The distribution of 5α-reductase and 3α(ß)-hydroxysteroid dehydrogenase activities in the hyperplastic human prostate gland.

41

3049

THE DISTHIBUTION OF ~a3EDlJCTASE AND ja(S)HYDHOXYSTEHOID DmOGENASE -TIC

ACTIVITIES IN THE

HUMAN PROSTATE GLAND.

F.K.,Habib, L. Beynon, G.D. Chisholm and A. Busuttil* Department of Surgery, University of Edinburgh Medical School, Teviot Place, Edinburgh, EH8 9AG, Scotland. and *Department of Pathology, Western General Hospital, Edinburgh, E& 2XlJ,Scotland. Received n-1-82 ABSTRACT This study examines the distribution of sa-reductase and ja(S)-hydroxysteroid dehydrogenase activities throughout the intact hyperplastic prostate gland and relate these measurements to the fibromuscular/epithelial composition and to the gross glandular The relative capacities of the stroma and epithelium morphology. to metabolize testosterone and dihydrotestosterone were also examined. The results indicate that under optimum reaction conditions an uneven distribution of sa-reductase and ja(p)-hydroxysteroid dehydrogenase could be measured across the prostate. These regional variations reflect true differences in metabolic activity and were independent of any morphological changes: caution is therefore advised when interpreting hormonal metabolic data obtained from singLe sampling of the gland. Our investigations also suggest that the capacity to metabolize testosterone was emnly distributed between stroma and epithelium and that both tissue components are primary sites for !&-reductase activity. The reductive ja(S)-hydroxysteroid dehyckcogenase was also found in both tissue types but the mean stromal activity was marginally higher than the levels measured in the epithelium. INTHODUCTION The aetiology of nodular prostatic hyperplasia is uncertain but the condition is generally believed to be due to hormonal imbalance (I), and the periurethral "inner gland group" is usually accepted as the site of origin (2). The measurement of enzyme activities in the prostate tissue may therefore be of value in the understanding of pathophysiological changes which occur in the gland and may also give important information on hormonal metabolism in

Volume

41, Number

1

S

TDEOXDb

January

1983

this tisaue. Recent studies suggest that the bulk of the androgen metabolizingenzpes are associatedwith the strad

tisaae

(3-S)

but there is also evidence fox their presencewithin the epithelium (a). !Pheprostate gland consists of two main comp3rtments,the peripheralcoqwrtment with a hi& gl3.ndu3,ar/stromal ratio apd the periurethrdlarea in which connectivetissue is more abundant (2). The proportions dtffedng

ofstroma 3ndepitheliumvarynotonlytithin

area8 of each compartmentbut also from patient to patient.

Thepopularityoft zwsurethr3J.resectioncauses problems to the biochemistin so faJ:88 it is not possible with even a large number of tz%nsurethr3Jchippingsto tell accuratelyfrom where they have been taken.

Consequently3ny

et&y

on

the enqnmtio properties

of these chippin& would be of little si&nificanceunless a better understandfngof hormonal metabolismthroughoutthe gland can be aohieved. When one rea;lises that the bulk of the earlier investigationswere confined to one 3ection per prostate the importance of a serial study become3 even more significant. The present study hopes to establishthe enzymaticpatterns in whole prostatesand relatothesemeasurement3to the stromal and epithelialdistribution and to the gross &xudula;rmorphology.

List of trivial nones and the abbreviationsused Testosterone= L7@-hydroxy-~-3ndro3ten-34x1e Sa-BELT= ~a-dJ&drote3so3terone=l7~-hgdrolg-~a-~st~-3-one 3a-3ndrost3nediol = 5a-androstane-ja, 17+diol 38-3udro3t3nedi01 = 5a-androstaw-3P, 17&d.iol Audroatenedione= &-azxkostene-3,?7-dione

S

1ZIIROTT)SB

43

S-Pecimens

Frostate glands were obtained from 8 patients (60-72 years) undergoingroutine retropnbicprostktectomysztdwere immediately t-ported to the laboratoryin ice cold physiologicalsaline solution (pEI7.4).

3 2 RIGHT LOBE

I

I

2 3 1 LEFTLOBE

1. Schematic representationof the procedureused for obtainingprostate sections from the left and ri@& lobes. Each lobe was split longitudin&lg along the uxethral axis into three or four even sections and'thesewere numbered in ascendingorder sectionswere then fxomthe urethra outwards. The longitu~ sliced transverselyinto sevexsJ.2 g samples snd were positionedin relation to their distance from the bladder.

Ffa.

On arrival the left and right lobes were separatedand these were in tuxn split in a plane pszsI.leK. to the urethral arfs into four even

S

TDROIDb

longitudinalsections. Each of these sectionawas subsequently sliced transverselyinto several 0.5 - 2g samples. The dissected sections were numbered accordingto their position in the prostate and in relation to the bladder base. The schematicdiagram (Fig. 1) illustratesthe procedurefor labellingthe various samples whereby the 1ongituM sectionswere numbered in ascendingorder from the urethra outwardswhilst the subsectionswere positionedin relation to their distance from the bladder. hromeachsmall section a segment was fixed in buffered formalinfor conventional histologictiexaminationto confirm the diagnosisand to assess the stromal/epithelial ratio; aeseemnentwas made visually at 80X magnificationon haematorylinand eosin stained sections. !L!he rem&nder was used for biochemicalanalysisbeing either used fresh or dipped in liquid nitrogen and stored at -3O'C until analysis. RadioactiveSteroids and Chemicals The radioactivesteroids]taed: [1,2,6,7-%I-testoeterone sp.act.83Ci mmol), [4-C ]-testosterone(sp.act.9 mCi/mmol>, 1,2 4 5,6,7 JH ] -Sa-d.thydrotestosterone (sp.act.122 Ci mm01 ee OS crone (sp.act. 51 mCi/mmol and 04-C I4_c4j-5a-dihydrot t t 1 androstenedione(sp.act.53mCi/mmol) were obtained from the ~adiochemicalCentre, Ameraham. [l&l ja-endroatanediol end [14cl jS-en&ostanediol were prepared b the sodium borohydride method aa previouslydescribed [y9j. The radioactivesteroidswere purifiedby paper chromatographybefore use. lion-radioactive steroidawere purchasedfrom SteraloidaLtd., Surrey, and all other chemicalsof AnaJar reagent gradevgrepurchasedfrom Sigma ChemicalCo. (St. Louis, Mssouri, U.S.A.), BLE ChemicalsLtd. Poole, Dorset, U.K. and Fisons ScientificApparatas (Lou&borough, U.K.). Separationof stroma and enithelium: The separationof prostate samples into stromal and gpithelialcomponentswae similar to that previouslydescribed (9). Briefly, minced tissue was homogenisedin three volumes of 10 mmol 1.5 mMEDPA, 1.0 mM dithiothreitol1 g/l Tris buffer pH 7.4, contW methyl-celluloseand the homogenatewas forced through a nylon gauze of 1.50um pore size. The obtained filtratewas centrifuged end the resultantpellet containingthe epithelialfractionwan resuspendedin buffer. The fwents retainedby the gauze consistedpredominantlyof stromal tissue and these were resuspended Fn the Tris buffer, homogenieedand filtered once more throu@ the nylon gauze. The purity of stroma and epithelialfractions, obtainedprior to the final homogenisationin the hris buffer, wae examined by li&t microscopyafter embeddingthese tissues in paraffin,sectioningand stainingusing standardhistological techniques. !Theefficiencyof the separationwas cross-checkedby the measurementof hydrorypmline and acid phosphataseconcentration in each of the components; a double antibody radioimmuneassay was used for monitoringthe human prostaticacid phosphatasewhilst the hydroxyprolinelevels were establishedby the method previously describedfor acticularcartUge tissue (10).

S The

TEIROXDS

Measurementof Enavme Activities

90 ul &Liquots of tissue componentpreparationsand homogenates of whole tissues were transferredto pre-heated incubationtubes containing50 I# of the radiolabelledandroSen and 5 x 10:z:M of the EARPH generatingsystem which usually consistedof 5 x 10 M RAW, 5 mM glxose-&phosphate and 0.5 units glucose-6-phosphate deQ&og,enase. !Pheincubationmixtures were identicalto those describedin ea&ier studiea for the measurementof enzyme activities wereincubatedfor under optimsl cogUtions (Y&2). Thentidures 30 mimites at 37 C in a sldcbg bath, and the reactionwas tezmdnated by the addition of 10 volumes of acetone contain& 12.5 u&l of -1abelledhormone steroidsas oarriersand tracer amounts of the C fox recoverypurposes. The ether extracted steroidswere separated by thin layer chromatographyon a silica gel (ITLC plates, Gelman Instrcrment Company,Ann Arbor, Michigan, rrSa>after thrae vol/vol);although this developmentsin benaene: ethanol (96:4; aUowed the oomplete separationof the 3a and .3@ andmstanediols from each other and from the other metabolitesproduced in our incubation mixture namely the testosterone,dihydrotestosterone and androstenedionewe have, in the present study, combinedthe 3a and 3 epimera and expressedour results as the products of the total (% androstanediolsformed. Following chromatographythe steroids were locslfsedby spray&&~with 1% vol/vol H SO in ethanol and heating at 100°C for 5 minutes, The zones oo&&ponding to the referencesteroidswere marked, zut with scissorsand directly immersed in 'counting vials containingthe Witon X-100 scintillation cocktail. Estimationof the $-reductase autivity was based on the formationof +&hy&rotestoaterone and ja(p)-androstanediol from testosterone. The 30;(#3)-hydromsteroid dehydrogenasewas, on the other hand, calculatedfrom the emonnt of 3&3)-androstanediol formed from dihydrotestosterone,All incubationswere performed in triplicateand the results of multiple determinationwere expressed as pm01 of metaboliteaformed per mg protein per 30 minute incubation. Prtdein concentrationswere measured by the Lowry et.al. method (11) and the countingefficiencyin our Packard Tri-Carb ip_uxd scintillation spe&Umeter system was 3946 for(3a)and$596 for C when both isotopes were counted simultaneously. Further details on the calculationof the e2zfmaticactivities,the separationand extractionproceduresand the quantitativeassessmentsof the m&abolites_ foxmed are to be found in our earlier studies. (9,12) StatisticalAna.lvsia In the present study the significancebetween two sample means was assessedby the Studentls 't' test method. BESULTS !I!he efficiencvof the stroma/euitheliumseuaxation Attemptswere made to check the efficiencyof the stroma/ epitheliumseparationprocedure. The concentrationof acid

S

TPROXD-

phosphataseand hydroryprclinewere measured in the separatedccmponents of five different samples. The levels of acid phosphatasewere 14.38 p&g

protein for the epitheliumand 0.096 pg/mg protein for the

stromal components; this suggestsan average of 0.6696 contsminationof the strcma by the glandulartissue. Conversely 5.996of the stromal hydrorgprolineconcentration(15.36 p&g

dry

weight tissue) were detected in the epithelium(0.93 pg/mg dry weight). The efficiencyof the separationwas further checkedby histological examination; these observationssupportedthe biochemicalanalysis and confinn cur earlier assessmentregardingthe efficiencyof the separationprocedure (9). Enzsme characterisationstudies The

followingstudies had been performed in order to optimize

the react-inn conditionsfor both the $-reductaee and 3a(P)hydrox~steroiddehydrcgenaseenzymes:(a) The effect of PH. The apparent activitiesfor each enzyme were tested over a pH range of 4.0 - 9.0. The optimumpH ranges for Sa-reductaseand 3a(P)-hydroxysteroid dehydrogenasewere similar, lying between pH 6.8 and 8.0. (b) Enzyme kinetics. The basal activitiesof Sa-reductaseand 3a(B)-hydrowsteroiddehydrogenasewere significantlyaugmented by the addition of co-factors; tb.iswas clearly shown in the earlier report by the same authors (9) in which it was demonstrated that theXABPHgenerating system was the most effectiveco-factor for both enzymes. Saturationof the $-reductase and 3a(p)hydroxysteroiddehydrogenasewas achieved at 5 x lo-’

M of

the

NADPlI

S

47

TZIEOXDLI

generatingsystem. We have also investigatedthe effect of substrate concentrations(0.005 - 2 UM) on each of the enzyme act3.vi.ties. Our data suggests that both enzymes obey the Michaelis - Menton kinetics and the apparentMichaelis constants(Km) extrapolatedfmm the regressionlines were 0.12 and 0.09 uM for the Sa-reductaseand ~(~)-~~~~e~id

dehydrogenaserespectively.

(c) Effect of time and protein concentration. The effect of time of incubationand protein concentrationon the !3x-rsductase and 3a~~)-~~ste~id

dehydrogenaseactivitieswere also investigated.

Figure 2 suggeststhat under our assay conditionsthe enzyme stability

lNC@,,,TlON

Fi

TIME hn)

INCUSATION TIME hnl

2. The effect of protein concentrationand time on the

-reductase(A) and 3a(P)-hydroxystemiddehydrogenase(B) 3+” in homogenatesof hyperplasticprostate tissue. The reaction 0.15;(n) mg/ml tissue mixtures contained .95 (4) 0.65 (0 ) NADP, $R glucose-6protein, 50 IN of (i3 ) a.ndrogen, 5 x 10-fkphosphate a.nd0.5 units glucose-6-phosphate dehydrogenasein a final volume of 500 ul Tris buffer, values representmeans of three determkations.

and the rate of metaboliteformationremain constantthroughout the 40 ereminedminutes. It was observed that in both sets of enzyme studies the formationof metabolitesincreasedlinearlywith protein concentrationup to 1.95 m&al.

All subsequentexperiments

reportedin this study were thereforeperformedunder identical conditionsof linearitywith regard to time and pmtein concentration. The distributionof ensvme activitiesin the EPE cfland and its relationshipto the stnxnal and enithelialcowosition Table

1

summarisesthe patterns of ensyme activitiesas measured

in one entire gland- Sectionswere also analysed for their glandular and fibromuscularcomposition. Clearly the mean $x-reductase activity for the right lobe (66.2 +6.5 is of the same order of magnitudeas that

pmol/mg protein/3Omin) measured in the left

lobe (56.2 = 15.5 pmol/mg protein/3Cmin) (P70.1).

Nonetheless

the data outlined in Table 1 suggest that there is also considerable regionalvariationbetween the segments of each lobe whereby some sectionsmanifest a higher activity than others. Moreover the measured activity appears to be independentof the local stromal/ epithelislcomposition. The results for the 3a(P)-hydrowstemid dehydrogenaseindicatethat there was also no signi.ficant difference between the mean enzyme activitiesfox the right (23.5 k4.3 pmol/mg protein/jOmin.)and left (22.2 + 6.4 pmol/mg protein/3Omin.)lobes (P)O.2)

but once more the activityvaried on a regionalbasis and

this was totally independentof the tissue compositionwithin the segment. Table 1 also suggeststhat the mean Sa-reductase/3a(p)-

S

49

TDSSOID6

Table 1. Distribution of $-reductsse (A) and 3a(S)-hydroxysteroid dehydrogenase (B) activities throughout the left (L) and right (R) Tissue homogenates (I:3 in lobes of a human hyperplastic gland. buffer) prepared from the sliced prostate sections were incubat d with 50 nM of the radio-labelled sndrogen in the presence of 5 x 10-' M The metabolic proNADPH generating system for 30 minutes at 37 C. ducts were separated by thin layer chromatography and the enzyme activities were estimated as detailed in the Material end Methods section. All samples were analysed in triplicate end values are expressed as pm01 metabolites formed/mg protein/30 min. incubation + S.E.M. The epithelial/stromal composition of each segment analyzed was assessed histologically.

Segment Number R R R R R

l-2 2-l 2-3

Mean+ S.E.M. l-l l-3 2-4

B

A

A/B

;:

;05

65.5 i 2.5

42.0 + 1.4

1.6

::

66:

86.0 61.3 16.4 + 1.1

36.0 16.8 + 0.4 1.2

3.6 2.4

$

g

57.0 92.5 57.8 f f. 2 3.4 7.3 5.2

23.0 + 17.5 18.8 + 0.5' 1.2 1.0 10.3 + 0.3

t:; 3.3

l-4

3-2 R 3-4 R 4-l R 4-3

L L L L L

s Volume of Stroma Epithelium

40

45.6 + 2.9

60

54.3 + 2.9

3 . 5 + 2.1 66.2 * + 6.5

23.5 * L 4.3

!&2 3.2 + 0.4

40 G

46.3 + 0.8 130.8 29.S + 0.9 5.3 -+

13.8 + 0.6 35.5 19.3 + 0.6 1.0

3.4 3.7

g 45

40.0 43.8 88.5 + 3.6 3.8 1.2 IL.5 + 0.3

53.5 11.8 17.8 + 0.7 0.4 2.6 3.6 + 0.1

;:; 2:s Ir.0

2-5

3-l L 3-4 L 3-6

Mean+ S.E.M.

44.2

+ 3.0

56.2 * ;fs15.1

22.2 * +. 6.4

2.9 2 0.4

* No signicant difference (PpO.1) H

No significant difference (P> 0.2)

dehydrogenase ratio was similar in both lobes (approx. = 3) suggesting that the capacity to reduce testosterone throughout the hypertropbied gland was on average three times as great as that manifested by the tissue for the metabolism of Sa-dihydrotestosterone.

S

TBEOXDS

Studies performedon a further4 @+I& confirmedin each case the above results. Relativemetabolismof testosteroneand dihvdrotestosterone in the stroma and enitheliumacross the nrostatenland Table 2 outlinesthe patternsof enzyme activitiesas measured in the stroma and epitheliumextractedfrom differentsections across one whole prostategland.

The resultshave been presented

as ratios of enzyme activitiesin order to highlightthe possible differences. The data indicatesthat the mean ratio of the Sareductaseto Jo(P)-hydroxysteroid dehydrogenasein the epithelium (4.6 + 1.7)

was statistically not differentfrom that observedin the

stroma (2.23f 0.16) (P) 0.2). The stroma and epithelialSareductasewere in most prostatesectionsat least twice as high as the corresponding3a(~)-~~o~ste~id

dehydrogenase. The results

also suggest that the stromal and epithelialcapacitiesto metabolize testosteronewere similaras is witnessedby the tissue Sa-reductase ratios (mean ratio 1.00 + 0.26).

The ?a(@)-hydrowsteroiddehydrogen-

ase activityin the two tissue types was also comparable(mean ratio = 1.29 ;tr 0.30).

Attemptsto establisha correlationbetween the

individualmetabolicvalues and the morphologyof the gland were not successful. Studies on eight other prostateglands yieldedsimilar results.

S

51

=x?m&OID=

Table 2. The relative activities of 3-reductase (A) and 3&)hydroqsteroid dehydrogenase (B) in tissue components taken from different psrts of the gland. Stroma (S) end epithelial (E) components extracted from segments obtained from the Left (L) snd right (R) lobes of the prostate were homogenised, incubate with 50 nM of radiolabelled androgen in the presence of 5 x 10 44 NADFB generating system and analyzed for enzyme activities as described in the Material end Methods section. The values shown below represent ratios of total enzyme activities expressed as pmol metabolites formed per mg protein in each fraction per 30 min. incubation. Each ratio represents the mean of triplicate assays. Segment Number

A/B in E

R 1-2 R 1-4 R 2-l

R 3-2 R 3-4 R 4-l L L L L

l-l

l-3 2-4

2.64 2.6 2.3 ;:t

S

B

A

2.4

1.9 2.4 2.5 2.2

1.2 1.2 1.4 0.33 0.18 1.52

1.24 0.55 0.19 1.70

2.4 1.9

2.6 3.1

3.00 O.W

2.70 0.25

0.17 0.16 1.47

2.30 0.83 1.88

1.00 ~0.26

1.29 LO.30

2.0

-

i::

L 3-4 L 3-6

22.6 7.5 2.9

Mean + S.E.M.

.6* 2.23+ Ll.7 ~0.16

2-5

S/E for

2.2

*No significant difference between the means (B)

0.2)

DISCUSSION In all previous biochemical studies (3-9, 12-13) the sampling cf the prostate was confined to one sliver of tissue per gland. In view however of the heterogeneous nature of the prostate end the variation in composition of tissue removed at surgery, the limitations of such sampling techniques are evident.In an attempt to overcome

S

52

'EDROXDI

these obstaclesand resolvethe dilemmaof the non-reproducibility of the enzyme studies,whole prostateswere serially sectionedthus allowingfor biochemicalinvestigationsand histologicalanalysis to be performedon every single sectionobtainedacross the prostate gland. In spite of the histologicaland biochemicalevidencethat the cross-overfrom stromal elementsinto the epitheliumis exceedingly negligible,the presentstudy confirmsthe presenceof the So-reductase in the glandulartissue. The mean glandular$-reductaae activitywas found to be of the same order of magnitudeas that measuredin the stroma suggestingthat both tissue componentsare primarysites for the formationof !$a-reduced metabolites. Our observationsindicate however that this activitywas not evenly distributedthroughoutthe entire prostategland.

Althoughthere were sectionsadjacentto the

predominantlystromalperiurethralregion8manifestinga capacityto metabolisetestosteronewhich was identicalto that in the glandular areas of the peripheralregions,the data in tables 1 and 2 reveal a wide range of enzymaticactivities. In view of the reliabilityand high reproducibility of the assay system (9, 12) one can merely attributethese differencesto regionalvariationsin enzyme patterns. This clearlyhighlightsthe need to carry out enzyme studieson multiple tissue samplesand underlinesthe limitationsof the single samplingper gland technique. The data on the 3a(p)-hydroxysteroid dehydrogenasefollowed the same generalpatternsassociatedwith the !?a-reductase namely

S

TIIEOID=

53

an uneven distribution of the ensyme activities throughout the prostate stroma end epithelial components; favoured a higher stromal activity.

the regional variations

The physiological significance

of these localized and preferential capacities to metabolise dihydrotestosterone is far from understood but these could account for the accumulation of Sa-DHT measured in the epithelium

(3). Clearly at

this stage one cannot eliminate the possible involvement of the 3a (S)-hydroxysteroid dehydrogenase in the aetiology of prostatic hyperplasia. ACKNOWLEJZFMENTS This project was generously supported by a grant from the Scottish Hospital adowment Research Trust.

1. ;:

4. 5. 6. 7. 8. 9. 10.

Isaacs, J.T. and Coffey, D.S., Endocrinology l& &5 (1981) McNeal, J.E., J. Ural. 102, 1008 (1972) Sirett, D.A.N., Cowan, S.K., Janeczko, A.E., Grant, J.K. and Glen, E.S., J. Steroid Biochem., 3, 723 (1980) Wilkin, R.P., Bruchovslcy,N., Shnitka, T.K., Rennie, P.S. and Comeau, T.L., Acta edocrinol., 24, 784 (1980) Acta Krieg, M., Klotzl, G., Kaufman, J. and Voigt, K.D., &docrinol., 96, 422 (1981) Harper, M.E., Pike, A., Peeling, W.B. and Griffiths, K., J. adocr., 60, 117 (1974) Jacobi, G.H. and Wilson, J.D., J. Clin. adocr. Metab., &, 107 (1977) Habib, F.K., Tesdale, Ann L., Chisholm, G.D. and Busuttil, A., The Prostate, 1, 117 (1980) Habib, F.K., Tesdale, Ann L., Chisholm, G.D. and Busuttil, A., J. Endocr., 91, 23 (1981) Videman, T., Eronen, I. and Candolin, T., Biochem. J., 200.

435 (1981) 11.

Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Rsndall, R.J., J. Biol. Chem., 12, 265 (1951) 12. Habib, F.K., Rafati, G., Robinson, M.R.G. and Stitch, S.R. J. Endocr., & 369 (1979) 13. Morfin, R.F., Charles, J.F. and Floch, H.H., J. Steroid Biochem., 11, 599 (1979)