Enhancement by extracts of mineralized tissues of protein production by human gingival fibroblasts in vitro

Arch oral Bid. Vol. 32, NO. 12, pp. 879-883, 1987

0003~9969/87 S3.00 + @OO

Printedin Great Britain

PergamonJournalsLtd

ENHANCEMENT BY EXTRACTS OF MINERALIZED TISSUES OF PROTEIN PRODUCTION BY HUMAN GINGIVAL FIBROBLASTS IN VIZ’XO M. J. SOMERMAN,‘~~ S. Y. ARCHER,’ T. M. HASSELL,‘,’ A. SHTEYER~and R. A. FOSTER’~ Departments of ‘Periodontics, 2Pharmacology, ‘Orthodontics and ‘Anatomy, Baltimore College of Dental Surgery, Baltimore, MD 21201, U.S.A. and sDepartment of Oral Surgery, Hadassah University, Jerusalem, Israel Summary-Non-confluent cell cultures were exposed to both guanidine and guanidine-EDTA extracts of cementum, dentine and alveolar bone, at concentrations from 2 to SOpgjml for 48 h. The cells were radioactively labelled during the last 24 h. Total protein production was measured via incorporation of radioactive proline; collagen production was estimated by digestion of the radioactive protein mixture with bacterial collagenam. All guanidine-EDTA extracts elicited statistically-significant increases in total protein production when compared to controls. At 50 pg/ml of extract, the increase in protein production was 340, 143 and 338 per cent for bone, cementum and dentine, respectively. Similar results were obtained for collagen production. Guanidine-EDTA extracts also stimulated an increase in the production of specific proteins, as ascertained by gel electrophoresis. In contrast, the guanidine extracts had no effect on either protein or collagen production. Thus the functions of gingival fibroblasts can be altered by proteins from associated mineralized tissues. Identification of such proteins and their biological functions would enhance knowledge of the mechanisms that regulate connective-tissue regeneration.

INTRODUCTION

Regeneration of the tooth-supportive connective tissue lost in periodontal disease may be a function of cells originating from residual healthy tissues (Melcher, 1976; Gould, Melcher and Brunette, 1980; Bowers, Schallhom and Melloning, 1982; Femyhough and Page, 1982; Nyman et al., 1982). To effect complete regeneration, such cells must undergo mitosis, migrate to the healing site and become active synthetically. The agents that trigger these events are not completely understood, but they may come from tissues to which the cells are exposed in vivo, i.e. alveolar bone, cementum and dentine. Extracellular matrix is required for adhesion, proliferation and subsequent differentiation of cells (Hay, 1981). In particular, subcutaneous implantation of bone matrix in rats results in local differentiation of fibroblasts to form endochondral bone (Urist et al., 1968). Studies of the specific components of bone matrix which may be responsible for triggering bone induction have identified boneassociated factors and proteins with unique characteristics. These include chondrogenic-stimulating proteins (Seyedin et al., 1983, 1985; Syftestad and Caplan, 1984), bone morphogenetic proteins (Sampath, Nathanson and Reddi, 1984; Sato and Urist, 1984), cell-proliferation proteins (Drivdahl, Howard and Baylink, 1982; Sampath, DeSimone and Reddi, 1982), and a bone-chemotactic factor (Somerman et al., 1983). We have sought to further characterize proteins associated with bone matrix, and to extend this research to the mineralized tissues of the mouth, e.g. cementum and dentine. Establishing the properties of crude protein extracts will provide a method for assaying the activities of mineralized tissue components during purification. Here we describe the effects of guanidine-EDTA extracts of alveolar

bone, cementum and dentine on synthetic activity by cultured human gingival fibroblasts. iW.KlXRLUR AND METHODS

Preparation of extracts The methods of obtaining extracts of bovine cementum, dentine and alveolar bone were similar to those described for the separation of other bonematrix proteins (Termine et al., 1981; Somerman et al., 1985a, b). Briefly, the extraction is sequentiak 4 M guanidine, followed by 4 M guanidine/OS M EDTA, separates mineralized-tissue matrices into mineralnon-associated and mineral-associated, soluble protein fractions, respectively. Upper and lower jaws from two-year-old steers kept on ice, were obtained immediately after death; teeth were extracted with chisels and hammer. Specimens of alveolar bone were also taken from around the teeth. The teeth (maxillary and mandibular, erupted molars and lower incisors) and bone were immediately placed in 0.02 M PBS, pH 7.4, containing protease inhibitors (0.05 M 6aminohexanoic acid, 0.005 M benxamidine HCl, 0.001 M phenylmethylsuphonyl fluoride) at 4°C. To obtain periodontal ligament, adherent soft tissue was first removed from the teeth with blunt scalers unddr the dissecting microscope. The cementum was than removed with sharp scalers, and both ligament and cementum were sequentially extracted in 0.05 M tris-HCl buffer @H 7.4; 50 ml/g wet weight) containing 4 M guanidine HCl and the protease inhibitors at 4°C for seven days, and then in tris buffer (same volume) containing 4 M guanidine HCl with 0.5 M EDTA until demineralized, usually after seven #days. The remaining root surface was then scaled extensively to preclude any contamination of cementum with dentine. Residual dentine was then removed 819

880

M. J.

thfERhf~N et al.

with bone cutters and guanidine and guanidineEDTA extracts obtained as for cementum. Extracts of alveolar bone were similarly obtained. All extracts were then concentrated by ultrafiltration (Amicon YM-10 filters), dialysed extensively against water, and lyophilized. The lyophilized materials, dissolved in appropriate media, were used in the assays.

material were reacted with TCA-tannic acid and FBS for 15 min, centrifuged, and the supematant scintillation counted. This provided a background value that was subtracted from the collagenase-treated series to yield a value taken to represent collagen production, expressed as collagenase-digestible protein.

Determination of total protein and collagen production

Heat stability of guanidine-EDTA extracts

The methods for culturing and labelling of human gingival fibroblasts have been reported in detail (Hassell and Stanek, 1983). Briefly, 2.0 x IO5 cells/Linbro test well were seeded in minimum essential medium (MEM) with 10 per cent fetal calf serum (FCS) and antibiotics, and allowed to attach overnight. These non-confluent cultures, in the log phase of growth, were exposed to extracts, at concentrations indicated in results, in culture medium for 24 h. Then each well was rinsed with 0.5ml of serum-free, proline-free MEM and pulse-labelled for 24 h in the same medium containing the specified extract and L-(5-‘HI-proline to make 20 pCi/ml. Concurrent with assays for protein production, further cultures were prepared to determine the effects of the agents on cell proliferation. These cells were exposed to the specific agents exactly as above except that during the last 24 h, no radioactive label was added. To determine cell number, the medium was removed and the cells rinsed twice with PBS. The attached cells were then removed by the addition of 0.08 per cent trypsin/0.04 per cent EDTA and the cells counted electronically by Coulter Counter. Immediately after the radioactive-labelling period the medium was placed into dialysis casings. Cells were harvested to include production of non-secreted protein. Cold 0.1 M acetic acid (1.5 ml) was added to the wells, the lysed cells scraped off and added to medium. Dialysis was done three times for 24 h against cold distilled water, then three times for 24 h against cold bulier (0.05 M tris-HCl, 0.2 M NaCl, 0.005 M CaCl,, pH 7.4). Buffer was then added to all samples so that the final volume was 5.0 ml. Samples (1.5 ml) of these samples were counted in a liquidscintillation spectrometer (Packard Tri-Carb) in 5.Oml of Aquasol(New England Nuclear). These counts represented total non-dialysable radioactive material, and were taken as a measure of total protein production. To estimate collagen production, 3.Oml of the dialysed material were reacted for 6 h in a shaking water bath (37”C), with ca 25 units of chromatographically-pm&d bacterial collagenase (Form III, Advance Biofactures; free of non-specific proteolytic activity) prepared in a 0.025 M tris-HCl, 0.33 M calcium-acetate buffer. pH 7.4, according to the manufacturer’s directions, in the presence of 1 mM Nethyhnaleimide (prepared in 0.05 M tris-HCl, 0.02 M NaCl, 0.005 M CaClr buffer at pH 7.4; Peterkofsky and Diegehnann, 1971). The reaction mixture was then promptly cooled on ice, and 400~1 of a 5 per cent trichloroacetic acid (TCA)-O.S per cent tannicacid solution added; 50 ~1 of FCS were also added to enhance protein precipitation. The mixture was kept at 4°C for 15 rnin, then centrifuged at 4°C for 15 min (10,OOOg). A sample of the supematant was counted in 15.0 ml of Aquasol-2. Further samples of dialysed

For heat-stability studies, 1 mg, 1 ml stocks of the protein extracts were prepared in DMEM and heated for 30 min at various temperatures. Stocks were diluted with fresh media to yield a final concentration of 50pg/ml and biosynthetic activity assayed as above. Electrophoresis Cells were seeded, exposed to guanidine-EDTA extracts of cementum, dentine or bone, at a concentration of 50pg/ml, and radioactively labelled as above. Each pooled medium and cell sample was dialysed extensively against distilled water and lyophilized. Lyophilized samples from cultures having equal cell numbers were dissolved in gel buffer. Thus, any variation in protein profiles found between samples would be based on increased protein production per cell. The nature of the proteins in the guanidine-EDTA extracts of bone, cementum and dentine, at a concentration of SOpg/lane, were evaluated by gel electrophoresis. After reduction with mercaptoethanol the protein samples were analysed by electrophoresis on SDS 4-20 per cent polyacrylamide gradient slab gels, according to the method of Laemmli (1970). The stacking gel was 4 per cent acrylamide. The gels were then lixed, dried and autofluorographed by the method of Bonner and Laskey (1974). Statistical analyses The statistical significance of differences in effects of the mineralized tissue extracts on protein and collagen production by fibroblasts was determined by analysis of variance and Duncan’s multiple range test. RESULTS Gel electrophoresis of the guanidine-EDTA extract of bone and cementum indicated several high-molecular-weight components (M, = 92,000200,000). There was also a number of components between 45,000 and 65,000 daltons, which stained more intensely for bone than for cementum proteins. lower molecular-weight components Several (M, = 14,ooo-43,000) were apparent, with more distinctive bands in the bone than in the cementum extract. The gel-electrophoresis profile of the guanidine-EDTA extract of dentine was similar to that of bone, except that the higher molecular-weight components (M, = 92,00&200,000) were not as distinct. These higher molecular-weight components were apparent in the guanidine extract of dentine (data not shown). The guanidine-EDTA extracts of bone, cementum and dentine elicited dramatic and statisticallysignificant increases in total synthetic activity of

881

Cell response to minaraiixed tissue proteins TOTAL PROTEIN PROOUCTION BY HUMAN GINGIVAL FIGROGLASTS

3.5

-

30

-

5

2.5

-

ui‘0

2.0

-

; u

I.5

-

IX

1.0

-

0.5

CoNlROL

6aNL EXTRYT

CEYCNTUN

MRUT

OENTINC LXIIACI

LX%i

Fig. 1. The effect of guanidine-EDTA extracts of alveolar bone, cementum, dentine and periodontal ligament (PDL) on protein production by cultured human gingival fibroblasts. Cells were exposed to 50pgjml of the extract for 48 h. During the second 24 h period [‘HI-proline was added to the cultures and total protein production measured via the incorporation of radioactive proline into media and cells. AU agents were evaluated in triplicate and the results expressed as CPM (counts/mm) per cell f standard deviation.

fibroblasts (Fig. 1). In contrast, the guanidine-EDTA extract of periodontal ligament had no significant effect above control levels on protein production (Fig. 1). The per cent increase in total protein production, when compared with controls (no agent), was 340, 143 and 338 per cent for bone, cementum and dentine extracts, respectively. Similar results coL~PmGLcTmBY

HUMAN

s

FIBROWS

600-

%? n h 500 X 2

GINGIVAL

-

400

-

300

-

Table 1. Effect of guanidine-EDTA extracts on ceil proliferation

200

-

Treatment

Cell number*

Control Bone Cementum Dentine

16,256 * 14,607 f 14.464 f 20,9 11 f

0 m

were obtained for collagen production (Fig. 2). The bone and cementum extracts, during the radioactive labelling period, had no significant effect on cell proliferation. In contrast, the dentine extract elicited a slight but significant increase in cell proliferation (Table 1). Therefore, the ability of extracts to enhance biosynthetic activity was not explained solely in terms of their cell-proliferation activity. These results were reproduced in seven separate experiments. The guanidine-EDTA extract of dentine enhanced protein production in a dose-dependent manner (Fig. 3). A slight but significant increase in protein production above the control level was observed at a concentration of 2 pg/ml; 100 pg/ml dentine extract did not significantly enhance protein production beyond that observed at SOpg/ml. In contrast, the guanidine extract of dentine at a concentration of 50pg/ml had no effect on protein production when compared to cells not exposed to agent. Similar dose-responses were observed with guamdine-EDTA extracts of cementum and bone, but the effects were not as dramatic. A minimal dose of 25 pg/ml was required in order to produce significant increases in total protein production. To insure that the guanidine-EDTA extracts had selective effects on protein production, other agents were evaluated for their effects on total protein and collagen production. Type I collagen, a major component in the guanidine-EDTA extracts of bone, cementum and dentine, as well as guanidine extracts of bone, cementum, dentine and periodontal ligament, at concentrations of 50pg/ml, had no effect above control levels on protein production by fibroblasts (data not shown). The guanidine-EDTA extract of dentine (50 pg/ml) was next evaluated for stability on heating. Exposure to temperatures of 60°C did not alter its ability to enhance fibroblast-protein production but, at lOO”C, the material was almost completely inactivated (Table 2). Gel-electrophoresis indicated that all three guanidine-EDTA extracts enhanced the production of specific proteins by human gingival fibroblasts during the 24 h labelling period. The heaviest labelled band, above 200 K mol. wt, was fibronectin, as verified by western blot. The two bands below

Fig. 2. The effect of guanidine-EDTA extracts of alveolar bone, cementum. dentine and periodontal ligament (PDL) on collagen production by cultured human gingival fibroblasts. Cells were exposed to SOpglml of the extract for 48 h. During the second 24 h period [‘H]-proline was added to the cultures and collagen production measured by digestion of the radioactive protein mixture (media and cells) with bacterial collagenaae. All agents were evaluated in triplicate and results expressed as CPM (counts/mitt) per cell f standard deviation.

1,687 666 1,258 508t

AU agents, at a concentration of 50 &nl, were evaluated in triplicate and results expressed as cell number f standard deviation. CThis data represents the cell number/ weU at the end of the radioactive Welling period for experiments shown in Figs 2 and 3. tp < 0.05 signifkantly different from control and guanidine_EDTA ex-

tract of bone and cementum.

M. J. !%MER~UV er al.

882

Table 2. Heat stability of guanidine-EDTA extract of dentinc Total protein production Treatment Counts/r& per well Control Dentine 37°C 60°C 100°C

55,009 f 1,992 113,282* 1,617 110,235 f 1,868 62,513 f 3,698

Cell number/well

Counts/min x lo-‘/106alls

21.036 f 794

2,615 f 95

29,531 f 768.. 27,169 2 1,248** 21,584 k 976.

3,836 f 55** 4,057 f 69.’ 2,896 f 171’

Dentine extract, at a concentration of 50 r&ml, was evaluated in triplicate and results expressed as mean f standard deviation. lp < 0.05 significantly different from control and guanidine-EDTA extract of dentine at 37 and 60°C. l*p < 0.05 significantly different from control and guanidine-EDTA extract of dentine at 100°C.

OU-EDTA LXTRACT

OiiENTlNL

(/q/ml)

Fig. 3. The effect of guanidine and guanidine_EDTA (GU-EDTA) extracts of dentine on protein production by cultured human gingival fibroblasts. Cells were exposed to extracts for 48 h. During the second 24 h period [‘HI_proline was added to the cultures and total protein production measured via the incorporation of radioactive proline into media and ah. Guanidine (GU) was added at a concentration of 5Opglml. Results are expressed as the ratio of CPM (counts/m@ in treated ah versus control cells&standard deviation. (All agents were evaluated in triplicate.) Similar resutls were observed in three separate experiments.

fibronectin were indicative of procollagen chains. A band below the procollagen chains was readily apparent only in cells exposed to the dentine extract. These findings were consistent in three gels from three separate experiments (data not shown).

We show that the proteins of mineralized tissue to which human gingival fibroblasts may be exposed in uiuo, influence the synthetic activities of these cells. This finding is in agreement with other investigations, which have shown that extracts of bone matrix regulate chondrocyte and osteoblast function (Urist et al., 1968; Ainsworth, Puzic and Anastassiades, 1977; Drivdahl et al., 1982; Sampath et al., 1982;

Seyedin et al., 1983; Sampath et al., 1984; Syftestad and Caplan, 1984). However, such research has not considered the effects of protein extracts of dentine or cementum on cell-synthetic activity. Evidence from our laboratory (Somerman, Shteyer and Bowers, 1984; Somerman et al., 1985b), and our present results, suggest that proteins from bone, cementum and dentine may each have unique properties. The dentine extract selectively stimulates the production of an unidentified protein of approx. 100 K, as seen by gel electrophoresis. In addition, lower doses of dentine extract than cementum or bone extract are required for enhancement of protein production by fibroblasts, while similar extracts of bone have a greater effect than those from cementurn. As each of these mineralized tissues has a different function and site, it is likely that each would have proteins unique to that tissue. That our extracts of the three mineralized tissues significantly affect the synthetic activity of human gingival fibroblasts in vitro suggests that the proteins are itiportant.in the normal homeostasis of connective tissues in duo. Moreover, such proteins from mineralized tissue may trigger fibroblasts to synthesize the specific matrix components required for regeneration of connective tissue lost to disease. Acknowledgements-We wish to thank Dr Richard Wynn and Dr John Bergquist for their support and advia in the preparation of the manuscript. Richard Merkhofer and Maria Perez-Mera for their excellent assistana in the preparation of the mineralized tissue extracts and JoAnn Walker for exallent secretarial skills. This research was supported by USPHS Grants DE-06671, DE-00123 and DE-07512 from NIDR.

REFERENCES

Ainsworth T., Puzic 0. and Anastassiades T. (1977) The effect of bone matrix on young connective tissue aUs in culture. Lob. Clin. h&d. 89, 81-91. Bonner W. M. and Laskey R A. (1974) A lb detection method for tritium-lab&d proteins and nucleic acids in polyacrylamide gels. Eur. Biochem. 4f& 83-88. Bowers G. M., Schallhom R. G. and Melloning J. T. (1982) Histologic evaluation of new attachment in human intrabony defects. A literature review. Periohntofogy !53,509. Drivdahl R., Howard G. and Baylink D. (1982) Extract of bone contain a potent regulator of bone formation. Biochbn. biophys. Acta 714, 25-33.

Cell response to mineralized tissue proteins Fernyhough W. and Page R. C. (1982) Attachment, growth and synthesis by human gingival tibroblasts on demineral&d or fibronectin treated normal and diseased tooth roots. Periodonrology !t4, 133-140. Gould T., Melcher A. H. and Brunette D. (1980) Migration and division of progenitor ailpopulations in periodontal ligament after wounding. Periodonr. Res. 15, 2042. Hassell T. M. and Stanek E. (1983) Evidence that healthy human gingiva contains functionally heterogeneous fibroblast subpopulations. Archr oral Biol. 28, 625-717. Hay E. D. (Ed.) (1981) Cell Biology of Extracellular Marrix. Plenum, New York. Laemmli U.K. (1970) Cleavage of structural proteins during assembly of the head of bacteriophage T4. Nature 117, 680-685. Melcher A. H. (1976) On the repair potential of periodontal tissues. Periodontology 47, 256-260. Nyman S., Gottlow J., Karting T. and Lindhe 1. (1982) The regenerative potential of the periodontal ligament. An experimental study in the monkey. C/in. Periodonr. 9, 257-265.

Peterkofsky B. and Diegelmann R. (1971) Use of a mixture of proteinase-free collagenases for the specific assay of radioactive collagen in the presence of other proteins. Biochemistry 10, 988-994, Sampath T. K., DeSimone D. P. and Reddi A. H. (1982) Extracellular bone matrix-derived growth factor. Expl Cell Res. 142-460. Sampath T. K., Nathanson M. and Reddi A. H. (1984) In vitro transformation of mesenchymal cells derived from embryonic muscle into cartilage in response to extracellular matrix components of bone. Proc. nam. Acad. Sci. U.S.A. 81, 3419-3423. Sato K. and Urist M. R. (1984) Bone morphogenetic

883

protein-induced cartilage development in tissue culture. Clin. Orthop. 183, 180-187. Seyedin S. M., Thompson A. Y., Rosen D. M. and Peiz K. A. (1983) In vitro induction of cartilage-specific macromolecules by a bone extract. Cell Biol.-97; 947-952. Sevedin S. M.. Thomas T. C.. Thomoson A. Y.. Rosen b. M. and Piez K. A. (1985)‘Purifi&ion and characterization of two cartilage-inducing factors from bovine demineralized bone. Proc. natn. Acad. Sci. U.S.A. 82, 2267-2271.

Somerman M. J., Hewitt A. T., Vamer H. H., Schiffnuann E., Termine J. and Reddi A. H. (1983) Identitication of bone matrix-derived chemotactic factor. Calc. Tics. Inr. 35, 481-485. Somerman M. J., Shteyer A. and Bowers M. R. (1984) Analysis of the organic composition of bovine cementum. Dent. Res. 63, 136OA. Somerman M. J., Archer S. Y., Foster R. A. and Hassell T. M. (1985a) Extracts of mineralized tissue enhana protein production by human gingival fibroblasts in vitro. Dent. Res. 64, 380A. Somerman M. J., Kagermeier A. S., Bowers M. R. and Fox P. C. (1985b) In vitro attachment of bacteria to extracts of amentum. Oral Path. 14, 793-799. Syftestad T. G. and Caplan A. I. (1984) A fraction from extracts of demineralized adult bone stimulates the conversion of mesenchymal cells into chondrocytes. Deal Biol. 104, 348-356. Termine J. R., Belcourt A. B., Conn K. M. and Kleinman H. K. (1981) Mineral and collagen-binding proteins of fetal calf bone. Biol. Chem. 256, 10,403-10,408. Urist M. R., Dowel1 T. A., Hay P. H. and Strates B. S. (1968) Inductive substrates for bone formation. Clin. Orthoped. 59, 59-68.