_____________________________________I~---"J~~

Immortalisation of Ovarian Granulosaand Theca Cells of theMarmoset Monkey Callithrix jacchusBettina Husen, Kai Lieder, Alexandra Marten, Angelika Jurdzinski, Kerstin Fuhrmann,Harald Petry, Wolfgang Luke and Almuth EinspanierDeutsches Primatenzentrum, D-Gottingen

SummaryIn view of the increasing need for laboratory primates inbiomedical research it is desirable to develop appropriateprimate-specific cell culture models that could prevent orsignificantly reduce the increasing use of primary cultures andexperiments with living animals. Follicular granulosa andtheca cells are essential for the control of hormone-dependentprocesses such as the ovarian cycle and pregnancy, butalso for the occurrence of hormone-dependent diseases.For this reason it is of great interest to know more about controlmechanisms existing in these follicular cell types and the effectof pharmacological or toxicological agents on them. Animmortalisation protocol for the two ovarian cell types of themarmoset monkey (Callithrix jacchus) has been developed.All cell lines established so far were examined with regard tothe maintenance of known, tissue-specific features (e.g.hormone responsiveness and enzyme expression). The resultsobtained indicate that it is worth while cloning and character-ising the cell lines in more detail so that they could be usedafter an adequate validation as a defined test system both forbasic research as well as for pharmacological or toxicologicalscreening.

Zusammenfassung: Immortalisierung von ovariellen Granu-losa- und Thekazellen des WeiBbiischelaffenUm einem steigenden Bedarf an Primaten als Versuchstiere ent-gegenzuwirken; ist es wiinschenswert, geeignete primaten-spezifische Zellkultursysteme zu entwickeln, die die zunehmendzu erwartenden Primarkultursysteme und Experimente amlebenden Tier verhindern oder erheblicb redurieren konnen.Granulosa- und Thekazellen sind essentiell fur die Steuerunghormonabhdngiger Prozesse wie z, B. Menstruationszyklus undSchwangerschaft, aber auch fur die Entstehung hormon-abhdngiger Erkrankungen. Darum besteht ein grofJes Interessedaran, mehr uber Kontrollmechanismen in diesen Zellen undiiber pharmakologische und toxikologische Effekte auf dieseZellen zu wissen. Es wurde ein Protokoll fur die Immorta-lisierung der beiden Zelltypen aus den Eierstocken vonWeifJbuschelaffen (Callithrix jacchus) entwickelt. Alle bisherentstandenen Zelllinien wurden auf Erhaltung bekanntergewebstypischer Eigenschaften (z.B. Hormonempfindlichkeitund Enzymausstattung) untersucht. Die bisher gewonnenenErgebnisse deuten darauf hin, dafJ es sehr lohnenswert ist,einige der Zelllinien zu klonieren und detaillierter ru charak-terisieren, so dass sie nach entsprechender Quaiitdtskontrollesowohl in der Grundlagenforschung als auch fur das pharma-kologische und toxikologische Wirkstoffscreening als definiertesTestsystem eingesetzt werden konnen.

Keywords: 3R, New World Monkey, ovary, immortalisation, in vitro system

1 Introduction

During the last decade there has been acontinuous increase in the total numberof primates used as laboratory animals.This is stated for example by the Annualreport on the use of laboratory animals,edited by the German government(Tierschutzbericht 2001 des deutschenVerbraucherministeriums, http://www.ver-brau cherminis teri um. de/tiers ch utz).Data about primates were selectedfrom this report and compiled in Figure 1which indicates that the increaseis caused mainly by rising numbers ofprimates needed for drug testing.

64

A further increase might also be ex-pected with continuous advances in newfields such as functional genomics.Drug screening is a field where substan-tial progress has been made in thedevelopment of suitable cell culturealternatives over the past years (Marx,1996; Spielmann, 1996). Unfortunatelythis statement does not pertain tosystems using cells from nonhumanprimates. From these animals onlyprimary cultures are in use whichprovide limited cell numbers and thestandardisation of which is hard toachieve. It is now time to prevent furtherexperiments with primary cultures by

the development of defined, differenti-ated permanent cell lines.For biomedical studies in the field of

reproductive biology the use of primatemodels is indispensable, because in thisrespect many physiological differencesare evident in primates compared toother mammalian species. Moreover,human tissue from healthy, pharma-cologically unimpaired ovaries is notreadily available in most cases. Animalmodels like the marmoset monkey (Fig.2) can provide tissue of defined physio-logical origin.This small New World monkey is a

widely used model species especially in

ALTEX 19, Supp!. 1102

____ ~~------------------------------------------------H-U-S-EN--ET--AL-.

~,

2~ ,-------------------------------,

UI2000~

E'2 1500CO

'0... 1000<1J.cE soo~t:

0

"'..... .~ •... ..••./' •......_rI•.. ....-...•. •...•... /

r•............. _ --+- _ _ _k- - -k- - -.6.

.6.- - -t: - -k- - -i'" "'"- •..- -+-- - -+

year

1991 1993 1995 1996 1997 1998 1999

_t'Jwlnumbe.r-_ druQt::5ting-+- basic research

-k- di<Jgno~tk:~: prophvta'ds. therapy

-0- h.;rti(:d~s-0- envrcnmentari~,s-<;::- other substances

Fig. 2: The marmoset monkey (Callithrix jacchus) is apopular primate model in reproductive biology research(Photograph: Courtesy of Alexander Schneiders).

Fig. 1: Number of primates used for animal experiments inGermany since 1991. Data are derived from the Annual reporton the use of laboratory animals 2001 (Tierschutzbericht 2001des deutschen Verbraucherministeriums).

been tried beforeand might be ofinterest for theestablishment ofcell lines from

reproductive biology research (Fraserand Lunn, 1999). It is characterised by ahigh fecundity and a short generation in-terval. Ovarian anatomy and physiology,including hormonal profiles during theovarian cycle show a strong similarityto the human (Hearn, 1983; Hillier et al.,1988). Furthermore, the marmoset mon-key is the primate model of choicein many toxicological studies (Smith etal., 2001).The development of permanent cell

lines as a defined primate-specific invitro system for research on reproductivebiology could serve several purposes. Inthe first place, it would facilitate studieson primate (including human) ovarianfunction, such as hormone synthesisor oocyte maturation, as well as studieson hormone dependent diseases of theprimate ovary. In the fields of pharma-cology or reproductive toxicology it wouldbe a more predictable test system thanothers using rodent tissues, either in vivoor in vitro. Furthermore, the evaluationof different immortalisation protocolswould help to find general strategies forthe immortalisation of cells derived fromNew World Monkeys, which has not

ALTEX 19, Suppl. 1/02

other tissues.Granulosa cells (GC) and theca cells

(TC) form the inner and outer layer ofthe ovarian follicle wall (Fig. 3). Inthe marmoset monkey during thecourse of every ovarian cycle up to threefollicles are selected to reach the maturestage of a preovulatory follicle. Only thesepreovulatory follicles are capable of hor-mone synthesis in response to a go-nadotrophic hormone stimulus from thepituitary. After the hormonal induction ofovulation GC and TC of the preovulatoryfollicles undergo a terminal differentiationto form the corpus luteum. This process isaccompanied by endocrine and morpho-logical changes and is called luteinisation.It is a prerequisite for the implantation ofthe conceptus and maintenance of earlypregnancy. GC and TC from preovulatoryfollicles have to be used to study thechanges accompanying luteinisation.In order to establish a permanent cell

line, primary cells have to be immor-talised, i.e. their limited life span asregular somatic cells has to be extended(reviewed by Hopfer et al., 1996).Immortalisation can take place sponta-neously, by certain mutations whichoccur often in rodent cells, but only

rarely in human and presumably onlyrarely in other primate cells. Immortali-sation can also be achieved experimen-tally by transfection (transfer of a foreignDNA) or by viral infection with certainoncogenes. Oncogenes code for proteinsthat deregulate the control of cellproliferation. Some authors reportedrecently on successful immortalisationby transfection with the catalytic subunitof the enzyme telomerase as an additionalmethod (Bodnar et al., 1998; Vaziri andBenchimol, 1998; Morales et al., J 999).In the project presented here the classicaloncogene approach was chosen, i.e.immortalisation by transfection of thewell-characterised large T antigen DNAof simian virus 40 (SV40LT). Thecorresponding protein is known to betargeted to the cell nucleus. There it isthought to bind and inactivate certainantioncogenes which code for proteinsthat ensure a correct cell cycle controland proliferation (Fig. 4).After the selection of an appropriate

immortalising gene, a suitable techniquehas to be found to transfer this DNA intothe cells. Classical methods utilisingagents such as calcium phosphate or thepolycation DEAE suffer from the disad-vantage of being rather cytotoxic. Mean-while several reagents are commerciallyavailable that work more gently andallow for transfection even in the presenceof serum in the cell culture medium.

65

_H_U_SE_N_E_T_A_L_. i~----"J~~

preovulatoryfollicles

A

Fig. 3: Microscopic anatomy of a marmoset monkey ovary shortly before ovulation (day7 of the follicular phase of the ovarian cycle). Two antral, preovulatory follicles arevisible. Inset: higher magnification view of the frame marked on one of the preovulatoryfollicles. A = follicular antrum, GC = granulosa cell layer, TC = theca cell layer,00 = Oocyte, BM = basement membrane. Hematoxylin-eosin staining, 20x magnification.

A common consequence of the immor-talisation of primary cells is that theydedifferentiate and undergo oncogenictransformation, i.e. they lose their formercell type-specific features. For this reasonimmortalised cells have to be charac-terised intensively to reveal conformityor emerging differences compared to thecorresponding primary cells. Characteri-sation studies in the present report havefocussed on cell-type specific steroidhormone receptor expression and thein vitro induction of enhanced prog-esterone secretion, a characteristicendocrine consequence of luteinisationin vivo.

2 Animals, materials and methods

2.1 AnimalsAdult, 3-6 year old marmoset monkeyswere maintained in family groups at theGerman Primate Centre, Gottingen,under standardised conditions(Einspanier et al., 1994). All animals hadbeen trained to be handled for routineblood collection. Blood samples wereretrieved twice a week over a period of

66

three months for assessment of serumprogesterone levels by a specific enzymeimmunoassay (Heistermann et al., 1993).This was done to check for regular ovar-ian cycles and to determine the expectedtime of ovulation. Three animals wereovariectomised and provided large antral,preovulatory follicles (> 2mm in diame-ter) for the preparation of primary cul-tures. Ovariectomy has to be performedregularly for reasons of colony manage-ment and was not carried out specificallyfor the retrieval of GC and TC.

2.2 Vector constructsThree different plasmid expressionvector constructs were used for trans-fection. Their compositions are schema-tically depicted in Figure 5.The vector pEGFP-Cl (Clontech,

Heidelberg, Germany) was used for thedetermination of transfection efficiency.pBluescript-SV40 (Courtesy of Dr. LisaWiesmiiller, Heinrich-Pette-Institut furExperimentelle Virologie und Immuno-logie, Hamburg, Germany, Wiesmiiller etal., 1996) and pMSSVLT (Schuermann et

rb107binding ;a~~I:I~~9

\ / signal

p53/binding

" /DNA binding

Fig. 4: Primary proteinstructure of large Tantigen of SV40LT andparts of the proteinwhich are thoughtto be involved in theimmortalizationprocess (modifiedfrom Modrow and

708 Falke 1998). Rb107and p53 are cellularanti oncogenes andtheir proper functionis disturbed byinteraction withSV40LT.

~~~:\ ~'y:~'-"O<: \0

pEGFP-C1(Clontech)

pBluescripl-SV40(\I\I\"esmilller et al. 1996)

pMSSVLT(Schuermann et al. 1990)

Fig. 5: Vectors usedfor transfection.pEGFP-C1 codes forenhanced greenfluorescent protein.pBluescript-SV40contains the completegenome of SV40-strain number 776including the largeT region, whilepMSSVLT containsonly the T-region ofthe SV40 genome.

ALTEX 19. Supp!. 1/02

;;~~ HUSEN ET AL.---~~~--------------------------------------------

~~

al., 1990) include either the completegenome of SV40 or only its large Tregion, respectively, and were used forimmortalisation.

2.3 Cell culturePrimary cultures were prepared ac-cording to an established protocol(Einspanier et al., 1995, 1997). Briefly,follicles were punctured and GC areaspirated from the antrum. TC were re-covered from the remaining part of thefollicle wall and were dissociatedmechanically and enzymatically. Thecells were washed in culture medium(M199, Invitrogen Life Technologies,Karlsruhe, Germany) supplemented with100 IU/ml Penicillin, 100 pg/ml Strepto-mycin (Invitrogen Life Technologies).12-well-culture dishes (Costar, Coming,USA) were precoated with fetal calf serum(FCS). For this purpose, lml FCS wasapplied per well, incubated overnight at37°C, then aspirated and washed withphosphate-buffered saline (PBS). PrimaryGC and TC were plated separately onthese dishes at a density of 100,000cells per well. After 48h, the differentimmortalisation protocols were applied.Electroporation was performed with

Gene Pulser II (BIORAD, Germany).DMEM serumfree, 4mm cuvette 500, 000celllml was tested with vector DNA-con-centrations ranging between 1.0 - 5.0 ug.The protocol for Effectenetv Trans-

fection Reagent (QIAGEN, Hilden,Germany) was optimised according tothe manufacturers recommendation.After transfection the cells were sub-

cultured in untreated dishes, in DMEM(Invitrogen Life Technologies), supple-mented with 10% FCS (PAA Laborato-ries, Linz, Austria). A volume of 2/3 ofthe culture supernatant was aspirated andreplaced by fresh medium three times aweek. When confluent, the cells werepassaged with 0.025% trysinlO.Ol% ED-TA (PAA Laboratories). Stock cultureswere grown in TI5CN tissue cultureflasks (Sarstedt, Ntimbrecht, Germany).The developing cell lines were named

according to the cell type (0 = granulosa,T = theca), animal number, immortalis-ing agent (SV = complete genome ofSV40 or LT = only large T-antigen fromSV40) and line number. For exampJe,line G1SV1 originates from granulosa

ALTEX 19. Suppl. l/02

cells of animal 1, transfected with com-plete genome of SV40.For the preparation of genomic DNA or

RNA cells were trypsinised and cen-trifuged according to standard protocols(Freshney, 1994). They were washed withPBS and the final pellet was stored frozen.

2.4 Nested PCRGenomic DNA was prepared with theDNeasy Mini Kit (QIAGEN). Oligonu-cleotide primers were designed to amplifya 699 bp fragment of the large T-DNA-sequence in the first round and an inter-nal 400 bp fragment in the second round(see scheme in Fig. 8 and Weber et al.,1994). In each round, 25 PCR cycleswere performed with Taq DNA-poly-merase using an annealing temperatureof 60°C. For amplification with thesecond primer pair the PCR-product ofthe first round was diluted 1:100.Genomic DNA from SVOp12 cells

(ATCC-Nr. CRL-8621) which are knownto contain SV40LT served as positivecontrol. Non-transfected cells served as anegative control.

2.5 Assessment of progesteronesecretion

For stimulation experiments cells wereseeded at a density of 100,000cell/ml/well in DMEMIl % FCS. Theywere allowed to adhere overnight. Then,8-bromo-adenosine 3'-5' -cyclic mono-phosphate (8-Br-cAMP, SIGMA-Aldrich, Munich, Germany) was addedat a final concentration of 1 mM accord-ing to experimental requirements. At dif-ferent points in time culture supernatantswere collected for progesterone measure-ment by enzyme immunoassay (Heister-mann et al., 1993). The cells were lysedwith 100 f.Il 0.3N NaOHIl % SDS andused for protein determination (Mark-well et al., 1981). Treatments werealways performed in triplicates. Theexperiments were repeated at least twice.

2.6 Reverse Transcriptase-PolymeraseChain Reaction (RT-PCR)

Total RNA was prepared with theRNeasy Midi Kit (QIAGEN) and wasdigested with DNase (PROMEGA,Mannheim, Germany) for 30 min at 37°C

TC

Fig. 6: Primary cultures before and 48h after transfection of pBluescript-SV40 usingEffectene™ transfection reagent. GC, TC = untransfected granulosa and theca cells,GC-SV, TC-SV: the same cells after transfection. The transfection protocol itselfhas no impact on cell morphology.

67

_H_U_S_EN__E_T_A_L. !~----"J~~

to remove residual genomic DNA. Thesynthesis of cDNA was performed using2 flg of purified RNA with Oligo-dT-Primers and SuperScriptII-Reverse 'Iran-scriptase (Invitrogen Life Technologies)according to manufacturers' instructions.From this reaction 1 fll was subjected toPCR analysis.Oligonucleotide primers for PCR were

designed according to the publishedsequences of the corresponding humancDNAs (estrogen receptor ex: Green et al.,1986, forward primer: GTG TAC CTGGAC AGC AGC AA, reverse primer:TCC AGG TAG TAG GGC ACC TG,fragment size: 275 bp; progesteronereceptor: Misrahi et al., 1987, forwardprimer: GGT CTA CCC GCC CTA TCTCA, reverse primer: GGC TTG GCT TTCATT TGG AA, fragment size: 397 bp).The PCR was performed for 35 cycles atan annealing temperatue of 55°C. Ampli-fication of ribosomal 26S protein mRNAwas used as an external standard to proveRNA integrity and loading (Krazeisen etal., 1999; Husen et al., 2001).PCR-products were visualised apply-

ing 10 ul-aliquots from each reaction on2% agarose gels stained with ethidiumbromide.

2.7 Immunofluorescence stainingFor immunofluorescence cells wereseeded at a density of 10,000 cells/wellin 8-well-slides (Nunc, Wiesbaden, Ger-many) and were allowed to adhereovernight. Cells were washed with phos-phate-buffered saline (PBS), fixed with0.4% formaldehyde for 20 min andpermeabilised by 0.2% Triton-X 100(SIGMA-Aldrich) for 3 min. Antibodieswere diluted in PBS/3% bovine serumalbumine and applied according to stan-dard protocols (Husen et al., 1999).Primary antibodies were directed toestrogen receptor (ER-BlO-A, 1:100,EUROMEDEX, Strasbourg, France)progesterone receptor (PRlOA9,Immunotech, Marseille, France) and 3B-hydroxy steroid dehydrogenase (l :500,courtesy of Dr. Ian Mason, University ofEdinburgh, U.K.). Secondary antibodieswere ALEXA Fluor" 546 rabbit anti-mouse IgG (1:500, MolecularProbeslMoBiTec, Gottingen, Germany).In negative controls the primary antibodywas replaced by non-immune IgG.

68

Staining was visualised with an Axio-phot fluorescence microscope usingOpenlab 3.0™ digital image analysis(Improvision, Coventry, Ll.K)

3 Results

3.1 ImmortalisationFor the establishment of permanent gra-nulosa and theca cell lines, preovulatoryfollicles were obtained from three fertilemarmoset monkeys shortly before theexpected onset of the gonadotropin surgethat is preceding ovulation. Primarycultures of granulosa and theca cellswere prepared and subjected to differentimmortalisation protocols.For the transfer of different vector

constructs into the cells we first triedelectro poration, but this techniqueproved lethal in all attempts. Cells wereno longer able to adhere to the culturesubstrate after this treatment. As a secondmethod of DNA-transfer, Effectene'Ytransfection reagent was used whichconsists essentially of non-liposomallipids. This method was successful.Morphologically no cytotoxic effectswere observed (Fig. 6) regardless of thetype of vector construct used. Primarycells, as well as cells 48 h after trans-fection had an epithelial-like phenotype.Transfection efficiencies were deter-

mined by transfection of an aliquot of thecells with the vector pEGFP-Cl, whichcontains the gene for enhanced greenfluorescent protein. Only cells whichhave taken up this vector are capable ofexpressing the green fluorescent proteinand can be detected under a fluorescencemicroscope. This fraction of the total cellnumber ranged between 5-10% (Fig. 7).According to the manufacturer this re-presents the expected success rate in pri-mary cells.After transfection of SV40LT the cells

were subcultured and considered as celllines according to Freshney (1994). Itwas assumed that within a period ofmaximally two months non-irnmor-talised cells should have died. At somestage (after 1-10 passages) most culturesunderwent the well-known phenomenonof crisis which is accompanied by astagnant proliferation. After at least 6passages, as soon as the cells had grown

to sufficient numbers, material forcharacterisation and cryopreservation ofthe obtained cell lines was collected.Freezing and thawing did not have animpact on cell viability or growth rate.All cell lines characterised here werepast their 17.-35. passage.

3.2 CharacterisationAfter 8 to 12 months of subculture oneGC line and six TC lines yielded enoughmaterial for characterisation studies.The successful internalisation of

SV40LT was demonstrated by nestedPCR of genomic DNA from the respec-tive cells. Interestingly, in most of thecell lines tested so far, SV40LT is notdetectable (Fig. 8). Nevertheless, it has tobe assumed that these cells are immor-talised, because they clearly have anextended life span (so far, up to one year,up to 40 passages) compared to cells inprimary culture (maximally 4 weeks,3 passages). After nine passages of sub-culture, one of the theca cell lines with

Fig. 7: Determination of transfectionefficiency by calculating the fraction ofcells that have taken up the vectorpEGFP-C1 and correspondingly expressgreen fluorescent protein.A: Phasecontrast microscopy,B: Fluorescence micrograph of the samesection of the culture dish.

ALTEX 19, Suppl. 1/02

~~ HUSEN ET AL.___ ~I~-------------------------------------"J~~

SV40-Genom

o5263. 2693

large T-anticen

1000bp --II>-

500bp --II>-

300bp --..

5263

4284 ••.••- .••4982

4505'-"4904

.-400bp

Fig. 8:Demonstration of theuptake of SV40-DNA by nested peR.Top: Scheme ofprimer positionsand amplifiedfragments. In thisexperiment noneof the immortalizedcell lines tested,but only the positivecontrol cell lineSVG contains theDNA-Fragment inquestion.

no detectable SV40LT (line TlSV2) wassubjected to a second transfection, eitherwith the plasmid vector pBluescript-SV40 or with pMSSVLT. In this case noperiod of crisis was observed with eitherimmortalising vector. Even after thissecond transfection SV40LT was stillnot detectable (Fig. 8, lines TlSVLTl,TlSVLT2, TlSVSV3, TlSVSV4).During luteinisation in vivo, granulosa

cells are stimulated to produce increasingamounts of progesterone. This processwas simulated in vitro by incubating thecells with 8-Br-cAMP, a cell membranepermeable analogue of the naturalcAMP. This molecule takes part in thephysiological signal transduction path-way leading to the stimulation of proges-terone secretion after ovulation. Cellscultured without 8-Br-cAMP had a lowbasal production rate of progesterone.With 8-Br-cAMP progesterone secretionwas markedly enhanced within a timeperiod of 48h of culture (Fig. 9).Before ovulation and luteinisation

primate GC express mRNA for estrogenreceptor ex which mediates cell prolife-ration in growing follicles via estrogens.Afterwards the expression of thishormone receptor is downregulated(Einspanier et al., 1997, Saunders etal., 2000). Similarly, estrogen receptorexpression could be inhibited inimmrnortalised GC by treatment with 8-

ALTEX 19, Supp!. 1102

Br-cAMP in vitro (Fig. lOA). TC havenot been tested yet.Progesterone receptor is expressed in

vivo only during luteinisation, in GCas well as in TC (Einspanier et al., 1997).In the immortalised cell lines examineda different expression pattern wasobserved. No progesterone receptormRNA expression was found in GlSVlcells with ("luteinised") or without 8-Br-cAMP ("non-Iuteinised") (Fig. lOA).Instead, progesterone receptor mRNAwas detectable in untreated TC where the

signal appears to be stronger in thosecells which were transfected only oncecompared to those which were trans-fected twice (Fig. lOB). Again treatmentof TC with 8-Br-cAMP was not yetperformed.Immunofluorescence staining revealed

expression of cell type-specific proteins(Fig. 11). The results correspondedclosely to the expression of the respec-tive mRNA. Unstimulated GlSVI cellsshowed a typical nuclear staining forestrogen receptor ex, but not for proges-

70

c: 60.Q- 50Q) •.••••... c:u .-Q) Q)In '0 40Q) •...c: c..e al 30Q)..E(;i CllQ) c: 20al •......•ec..

10

00 8 16 24

incubation time [h]

32 40

--0- control ----1 mMcAMP

Fig. 9:Stimulation ofprogesteronesecretion ingranulosa cell lineG1SV1. Cells wereincubated with(filled squares) orwithout (opensquares) S-Sr-cAMPfor 48h andprogesterone contentof the culturesupernatant wasassessed at differentpomts of time.48

69

_H_U_S_EN__ET__A_L. !~----"J~~

terone receptor. A moderate immunore-action in the cytoplasm of the cells wasdetectable for the progesterone synthe-sising enzyme 3~-hydroxysteroid dehy-drogenase.

4 Discussion

Permanent cell lines from GC and TC ofprimates would provide an extremelyuseful experimental test system forbiomedical studies. Reproductive physio-logy of rodents is too different from thatof primates including humans. For ex-ample, rodents have extremely shortovarian cycles (4-5 days versus 25-40days in primates, 28 days in humans).Their follicular development is con-trolled by different regulatory factors. Afunctional corpus luteum is formed onlywhen conception has occurred. Conse-quently, immortalised rodent cells arenot comparable to those of primates.From humans, sufficient amounts ofovarian cells for primary cultures can beobtained only from in vitro fertilization(IVF) programs. In this case onlyluteinised GC are available, because forsuccessful oocyte retrieval the womenhave to be pretreated with gonadotro-phic hormones. Since these luteinised GCbehave already like luteal cells, they arenot appropriate to study the complexprocess of luteinisation that is crucial tofemale fertility. Human TC are not avail-able at all by IVF techniques, althoughthey play an important part in theformation of a functional corpus luteumduring luteinisation. Only tissue fromprimates like the marmoset monkeywhich lives in a controlled environmentand in which the ovarian cycle can bemonitored exactly meets all require-ments of a defined starting material forthe preparation of cell cultures.Many authors have imrnortalised GC

from different species and from differentstages of GC development (rat: Amster-dam et al., 1988; Suh and Amsterdam,1990 and others, pig: Chedrese et al.,1998; Gillio-Meina et aI., 2000; human:Rainey et aI., 1994; Lie et aI., 1996;Hosokawa et al., 1998). The results of thestudies in rats and pigs indicate that byimmortalisation it is possible to maintainGC at specific differentiation stages cor-

70

1000bp+500bp••••••300bp+

A

1000bp+-

500bp••••••300bp+

B

iO\~~~1-~<J.,~~S~0~\~ vO~~""'''S~'''\~'\~~l

r·PR

~~~~~~~======~~.~,26S

+268

Fig. 10:Detection of estrogen receptorex mRNA (ERa) and pro-gesterone receptor mRNA (PR)by RT-PCR. A: Granulosa cellline G1SV1 with and without8-Br-cAMP (cAMP) as a stimulusfor in vitro luteinisation. Asknown from in vivo experimentsestrogen receptor is present,but down-regulated afterluteinisation, progesteronereceptor is not detectable. B:Progesterone receptor mRNA inG1SV1 and different theca celllines. A positive signal isevident only in theca cells.Ribosomal 265 protein (265)was amplified as ahousekeeping gene servingas an external standard.

Fig. 11: Immunofluorescence staining of unstimulated G15V1 cells. As known from thesituation in vivo, a strong positive immunoreaction is observed for estrogen receptor.(A, nuclear staining) and 3~-hydroxysteroid dehydrogenase (C, cytoplasmic staining), but notfor progesterone receptor (8). There is no reaction in the control slide (0, primary antibodyreplaced by non-immune IgG).

ALTEX 19, Supp!. 1/02

_____ !~---------------------------------------------------H-U-S-E-N-E-T-AL--.

~~

responding to the developmental stageof the follicle they were isolated from.With granulosa cell lines from variousstages of follicular development thesemay be studied and characterised morethoroughly. This is again particularlyimportant for the study of the luteinisa-tion pathway. Similar possibilities shouldbe expexted with theca cell lines. Yet,no one has immortalised TC from pre-ovulatory follicles. None of the ovariancell lines reported has been validated asa defined in vitro test system.In vivo GC and TC are dependent on

each other to fulfil their functions duringluteinisation. Their typical enzymescomplement each other for properhormone synthesis. It is important toestablish permanent cell lines from bothcell types. Only this offers the possibilityof exposing one cell type with condi-tioned medium of the other one anddeveloping cocultures or even organotyp-ic cultures at a later stage of the project.The results of this study show that it

is possible to immortalise primarygranulosa and theca cells from NewWorld monkeys by transfection with theclassical oncogene SV40LT. Unex-pectedly, SV40LT-DNA could not bedetected in most of the developing celllines by nested PCR. In addition, by aquantitative PCR approach with theLightcycler instrument the expressionof the catalytic subunit of telomerase, anenzyme which is known to be expressedin many immortalised cell lines wasstudied (Harley and Villeponteau,1995). Again none of the tested celllines was positive for telomerase (datanot shown). The observation of lack ofSV40LT contradicts current unterstand-ing. Thus, the immortalisation mecha-nism in these cells is still a matterof speculation. It may be suspectedthat SV40LT protein was expressedfrom the vectors for some time directlyafter transfection, so that irreversibleinteractions with genes controlling cellproliferation were possible.Some specific features of the newly

immortalised cells were assessed. One ofthe most important functions of GC, thecapacity to secrete progesterone could bedemonstrated reproducibly. It could beinduced by a signalling molecule that isinvolved in the process of luteinisation

ALTEX 19, Suppl. 1/02

In vivo. In further experiments otherfactors involved in the hormonal sig-nalling pathway such as gonadotrophins,steroid hormones or extracellular matrixwill be tested.The mRNA and protein expression of

steroid hormone receptors and theprogesterone-synthesizing enzyme 3P-hydroxysteroid dehydrogenase wereexamined as cell-type and differentiationspecific markers (Einspanier et aI.,1997). This was studied with and withoutstimulation of progesterone secretion by8-Br-cAMP. With the exception ofprogesterone receptor, the expressionpatterns were exactly as known from invivo experiments (Einspanier et aI.,1997; Suzuki et aI., 1994). Progesteronereceptor expression should have beenincreased in GC with enhanced proges-terone secretion, but this was not the casein cell line G1SV1. The examination offurther granulosa cell lines is needed toclarify this issue.It has to be considered that at the

moment the cell lines still representheterogeneous populations. For theestablishment of a defined cell line theyhave to be cloned and characterised withthe same methods again. Generally, it isaimed at developing several cell lines ofeach cell type, to be able to cover thedifferent phenotypes of these cells thatoccur in vivo. According to the firstcharacterisation studies presented here,the cell lines tested so far, have retainedseveral cell-type specific features. Theyare a promising basis for the futuredevelopment of a defined in vitro testsystem.

References

Amsterdam, A., Zaubermann, A., Meir,G. et a1. (1988). Cotransfection ofgranulosa cells with simian virus 40and Ha-RAS oncogene generatesstable lines capable of inducedsteroidogenesis. Proc. Nat. Acad. Sci.85,7582-7586.

Bodnar, A. G., Ouellette, M., Frolkis, M.et a1. (1998). Extension of life-span byintroduction of telomerase into normalhuman cells. Science 279, 349-352.

Chedrese, P. J., Rodway, M. R., Swan, C.L. and Gillio-Meina, C. (1998).

Establishment of a stable steroidogenicporcine granulosa cell line. 1. Mol.Endocrinol. 20, 287-292.

Einspanier, A., lvell, R., Rune, G. andHodges, J. K. (1994). Oxytocin geneexpression and oxytocin immuno-reactivity in the ovary of the commonmarmoset monkey (Callithrixjacchus). BioI. Reprod. 50, 1216-1222.

Einspanier, A., Ivell, R. and Hodges, J.K. (1995). Oxytocin: a follicularluteinization factor in the marmosetmonkey. Adv. Exp. Med. Biol. 395,517-522.

Einspanier, A., Jurdzinski, A., andHodges, J. K. (1997). Changes in theintraovarian oxytocin, steroid recep-tors and 3p hydroxysteroid dehydroge-nase in the preovulatory ovary of themarmoset monkey. Biol. Reprod. 57,16-26.

Fraser, H. M. and Lunn, S. F. (1999).Nonhuman primates and female repro-ductive medicine. In G.F. Weinbauerand R. Korte (eds.), Reproduction innonhuman primates. A model systemfor human reproductive physiologyand toxicology (27-59). MUnster: Wax-mann Verlag.

Freshney, R. 1. (1994). Culture of animalcells. A manual of basic technique, 3'"edition, New York: Wiley Liss, Inc.

Gillio-Meina, C., Swan, C. L., Crellin.N.K. et a1. (2000) Generation of stablecell lines by spontaneous immortaliza-tion of primary cultures of porcinegranulosa cells. Mol. Reprod. Dev. 57,366-374.

Green, S., Walter, P., Kumar, V. et al.(1986). Human oestrogen receptorcDNA: sequence, expression andhomology to v-erb-A. Nature320(6058), 134-139.

Harley, C. B. and Villeponteau, B.(1995). Telomeres and telomerase inaging and cancer. Curr. Opin. Genet.Dev. 5, 249-255.

Hearn, J. P. (1983). The commonmarmoset (Callithrix jacchus). In J. P.Hearn (eds.), Reproduction in NewWorld Primates, New Models inMedical Science (181-215). Lancaster:MTP Press Ltd.

Heistermann, M., Tari, S. and Hodges, J.K. (1993). Measurement of fecalsteroids for monitoring ovarianfunction in New World primates,

71

!~-----------------------------------------------------~~HUSEN ET AL.

Callitrichidae. J. Reprod. Fertil. 99,243-251.

Hillier, S. G., Harlow, C. R., Shaw, H. J.et al. (1988). Cellular aspects of pre-ovulatory folliculogenesis in primateovaries. Hum. Reprod. 3(4), 507-511.

Hopfer, U., Jacobberger, .T.W., Gruenert,D. C. et al. (1996). Immortalization ofepithelial cells. Am. J. Physiol. 270(1),Cl-ClI.

Hosokawa, K., Dantes, A., Schere-Levy,C. et al. (1998). Induction ofAd4BP/SF-1, steroidogenic acuteregulatory protein, and cytochromeP4S0scc enzyme system expression innewly established human granulosacell lines. Endocrinology 139(11),4679-4687.

Husen, B., Giebel, J. and Rune, G. M.(1999). Expression of integrin subunitsas, a6 and ~1 in the testes of thecommon marmoset. Int. J. Andral. 22,374-384.

Husen, B., Adamski, J., Rune, G.M.,Einspanier, A. (2001). Mechanisms ofestradiol inactivation in primateute-rus. Mol. Cell. Endocrinol. 171(1-2), 179-186.

Krazeisen, A., Breitling, R., Irnai, K etal. (1999). Determination of cDNA,gene structure and chromosomallocalization of the novel human 17~-hydroxy steroid dehydrogenase type 7.FEBS Lett. 460(2), 373-379.

Lie, B., Leung, E., Leung, P. C. K andAuersperg, N. (1996). Long-termgrowth and steroidogenic potential ofhuman granulosa-lutein cells immor-talized with SV40 large T antigen.Mol. Cell. Endocrinol. 120, 169-176.

Markwell, M. A. K,Haas, S. M., Tolbeit,N. E. and Bieber, L. L. (1981). Proteindetermination in membrane andlipoprotein samples: manual and auto-mated procedures. Met. Enzymol. 72,296-299.

Marx, U. (1996) Stand und Weiteren-twicklung von Zellkulturmethoden. InF. P. Gruber, H. Spielmann (eds.)Alternativen zu Tierexperimenten -Wissenschaftliche Herausforderungund Perspektiven (69-97). Heidelberg,Berlin, Oxford: Spektrum Akade-mischer Verlag.

Misrahi, M., Atger, M., d'Auriol, L. et al.(1987). Complete amino acid sequenceof the human progesterone receptor

72

deduced from cloned cDNA. Biochem.Biophys. Res. Commun. 143 (2), 740-748.

Modrow, F. and Falke, D. (1998).Molekulare Yirologie. Heidelberg,Berlin, Oxford: Spektrum Akade-mischer Verlag.

Morales, C. P., Holt, S.E., Ouellette, M.et al. (1990). An expression vectorsystem for stable expression ofoncogenes. Nucleic Acids Res. 18(16),4945-4946.

Rainey, W. H., Sawetawan, C., Shay, J.W. et al. (1994). Transformation ofhuman granulosa cells with the E6 andE7 regions of human papillomavirus.J. Clin. Endocrinol. Metab. 78(3),70S-710.

Saunders, P. T. K, Millar, M. R.,Williams, K. et al. (2000). Differentialexpression of estrogen receptor-a and-~ and androgen receptor in the ovariesof marmosets and humans. Biol. Re-prod. 63, 1098-1I0S.

Schuermann, M. (1990). An expressionvector system for stable expression ofoncogenes. Nucleic Acids Res. 18,494S-4946.

Smith, D., Trennery, P., Famingham, D.and Klapwijk, J. (2001). The selectionof marmoset monkeys (Callithrixjacchus) in pharmaceutical toxicology.Lab. Anim. 35(2),117-130.

Spielmann, H. (1996) Altemativen in derToxikologie. In F. P. Gruber, H.Spielmann (eds.) Alternativen zu Tier-experimenten WissenschaftlicheHerausforderung und Perspektiven(108-126). Heidelberg, Berlin, Oxford:Spektrum Akademischer Verlag.

Suh, B. S. and Amsterdam, A. (1990).Establishment of highly steroidogenicgranulosa cell lines by cotransfectionwith SV 40 and Ha-ras oncogene:induction of steroidogenesis by cyclicadenosin 3'-5' -monophosphate and itssuppression by phorbolester.Endocrinology 127(5), 2489-2S00.

Suzuki, T., Sasano, H., Kimura, N. et al.(1994). Immunohistochemical distri-bution of progesterone, androgen andoestrogen receptors in the humanovary during the menstrual cycle:relationship to expression of stero-idogenic enzymes. Hum. Reprad. 9(9),1589-1S9S.

Weber, T., Turner, R. w., Frye, S. et al.

(1994). Specific diagnosis of progres-sive multifocal leukoencephalopathyby polymerase chain reaction. J. Infect.Dis. 169(5), 1138-1141.

Wiesmliller, L., Cammenga, J. andDeppert, W. W. (1996). In vivo assay ofp53 function in ho-mologous recombi-nation between SV40 chromosomes. J.Viral. 70(2), 737-744.

Vaziri, H. and Benchimol, S. (1998).Reconstitution of telomerase activityin normal human cells leads to elonga-tion of telomeres and extended replica-tive lifespan. CurroBioi. 8, 279-282.

AcknowledgementsThis work was supported by the founda-tion SET (Stiftung zur Forderung der Er-forschung von Ersatz- und Erganzungs-methoden zur Einschrankung vonTierversuchen). We are grateful to Prof.J. K Hodges for critical reading of themanuscript.

Correspondence toDr. Bettina HusenAbteilung ReproduktionsbiologieDeutsches PrimatenzentrumKeUnerweg 4D-37077 GottingenTel. +49-SS1-38S1-132Fax: +49-551-38S1-288E-mail: bhusen@www.dpz.gwdg.de

ALTEX 19, Suppl. 1/02