5, and 45 kDa (indicated with arrowheads) are apparent in lane 3. Although the bands of 45 kDa were revealed in most instances, they were often too weak to be repro- duced photographically. The bands of higher or lower molecular weight than these three bands may suggest a precursor and broken down prod- ucts of the enzyme.

COLLAGENASE SYNTHESIS BY CLONED RABBIT PULP CELLS

FIG. 3. Immunofluorescent localization of collagenase in RP cells. A, RP cell culture without stimulation; B, RP cell culture stimulated with PMA (10 rig/ml). A set of appropriate control stainings gave no specific fluorescence.

technique using the polyclonal antibody: sheep anti-rabbit fibroblast collagenase antiserum (Fig. 3). Collagenase im- munofluorescence observed in RP cells which had not been stimulated with PMA was as intense as in those stimulated with PMA. The control experiments using nonimmune sheep serum revealed no immunofluorescence (results not shown).

In Northern blot analysis, the probe was shown to rec- ognize collagenase mRNA isolated from RP cells. The probe, rabbit fibroblast collagenase cDNA, hybridized strongly with the RNA samples, forming a distinct band at 2.7 kb and a lesser band at 5 kb (Fig. 4).

TABLE I

Inhibition of RP Cell Collagenase, and Western Blot Reaction by IgGs Prepared from Hybridoma Culture Media

and Polyclonal Antiserum

IgGs

Monoclonal 2A6 F9 2A2 2E2

Control Nonimmune (mouse)

Polyclonal Control

Nonimmune (sheep)

Collagenase inhibition

(enzyme units neutralized per mg of antibody)

5.56 1.79

15.6 3.85

0 72.6

0

Western blot (approximate

molecular mass, kDa)

57, 52.5, 45 -a

57, 52.5, 45 57, 52.5, 45

-

57, 52.5, 45

’ No positive reaction was observed.

Hybridization of RP cell and RF cell collagenase mRNA was carried out by probing with the 3”S-labeled rabbit collagenase insert from the H-9 cDNA clone. The auto- radiography demonstrated the strong specific hybridiza- tion of this insert to the collagenase mRNA in RP cells and PMA-stimulated RF (Fig. 5). PMA-stimulated RP cells revealed even higher hybridization than nonstimu- lated RP cells. On the other hand, nonstimulated RF re-

28s

ias -- -- 2kb

5kb

5s

1 2 3

FIG. 4. Northern blot of RNA isolated from RP cells with rabbit fibroblast collagenase cDNA, clone H9. Lane 1, RP cell total RNA, 20 Kg; lane 2, RP cell total RNA, 10 wg; and lane 3, RP cell poly(A)+RNA, 4 Gg. The hybridization yielded an extremely strong band at approxi- mately 2.7 kb.

492 GOLDBERG ET AL.

FIG. 5. Hybridization of RP cell and RF collagenase mRNA with rabbit fibroblast collagenase cDNA, clone H9. A, RP cell culture without stimulation, reacted with collagenase cDNA. Note that abundant silver grains are present associated with cells and cell clusters. B, Dark field photomicrograph of the same area of the above specimen. C, RP cell culture similar to the above but reacted with control DNA. Note that random and sparse silver grains are present without association with cells. D, Dark field photomicrograph of the same area of the above specimen. E, PMA(10 rig/ml)-stimulated RF culture reacted with collagenase cDNA as a reference for RP cell cultures. Distinct cytoplasmic localization of silver grains is noted. Similar RF cultures reacted with control DNA revealed random and sparse silver grains as above (D).

vealed little or no reaction (results not shown). Thus, the DISCUSSION hybridization correlated with the production of functional collagenase by the cells in culture (Fig. 1). RP cells, when The RP cloned cell line was established by Sato and cultured on glass slides, did not spread evenly and often Ozawa (6) from pulp tissue of a lo-week-old New Zealand formed clumps which made it difficult to localize silver white rabbit. They demonstrated that RP cells formed grains in terms of cell morphology (Figs. 5A and 5B). In dentin-like calcified tissue in uitro; therefore, the cells contrast, rabbit fibroblasts formed a smooth monolayer were previously designated as “odontoblast-like” cells (5, on the glass slides; thus cytoplasmic localization of silver 6). Nonetheless, the RP cell line has not been character- grains was more apparent in these cells (Fig. 5E). The ized enough to define its exact cell type. We have at- experiments using the labeled control probe were char- tempted to induce mineralization in this RP cell culture acterized by the sparse and random distribution of silver under various conditions using an osteoblastic cell line grains (Figs. 5C and 5D). as a reference (17), but without success. To our surprise,

COLLAGENASE SYNTHESIS BY CLONED RABBIT PULP CELLS 493

FIG. 5--C’ontinued

moreover, these RP cells were found to continuously pro- duce high levels of collagenase in cultures sustained for long periods of time (5, 8). We, therefore, suggest that this cell line represents a particular cell type distinct from, yet derived from, rabbit pulp cells.

Unlike many other studies of collagenase synthesis in various types of fibroblasts, in this study we cultured RP cells in medium supplemented with 10% FBS, and isolated collagenase from the medium by heparin-affinity chro- matography. RP cells produced larger amounts of colla- genase in serum-supplemented medium than in serum- free medium. This was also the case for rabbit skin fibro- blasts (results not shown). Although it affects subsequent purification studies, the presence of serum in the culture medium appeared to enable cells to synthesize collagenase under more favorable conditions and for longer periods of time.

The control (nonstimulated) RP cell cultures produced amounts of collagenase comparable to those synthesized by fibroblasts that were stimulated with PMA (100 ng/ ml) under otherwise similar culture conditions, including the presence of serum (Fig. 1). On the other hand, primary cultures of fibroblasts isolated from bovine dental pulp were shown to produce almost negligible amounts of col- lagenase without stimulation (4). In view of their differ- ences in collagenase production, it is therefore conceivable that RP cells represent certain dental pulp cells distinct from fibroblasts.

Collagenase synthesis in BALB/c3T3 fibroblasts was markedly stimulated by mycoplasm contamination of the cultures (18). This finding suggests that mycoplasm in- fection may be implicated in certain disease states, such as rheumatoid arthritis, where elevated collagenase ac- tivity is often demonst,rated. The negative results of my- coplasm screening of the present culture, however, ruled

out the possibility that, the relatively high levels of col- lagenase synthesis in RP cell cultures could be due to such contamination.

In a previous study (5), RP cell collagenase in its most purified state revealed three discrete bands at approximate molecular masses of 45, 52.5, and 57 kDa. Since these bands were all observed in Western blot analysis using both monoclonal and polyclonal antibodies (Fig. 2 and Table I), all three bands represent RP cell collagenase. The band at 45 kDa was previously shown to be an ac- tivated form of the enzyme (8).

The results, various degrees of collagenase inhibition and differing Western blot patterns, suggest either that the four monoclonal antibody preparations were directed to epitopes which were at different sites relative to the active center of the molecule, or that they were directed to the same epitope with varying affinities, or both. The results also suggest that some epitopes were exposed only after the reduction of the antigen for the Western blot reaction (Table I and Fig. 2).

When control or PMA-treated RP cells were stained with sheep anti-rabbit fibroblast collagenase antiserum as the first antibody, distinct perinuclear collagenase flu- orescence was observed (Fig. 3). Monensin, a monovalent cation ionophore, has often been used, particularly with osteoblastic cells whose rate of collagenase synthesis is rather low to enhance immunostaining by blocking en- zyme transport and thereby increasing the intracellular accumulation of collagenase (19). The strong intracellular collagenase immunostaining in RP cells without either monensin or a particular stimulating agent indicates that RP cells are able to synthesize and release large quantities of collagenase in culture.

The rabbit synovial fibroblast collagenase cDNA (clone H9) was used to detect a molecular signal for collagenase both in isolated RP cell RNA, by Northern blot analysis, and in cultures of RP cells. The strong hybridization demonstrated in these experiments between this probe and the collagenase mRNA of RP cells (Figs. 4 and 5), along with other immunological evidence (Figs. 2 and 3), clearly indicates that RP cell collagenase is identical with rabbit fibroblast collagenase, although RP cells and fi- broblasts differ somewhat from each other as observed above.

Although the phenotypes of RP cells pertinent to odontoblasts remain to be studied and RP cells cannot at present be considered odontoblasts, it is apparent that the RP cell line is an ideal source of rabbit fibroblast type collagenase because of its rapid and easy propagation, and of its prolonged synthesis of high levels of collagenase in a routine culture system without particular stimulation. In view of the wide prevalence of the rabbit model in various areas of research, including inflammation and wound healing, this RP cell line appears to provide a useful in vitro reference for the study of the cellular and molec- ular biology of collagenolysis in the rabbit system.

494 GOLDBERG ET AL

ACKNOWLEDGMENTS The authors thank Dr. A. Sato of Tokyo Medical and Dental Uni-

versity for the gift of RP cells, and Dr. C. E. Brinckerhoff of Dartmouth Medical School for her gifts of rabbit collagenase cDNA clone H9 and monospecific collagenase antiserum. The authors also express their im- mense gratitude to Dr. David Wang and the other members of his lab- oratory in this school for patiently and generously sharing their knowl- edge and experience in the area of molecular biology.

REFERENCES 1. Woolley, D. E. (1984) in Extracellular Matrix Biochemistry (Pies,

K. A. and Reddi, A. H., Eds.), pp. 119-157, Elsevier, New York. 2. Goldberg, I. G., Wilhelm, S. M., Kronberger, A., Bauer, E. A., and

Eisen, A. Z. (1986) J. Biol. C&m. 261, 6600-6605. 3. Brinckerhoff, C. E., and Fini, E. (1989) in Collagen (Nimni, M. E.,

Ed.), pp. 65-84, CRC Press, Boca Raton, FL. 4. Kishi, J. and Hayakawa, T. (1985) Biomed. Res. 6,35-40. 5. Sakamoto, S., Baird, S. A., and Sakamoto, M. (1985) B&hem. Bio-

phys. Res. Commun. 133,7099716. 6. Sato, A., and Ozawa, K. (1975) Jpn. J. Oral Biol. 17, 186-193. 7. Gotoh, Y., Saito, S., and Sato, A. (1979) Biochim. Biophys. Acta

587, 253-262. 8. Sakamoto, S., Bolender, Y., and Sakamoto, M. (1990) in Dynamic

Aspects of Dental Pulp (Inoki, R., Kudo, T., and Olgart, L. M., Eds.), pp. 221-237, Chapman and Hall Ltd., London.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

Freshney, R. I. (1987) Culture of Animal Cells, 2nd ed., A. R. Liss, Inc., New York.

Laemmli, U. K. (1970) Nature 227, 680-685.

Towbin, H., Staehelin, T., and Gordon, J. (1979) Proc. N&l. Acad. Sci. USA 76,4350-4354.

Brinckerhoff, C. E., Beloit, M. C., and Culp, W. J. (1985) Proc. Natl. Acad. Sci. USA 82, 1916-1920.

Sakamoto, M., and Sakamoto, S. (1984) Biomed. Res. 5, 29-38.

Davis, L. G., Dibner, M. D., and Battey, J. F. (1986) Basic Methods in Molecular Biology, Elsevier, New York.

Gross, R. H., Sheldon, L. A., Fletcher, C. F., and Brinckerhoff, C. E. (1984) Proc. NatL Acad. Sci. USA 81, 1981-1985.

Zeller, R., Watkins, S., and Rogers, M. (1988) in Current Protocols in Molecular Biology (Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G., Smith, J. A., and Struhl, K., Eds.), Vol. 2, pp. 14.0.3-14.5.4., Wiley, New York.

Sakamoto, S., Sakamoto, M., Goldberg, L., Colarusso, L., and Gotoh, Y. (1989) Biochem. Biophys. Res. Commun. 162, 773-780.

Kluve, B., Merrick, W. C., Stanbridge, E. J., and Gershman, H. (1981) Nature 292, 855-857.

Cowen, K. S., Sakamoto, M., and Sakamoto, S. (1985) Biochem. Int. 11,273-280.