J Clin PeriodoiiUil mS: 25: ^7-7,?Prinlt'tl ill Denmark . All rights reserved

Copyright © Munksgaard 1998

~' mm

pcriodDiitologyISSN 03U.=l-6979

Effects of the anti-androgenfinasteride on 5a-reductaseactivity in human gingivalfibroblasts in response tominocycline

M.Soory and H.VirdiDepartment of Periodontology, King's CollegeSchool of Medicine & Dentistry, London SE59RW, United Kingdom

Soory M, Vinli H: Effects of the anti-amlrogen finusleride on 5a-reductase activityin human gingival fibroblasts in respon.se to minocrctinc. J Clin Periodonlol 1998:25: 67-73. © Munksgaard. 1998.

Abstract. In addition lo their antimicrobial properties, tetracyclines have anti-inflamniiitory and pro-anabolic elTects on the reparatory potential of connectivetissue and bone. The physiologically active androgen 5a-dihydrotestosterone(DHT) implicated in matrix synthesis is fofmed in gingivae from androgen sub-strates. The aim of this investigation is to study the androgen metabolic responseof gingivae to minocyeline. in ihe presence or absence of the anti-androgen ftnas-teride. Chronically inflamed gingival tissue derived from 12 subjects aged 30-50years and passaged fibroblasts derived from this source, were used for the ex-periments. Duplicate incubations were performed in Eagle's MEM with 14C-testosierone/14C-4-androstenedione in the presence or absence of minocycline(5-60 /ig/ml) or finasteride for 24 h. The androgen substrate l4C-testosteronewas metabolised rnainly to DHT and 4-androstenedione. while 14C-4-andros-tenedione was converted mainly to DHT and testosterone. Minocycline at 20-30/(g/ml stimulated the formation of these metabolites from both substrates byl3-25'^o. In the tissue incubations ihere were 3- and 2-fold increases in DHT and4-androstenedione formation (//^12: /)<O,OI). The anli-androgen finasteridecaused significant inhibition of 5a-reductase activity on both substrates at O.I &1.0 /(g/ml with total inhibition at 10 & 50 /ig/ml (/i^3; p<0.0\). Minoeycline-induced stimulation of 5a-reductase activity was also inhibited by finasteride(7;-4; /)<0.02), Since finasteride inhibition of 5a-reductase activity is specific forthe type 2 isoenzyme associated with anabolic functions of target tissue, thisenzyme activity may contribute to some of the cited anabolic tissue responses tominocycline.

Key words; androgens; finasteride; fibroblasts;minocycline

Accepted for publication 12 March 1997

Anti-microbial therapy has been usedto help augtnenl the effects of mechan-ical treatment aimed at the removal ofsubgingival calculus and plaque con-stituents (Geneo 1991). In vivo studieshave shown that the semi-synthetic te-tracycline analogue minocycline is moreeffective than the parent compoundagainst several bacteria commonlyfound in plaque (O'Connor et al. 1990),A randomised double-bhnd multicenter

study carried out in patients with mod-erate to severe chronic adult peri-odontal disease demonstrated that ad-junctive administration of minocyclinegel following scaling,/root planing wasmore effective than scaling and rootplaning followed by placebo (vehicle)application (Van Steenberghe et al.1993). Sustained delivery of tetracyclinehydrochloride frotn Actisite fibre hasbeen fYiund to be effective, in main-

taining a substantial concentration ofthe drug in the pocket environment, forthe duration of treatment (Litch et al.1996).

Rompen et al. (1993) showed thatsurface conditioning of dentine withminocycline hydrochloride significantlyimproved the initial attachment of peri-odontal ligament fibroblasts: incuba-tion of fibroblasts with the drug also in-creased their attachment to untreated

68 Soorv £ Virdi

dentine, indicating that minocycline candirectly influence the attachment prop-erties of fibroblasts in addition to con-ditioning dentinal surfaces. The subjectof tetracyclines in the management ofperiodontal diseases has been exten-sively reviewed recently {Seymour &Heasman 1995).

In addition to their antimicrobialproperties, tetracyclines have matrixstimulatory effecls on connective tissueand bone. They have been shown tonormalise skin collagen mass in dia-betic rats (Schneir et al. 1990) and alsoinhibit collagenascs associated with in-flarnmatory degradation of connectivetissue (Golub et al, 1991), Tetracyclineanalogues as proteinase inhibitors, arecurrently closer to being used clinically,than any other agents in periodonlaltherapy (Ryan et al, 1996), Minocyclineand other tetracyclines are known tostimulate alkaline phosphatase activity,thus enhancing the expression of themature osteoblast phenotype (Sasaki etal, 1991),

The anabolic effects of 5a-dihydrole-stosterone on bone and connectivetissue are well documented. For ex-ample, the androgens have a very speci-fic role on the acidic glycosaminogly-cans in rat gingivae by doubling the hy-aluronic acid concentration. Receptorsfor the biologically active androgen 5a-dihydrolesiosterone (DHT) have beenidentified in gingivae (Southren et al.1978) and in human osteoblasts (Kas-perk et al. 1989). Our previous studieshave demonstrated that 5«-reductaseactivity is elevated in human gingivaeand periodontal ligament in response togrowth factors associated with in-flammatory repair (Kasasa & Soory1995, 1996a, b).

The 4-azasteroid compound finas-teride is a specific inhibitor of the 5a-reductase type 2 isoenzyme associatedwith target tissue such as the prostate.The type 2 enzyme is more abundantthan the type 1 enzyme in male repro-ductive tissues, while the type 1 enzymepredominates in peripheral tissues. Thedifferences in substrate affinities andthe tissue distribution of the isoenzymessuggest that type 2 plays an anabolicrole and type 1 a catabolic role. Thesedifferences are reflected in their e,\press-ion levels in the prostate, pH optimasand their sensitivity to 4-azasteroid in-hibitors (Andersson et al. 1991, Jenkinset al. 1992). The rat type 2 isoenzymehas a 10-20-fold higher affinity for tes-tosterone, androstenedione, progester-

one and corticosterone than the type 1isoenzyme and plays an anabolic role.Thus the type 2 isoenzyme may belargely responsible for the paracrineand autocrine generation of DHT inandrogen target tisues, ln contrast, thetype 1 5tt-reductase isoenzyme appearsto have a cataboHc role in androgen me-tabolism, presumably involving the in-activation of testosterone in nonandro-gen target tisues. This occurs particu-larly in the liver, where Ihe type 1isoenzytne is exclusively expressed(Normington & Russell 1992). Otherstudies support the view that gingivaeare target tissues for androgen metab-olism (Ojanotko-Harri et al. 1986),Similarly Kasperk et al. (1989) demon-strated that androgens directly stititu-lated osteoblasts in vitro. In view oi theabove findings, it was relevant to ident-ify the type of isoenzyme activity in thegingivae with regard to the cited ana-bolic functions.

The aim of this investigatioti wastherefore to study 5«-reductase activityon 2 androgen substrates in gingivaeand cultured gingival fibroblasts in re-sponse to minocycline and the effect ofthe anli-androgen finasteride on thisconversion, to elucidate the anabolicpotential of minocycline in gingivae.

Material and Methods

The radioisotopcs l4C-testosterone and14C-4-androstenedione (specific radio-activity 58 mCi/mmol, were obtainedfrom Amersham International, Amer-sham. Bucks., UK. Organic solvents(benzene, acetone) for thin layer chrom-atography (TLC). ethyl acetate for ex-traction of metabolites and chloroformto re-dissolve the dried extract were allprovided by Merck Ltd., Dagenham,Essex.

The incubation medium used wasEagle's Minimutn Essential Mediutn(MEM) with 10% foetal calf serum, L-glutamine, antibiotic/antimycotie solu-tion and sodium bicarbonate whichwere all provided by Gibco Ltd., Pais-ley. Scotland, Minocycline and finas-teride used in the incubations were ob-tained from the Pharmacy at King'sCollege School oi Medicine and Den-tistry. For the derivatisation of steroids,O- (pentafluorobenzyl) - hydroxyl-amine - HCI (Elorox reagent) and N.0-bis (trimethylsilyl) trifluoroacetamidewere obtained from Pierce and Warriner(UK Ltd,, Chester).

Chronically inflamed gingival tissue

was obtained from periodontal patientsaged 30-50 years af̂ ter completion of in-itial phase therapy, comprising scalingand root planing. The tissue was iso-lated during periodontal surgical pro-cedures. Our previous work has estab-lished the response of human gingivalfibroblasts derived from a healthysource to androgen substrates at base-line and in response to the anti-convuls-ant drug phenytoin (Sooriyamoorthy etal. 198!S). compared with the responseof fibroblasts passaged from inflamedgingival tissue (Sooriyamoorthy &Gower 1989 Sooriyamoorthy et al.1990. Soory & Gower 1990), In theseinvestigations fibroblasts derived froman inflamed tissue source demonstratedheightened metabolic activity with an-drogen substrates, both at baseline andin response to stimulants comparedwith fibroblasts passaged from non-in-Hamed tissue.

It has been shown by others (Rappa-port et al. 1976) that the metabolisrn oftestosterone in inflamed gingivae/fibroblasts from this source is equallyprevalent amongst males and females,whereas in tissue or cells from a healthysource testosterone is tiictabolised prc-dotninantly in males: the predotninantcirculating hormone in females is 4-att-drostenedione which is converted to tes-tosterone via 17(i-hydroxysteroid de-hydrogenase activity and subsequentlyto DHT as a result of 5a-reduction.

The gingival tissue was washed inphysiological saline, placed in Hank'sbalanced salt solution and minced inEagle's MEM. The wet weight was es-tablished and duplicate incubationswere performed for 24 h in I ml Eagle'sMEM with 0.05 pC\ 14C-testosteroneand optitTtai stimulatory concentrationsof tninocycline in a hutiiidified COitissue culture incubator at 37°C.

Gingivnl tissue was obtained as de-scribed above and plated in 25 em-flasks in Eagle's MEM, until primarycultures were established. Serial passa-ging of primary cultures was carriedout by partial digestion with trypsinsolution (0.25%). Cells of the 5th-10thpassage were used for the experimentsThe contents of a fully confluent 25 cm-flask (2x 10" cells) were distributed into24 wells of a multiwell dish and dupli-cate incubations were performed with14C-testosterone and 14C-4-andros-tenedione respectively as initial sub-strates, in Eagle's MEM. The cells wereallowed to become fully confluent in themultiwell dishes prior to experimen-

Minocycline and 5a-reductuse activity 69

tation to overcome mitogenic effects(Kahariet al. 1987).

The androgen substrates 14C-testos-terone/l4C-4-androstenedione togetherwith serial concentrations of minocyc-line and finasteride were used indepen-dently to establish optimal stitnulatory/inhibitory concentrations respectivelyfor 5a-reductase activity. Experimentswere performed in duplicate with cell-lines derived from several subjects, withindependent controls for each test in-cubation. There was no pooling of cellsderived from individual cultures.

Extraction and quantification ofmetabolites

At the end of the incubation period, themedium was removed frotn the multi-well dish, extracted with ethyl acetate(1.5mix3) after the addition of carrierSteroids, and reduced to small bulk in avortex evaporator (Gyrovap. VA.Howe. Ltd., Banbtiry. Oxon. UK). Theresidue was dissolved in 100 fi\ CHChand subjected to thin layer chromato-graphy in a benzene, acetone solventsystem (4:1 v,'v) for the separation ofmetabolites formed lYom 14C-testoster-one and I4C-4-androstenedione. re-spectively. TLC plates were ihenscanned for the quantification of separ-ated metabolites using a Berthold linearanalyser. Identification of the fortnedmetabolites was confirmed using themobility of cold standards and estab-lished further, using gas chromatogra-phy tnass spectrometry (g.e-m.s).

Characterisation of DHT using g.c-m.s

As DHT is the most significant biolo-gically active metabolite in stimulatingfibroblast synthetic activity it was rel-evant to confirrn its identity Under theconditions described above, several in-cubations were performed with unla-belled testosterone (10"^ mol/I), Afterextraction, the identity of 5a-DHT asa metabolite in the dried e.xtracts wasconfirmed by combined gas chromato-graphy-mass spectrometry (g.c-m.s.courtesy Professor A.I. Mallet, St.Thomas' Hospital. London, UK), afterderivatisation to pentafluorobenzylox-ime trimethyl silyl ether (PEBO/TMS).The derivatised biological material hada molecular ion (557) and mass spectralfragmentation pattern identical to thoseof authentic PFBO/TMS ether of 5a-DHT, but at lower levels, due to smallerconcentrations of the steroid. These

procedures have been described in de-tail (Soory 1995).

Establishing the optimal stimulatoryconcentration of minocyctine for 5a-reductase activity in human gingivalfibroblasts

Cultured gingival fibroblasts of the 8thpassage were incubated in duplicatewith 14C-testosterone in Eagle's MEMin (he presence or absence of minocyc-line at serial concentrations of 5, 10, 15,20, 25, 30, 35. 40, 50. 60 and 70 //g/mlfor 24 h. At the end of the incubationperiod, the medium was extracted andanalysed for steroid metabolites as de-scribed above.

Effects of minocycline on 5a-reductaseactivity in human gingival tissue

Duplicate incubations were performedfor 24 h. with human gingival tissue de-rived from 12 subjects after establishingthe wet weight, in Eagle's MEM withl4C-testosterone as substrate, in thepresence or absence of optimal stimu-latory concentrations of minocycline.This was detertnined frotn the above ex-pcritnent. Each duplicate test incuba-tion had its own duplicate controlswithout the drug. At the end of the in-cubation period the incubate was ana-lysed for metabolites as describedabove.

Establishing the optimal inhibitoryconcentration of the anti-androgenfinasteride on 5(i-reductase activity infibroblasts. using 2 androgen substrates14C-testosterone and 14C- 4-androstenedione

Independenl incubations of culturedgingival fibroblasts of the 5th lOthpassage derived from three subjects,were performed in Eagle's MEM withthe anti-androgen flnasteride in serialconcentrations (0,01, 0,1. l.D. 10and50//g/ml) tor 24h, Both androgen sub-strates 14C-testosterone and 14C-4-an-drostenedione were used separately.This was followed by analysis of formedmetabolites.

Effects of 2 inhibitory concentrations offinasteride on minocycline-induced 5»-reductase activity, using 2 androgensubstrates

Hutnan gingival fibroblasts derivedfrotn 4 subjects were incubated in dupli-

cate using 14C-testosterone/i4C-4-and-rostenedione as substrate in the pres-ence or absence of optimal stitnulatoryconcentrations of minocycline (25 fig/'ml) established in the above experimentand 2 inhibitory concentrations of fin-asteride (0.1 and 1,0 /;g/ml) as deter-mined from the above investigation.The combined elTect of minocyclineand finasteride on the metabolism ofandrogen substrates was also studied incomparison with controls for each testgroup, using substrate only.

ResultsEstablishing the optimal stimulatoryconcentration of minocycline for 5«-reductase activity in human gingivalfibroblasts

When serial concentrations of mino-cycline were incubated with 14C-ies-tosterone as substrate there was effec-tive cotiversion of the substrate to 4-androstenedione due to 17/y-hydroxy-steroid dehydrogenase activity andDHT as a result of 5a-reductasc activ-ity in control incubations, as well asin the presence of the drug. Therewere 55'M.. 57!'(p and 49'̂ ^ increases inthe formation of DHT at minoeycline

ia 15 20 2S 30 35 10 SO 60 '0

Fi};. I. Esuiblishirig optimal stimulatory coti-centrations of minocycline for 5«-reduLtaseactivity in human gingiviil fibroblasts. Aver-age Viilues o( duplicate incubations areshown for DHT and 4-A, at each of the con-centrations ol' minocycline.

70 Soory & Virdi

1000

SOO

T

m Conrrol • MInocy

Fig. 2. Effects of minocycline on 5i-(-re-ductase activity in human gingival tissue.Median values and standard deviations ofduplicate incubations performed with tissuederived from 12 subjects are shown i.n=\2:

Fig, y. Fstablishing the optimal inliibitoryconcentration of the anti-androgen finasier-ide on 5«-reductase activity in fibroblasls.using 14C-tes(oslerone as substrate. Medianvalues and standard deviations of duplicateineubations performed wilh 3 cell-lines areshown (rt=3;/7<0,0!).

concentrations of 20. 25 and 30 //g/tnlwith optimal synthetic activity at 25/ig/iiil. Similarly the extent of forma-tion of 4-aiidrostencdione was in-creased by 23'!̂ i, 35'yii and A1"A< at min-ocycline concentrations of 20, 25, and30 /ig/'niL (Fig. I). Mean values ofdtiplicate inctibalions lor controls andeach of the concentrations studied areshown.

Effect of minocycline on 5ii-reductaseactivity in human gingival tissue

When gingival tissue was incubatedwith optimal stimulatory concen-tnitions of minocychnc, there was a 3-fold increase in the formation of DHTover control itictihalioiis. Similarly theformatioti of 4-iindrostenedione was in-creased by 2 fold (Fig. 2; /)=I2;

Establishing the optimal inhibitoryconcentration of the anti-androgenfinasteride on 5(i-reductase activity infibroblasts, using 2 androgen substrates14-C-testosterone and 14C-4-androstenedione

When serial concentrations of the anti-androgen finastcfide (0,01. 0,1. 1,0, 10.0and 50 //g/ml) were incubated with twoandrogen substrates. 5« -reductase ac-tivity was signilicantly inhibited athigher concentrations. When l4C-tes-tosterone was used as substrate, therewere 28% ( ; J<0 ,05 ) and 7,5-fold(/j<0.0!) decreases at O.I and 1,0/(g/tnl.with tottil inhibition at 10 and 50 /(g ml(wilcoxon signed rank test for paired ob-servations: /;-3;/?<0. 01); l7B-hydroxy-steroid dehydrogenase activity leading tothe formation of 4-androstenedionefrom this substrate showed no significantdilTercnce (Fig. 3), Median values ofduplicate incubations set up with 3 cell-lines are represented here.

Similarly, with 14C-4-androstene-dione as substrate, there were 14'%i(P<i). I) and 6.3-rold ( /J<0,0I) decreasesin 5«-reductase activity at finasterideconcentrations of 0.1 and 1 /ig/ml, withtotal inhibition at 10 and 50 //g/tnl(Wilcoxoii signed rank test for paii"edobservations: / ; - 3 ; / J < 0 , 0 1 ) . Again I7B-hydroxysteroid dehydrogenase activityresulting in the formation of lestoster-one frotn l4C-4-androstenedioneshowed no significant change comparedto control incubations (Fig. 4), Medianvalues of duplicate incubations derivedfrotn 3 cell-lines are shown.

Effects of 2 inhibitory concentrations offinasteride on minocycline-induced 5u-reductase activity, using 2 androgensubstrates

When l4C-testosterone/l4C-4-andros-tenedione were used as substrates incu-bated with human gingival fibroblasts.5a-reductase and 17//-hydroxysteroiddehydrogenase activity resulted in theformation of DHT and 4-androstene-dione respectively. There was no signifi-cant change in the yields oi 4-andros-tenedione in response to finasteridewhich caused specific inhibition of DHTformation.

There were 6 duplicate experimentsfor each androgen substrate (14C-T/14C-4-A1 comprising controls, incuba-tions with tninocycline (25 /(g/ml)/linas-teride at two concentrations (O.l/l /(g/ml) and combined incubations of mino-cyciine with each concentration oi' tina-steride. When l4C'-testosterotie was usedas substrate, there was effective conver-sion of the substrate to DHT and 4-and-rostenedione. There was a \y'A, increasein 5«-reduclase activity compared withcontrols ( / J<0 .02) when minocycline was

10

so

Fig. 4. FsUiblisbing the optimal inhibitoryconcentration of linasteride on 5a-reductaseactivity in fibroblasts. using 14C-4-andros-tenedione as substrate. Median values andstandard deviations derived from duplicateincubations of 3 eell-lines are shown (f' = .̂ ;

Afinocvcline and 5a-reductasi' activltv 71

60

Fina

M +Fina

Finb Fin

b

M

Fig. ^. Effects of 2 inhibitory concentrations of tinasteride on miiuKycline-induced 5«-re-ductase activity, usitig 14C-lestosterone us substrate. Median values derived from duplicateincubations of 4 cell-lines and standard deviations are shown (M;25 /;g/ml: Fin,a: 0,1 /;g''ml;Fin.b: 1,0 //g/'ml).

7

Fm a M • Fin a Finb M * Fm b

Fig. 6. Effects of 2 inhibitory concentralions of finasteride ott minoc>'cline- indttced 5«-re-ductase activity, using 14C-4-antirostcncdione as stibstrate. Median values derived from dupli-cate incubations of 4 cell-Mnts and standard deviations are shown.

present in the medium. Finasteride at 0.1//g/ml caused 2.5-fold inhibition of DHTsynthesis compured with controls(/)<0.0!); when minocyciine (25 /(g./ml)and finasteride al ihis concentrationwere present in the medium, the extent ofinhibition was decreased by 10'!̂ .(/X0.02),

Similarly (Inasteride at a concen-tration of 1 /(g/ml caused 9-fold inhi-bition of 5a-reductase activity(/7<0,01), with a 50"A' increase in theyield of DHT (Wilcoxon signed ranktest for paired observations: /i=4;p<0.0\). when minocyciine was pres-ent in the medium (Fig. 5). Medianvalues of duplicate incubations derivedfrom 4 cell-lines are represented, eachwith its own control.

There was no significant change inthe yields of 4-androstenedione.

These experimental conditions are

the same for the following study, usingl4C-4-androstenedione as substrate.

l4C-4-androstenedione was con-verted mainly to DHT and testosterone(Fig, 6), There was a 25"-'!> increase inthe synthesis of DHT in response to mi-nocyeline (/)<0.01). Finasteride causeda 25*'/] inhibition in 5n;-reductase activ-ity over control incubations at 0,1 /ig/ml (/;<(),01): the extent of this inhi-bition was reduced by lQ"Ai when mino-cyciine was present in combination(/j<0.02). Finasteride at the higher con-centration of 1.0 /(g/ml inhibited 5a-re-ductase activity by 3,3-fold (/)<0.01)with further reduction in the presenceof minocyciine.

Discussion

In addition to their anti-microbialproperties and inhibition of matrix

tnetalloproteinases. the tetracyclines ex-ert pro-anabolic effects on bone andconnective tissue (Schneir et al, 1990,Sasaki et al, 1991); these properties areof therapeutic value in the treatment ofinflammatory periodontal disease. Inview of the matrix stimulatory func-tions of DHT it was interesting to ob-ser\e ihe eftecls of minocyciine and spe-cilic inhibition of the 5o;-reductase type2 isoenzyme by finasteride on this me-tabolism in fibroblasts in vitro.

In this investigation, minocyciinecaused stimulation of 5«-reductase ac-tivity on androgen substrates, in mostcell-lines studied, indicating a possiblemechanism for some of its effects onconnective tissue repair. Minocyelinecan mediate the release of growth fac-tors such as IGF (Schneir et al. 1990).which in turn trigger androgen meta-bolic pathways (Kasasa & Soory 1995)by ligand-dependent/ligand-indepen-dent mechanisms (Ignar-Trowbridge etal, 1993), Although the extent of stim-ulation was small, it is significantphysiologically in the context of endo-crinology, to cause some of the effectscited in the literature. Variations be-tween cell-lines in [heir metabolic ca-pacity could reflect selection of differentsub-populations of fibroblasts in thecell-culture dish, from the originalsample, some being more responsivethan others.

Several mechanisms such as inhi-bition of matrix metalloproteinases(Golub et al. 1995),, and pro-anaboliceffects on connective tissue and bonematrices (Schneir et al. 1990) have beenpostulated for the healing potential oftetracyclines.. in addition to their anti-microbial function. One aspect of theiranabolic potential has been investigatedin this study. Tissue incubations withminocyciine resulted in greater stimula-tion of 5a-reductase activity, partly dueto the greater number of cells/gratnweight of tissue compared with 50.000cells/well in each of the cultured cell in-cubations.. Any mechanisms involvingsecondary stimulation of growth fac-tors by minocyciine are likely to bemore active in the tissue incubationscompared with cells which have under-gone serial passaging. Other cell typesfound in human gingival tissue, such asplatelets and macrophages are knownto metabolise androgens and could con-tribute to the elevated yield (Cod-dington et al, I9fi8. Milewich & Whis-etiant I9S2),

Other studies have demonstrated a

72 Soory <& Virdi

lesser metabolic yield associated withpassaged cells in comparison with fresh-ly isolated tissue (Nohutcu et al. 1996),Freshly isolated periodontai ligamenttissue was compared with cells in cul-ture for expression of proteins associ-ated with mineralised tissues, such asbone sialoprotein. osteopontin and typte1 collagen; greater expression was notedfor freshly isolated tissue comparedwith cultured cells,

Finasteride is a specific inhibitor ofthe 5a-reductase type 2 isoenzymewhich is associated with metabolic ac-tivity in target tissue such as the pros-tate, mediating anabolic functions.compared with the catabolic effects ofthe 5a-rcductase type I isoenzyme. Itwas relevant that 5a-reductase activityin the human gingival model used inthese experiments was effectively inhib-ited by finasteride, confirming the pres-ence of the type 2 isoenzyme. This issuggestive of target tissue activity ingingivae, associated with anabolic func-tions,, which is supported by previousstudies carried out by other workers(Southren et al. 1978). Cell viability wasconfirmed by the trypan blue exclusiontest, indicating that the effects of finas-teride were due to enzyme inhibitionand not due to eell death.

These effects of finasteride wereshown using androgen substrates, tes-tosterone and 4-androstenedione intheir conversion to DHT, while therewas not mueh change in 17/y-hydroxy-steroid dehydrogenase activity, resultingin the formation of 4-androstenedione/testosterone respectively from each sub-strate. This confirms the specificity ofthe anti-androgen as a 5a-reductasetype 2 isoenzyme inhibitor (Figs. 3, 4).The specific inhibition of the 5a-re-ductase lype 2 isoenzyme by finasteridein the human gingival model is indica-tive of possible matrix reparatory func-tions which may be attributed to theDHT formed in these tissues.

The slightly smaller inhibitory effectsof finasteride in the presence of mino-cycline {Figs, 5. 6), demonstrate a com-pensatory mechanism which may bepertinent in an environment of in-ilammation: this effect seemed morepronounced in the cell lines which wereless responsive to minocyciine (Fig. 6).The nullifying effects of growth pro-moting substances on the inhibitory ca-pacity of other components present hasbeen reported previously. F'or example,the inhibitory effect of LPS on the pro-liferation of gingiva! fibroblasts has

been shown to be reduced by the simul-taneous presence of the growth factorPDGF(Bartold et al. 1992),

This investigation demonstrates thatminocyciine can stimulate 5a-reductaseactivity in gingivae resulting in the syn-thesis of the physiologically active an-drogen DHT associated with anabolicfunctions. Specific inhibition of the type2 isoenzyme, which is associated withtarget tissue activity in this context, hasbeen shown in comparison with base-line androgen metabolism and in thepresence of minocyciine. These findingssuggest that some of the matrix repara-tory functions of minocyciine in theperiodontium may be attributed to tar-get tissue activity via the androgenmetabolic pathway.

Acknowledgements

The authors wish to acknowledgeProcter & Gamble, Oral & Dental Re-search Trust, UK for their consumablesgrant and Dr. 1. Tavares (Department ofSurgery. Rayne Institute, King's CollegeSchool of Medicine), for usage of theradioisotope scanner in his laboratory.

Zusammenfassung

IVirkung de.s Anti-Aiulrot^e'J.s Ftna.sierid aufllie 5-a-Reduktii.seakiivi!at \on metischlichengingivalen Fibrohlasteti. die aid Mi'tocyetiureagierenZusatzllch zu ihren anlimikrobiellen Eigen-schaflcn haben die Tetracycline eitie antiin-tlammalorische und pro-anabole Wirkungtiiif das ReparationspotetilittI von Bindcgt'we-be Lind Mnochcn, Das pliysiologiscb aklivcAndrogen 5-a-Dihydrotcstosteron (DHT),welches an der Matrixsynlhese beteiligt ist,wird in der Gingivn aiis atidrogenen Siibslra-ten gebildet. Das Ziel dieser Untersuchungist es, die androgenc Sioffwechselreaktion derGingiva unter Minocyclin. bei Vorhanden-sein oder Fehlen von Finastcrid. zu sludie-ren, Chronisch enlziindcics Gingiva gewebevoti 12 Patienten im Alter zwisehen 30 und50 Jahren und passagierte Fibroblasten vondieser Quelle wurden Itir die Experimente be-nutzt, Doppelle lnkubalion in Eagle MEMmit 14C-Testosleron/14C-4-Androstenedionmit oder ohne Minocyclin (5-60 /(g/ml) oderFinaslerid wurde I'llr 24 Stunden durchge-fuhrt. Das androgene Substrat 14C-Testoste-ron wurde hauptsachlich zu DHT und 4-An-drostenedion metabolisiert, wahrend I4C-4-Androstenedion bauptsachlich zu DHT undTestosteron umgebaul wurde. Minocyclin bei20-30 /(g/ml stimulierie die Bildung dieserMetdboliten aus beiden Substraten um 13-25"^, In den Gewebekuliuren gab es 3- und2-rache Erhobungen der DHT- und 4-Andro-stenedion-Bildung (/i= i2; p<O.QI), DasAnli-Androgens Finasterid verursachte eine

signifikante Hemmung der 5-(r-Reduktaseak-livkat mit beiden Substraten bei 0,1 and 1,0/(g/ml mit einer [otaten Hemmung bei 10 and50 //g/nil (« = _•': /)<0,0l). Die von Minocyclininduzierie Stimulation der 5-n-Reduktaseak-livitat wurde auch von Finasterid (« = 4;/)<0,02) gehemmt. Da die Hemmung der 5-K-RedukiaseaktivitiU durch Finasterid spezi-fisch fiir das Typ-2-!soenzym isl. welcbes mitder anabolcn Funktion des Zielgewcbes ver-bunden ist. koniile diese Enzymaktivitat zueinigen der ziticrlen anabolen Gewebereak-lionen von Minocyclin beitragen.

Resume

EJfet de lantiandrugemefinasteride sur i'aeli-vile 5a-reduclase dans les fibrohlastes gingi-viit4.\ huinain.s en repon.se a la niinocyclineEn plus de leurs proprietes aniimicrobicnnes.ies tetracyclines ont des effets antl-inflam-matoires et anabolisants sur le potentiel dereparation du lissu conjonctif et de Tos, L'an-drogcnc 5«-dihydro[estosterone (DHT) pby-siologiquement actif qui participe a Ui syn-Ihese de la mairicc es! forme dans la geneivea partir de substrats androgenes, Le presenttravail se propose d'etudier la reponse tneta-bolique androgene de la gencive a la minocy-ciine. en presence et en I'absence de i'antian-drogene finasteride, Des tissus gingivaux aveeinflammation chroniquc. provenant dc 12sujets ages de 30-50 ans et des fibroblaslescultivcs a partir de cette source avec des se-ries de passages ont servi pour ces experien-ces. Des incubations ont etc pratiquees endouble dans le milieu MEM d'Eagle aveetestosterone-14C/androstenedione-4-14C enpresence ou en I'absence de minocyciine (5 -60 /(g/'ml) ou de finasteride. pendant 24 h. Lesubstrat androgene, la testosterone-14C. a etemetaboliscc principalement en DHT et enandrostenedione-4. tandis que Fandroslene-dione-4-14C citait transformee principale-ment en DHT et testosterone. La minocycii-ne a 20 30 /jg/ml stimulait de 13 -2S"/t> la for-mation de ees metabolites a partir des deuxsubstrats, Dans les incubations de tissus. laformation de DHT et d"atidrostenedione-4augmentait de 3 lois el de 2 fois (n=\2;/)<().01), L'anti-androgene tiiiasteride causaitune inhibition signilicalive de I'activitc 5a-re-ductase sur les deu,K substrats a 0,1 et 1.0 fig/ml. avec inhibition tolale a 10 et 50 /tg/m\('1 = 3; ;j<0.01). La stimulation de l'activite5ct-reductase induite par In minoeyeline elaitaussi inhibee par le finasteride (/i=4;p<0.02). Etant donne que l"inhibition par lefinasteride de I'aetivite 5a-reductase est spe-cifique de I'isozyme de type 2 associee avecdes Ibnclions anaboliques du tissu cible. cetteactivite enzymatique peut contribuer a cer-taines des reponses anaboliques tissulaires ala minocyciine citees.

References

Andersson. S,, Berman. D, M,, Jenkins, E,P & Russell, D, W. (1991) Deletion of ster-oid 5a-reductase 2 gene in male

Minocycline and 5a-reductase activity 73

pseudohermaphroditistn, .\ttiure 354, 159-161,

Bartold. M, P. Narayatiati, A, S, & Page. R,C. (1992) Platelet derived growth factor re-duces the inhibitory efTects of lipopolysac-charide oti gingival fibrobliisl prolifer-ation, Jiiuriiui of Periodontai Research Tl,499-505,

Coddington, C. C , Letterie. G, S.. Klein, T,A. &. Winkel. C, A. (1988) Androgen me-tabolism by hurnan peritoneal macro-phages. Sicrnids 51. 143- 161,

Genco. R, J, 119911, Using anti-microbialagents to manage periodontal diseases,Joitrmii of till' .American Dented Associ-cilion. 122. 3()-38.

Golub, L, M., Sorsa. T, Lee. H-M., Ciancio,S., Sorbi, D., Ramarnurthy. N, S. et al..(1995). Doxycycline inhibits neutrophil(PMN) type matrix metalloproteinases inhuman adult periodontitis gingiva.Journal of Cimicai Periodontoiogy 22. 100109.

ignar-Trowbridge. D, M.. Teng. C, T,, Ross.K. A,. Parker. M, G,. Korach. K, S, MeLachlan. J, A. (1993) Peptide growth fac-tors elicit estrogen receptor-dependenttranscriptional activation of an estrogenresponsive element. Moiecular Endocrin-ology 7, 992-998.

Jenkins, E. P. Andersson. S,. Imperato-McGinley, J.. Wilson. J, D. & Russell. D.W (1992) Genetic and pharmacologicalevidence for more than one human steroid5a-reductase. Journai of Ciinicai invcsliga-lion 89. 293 3U0,

Kahari, V, M,. Heino J, & Vuorio, E, (1987)Interleukin-l increases collagen produc-tion and mRNA levels in cultured skinfibroblasts. Biociiini Biophys .4cni 929.142-147.

Kasasa. S, C, & Soory. M, (1995) The infiti-ence of insulin-like growth factor (IGh) onC,y steroid conversions by human gingivaand in cultured fibroblasts, Journai of peri-odonlology 66. 966-972.

Kasasa. S, C, & Soory. M, (1996a) The effectof interIeukin-1 (IL-1) on androgen metab-olism in human gingival tissue (HGT) andperiodontal ligament (PDL), Journal ofCiinicai Periodontoiogy 23., 419^24.

Kasasa. S, C, & Soory, M, (1996b) The syn-thesis of 5«-dihydrotcstosterone from an-drogens by human gingival tissues andfibroblasts in culture in response to TGF-fi and PDGF. Journal of Pcriodonlai Re-5rarWi 31. 312-321,

Kasperk. C. H.. Wergedat. J, E.. Farley. J. R,.

Linkhart. T, Turner. R. T, & Baylink, D. J.(19S91 Androgens directly stimulate prolif-erating bone cells in vitro, Endocrinoiogy124. 1576 157X,

Litch, J. M.. Encarnaciou. M. Chen, S,.Leonard. J. & Burkoth, T L, (1996) Useof the polymeric matrix as internal stan-dard for qtjantitation of in vivo delivery oftetracycline HCI from Actisite tetracyclinefibre during periodontal treatment,Journai of Periodonlai Re.searcli 31, 540-544,

Milewich. L & Whisenant. M, G, (1982) Me-tabolism of androstenedione by humanplatelets. A source of poteni androgens,Jourmd oj Ciinicai Endocrinology and Me-taboiism 54. 969-974.

Nohutcu. R, M.. Me Cauley. L, K,, Shigeya-ma. Y, & Somerman. M, J, (1996) Express-ion of mineral associated proteins byperiodontal ligament cells: in vitro versusex vivo, Juurnid oj Periodontai Research31,369-372.

Normington. K, & Russell, D, W, (1992)Tissue distribution and kinetic character-istics of rat steroid 5a-re(Juctase iso-en-zymes, Journai of Biological Chemistry267, 19548 19554.

O Connor. B,C,. Newman, H,N, & Wilson,M, (1990) Susceptibility and resistance ofplaque bacteria lo minocyctine. Journal ofPeriodonlology 61. 228 233,

Ojanotko-Harri, A., Hurrtia. H. 8L Harri. M,P (1986) Role of granulation tissue andfibroblasts in gingival testosterone metab-olism in the rat in vitro. Journai of SteroidBiochemistry 15. 689 693,

Rappaport. S. C . Vittek. J., Altman. K..Gordon. G. G-. Southren. A. L. (1976) Sexdifferences in Ihe metabolism of androgensby human gingiva. Journal of Dentai Re-search 55, 1371.

Rompen, E, H.. Kohl, J.. Nusgens. B, & Lap-iere. C, M, (1993) Kinetic aspects of gingi-val and periodontal ligament fibroblastattachment to surface-conditioned den-tine. Journal of Dental Research 72. 607612,

Ryan. M. E,. Ramamurthy, S., Golub, L, M,(1996) Matrix metalloproteinases andtheir inhibition in periodontal treatment.Current Opinion in Periodonloiogy 3. 85-96-

Sasaki. T. Kaneko. H,. Ramamurthy N, S, &Golub. L, M, (1991) Teiracyclinc adminis-tration restores osteoblast structure andfunction during experimental diabetes.Anatomical Records 231, 25 34,

Schneir. M,, Ramamurihy. N. S, & Golub. L(1990), Minocycline treatment of diabeticrats normalizes skin collagen productionand mass. Matrix:: Coiiagen and ReiaicdResearch 10. 112-123.

Seymour, R. A. & Heasman. P A. (1995) Te-tracyclines in Ihe management of peri-odontal diseases. A review. Journai of Clin-ical Periodontology 22, 22 35,

Sooriyamoorthy. M, & Gower, D, B, (1989).Phenytoin stimulation of 5«-reductase ac-tivity in inflamed gingival fibroblasts.Medicai Science Research 17, 989-990,

Sooriyamoorlhy. M,. Gower. D. B. & Eley, B,M, (1990), Androgen metabolism in gingi-val hyperplasia induced by nifedipine andcyclosporin. Journal or Periodonial Re-search 25, 25-30.

Sooriyamoorthy. M.. Harvey. W, & Gower.D B, (1988), The use of human gingivalfibroblasts in culture for studying the ef-fects of phenytoin on testosterone metab-olism. Archives of Orai Bioiogy 33. 353-359.

Soory. M (1995), Bacterial steroidogenesis byperiodontal pathogens and the effect ofbacierial enzymes on steroid conversionsby human gingival fibroblasts in ciiiture.Journal oj Periodonlal Research 30, 124-131,

Soory. M. & Gower. D,B, (1990), Investiga-tion of androgen metabolism by hyper-plastic tissue fibroblasts in relation tophenytoin. Medical Science Re.search 18,835-837.

Southren. A, L,. Rappaport. S. C. tS: Vittek..1, (I97H), Specific 5H-dihydrotestosteronereceptors in human gingiva. Journal ofClinical Endocrinoiogy and Mciaholism 47.1378-1382.

Van Steenberghe. D,. Bercy. P. Kohl. J.. DeBoever. J.. Adriaens. P. Vanderfaeiliie. A.el al,. (1993), Sub-gingival minocycline hy-drochloride ointment in moderate to se-vere chronic adult periodontitis; A ran-domised double-blind vehicle controlledmulticenter study. Journai of Perioikmio-logy 64. 637 644.

Address:

Mena SooryDepartment of PeriodontologyKing's College Dental SchoolCaldecot RoadEondort SE5 9 RWUK