- Email: [email protected]
The uptake of two androgen substrates by cultured human gingival fibroblasts in response to minocycline and metabolic studies using a cell-free system
Archives of Oral Biology 44 (1999) 215±222
The uptake of two androgen substrates by cultured human gingival ®broblasts in response to minocycline and metabolic studies using a cell-free system M. Soory*, H. Virdi Department of Periodontology, G.K.T Dental Institute (King's College Hospital), London, SE5 9RW, UK Accepted 17 November 1998
Abstract The uptake of two androgen substrates by cultured ®broblasts and the anabolic response of homogenates of cultured gingival ®broblasts to minocycline were investigated. Monolayer cultures of con¯uent ®broblasts were incubated with [14C]testosterone/[14C]4-androstenedione for timed intervals in the presence or absence of optimal concentrations of minocycline. The intracellular uptake of androgens was quanti®ed by radio-isotope counts on cell digests. Con¯uent gingival ®broblasts were homogenized by snap freezing/rapid thawing and duplicate incubations were made in phosphate-buered saline (pH 6.5) with radiolabelled androgens, in the presence or absence of minocycline, for 24 h. At the end of the incubation period the buer was extracted for radioactive metabolites, analysed and quanti®ed with a radio-isotope scanner. There were 30% increases in the uptake of androgen substrates in the presence of minocycline (n=3; p<0.01; one-way ANOVA). With the metabolic studies there were 2±3-fold increases in the formation of dihydrotestosterone from [14C]testosterone and [14C]4-androstenedione, respectively (n=4; p<0.001; one-way ANOVA), and 4-fold/2-fold increases in the formation of 4-androstenedione/ testosterone from these substrates (n=4; p<0.001) in response to an optimal concentration of 20 mg/ml of minocycline, compared with control incubations. The presence of minocycline in the incubate signi®cantly increased the activity of the steroid-metabolizing enzymes. This increase might result from increased intracellular availability of steroid substrate and enhanced metabolic activity. Homogenates of cultured gingival ®broblasts are a useful model for studying the anabolic potential of minocycline in gingiva, using C19 steroid substrates. # 1999 Elsevier Science Ltd. All rights reserved. Keywords: Uptake; Metabolism; Androgens; Fibroblasts; Cell-free system; Minocycline
1. Introduction The subgingival micro¯ora is important in the initiation and progression of chronic periodontal dis-
Abbreviations: DHT, 5a-dihydrotestosterone, MEM, minimal essential medium, PBS, phosphaye-buered saline. * Corresponding author. Fax: 0171-346-3185
eases. Colonies of Gram-negative anaerobes are consistently isolated from diseased sites of the periodontium and are implicated in their aetiology (Moore et al., 1982). Hence antimicrobial agents such as tetracycline and metronidazole have been used as adjuncts in the treatment of in¯ammatory periodontal diseases, particularly in cases of aggressive disease in young patients and in recalcitrant forms (Genco, 1981; Kornman and Karl, 1982). Minocycline, the semisyn-
0003-9969/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved. PII: S 0 0 0 3 - 9 9 6 9 ( 9 8 ) 0 0 1 2 1 - 6
216
M. Soory, H. Virdi / Archives of Oral Biology 44 (1999) 215±222
thetic analogue of tetracycline, reduced pocket probing depths and the number of periodontal pathogens in the adjunctive treatment of moderately severe adult periodontitis (van Steenberghe et al., 1993). Minocycline was more eective than tetracyclines against several bacteria commonly found in dental plaque (O'Connor et al., 1990). In addition to their antimicrobial properties, tetracyclines have anti-in¯ammatory (Ryan et al., 1996) and proanabolic eects on connective tissue (Schneir et al., 1990) and bone matrices (Sasaki et al., 1991). When dentine granules were conditioned with minocycline hydrochloride, the subsequent attachment of human periodontal ligament cells to the granules improved (Rompen et al., 1993). The chemically modi®ed tetracyclines, in which the antimicrobial moiety has been removed, retaining the anti-in¯ammatory and pro-anabolic eects, have been used in this capacity for the treatment of autoimmune diseases (Golub et al., 1987; Breeveld et al., 1990). Our previous studies have demonstrated increased formation of the biologically active androgen DHT in response to stimulants that increase protein turnover, such as growth factors and drugs implicated in causing gingival overgrowth (Kasasa and Soory, 1996; Soory and Kasasa, 1997), indicating a link between biologically active androgens and anabolic eects on connective-tissue matrices. In view of the proanabolic eects of androgen metabolites such as DHT on the synthetic functions of ®broblasts (Sultan et al., 1986) and osteoblasts (Kasperk et al., 1989), it was relevant to investigate the anabolic eects of minocycline, using androgen substrates. Our aim now was to study the uptake of androgen substrates by cultured gingival ®broblasts and their metabolic pro®les in a cell-free system. 2. Materials and methods The radioisotopes [14C]testosterone and [14C]4androstenedione (spec. act. 58 mCi/mmol) were obtained from Amersham International, Amersham, Bucks. Organic solvents (benzene, acetone) for thinlayer chromatography, ethyl acetate for extraction of metabolites and chloroform to redissolve the dried extract were all provided by Merck Ltd., Dagenham, Essex. Incubations were done in PBS at pH 6.5 for experiments with homogenates of cultured ®broblasts (constituents were from Sigma Chemical Co. Ltd., Poole, Dorset); the incubation medium for the uptake experiments was Eagle's MEM with 10% fetal bovine serum, L-glutamine, antibiotic/antimycotic solution and sodium bicarbonate, which were obtained from Gibco Ltd., Paisley, Scotland. The pharmacy at King's
College School of Medicine and Dentistry provided the minocycline. For the derivatization of steroids, O-penta¯uorobenzyl hydroxylamine hydrochloride (Florox reagent) and N, )-bis trimethylsilyl tri¯uoroacetamide were obtained from Pierce and Warriner (UK) Ltd, Chester. Chronically in¯amed gingival tissue was taken from patients (aged 30±50 years) with chronic peridontal disease during surgical treatment for pocket elimination, after the completion of initial treatment consisting of scaling and root planing. Our previous work established that cells grown from an in¯amed tissue maintain their characteristics over serial passaging, demonstrating a heightened androgen-metabolic pro®le compared with cells derived from healthy gingiva, both at baseline and in response to stimulants (Sooriyamoorthy et al., 1988; Sooriyamoorthy and Gower, 1989; Sooriyamoorthy et al., 1990; Soory and Gower, 1990). Hence the experimental model here consisted of ®broblasts passaged from chronically in¯amed tissue, thus simulating conditions in the in¯amed periodontium and events leading to repair. Clones of ®broblasts respond to in¯ammatory stimuli, altering their phenotype to meet functional demands. The metabolism of androgen substrates in cells derived from a healthy tissue occurs predominantly in males, while 4-androstenedione, the main circulatory androgen in females, is preferentially metabolized by cells from this source. In in¯amed gingiva, substrate preference and the extent of androgen metabolism is similar in both males and females (Rappaport et al., 1976). Therefore the present experimental model, comprising cells/homogenized cells derived from in¯amed tissue, will incorporate a spectrum of anabolic activity from the in¯amed periodontium, representing males and females. 2.1. Cell culture The gingival tissue was washed in physiological saline, placed in Hank's balanced salt solution and minced in Eagle's MEM into 1-mm3 pieces. It was then plated in 25-cm2 ¯asks with Eagle's MEM until primary cultures were established. The contents of a fully con¯uent 25-cm2 ¯ask (2.2 106 cells) were distributed into 24 wells of a multiwell dish and triplicate incubations were performed with [14C]testosterone for timed periods of 5, 15, 30 min, 1, 2, 3, 4, 6 and 8 h, in the presence or absence of optimal concentrations of minocycline. These optima had been established previously, using a series of ®broblast cell lines (Soory and Virdi, 1998), and found to be 20 mg/ml. At each incremental period, the medium was removed over crushed ice to minimize the passage of substrate taken up by the cells and washed twice in physiological saline.
M. Soory, H. Virdi / Archives of Oral Biology 44 (1999) 215±222
217
The cells at the base of the multiwell dishes were then digested with 1N NaOH and solvent-extracted with ethyl acetate (2 ml2) to isolate the radiolabelled testosterone taken up by them. It was then evaporated and spotted on a thin-layer chromatographic plate for the purpose of quantifying the uptake of radiolabelled steroid at dierent intervals, using a radio-isotope scanner. In order to clarify the possibility of metabolism of radiolabelled substrate during the periods studied, the medium was also analysed for steroid metabolites. This was done for one set of experiments with [14C]testosterone as substrate, in the presence or absence of minocycline. For each of the timed periods, the medium was isolated, solvent-extracted and the metabolites separated by thin-layer chromatography. The chromatographic plates were subjected to radioisotope scanning for quanti®cation of metabolites. The analytical details are described fully below. 3. Metabolic studies These experiments were done with cell lines derived from several patients, with independent controls for each line studied. Cells derived from individual cultures were not pooled. Gingival ®broblasts were serially passaged and cells of the 5th±9th passages were used for the experiment. They were grown out in 25cm2 ¯asks and, when con¯uent, the contents of each ¯ask were used for duplicate incubations after homogenization of the con¯uent cells. Homogenization was begun by removing the incubation medium, washing the cell layer in PBS (pH 6.5) and subjecting the ¯ask to snap freezing and thawing in liquid nitrogen repeatedly in a small volume of the buer until the cells were fragmented. They were then homogenized with a fused pipette; incubations were made in 24-well multiwell dishes, using 1 ml PBS per incubation, with radiolabelled substrates [14C]testosterone/[14C]4-androstenedione and serial concentrations of minocycline of 5, 10, 20, 30 and 40 mg/ml. At the end of a 24-h incubation, the homogenates were solvent-extracted for steroid metabolites with ethyl acetate (2 ml3) after the addition of cold carrier steroids and evaporated to dryness in a vortex evaporator (Gyrovap, VA Howe Ltd., Banbury, Oxon). The samples were then solubilized in 100 ml chloroform and subjected to thin-layer chromatography in a benzene/acetone solvent system (4:1, v/v) to separate the metabolites from the two radiolabelled androgen substrates. The chromatographic plates were then scanned in a radio-isotope scanner for the quanti®cation of the separated metabolites. The formed metabolites were tentatively identi®ed by establishing the mobility of
Fig. 1. Uptake of [14C]testosterone by human gingival ®broblasts. Median values and SD for substrate uptake by the cells are shown [14C-T(C)]. Similarly, the metabolites isolated from the medium (Met) at timed periods are shown at the base of the graph.
cold standards, using an iodine tank to disclose them on the thin-layer chromatographic plates, to show correspondence with the radioactive metabolites identi®ed during radio-isotope scanning. Identi®cation was con®rmed by comparing the fragmentation pattern of derivatized samples with authentic standards, using gas chromatography±mass spectrometry. 4. Characterization of DHT by gas chromatography± mass spectrometry It was important to con®rm the identity of 5a-DHT,
218
M. Soory, H. Virdi / Archives of Oral Biology 44 (1999) 215±222
14
Fig. 2. Uptake of [ C]testosterone by human gingival ®broblasts in the presence of minocycline. Values for substrate uptake by the cells are shown [14C-T (M)], as in Fig. 1. The metabolites (Met) isolated from the medium are shown at the base of the graph.
being the physiologically active androgen in healing responses. Several incubations were made under the conditions described above, using unlabelled testosterone (10ÿ6 M). After extraction, the identity of 5a-DHT as a metabolite in the dried extracts was con®rmed by gas chromatography±mass spectrometry (courtesy of Professor A.I Mallet, St. Thomas' Hospital, London, UK), after derivatization to penta¯uorobenzyloxime trimethyl silyl ether. The derivatized biological material had a molecular ion (557) and mass-spectral fragmentation pattern identical to those of the authentic penta¯uorobenzyloxime trimethyl silyl ether of 5a-DHT, but at lower levels, owing to the smaller concentrations of
Fig. 3. Uptake of [14C]4-androstenedione by human gingival ®broblasts in the presence or absence of minocycline. Uptake of [14C]4-androstenedione by ®broblasts [14C-4-A(C)] compared with uptake in the presence of minocycline [14C-4A(M)] at timed intervals. Median values and SD are shown.
the steroid. These procedures were described in detail by Soory (1995). 5. Results 5.1. Uptake of [14C]testosterone/[14C]4androstenedione by gingival ®broblasts in the presence or absence of minocycline 5.1.1. [14C]testosterone Control incubations showed a gradual increase in the uptake of radiolabelled testosterone by the gingival ®broblasts over a period of 1±3 h by 20±40% (Fig. 1;
M. Soory, H. Virdi / Archives of Oral Biology 44 (1999) 215±222
219
Fig. 4. Metabolism of [14C]testosterone by homogenates of human gingival ®broblasts in response to minocycline. Duplicate incubations with [14C]testosterone (CH) in the presence or absence of serial concentrations of minocycline (MH: 5, 10, 20, 30 and 40 mg/ml) for 24 h. Median values (pmol/ ml) and SD are shown.
Fig. 5. Metabolism of [14C]4-androstenedione by homogenates of human gingival ®broblasts in response to minocycline. As Fig. 4, with [14C]4-androstenedione (CH) as substrate in the presence or absence of minocycline (MH). Median values (pmol/ml) and SD are shown.
n=3; p<0.01, one-way ANOVA) when compared with the initial uptake at 5 min. This increase remained fairly similar at 4, 6 and 8 h. In the presence of minocycline, the initial uptake of [14C]testosterone was similar to that of control incubations at 5 min, with gradual increases of 25% at 15 min, 40% at 30 min and 57% at 1 h (Fig. 2; n=3; p<0.01) over the value at 5 min. There was a further increase in uptake at 3 h, resulting in a two-fold increase over that in the 5 min period (n=3, p<0.001), which was maintained at 4, 6 and 8 h. If Figs. 1 and 2 are compared, it will be seen that the maximum uptake of radiolabelled testosterone
at 3 h was 30% greater when minocycline was present in the medium (n=3; p<0.01). The medium was analysed for the metabolites formed during the timed periods from the uptake of [14C]testosterone in the presence or absence of minocycline. DHT and 4-androstenedione were the main steroids formed and are shown as metabolites in Figs. 1 and 2. There was some metabolic conversion between 3±8 h from [14C]testosterone (Fig. 1) and between 30 min and 8 h in the presence of minocycline (Fig. 2). It is possible that the uptake was still maintained adequately, due to passive diusion from the substrate
220
M. Soory, H. Virdi / Archives of Oral Biology 44 (1999) 215±222
pool in the medium, when this metabolic conversion occurred. 5.1.2. [14C]4-androstenedione When [14C]4-androstenedione was used, its uptake by human gingival ®broblasts was increased by 22% at 3 h (Fig. 3; n=3, p<0.01, one-way ANOVA) compared to that at the 5 min period, and was maintained at this value at 4, 6 and 8 h. The initial uptake in the presence of minocycline was similar to that of control incubations at 5 min, with approx. 35% increases at 15, 30 and 1 h; subsequently, at 2 h and 3 h, there was a gradual increase of 40±46% (n=3; p<0.01). This increase was maintained at 4, 6 and 8 h. The maximum uptake at 3 h was about 30% greater in the presence of minocycline than in controls (n=3; p<0.01). 5.2. Metabolism of [14C]testosterone/[14C]4androstenedione by homogenates of gingival ®broblasts in response to minocycline When homogenates of human gingival ®broblasts were incubated with the radiolabelled androgen substrates testosterone or 4-androstenedione, they were metabolized by the enzymes 5a-reductase and 17b-hydroxysteroid dehydrogenase to DHT and 4-androstenedione or testosterone, respectively. There were 2±3fold increases in the formation of DHT from [14C]testosterone and [14C]4-androstenedione, respectively, at a minocycline concentration of 20 mg/ml (Figs. 4 and 5; n=4, p<0.001; one-way ANOVA), with gradual increases at suboptimal concentrations and maintained beyond optimal stimulation at 30 and 40 mg/ml. Similarly, there was a four-fold increase in the formation of 4-androstenedione from [14C]testosterone as substrate (Fig. 4; n=4; p<0.001) and a two-fold increase in the formation of testosterone from [14C]4-androstenedione as substrate, at an optimal concentration of 20 mg/ml, maintaining high yields at 30 and 40 mg/ml (Fig. 5; n=4; p<0.001). 6. Discussion Both androgens, [14C]testosterone and [14C]4-androstenedione, were taken up by con¯uent monolayer cultures of human gingival ®broblasts over a period of 5 min to 8 h. The presence of minocycline in the medium seemed to enhance this process for both androgens, but they displayed slightly dierent patterns of uptake. In order to demonstrate the implications of substrate metabolism during the periods used to study uptake of substrate by the cells, the medium was analysed when [14C]testosterone was used as substrate in the presence or absence of minocycline. There was a small amount
of metabolism of substrate, as shown in Figs. 1 and 2. This might not have made a signi®cant dierence to the pattern of uptake of steroid substrate, as the cells were in constant contact with a reservoir of substrate; steady uptake by passive diusion could counteract any metabolic loss into the medium. It is interesting that the metabolites were formed at an earlier stage when minocycline was also present in the medium than with [14C]testosterone alone. This observation further reinforces our ®ndings from the metabolic studies concerning the enhanced expression of steroid-metabolizing enzymes in the presence of minocycline. Steroid hormones share a common general mechanism of action. They are lipophilic and non-ionic, and are therefore transported mainly by simple diusion, as they encounter no obstruction at the lipid-bilayered plasmalemma. They then freely enter the cells, being captured and preferentially retained only in target cells possessing receptor molecules at nuclear sites (Szego, 1994). This would account for the relative ease with which optimal intracellular concentrations of steroid hormones are achieved. Enhanced uptake of these hormones by ®broblasts in response to minocycline could occur as a result of an increased size or number of plasmalemmal pores. When cell-free preparations of ®broblasts were incubated with radiolabelled androgens, the substrates were metabolized to the biologically active androgen DHT as a result of 5a-reductase activity; 17b-hydroxysteroid dehydrogenase activity resulted in the formation of 4-androstenedione or testosterone from each of the substrates used. There was considerable enhancement of these enzyme activities in the presence of optimal concentrations of minocycline in the cellfree system, with gradual increases at suboptimal concentrations and not much change beyond optimal stimulation. This pattern of stimulation appears to be consistent with that seen in protein studies investigating the eects of minocycline on matrix synthesis. The metabolic yields in the present investigation were signi®cantly higher than with the stimulation of these enzyme systems in intact con¯uent monolayers of ®broblasts (Soory and Kasasa, 1997). It is possible that intracellular factors known to enhance these enzyme activities are in close proximity to each other and not limited by the constraints of membrane barriers on subcellular organelles. For example, a phospholipid environment enhances 5a-reductase activity. Cook and Robair (1985) show that treatment of the membrane fraction of the epididymis with phospholipase A2 and C increased the activity of 5a-reductase. Phosphatidyl cholines were always found to activate 5a-reductase both in epididymis and liver. This mechanism is applicable to the en-
M. Soory, H. Virdi / Archives of Oral Biology 44 (1999) 215±222
vironment of the in¯amed periodontium, as during in¯ammatory episodes increased amounts of phospholipases A and C are synthesized by micro-organisms from dental plaque, including spirochaetes (Siboo et al., 1989). The amounts of phospholipase C are also high in leucocytes, which release this enzyme during cell lysis (Siboo et al., 1989). These events could modulate 5a-reductase activity in the in¯amed periodontium. Stimulation of 5a-reductase activity by growth factors or certain drugs results in fairly rapid selectivity, ligand speci®city and correspondence of biochemical and cytostructural responses in the early stages of hormone action. There is increasing evidence of vesicular translocation of eectors to perinuclear sites within a short period of exposure of target cells to the relevant hormone. The generalized nature of this transmission in target cells appears to signify an important pattern in staging a suitable cellular response (Szego, 1994), such as matrix synthesis in ®broblasts. Using a cellfree system of cultured ®broblasts to demonstrate the eect of minocycline on the metabolism of steroid hormones encapsulates these processes by overcoming the membrane barriers of organelles. Such a pattern of perinuclear aggregation, sometimes extending to nuclear capping of lysosomal elements, is also characteristic of rapidly growing tissues in the presence of growth factors/other matrix stimulants. These hormone-labilized lysosomes are very eective in promoting events at a nuclear level by ®rst breaching the nuclear envelope with enzymes such as lysosomal cathepsin B. This breach results in the regulation of speci®c nuclear-protein import and bidirectional molecular tracking across the pore complex (Gerace, 1992). Additionally, phosphorylation of certain nearby residues, with consequent conformational changes, might also occur under the in¯uence of lysosomal protein kinase. All the above stages required for a cellular hormonal response to stimulants are obviated in a cell-free system devoid of cytostructural barriers. This may explain the eective metabolic response of ®broblasts in a cellfree system to optimal concentrations of minocycline. That system has proved to be an eective model for demonstrating androgen metabolic responses in ®broblasts and the potential of minocycline in mediating anabolic repair processes by stimulating 5a-reductase activity. In¯ammatory periodontal disease results from interaction between bacterial insult and host response. Antimicrobials such as tetracycline, and their semisynthetic analogues such as minocycline and doxycycline, have been used as adjunctive therapeutic agents. They are eective adjuncts due to their anti-in¯ammatory and proanabolic functions in addition to their antimicrobial properties. The chemically modi®ed tetracyclines, in which the antimicrobial moiety has been
221
removed, overcome the risks of microbial resistance while retaining proanabolic and anti-in¯ammatory properties. These agents may be useful in curtailing the uncontrolled in¯ammatory process triggered by microbial activity in aggressive forms of periodontal diseases, which can be damaging. They could also contribute to anabolic repair in connective tissue and bone. Karimbux et al. (1998) show that the non-antimicrobial, chemically modi®ed tetracycline-1, restored type 1 and type XII collagen mRNA expression to control levels, in the in¯amed periodontal tissues of adult male Sprague±Dawley rats. This eect may be the result of stimulation of the synthetic capacity of ®broblasts and inhibition of matrix metalloproteinase activity. Similarly, the eects of minocycline on 5a-reductase activity demonstrated here in a cell-free model could have implications in vivo for the healing periodontium via androgen-mediated, connective-tissue repair. Acknowledgements We wish to acknowledge Dr I. Tavares for use of the radioisotope scanner at the Department of Surgery, Rayne Institute, King's College Medical School, London. References Breeveld, F.C., Dijkmans, B.A.C., Mattie, H., 1990. Minocycline treatment for rheumatoid arthritis: An open dose ®nding study. J. Rheumatol 17, 43±46. Cook, G.M., Robaire, B., 1985. Modulation of epididymal 5a-reductase in vitro by the phospholipid environment. J. Biol. Chem 260, 7489±7495. Genco, R.J., 1981. Antibiotics in treatment of human periodontal diseases. J. Periodontol 52, 545±558. Gerace, L., 1992. Molecular tracking across the nuclear pore complex. Curr. Opinion Cell Biol 4, 637±645. Golub, L.M., McNamara, T.F., D'Angelo, G., Greenwald, R.A., Ramamurthy, N.S., 1987. A non antibacterial chemically modi®ed tetracycline inhibits mammalian collagenase activity. J. Dent. Res 66, 1310±1314. Karimbux, N.Y., Ramamurthy, N.S., Golub, L.M., Nishimura, I., 1998. The expression of collagen I and XII mRNAs in porphyromonas gingivalis-induced periodontitis in rats: The eect of doxycycline and chemically modi®ed tetracycline. J. Periodontol 69, 34±40. Kasasa, S.C., Soory, M., 1996. The synthesis of 5a-dihydrotestosterone from androgens by human gingival tissues and ®broblasts in culture in response to TGF-b and PDGF. J. Periodont. Res 31, 312±321. Kasperk, C.H., Wergedal, J.E., Farley, J.R., Linkhart, T., Turner, R.T., Baylink, D.J., 1989. Androgens directly stimulate proliferating bone cells in vitro. Endocrinol 124, 1576±1578.
222
M. Soory, H. Virdi / Archives of Oral Biology 44 (1999) 215±222
Kornman, K.S., Karl, E.H., 1982. The eect of long-term low dose tetracycline therapy on the sub-gingival ¯ora in refractory adult periodontitis. J. Periodontol 53, 604±610. Moore, W.E.C., Ranney, R.R., Holdman, L.V., 1982. Subgingival ¯ora in periodontal disease: Cultural studies. In: Genco, R.J., Mergenhagen, S.E. (Eds.), Host±Parasite Interaction in Periodontal Disease. ASM Publications, Washington DC, pp. 13±26. O'Connor, B.C., Newman, H.N., Wilson, M., 1990. Susceptibility and resistance of plaque bacteria to minocycline. J. Periodontol 61, 228±233. Rappaport, S.C., Vittek, J., Altman, K., Gordon, G.G., Southren, A.L., 1976. Sex dierences in the metabolism of androgens by human gingiva. J. Dent. Res 55, 1371. Rompen, E.H., Kohl, J., Nusgens, B., Lapiere, C.M., 1993. Kinetic aspects of gingival and periodontal ligament ®broblast attachment to surface conditioned dentine. J. Dent. Res 72, 607±612. Ryan, M.E., Ramamurthy, N.S., Golub, L.M., 1996. Matrix metalloproteinases and their inhibition in periodontal treatment. Curr. Opinion Periodontol 3, 85±96. Sasaki, T., Kaneko, H., Ramamurthy, N.S., Golub, L.M., 1991. Tetracycline administration restores osteoblast structure and function during experimental diabetes. The Anat. Rec 231, 25±34. Schneir, M., Ramamurthy, N.S., Golub, L., 1990. Minocycline treatment of diabetic rats normalises skin collagen production and mass: possible causative mechanisms. Matrix 10, 112±123. Siboo, R., Al, Jobuiri W., Gornitsky, M., Chan, E.C.S., 1989. Synthesis and secretion of phospholipase C by oral spirochaetes. J. Clin. Microbiol 27, 568±570. Sooriyamoorthy, M., Gower, D.B., 1989. Hormonal in¯uences on gingival tissue; relationship to periodontal disease. J. Clin. Periodontol 16, 201±208.
Sooriyamoorthy, M., Gower, D.B., Eley, B.M., 1990. Androgen metabolism in gingival hyperplasia induced by nifedipine and cyclosporine. J. Periodont. Res 25, 25±30. Sooriyamoorthy, M., Harvey, W., Gower, D.B., 1988. The use of human gingival ®broblasts in culture for studying the eects of phenytoin on testosterone metabolism. Archs. Oral Biol 33, 353±359. Soory, M., 1995. Bacterial steroidogenesis by periodontal pathogens and the eect of bacterial enzymes on steroid conversions by human gingival ®broblasts in culture. J. Periodont. Res 30, 124±131. Soory, M., Gower, D.B., 1990. Investigation of androgen metabolism by hyperplastic tissue ®broblasts in relation to phenytoin. Med. Sci. Res 18, 835±837. Soory, M., Kasasa, S.C., 1997. The eects of epidermal growth factor, interleukin-1 and phenytoin, alone and in combination on C19 steroid conversions in ®broblasts. J. Periodontol 68, 819±826. Soory, M., Virdi, H., 1998. Eects of the anti-androgen ®nasteride on 5 a -reductase activity in human gingival ®broblasts in response to minocycline. J. Clin. Periodontol 25, 67±73. van Steenberghe, D., Bercy, P., Kohl, J., De Boever, J., Adriaens, P., Vanderfaeillie, A., et al., 1993. Subgingival minocycline hydrochloride ointment in moderate to severe chronic adult periodontitis: A randomised double blind, vehicle controlled multicentre study. J. Periodontol 64, 637±644. Sultan, C., Loire, C., Kern, P., 1986. Collagene et hormones steroids. Ann. Biol. Clin 44, 285±288. Szego, C.M., 1994. Cytostructural correlates of hormone action: new common ground in receptor mediated signal propagation for steroid and peptide agonists. Endocrine 8, 1078±1093.
The uptake of two androgen substrates by cultured human gingival ®broblasts in response to minocycline and metabolic studies using a cell-free system M. Soory*, H. Virdi Department of Periodontology, G.K.T Dental Institute (King's College Hospital), London, SE5 9RW, UK Accepted 17 November 1998
Abstract The uptake of two androgen substrates by cultured ®broblasts and the anabolic response of homogenates of cultured gingival ®broblasts to minocycline were investigated. Monolayer cultures of con¯uent ®broblasts were incubated with [14C]testosterone/[14C]4-androstenedione for timed intervals in the presence or absence of optimal concentrations of minocycline. The intracellular uptake of androgens was quanti®ed by radio-isotope counts on cell digests. Con¯uent gingival ®broblasts were homogenized by snap freezing/rapid thawing and duplicate incubations were made in phosphate-buered saline (pH 6.5) with radiolabelled androgens, in the presence or absence of minocycline, for 24 h. At the end of the incubation period the buer was extracted for radioactive metabolites, analysed and quanti®ed with a radio-isotope scanner. There were 30% increases in the uptake of androgen substrates in the presence of minocycline (n=3; p<0.01; one-way ANOVA). With the metabolic studies there were 2±3-fold increases in the formation of dihydrotestosterone from [14C]testosterone and [14C]4-androstenedione, respectively (n=4; p<0.001; one-way ANOVA), and 4-fold/2-fold increases in the formation of 4-androstenedione/ testosterone from these substrates (n=4; p<0.001) in response to an optimal concentration of 20 mg/ml of minocycline, compared with control incubations. The presence of minocycline in the incubate signi®cantly increased the activity of the steroid-metabolizing enzymes. This increase might result from increased intracellular availability of steroid substrate and enhanced metabolic activity. Homogenates of cultured gingival ®broblasts are a useful model for studying the anabolic potential of minocycline in gingiva, using C19 steroid substrates. # 1999 Elsevier Science Ltd. All rights reserved. Keywords: Uptake; Metabolism; Androgens; Fibroblasts; Cell-free system; Minocycline
1. Introduction The subgingival micro¯ora is important in the initiation and progression of chronic periodontal dis-
Abbreviations: DHT, 5a-dihydrotestosterone, MEM, minimal essential medium, PBS, phosphaye-buered saline. * Corresponding author. Fax: 0171-346-3185
eases. Colonies of Gram-negative anaerobes are consistently isolated from diseased sites of the periodontium and are implicated in their aetiology (Moore et al., 1982). Hence antimicrobial agents such as tetracycline and metronidazole have been used as adjuncts in the treatment of in¯ammatory periodontal diseases, particularly in cases of aggressive disease in young patients and in recalcitrant forms (Genco, 1981; Kornman and Karl, 1982). Minocycline, the semisyn-
0003-9969/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved. PII: S 0 0 0 3 - 9 9 6 9 ( 9 8 ) 0 0 1 2 1 - 6
216
M. Soory, H. Virdi / Archives of Oral Biology 44 (1999) 215±222
thetic analogue of tetracycline, reduced pocket probing depths and the number of periodontal pathogens in the adjunctive treatment of moderately severe adult periodontitis (van Steenberghe et al., 1993). Minocycline was more eective than tetracyclines against several bacteria commonly found in dental plaque (O'Connor et al., 1990). In addition to their antimicrobial properties, tetracyclines have anti-in¯ammatory (Ryan et al., 1996) and proanabolic eects on connective tissue (Schneir et al., 1990) and bone matrices (Sasaki et al., 1991). When dentine granules were conditioned with minocycline hydrochloride, the subsequent attachment of human periodontal ligament cells to the granules improved (Rompen et al., 1993). The chemically modi®ed tetracyclines, in which the antimicrobial moiety has been removed, retaining the anti-in¯ammatory and pro-anabolic eects, have been used in this capacity for the treatment of autoimmune diseases (Golub et al., 1987; Breeveld et al., 1990). Our previous studies have demonstrated increased formation of the biologically active androgen DHT in response to stimulants that increase protein turnover, such as growth factors and drugs implicated in causing gingival overgrowth (Kasasa and Soory, 1996; Soory and Kasasa, 1997), indicating a link between biologically active androgens and anabolic eects on connective-tissue matrices. In view of the proanabolic eects of androgen metabolites such as DHT on the synthetic functions of ®broblasts (Sultan et al., 1986) and osteoblasts (Kasperk et al., 1989), it was relevant to investigate the anabolic eects of minocycline, using androgen substrates. Our aim now was to study the uptake of androgen substrates by cultured gingival ®broblasts and their metabolic pro®les in a cell-free system. 2. Materials and methods The radioisotopes [14C]testosterone and [14C]4androstenedione (spec. act. 58 mCi/mmol) were obtained from Amersham International, Amersham, Bucks. Organic solvents (benzene, acetone) for thinlayer chromatography, ethyl acetate for extraction of metabolites and chloroform to redissolve the dried extract were all provided by Merck Ltd., Dagenham, Essex. Incubations were done in PBS at pH 6.5 for experiments with homogenates of cultured ®broblasts (constituents were from Sigma Chemical Co. Ltd., Poole, Dorset); the incubation medium for the uptake experiments was Eagle's MEM with 10% fetal bovine serum, L-glutamine, antibiotic/antimycotic solution and sodium bicarbonate, which were obtained from Gibco Ltd., Paisley, Scotland. The pharmacy at King's
College School of Medicine and Dentistry provided the minocycline. For the derivatization of steroids, O-penta¯uorobenzyl hydroxylamine hydrochloride (Florox reagent) and N, )-bis trimethylsilyl tri¯uoroacetamide were obtained from Pierce and Warriner (UK) Ltd, Chester. Chronically in¯amed gingival tissue was taken from patients (aged 30±50 years) with chronic peridontal disease during surgical treatment for pocket elimination, after the completion of initial treatment consisting of scaling and root planing. Our previous work established that cells grown from an in¯amed tissue maintain their characteristics over serial passaging, demonstrating a heightened androgen-metabolic pro®le compared with cells derived from healthy gingiva, both at baseline and in response to stimulants (Sooriyamoorthy et al., 1988; Sooriyamoorthy and Gower, 1989; Sooriyamoorthy et al., 1990; Soory and Gower, 1990). Hence the experimental model here consisted of ®broblasts passaged from chronically in¯amed tissue, thus simulating conditions in the in¯amed periodontium and events leading to repair. Clones of ®broblasts respond to in¯ammatory stimuli, altering their phenotype to meet functional demands. The metabolism of androgen substrates in cells derived from a healthy tissue occurs predominantly in males, while 4-androstenedione, the main circulatory androgen in females, is preferentially metabolized by cells from this source. In in¯amed gingiva, substrate preference and the extent of androgen metabolism is similar in both males and females (Rappaport et al., 1976). Therefore the present experimental model, comprising cells/homogenized cells derived from in¯amed tissue, will incorporate a spectrum of anabolic activity from the in¯amed periodontium, representing males and females. 2.1. Cell culture The gingival tissue was washed in physiological saline, placed in Hank's balanced salt solution and minced in Eagle's MEM into 1-mm3 pieces. It was then plated in 25-cm2 ¯asks with Eagle's MEM until primary cultures were established. The contents of a fully con¯uent 25-cm2 ¯ask (2.2 106 cells) were distributed into 24 wells of a multiwell dish and triplicate incubations were performed with [14C]testosterone for timed periods of 5, 15, 30 min, 1, 2, 3, 4, 6 and 8 h, in the presence or absence of optimal concentrations of minocycline. These optima had been established previously, using a series of ®broblast cell lines (Soory and Virdi, 1998), and found to be 20 mg/ml. At each incremental period, the medium was removed over crushed ice to minimize the passage of substrate taken up by the cells and washed twice in physiological saline.
M. Soory, H. Virdi / Archives of Oral Biology 44 (1999) 215±222
217
The cells at the base of the multiwell dishes were then digested with 1N NaOH and solvent-extracted with ethyl acetate (2 ml2) to isolate the radiolabelled testosterone taken up by them. It was then evaporated and spotted on a thin-layer chromatographic plate for the purpose of quantifying the uptake of radiolabelled steroid at dierent intervals, using a radio-isotope scanner. In order to clarify the possibility of metabolism of radiolabelled substrate during the periods studied, the medium was also analysed for steroid metabolites. This was done for one set of experiments with [14C]testosterone as substrate, in the presence or absence of minocycline. For each of the timed periods, the medium was isolated, solvent-extracted and the metabolites separated by thin-layer chromatography. The chromatographic plates were subjected to radioisotope scanning for quanti®cation of metabolites. The analytical details are described fully below. 3. Metabolic studies These experiments were done with cell lines derived from several patients, with independent controls for each line studied. Cells derived from individual cultures were not pooled. Gingival ®broblasts were serially passaged and cells of the 5th±9th passages were used for the experiment. They were grown out in 25cm2 ¯asks and, when con¯uent, the contents of each ¯ask were used for duplicate incubations after homogenization of the con¯uent cells. Homogenization was begun by removing the incubation medium, washing the cell layer in PBS (pH 6.5) and subjecting the ¯ask to snap freezing and thawing in liquid nitrogen repeatedly in a small volume of the buer until the cells were fragmented. They were then homogenized with a fused pipette; incubations were made in 24-well multiwell dishes, using 1 ml PBS per incubation, with radiolabelled substrates [14C]testosterone/[14C]4-androstenedione and serial concentrations of minocycline of 5, 10, 20, 30 and 40 mg/ml. At the end of a 24-h incubation, the homogenates were solvent-extracted for steroid metabolites with ethyl acetate (2 ml3) after the addition of cold carrier steroids and evaporated to dryness in a vortex evaporator (Gyrovap, VA Howe Ltd., Banbury, Oxon). The samples were then solubilized in 100 ml chloroform and subjected to thin-layer chromatography in a benzene/acetone solvent system (4:1, v/v) to separate the metabolites from the two radiolabelled androgen substrates. The chromatographic plates were then scanned in a radio-isotope scanner for the quanti®cation of the separated metabolites. The formed metabolites were tentatively identi®ed by establishing the mobility of
Fig. 1. Uptake of [14C]testosterone by human gingival ®broblasts. Median values and SD for substrate uptake by the cells are shown [14C-T(C)]. Similarly, the metabolites isolated from the medium (Met) at timed periods are shown at the base of the graph.
cold standards, using an iodine tank to disclose them on the thin-layer chromatographic plates, to show correspondence with the radioactive metabolites identi®ed during radio-isotope scanning. Identi®cation was con®rmed by comparing the fragmentation pattern of derivatized samples with authentic standards, using gas chromatography±mass spectrometry. 4. Characterization of DHT by gas chromatography± mass spectrometry It was important to con®rm the identity of 5a-DHT,
218
M. Soory, H. Virdi / Archives of Oral Biology 44 (1999) 215±222
14
Fig. 2. Uptake of [ C]testosterone by human gingival ®broblasts in the presence of minocycline. Values for substrate uptake by the cells are shown [14C-T (M)], as in Fig. 1. The metabolites (Met) isolated from the medium are shown at the base of the graph.
being the physiologically active androgen in healing responses. Several incubations were made under the conditions described above, using unlabelled testosterone (10ÿ6 M). After extraction, the identity of 5a-DHT as a metabolite in the dried extracts was con®rmed by gas chromatography±mass spectrometry (courtesy of Professor A.I Mallet, St. Thomas' Hospital, London, UK), after derivatization to penta¯uorobenzyloxime trimethyl silyl ether. The derivatized biological material had a molecular ion (557) and mass-spectral fragmentation pattern identical to those of the authentic penta¯uorobenzyloxime trimethyl silyl ether of 5a-DHT, but at lower levels, owing to the smaller concentrations of
Fig. 3. Uptake of [14C]4-androstenedione by human gingival ®broblasts in the presence or absence of minocycline. Uptake of [14C]4-androstenedione by ®broblasts [14C-4-A(C)] compared with uptake in the presence of minocycline [14C-4A(M)] at timed intervals. Median values and SD are shown.
the steroid. These procedures were described in detail by Soory (1995). 5. Results 5.1. Uptake of [14C]testosterone/[14C]4androstenedione by gingival ®broblasts in the presence or absence of minocycline 5.1.1. [14C]testosterone Control incubations showed a gradual increase in the uptake of radiolabelled testosterone by the gingival ®broblasts over a period of 1±3 h by 20±40% (Fig. 1;
M. Soory, H. Virdi / Archives of Oral Biology 44 (1999) 215±222
219
Fig. 4. Metabolism of [14C]testosterone by homogenates of human gingival ®broblasts in response to minocycline. Duplicate incubations with [14C]testosterone (CH) in the presence or absence of serial concentrations of minocycline (MH: 5, 10, 20, 30 and 40 mg/ml) for 24 h. Median values (pmol/ ml) and SD are shown.
Fig. 5. Metabolism of [14C]4-androstenedione by homogenates of human gingival ®broblasts in response to minocycline. As Fig. 4, with [14C]4-androstenedione (CH) as substrate in the presence or absence of minocycline (MH). Median values (pmol/ml) and SD are shown.
n=3; p<0.01, one-way ANOVA) when compared with the initial uptake at 5 min. This increase remained fairly similar at 4, 6 and 8 h. In the presence of minocycline, the initial uptake of [14C]testosterone was similar to that of control incubations at 5 min, with gradual increases of 25% at 15 min, 40% at 30 min and 57% at 1 h (Fig. 2; n=3; p<0.01) over the value at 5 min. There was a further increase in uptake at 3 h, resulting in a two-fold increase over that in the 5 min period (n=3, p<0.001), which was maintained at 4, 6 and 8 h. If Figs. 1 and 2 are compared, it will be seen that the maximum uptake of radiolabelled testosterone
at 3 h was 30% greater when minocycline was present in the medium (n=3; p<0.01). The medium was analysed for the metabolites formed during the timed periods from the uptake of [14C]testosterone in the presence or absence of minocycline. DHT and 4-androstenedione were the main steroids formed and are shown as metabolites in Figs. 1 and 2. There was some metabolic conversion between 3±8 h from [14C]testosterone (Fig. 1) and between 30 min and 8 h in the presence of minocycline (Fig. 2). It is possible that the uptake was still maintained adequately, due to passive diusion from the substrate
220
M. Soory, H. Virdi / Archives of Oral Biology 44 (1999) 215±222
pool in the medium, when this metabolic conversion occurred. 5.1.2. [14C]4-androstenedione When [14C]4-androstenedione was used, its uptake by human gingival ®broblasts was increased by 22% at 3 h (Fig. 3; n=3, p<0.01, one-way ANOVA) compared to that at the 5 min period, and was maintained at this value at 4, 6 and 8 h. The initial uptake in the presence of minocycline was similar to that of control incubations at 5 min, with approx. 35% increases at 15, 30 and 1 h; subsequently, at 2 h and 3 h, there was a gradual increase of 40±46% (n=3; p<0.01). This increase was maintained at 4, 6 and 8 h. The maximum uptake at 3 h was about 30% greater in the presence of minocycline than in controls (n=3; p<0.01). 5.2. Metabolism of [14C]testosterone/[14C]4androstenedione by homogenates of gingival ®broblasts in response to minocycline When homogenates of human gingival ®broblasts were incubated with the radiolabelled androgen substrates testosterone or 4-androstenedione, they were metabolized by the enzymes 5a-reductase and 17b-hydroxysteroid dehydrogenase to DHT and 4-androstenedione or testosterone, respectively. There were 2±3fold increases in the formation of DHT from [14C]testosterone and [14C]4-androstenedione, respectively, at a minocycline concentration of 20 mg/ml (Figs. 4 and 5; n=4, p<0.001; one-way ANOVA), with gradual increases at suboptimal concentrations and maintained beyond optimal stimulation at 30 and 40 mg/ml. Similarly, there was a four-fold increase in the formation of 4-androstenedione from [14C]testosterone as substrate (Fig. 4; n=4; p<0.001) and a two-fold increase in the formation of testosterone from [14C]4-androstenedione as substrate, at an optimal concentration of 20 mg/ml, maintaining high yields at 30 and 40 mg/ml (Fig. 5; n=4; p<0.001). 6. Discussion Both androgens, [14C]testosterone and [14C]4-androstenedione, were taken up by con¯uent monolayer cultures of human gingival ®broblasts over a period of 5 min to 8 h. The presence of minocycline in the medium seemed to enhance this process for both androgens, but they displayed slightly dierent patterns of uptake. In order to demonstrate the implications of substrate metabolism during the periods used to study uptake of substrate by the cells, the medium was analysed when [14C]testosterone was used as substrate in the presence or absence of minocycline. There was a small amount
of metabolism of substrate, as shown in Figs. 1 and 2. This might not have made a signi®cant dierence to the pattern of uptake of steroid substrate, as the cells were in constant contact with a reservoir of substrate; steady uptake by passive diusion could counteract any metabolic loss into the medium. It is interesting that the metabolites were formed at an earlier stage when minocycline was also present in the medium than with [14C]testosterone alone. This observation further reinforces our ®ndings from the metabolic studies concerning the enhanced expression of steroid-metabolizing enzymes in the presence of minocycline. Steroid hormones share a common general mechanism of action. They are lipophilic and non-ionic, and are therefore transported mainly by simple diusion, as they encounter no obstruction at the lipid-bilayered plasmalemma. They then freely enter the cells, being captured and preferentially retained only in target cells possessing receptor molecules at nuclear sites (Szego, 1994). This would account for the relative ease with which optimal intracellular concentrations of steroid hormones are achieved. Enhanced uptake of these hormones by ®broblasts in response to minocycline could occur as a result of an increased size or number of plasmalemmal pores. When cell-free preparations of ®broblasts were incubated with radiolabelled androgens, the substrates were metabolized to the biologically active androgen DHT as a result of 5a-reductase activity; 17b-hydroxysteroid dehydrogenase activity resulted in the formation of 4-androstenedione or testosterone from each of the substrates used. There was considerable enhancement of these enzyme activities in the presence of optimal concentrations of minocycline in the cellfree system, with gradual increases at suboptimal concentrations and not much change beyond optimal stimulation. This pattern of stimulation appears to be consistent with that seen in protein studies investigating the eects of minocycline on matrix synthesis. The metabolic yields in the present investigation were signi®cantly higher than with the stimulation of these enzyme systems in intact con¯uent monolayers of ®broblasts (Soory and Kasasa, 1997). It is possible that intracellular factors known to enhance these enzyme activities are in close proximity to each other and not limited by the constraints of membrane barriers on subcellular organelles. For example, a phospholipid environment enhances 5a-reductase activity. Cook and Robair (1985) show that treatment of the membrane fraction of the epididymis with phospholipase A2 and C increased the activity of 5a-reductase. Phosphatidyl cholines were always found to activate 5a-reductase both in epididymis and liver. This mechanism is applicable to the en-
M. Soory, H. Virdi / Archives of Oral Biology 44 (1999) 215±222
vironment of the in¯amed periodontium, as during in¯ammatory episodes increased amounts of phospholipases A and C are synthesized by micro-organisms from dental plaque, including spirochaetes (Siboo et al., 1989). The amounts of phospholipase C are also high in leucocytes, which release this enzyme during cell lysis (Siboo et al., 1989). These events could modulate 5a-reductase activity in the in¯amed periodontium. Stimulation of 5a-reductase activity by growth factors or certain drugs results in fairly rapid selectivity, ligand speci®city and correspondence of biochemical and cytostructural responses in the early stages of hormone action. There is increasing evidence of vesicular translocation of eectors to perinuclear sites within a short period of exposure of target cells to the relevant hormone. The generalized nature of this transmission in target cells appears to signify an important pattern in staging a suitable cellular response (Szego, 1994), such as matrix synthesis in ®broblasts. Using a cellfree system of cultured ®broblasts to demonstrate the eect of minocycline on the metabolism of steroid hormones encapsulates these processes by overcoming the membrane barriers of organelles. Such a pattern of perinuclear aggregation, sometimes extending to nuclear capping of lysosomal elements, is also characteristic of rapidly growing tissues in the presence of growth factors/other matrix stimulants. These hormone-labilized lysosomes are very eective in promoting events at a nuclear level by ®rst breaching the nuclear envelope with enzymes such as lysosomal cathepsin B. This breach results in the regulation of speci®c nuclear-protein import and bidirectional molecular tracking across the pore complex (Gerace, 1992). Additionally, phosphorylation of certain nearby residues, with consequent conformational changes, might also occur under the in¯uence of lysosomal protein kinase. All the above stages required for a cellular hormonal response to stimulants are obviated in a cell-free system devoid of cytostructural barriers. This may explain the eective metabolic response of ®broblasts in a cellfree system to optimal concentrations of minocycline. That system has proved to be an eective model for demonstrating androgen metabolic responses in ®broblasts and the potential of minocycline in mediating anabolic repair processes by stimulating 5a-reductase activity. In¯ammatory periodontal disease results from interaction between bacterial insult and host response. Antimicrobials such as tetracycline, and their semisynthetic analogues such as minocycline and doxycycline, have been used as adjunctive therapeutic agents. They are eective adjuncts due to their anti-in¯ammatory and proanabolic functions in addition to their antimicrobial properties. The chemically modi®ed tetracyclines, in which the antimicrobial moiety has been
221
removed, overcome the risks of microbial resistance while retaining proanabolic and anti-in¯ammatory properties. These agents may be useful in curtailing the uncontrolled in¯ammatory process triggered by microbial activity in aggressive forms of periodontal diseases, which can be damaging. They could also contribute to anabolic repair in connective tissue and bone. Karimbux et al. (1998) show that the non-antimicrobial, chemically modi®ed tetracycline-1, restored type 1 and type XII collagen mRNA expression to control levels, in the in¯amed periodontal tissues of adult male Sprague±Dawley rats. This eect may be the result of stimulation of the synthetic capacity of ®broblasts and inhibition of matrix metalloproteinase activity. Similarly, the eects of minocycline on 5a-reductase activity demonstrated here in a cell-free model could have implications in vivo for the healing periodontium via androgen-mediated, connective-tissue repair. Acknowledgements We wish to acknowledge Dr I. Tavares for use of the radioisotope scanner at the Department of Surgery, Rayne Institute, King's College Medical School, London. References Breeveld, F.C., Dijkmans, B.A.C., Mattie, H., 1990. Minocycline treatment for rheumatoid arthritis: An open dose ®nding study. J. Rheumatol 17, 43±46. Cook, G.M., Robaire, B., 1985. Modulation of epididymal 5a-reductase in vitro by the phospholipid environment. J. Biol. Chem 260, 7489±7495. Genco, R.J., 1981. Antibiotics in treatment of human periodontal diseases. J. Periodontol 52, 545±558. Gerace, L., 1992. Molecular tracking across the nuclear pore complex. Curr. Opinion Cell Biol 4, 637±645. Golub, L.M., McNamara, T.F., D'Angelo, G., Greenwald, R.A., Ramamurthy, N.S., 1987. A non antibacterial chemically modi®ed tetracycline inhibits mammalian collagenase activity. J. Dent. Res 66, 1310±1314. Karimbux, N.Y., Ramamurthy, N.S., Golub, L.M., Nishimura, I., 1998. The expression of collagen I and XII mRNAs in porphyromonas gingivalis-induced periodontitis in rats: The eect of doxycycline and chemically modi®ed tetracycline. J. Periodontol 69, 34±40. Kasasa, S.C., Soory, M., 1996. The synthesis of 5a-dihydrotestosterone from androgens by human gingival tissues and ®broblasts in culture in response to TGF-b and PDGF. J. Periodont. Res 31, 312±321. Kasperk, C.H., Wergedal, J.E., Farley, J.R., Linkhart, T., Turner, R.T., Baylink, D.J., 1989. Androgens directly stimulate proliferating bone cells in vitro. Endocrinol 124, 1576±1578.
222
M. Soory, H. Virdi / Archives of Oral Biology 44 (1999) 215±222
Kornman, K.S., Karl, E.H., 1982. The eect of long-term low dose tetracycline therapy on the sub-gingival ¯ora in refractory adult periodontitis. J. Periodontol 53, 604±610. Moore, W.E.C., Ranney, R.R., Holdman, L.V., 1982. Subgingival ¯ora in periodontal disease: Cultural studies. In: Genco, R.J., Mergenhagen, S.E. (Eds.), Host±Parasite Interaction in Periodontal Disease. ASM Publications, Washington DC, pp. 13±26. O'Connor, B.C., Newman, H.N., Wilson, M., 1990. Susceptibility and resistance of plaque bacteria to minocycline. J. Periodontol 61, 228±233. Rappaport, S.C., Vittek, J., Altman, K., Gordon, G.G., Southren, A.L., 1976. Sex dierences in the metabolism of androgens by human gingiva. J. Dent. Res 55, 1371. Rompen, E.H., Kohl, J., Nusgens, B., Lapiere, C.M., 1993. Kinetic aspects of gingival and periodontal ligament ®broblast attachment to surface conditioned dentine. J. Dent. Res 72, 607±612. Ryan, M.E., Ramamurthy, N.S., Golub, L.M., 1996. Matrix metalloproteinases and their inhibition in periodontal treatment. Curr. Opinion Periodontol 3, 85±96. Sasaki, T., Kaneko, H., Ramamurthy, N.S., Golub, L.M., 1991. Tetracycline administration restores osteoblast structure and function during experimental diabetes. The Anat. Rec 231, 25±34. Schneir, M., Ramamurthy, N.S., Golub, L., 1990. Minocycline treatment of diabetic rats normalises skin collagen production and mass: possible causative mechanisms. Matrix 10, 112±123. Siboo, R., Al, Jobuiri W., Gornitsky, M., Chan, E.C.S., 1989. Synthesis and secretion of phospholipase C by oral spirochaetes. J. Clin. Microbiol 27, 568±570. Sooriyamoorthy, M., Gower, D.B., 1989. Hormonal in¯uences on gingival tissue; relationship to periodontal disease. J. Clin. Periodontol 16, 201±208.
Sooriyamoorthy, M., Gower, D.B., Eley, B.M., 1990. Androgen metabolism in gingival hyperplasia induced by nifedipine and cyclosporine. J. Periodont. Res 25, 25±30. Sooriyamoorthy, M., Harvey, W., Gower, D.B., 1988. The use of human gingival ®broblasts in culture for studying the eects of phenytoin on testosterone metabolism. Archs. Oral Biol 33, 353±359. Soory, M., 1995. Bacterial steroidogenesis by periodontal pathogens and the eect of bacterial enzymes on steroid conversions by human gingival ®broblasts in culture. J. Periodont. Res 30, 124±131. Soory, M., Gower, D.B., 1990. Investigation of androgen metabolism by hyperplastic tissue ®broblasts in relation to phenytoin. Med. Sci. Res 18, 835±837. Soory, M., Kasasa, S.C., 1997. The eects of epidermal growth factor, interleukin-1 and phenytoin, alone and in combination on C19 steroid conversions in ®broblasts. J. Periodontol 68, 819±826. Soory, M., Virdi, H., 1998. Eects of the anti-androgen ®nasteride on 5 a -reductase activity in human gingival ®broblasts in response to minocycline. J. Clin. Periodontol 25, 67±73. van Steenberghe, D., Bercy, P., Kohl, J., De Boever, J., Adriaens, P., Vanderfaeillie, A., et al., 1993. Subgingival minocycline hydrochloride ointment in moderate to severe chronic adult periodontitis: A randomised double blind, vehicle controlled multicentre study. J. Periodontol 64, 637±644. Sultan, C., Loire, C., Kern, P., 1986. Collagene et hormones steroids. Ann. Biol. Clin 44, 285±288. Szego, C.M., 1994. Cytostructural correlates of hormone action: new common ground in receptor mediated signal propagation for steroid and peptide agonists. Endocrine 8, 1078±1093.