Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

A role of inhibin as a tumor suppressor in Sertoli cells: down-regulation upon aging and repression by a viral oncogene


Inhibin, a member of the TGF-β superfamily, is synthesized in the testis by Sertoli cells and exerts an endocrine regulatory function on pituitary hormone synthesis. A distinct local function has been proposed, negatively controlling cellular growth in the testis (tumor suppressor activity). A critical test for the identification of a tumor suppressor is the reversal of transformed growth properties upon re-expression of the gene in tumor-derived cell lines. Sertoli cell-derived tumoral lines were previously established from tumors that develop in elderly transgenic males which express in the testis the large T antigen of polyoma virus. Both the tumors and the cells in culture exhibited reduced levels of the inhibin α subunit mRNA. Stable transfectants were generated, in which this subunit was expressed from a heterologous promoter. All of them exhibited a strict inhibition of growth at confluency. On the other hand, in addition to an aging-related decrease in inhibin synthesis, the α subunit gene was down regulated in vivo in cells expressing the viral protein. The conjunction of these two factors accounts for the age-related occurrence of testicular cancers in the transgenic model, again pointing to inhibin as a potent cell growth regulator in the seminiferous epithelium.


Inhibin and activin, members of the transforming growth factor β family, share common subunits designated βA and βB. Activins are homo- and heterodimers of the β subunits. Inhibin is a heterodimer made of a distinct subunit, α, associated with either βA or βB (Chen, 1993; Chen et al., 1992 for reviews). The major site of inhibin synthesis in the testis is the Sertoli cell (Skinner, 1993 for review), while the β subunits are made in various cell types at a constant rate. The rate of α chain synthesis determines the ratio of the two hormones, which exert opposite roles on pituitary activity. In addition to this endocrine function, a paracrine or autocrine role has been considered for inhibin in the testis. Both in vivo (van Dissel-Emiliani et al., 1989) and in isolated seminiferous tubules (Hakovirta et al., 1993), it was reported to inhibit DNA synthesis and reduce the number of spermatogonia. The most compelling evidence, however, was provided by studies on mice homozygous for a deletion in the Inha gene encoding the α subunit (hereafter designated α-inhibin). The mutant mice develop gonadal tumors at an early age, suggestive of a role of inhibin as a tumor suppressor (Matzuk et al., 1992).

A critical test for tumor suppressor activity is the restoration of normal growth properties in culture upon re-expression of the wild type gene in tumoral cells (Huang et al., 1988). Few tumoral cell lines have been derived from testicular cells. We reported previously the constant occurrence of tumors of Sertoli cell origin in the testes of transgenic males expressing the Large T antigen of polyoma virus (PyLT), and the subsequent derivation of permanent cell lines (Paquis-Flucklinger et al., 1993). The observation of a much reduced rate of α-inhibin synthesis in these lines as well as in the tumors led us to inquire whether a change in the growth pattern of the tumoral lines would follow the resumption of expression of α-inhibin from a transfected construct.

In these transgenic families, testicular tumors appear characteristically at old ages. All the males show normal histological structures, as well as normal fertility during the first 14 – 15 months of their lives. All of them, however, later develop fast growing bilateral tumors, starting with the uncontrolled proliferation of Sertoli cells. The striking association of these malignancies with the aging process suggested a causal link involving changes in hormonal balance. Our present results point to a decrease in α-inhibin synthesis as a likely triggering event, and thus, to an important function of inhibin as a cell growth regulator in the testis.


Expression of the Inha gene (α subunit of inhibin) reinstates a normal growth control in a tumor-derived Sertoli cell line

Our initial aim was to establish directly and further precise a possible role of inhibin as a tumor suppressor in the testis, as initially suggested by the high incidence of testicular cancer in mice in which the gene encoding the α subunit had been deleted (Matzuk et al., 1992). We used the murine cell line 45T-1, established from a Sertoli cell tumor (Paquis-Flucklinger et al., 1993). These cells show the typical characteristics of a deregulated growth control, reaching high densities in multilayered post-confluent cultures. Northern blot analysis evidenced a reduced accumulation of α-inhibin mRNA in the tumor, as compared with normal Sertoli cells in primary cultures (Figure 1a). Accordingly, semi-quantitative RT – PCR assays detected low levels of Inha RNA in 45T-1 cells (Figure 1b).

Figure 1

Reduced levels of inhibin α subunit (Inha) mRNA in a Sertoli tumor and in the tumor derived cell line 45T-1 compared to normal Sertoli cell RNA. (a) Northern blot hybridization of RNAs prepared from purified Sertoli cells in primary culture (lane 1) and from the testicular tumor of Sertoli origin developed by an 18-month-old PyLT transgenic male (lane 2). Hybridization was performed in succession with probes for the Inha cDNA sequence, for the cDNA of KL (Kit ligand, Steel factor) and for the ubiquitous ribosomal S26 protein cDNA. KL is characteristic of either normal or tumoral Sertoli cells, and expressed in a cycle-dependent manner in normal Sertoli cells (Vincent et al., 1998) and upregulated in this Sertoli tumor (our unpublished observations). S26 RNA is ubiquitously expressed in mouse cells (Vincent et al., 1993) and is used for normalization. (b) Analysis of amplification products obtained after reverse transcription and semi-quantitative amplification with primers A and B for α-inhibin (257 bp), G and H for HPRT (354 bp) (see Materials and methods), using 1 μg of total RNA extracted from 45T-1 cells (lane 1), a B6D2F1 mouse testis (lane 2), three independent neor clones derived from 45T-1 (45T-1[co]: lanes 3 – 5) and three neor-Inha+ clones (45T-1[inh]: lanes 6 – 8)

To check whether this reduced level of expression plays a determinant role in the maintenance of the transformed state, clonal derivatives were established, which express the gene at higher levels. A plasmid construct was generated, in which the complete Inha coding region was inserted downstream of the early cytomegalovirus promoter and of the neor selectable gene (pCDNAinh, see Materials and methods). Cultures were transfected in parallel with the Inha expression vector and with the empty neor plasmid. After selection in G418 containing medium, three independent clones were randomly picked in the two series, noted 45T-1[inh] and 45T-1[co], respectively. Reverse transcription and semi-quantitative RT – PCR assays verified the expected increased accumulation of Inha transcripts in all three 45T-1[inh] clones as compared with the controls (Figure 1b).

The 45T-1[inh] and control clones grew at the same rate (Figure 2a). After they reached confluency, the neor control cultures, as the original 45T-1 cells, remained in an actively dividing state, forming dense multilayered and morphologically disorganized cultures (Figure 2b). In clear contrast, all three α-inhibin expressers showed a strict contact inhibition of growth. In spite of regular medium changes, they remained for up to 1 month as healthy monolayers of cells organized in parallel arrays (Figure 2c). This strict contact inhibition of growth is similar to that of freshly explanted Sertoli cells in primary cultures (Figure 2d). As initially established for the paradigmatic Rb tumor suppressor (Huang et al., 1988), the observed suppression of transformed growth properties in culture strongly argues for a negative regulation of cell growth exerted by the product of the Inha gene.

Figure 2

Growth inhibition at confluency in 45T-1[inh] transfectants. (a) Growth of the 45T-1[co] and 45T-1[inh] cell lines. Cells were seeded at day 0 at a density of 2×104 cells per well and medium was changed every day. Cells were counted at the indicated times. Upper panel: the two cell types show identical doubling time in non-confluent cultures; open symbols: 45T-1[inh], closed symbols: 45T-1[co]. Lower panel: in confluent cultures, growth of 45T-1[inh] cells is arrested at lower cell densities; closed bars: 45T-1[co]; open bars: 45T-1[inh]. The values reported correspond to the average of three distinct experiments performed on three independent clones of each type. Standard error of the mean (s.e.m.) indicated in each series. (b) Giemsa staining of cultures of 45T-1[co] (b), 45T-1[inh] (c) and of primary Sertoli cells (d) 1 week after they reached confluency. Final enlargement: 400×

The conjunction of an age-related decrease in α-inhibin synthesis and of a further reduction upon expression of the viral oncogene coincides with the onset of tumor growth in the male PyLT mouse

The 45T-1 cell line had been initially established from the Sertoli cell tumors that arise in the elderly PyLT transgenic males. Their strict age-dependence occurrence suggested that a change in the hormonal status of the animal was critical in the onset of tumoral growth. Since a decrease in inhibin concentration in the sera of aging individuals has been documented in humans (Guerra et al., 1989; Pal et al., 1991), we measured the levels of Inha expression in male mice at various ages. A significant reduction was consistently observed in the accumulation of mRNA in elderly mice (14 months) as compared with the young adults (5 months) (Figure 3). This was the case both in the normal B6D2 and in the PyLT transgenic lineages. It was also noted, however, that, in all transgenic males, the levels of Inha RNA were reduced by at least 50% compared to the age-matched normal controls (Figure 3).

Figure 3

Aging-related decrease of inhibin α subunit expression in the testis and down regulation in PyLT transgenic mice. (a) Expression analysis was performed on testis mRNA by Northern blot hybridization with specific α-inhibin or GAPDH cDNA probes. Lanes 1 and 3: B6D2F1 normal mice, lanes 2 and 4: PyLT transgenic mice; lanes 1 and 2: 5-month-old mice, lanes 3 and 4: 14-month-old mice. (b) Same experiment as in (a), with densitometric scanning values normalized against the GAPDH controls. Lane 1: B6D2F1, 5 months, lane 2: B6D2F1, 14 months, lane 3: PyLT, 5 months, lane 4: PyLT: 14 months. Mean±s.e.m. of at least three independent experiments in each series. Student's t-test for paired data: P<0.05 between 1 and 2, 1 and 3, and 3 and 4. (c) Expression analysis was performed on testis mRNA isolated from a 1-year-old mouse by RNase protection assays. Upper panel: Inha RNA levels are higher in normal testis (lane 2) than in the testes of three PyLT transgenic mice (lanes 3 – 5). The undigested probe is shown in lane 1. Lower panel: RNAse protection assays on WT1 RNA, a marker of adult Sertoli cell (Pelletier et al., 1991), used for normalization of densitometric scanning measurement. (d) Signals were normalized against WT1 controls in two distinct experiments performed on three control animals and eight PyLT transgenic mice

The emerging picture would therefore include an age-related decrease of α-inhibin synthesis, which, by itself, would not be sufficient to generate neoplastic developments, but would be potentiated in the transgenic mouse by an additional down regulation due to the viral oncogene.

The observed effect may involve the secretion of inhibin in the medium and the subsequent activation of the receptor. To investigate this point, we established mixed cultures of the 45T-1[inh] and control clones starting with different ratios of the two cell types. One week after confluency, all cultures showed multilayered regions, whose extent correlated with the ratio of control to 45T-1[inh] cells. Even in the presence of a 100 – 1000-fold excess of the inhibin producers, 45T-1[co] cells generated disorganized multilayered foci (Figure 4). We also checked whether the growth pattern of the tumor cells could be modified by addition of human recombinant inhibin to the culture medium. The human hormone is known to be active on testicular mouse cells in organ culture (Hangoc et al., 1992; Jakubowiak et al., 1991), but no effect was observed on the growth of 45T-1 at concentrations ranging from 50 – 200 ng/ml (Figure 4). These two series of results therefore concur to suggest a juxtacrine or intracrine mode of action rather than one implying the release of a diffusible factor.

Figure 4

Neither a large excess of 45T-1[inh] nor the addition of recombinant inhibin affects the growth properties of 45T-1[co]. (a – d): mixtures of different ratio of 45T-1[inh] and 45T-1[co] were seeded. Giemsa staining of 45T-1[inh] (a), 45T-1[co] foci (arrows) in 45T-1[inh] (b,c) (cells were seeded at ratio of 1 45T-1[co] cell for 100 45T-1[inh]; (d) 45T-1[co]. Final enlargement: 100×. (e – h) Giemsa staining of untreated 45T-1[inh] (e), 45T-1[co] (g), or treated during 1 week after confluency with 200 ng/ml human recombinant inhibin (f and h). Final enlargement: 200×

PyLT expression results in the down regulation of the Inha promoter in primary culture Sertoli cells

A 452 bp minimal Inha promoter has been previously shown to be sufficient to drive the expression of a reporter gene in transfected granulosa cells (Su and Hsueh, 1992). After PCR amplification and cloning, this fragment was inserted upstream of the LacZ reporter in the pNASSβ vector (see Materials and methods). The resulting construct (pInh-lacZ) efficiently expressed β-galactosidase when transfected into primary culture Sertoli cells. As shown in Figure 5, cotransfection of increasing amounts of either pPyLT1 DNA encoding only the Large T protein (Zhu et al., 1984), or of pPY1 DNA (Rassoulzadegan et al., 1981), which encodes all three viral early proteins (Large T, Middle T and Small T), resulted in a decreased expression of the Inha-directed reporter. This negative regulation, either direct or indirect, of the expression of the inhibin α subunit, explains its decreased expression in the testes of the PyLT transgenic males.

Figure 5

PyLT expression down-regulates the inhibin α subunit promoter activity in primary Sertoli cell cultures. β-galactosidase activity was measured 72 h after transfection of 0.5 μg of pInh-lacZ plasmid (see Materials and methods) into primary culture Sertoli cells. This value was referred to as 100%. Activities measured after transfection of the indicated DNA mixtures are represented (mean±s.e.m. of triplicate experiments). The histogram shows the relative activity in non transfected cells (lane 1), in cells transfected with pInh-lacZ alone (lane 2) or together with increasing amounts of pPyLT1 plasmid DNA (lanes 3 – 5: 0.1, 0.25 and 0.5 μg), in cells transfected with 0.5 μg of pInh-lacZ and 0.5 μg of pPY1 DNA (lane 6), in cells transfected with 0.5 μg of the control vector (pNASSβ) and increasing amounts of pPyLT1 vector (lanes 7 – 9: 0, 0.25 and 0.5 μg of pPyLT1 respectively). All transfection mixtures were completed to a total of 1 μg with pBR322 DNA


The hypothesis that inhibin acts as a tumor suppressor in the testis was initially based on the observation that mice homozygous for a null allele of the Inha gene develop testicular tumors at a young age (Matzuk et al., 1992). Our results further document the fact that, in the mouse testis, a decreased expression of inhibin in Sertoli cells leads to uncontrolled cellular growth. Experimental Sertoli cell tumors induced in transgenic mice by a viral oncogene, and a derived cell line were found to express low to undetectable amounts of Inha mRNA. Transfectants in which Inha expression was resumed from a recombinant gene recovered a growth control in culture similar to that of normal Sertoli cells, with a strict inhibition of proliferation past confluency. An independent indication for a growth-controlling function of α-inhibin is that the onset of tumoral development in the aging transgenic males of the families coincides with a physiological decrease in inhibin synthesis, with, in the transgenic model, an additional down-regulation of the Inha promoter by the viral oncogene.

Neither the addition of recombinant inhibin to the culture medium of the tumoral cell line, nor its coculture in the presence of a large excess of the inhibin producer, modified its growth pattern. This negative result would obviously require confirmation by other means, and tools for blocking either the activity or the synthesis of the inhibin receptor would be of interest. Taking them at face value, our observations rather suggest that inhibin acts in an autocrine manner, possibly via an intracrine loop, as do a number of other growth controlling agents, including granulocyte macrophage colony stimulating factor and erythropoietin (Pech et al., 1993), interleukine 3 (Dunbar et al., 1989), interleukine 6 (Barut et al., 1993) and platelet derived growth factor (Keating and Williams, 1988).

An alternative explanation for the lack of effect of external inhibin could be, however, that, rather than inhibin itself, the active compound might be (one of the) activin(s). Activins are dimers of the β subunits. The rate of formation of the α/β heterodimeric inhibin, which depends on the level of α subunit, ultimately determines the available amount of the constitutively synthesized β subunits. In other words, the changes in growth regulation observed in culture as well as during in vivo tumorigenesis might reflect a growth promoting function of activin rather than a growth suppressing function of inhibin. Activin was indeed shown to stimulate the proliferation of rat Sertoli cells, but only in combination with FSH and during a short period of the postnatal development (Boitani et al., 1995). This hypothesis is currently being tested, but with, so far, only negative results. We asked whether media conditioned by growth of the tumoral line would induce growth post confluency of the 45T-1[Inh] clones. As in the reverse experiment (Figure 4), however, these conditioned media had no effect. We also tested the possible effects of recombinant activin, again with negative results. Taken together, the present data therefore suggest that, in addition to its long distance endocrine effects on pituitary secretion, inhibin acts, possibly via an intracrine loop, to suppress the growth of Sertoli cells.

The observed effect of PyLT on Inha promoter expression may result from protein-protein interactions with either the retinoblastoma (Dyson et al., 1990) or other regulatory proteins. Alternatively, T antigen might bind to one or several site(s) within the Inha promoter. Binding might either result in transcriptional repression, as it is the case in the viral promoter (Cowie and Kamen, 1984), or it could compete with that of a positive regulator. Such a model is in part supported by the observation that the region of the T antigen binding sites in the viral promoter shows significant similarity with a sequence of the Inha promoter, including the (A/G)GGC repeates constitutive of T antigen binding sites (Gaudray et al., 1981; Tjian, 1978) and at a comparable position relative to the transcription start site (Figure 6). Moreover, the same sequence elements are found in the promoter of another gene expressed in Sertoli cells, that of the urokinase plasminogen activator (m-uPA), and within an even more extended region of similarity with Inha (Grimaldi et al., 1996). The hypothesis that binding of a Sertoli-specific positive regulator in this region is competed out by T antigen is currently being tested. A first prediction is that mutating the putative T antigen binding sites in the Inha promoter should reduce its level of expression in Sertoli cells, the residual expression, if any, being insensitive to PyLT. It seems to be corroborated by preliminary results with a mutant Inha promoter with the two GGGGC boxes mutated into GGTGC, a sequence which is not recognized with high affinity by T antigens (DeLucia et al., 1983). We are also searching for a Sertoli-specific protein that would bind this region of the Inha promoter. Candidate cDNAs have been identified by screening of an expression library and their cellular distribution, DNA binding ability and effect in trans on Inha expression are currently under study.

Figure 6

Sequence similarities between the polyoma virus early promoter and two mouse promoters expressed in Sertoli cells. The (−103,−69) region of the Inha promoter (Su and Hsueh, 1992) is aligned with the promoter of the murine urokinase plasminogen activator gene (`m-uPA'), in a region which is itself highly similar to the human urokinase plasminogen activator gene (not shown, Grimaldi et al., 1996). The sequence of the T antigen binding site in the polyoma virus early promoter (`PyE') is shown on top (Cowie and Kamen, 1984). Stars indicate identical nucleotides and the (A/G)GGC boxes characteristic of the T antigen binding sites are underlined

Materials and methods

Cell lines, transfection and culture conditions

All cell lines were grown routinely at 32°C in Dulbecco Modified Eagle's Medium (DMEM) supplemented with 10% fetal calf serum (GIBCO Laboratories, Grand Island, NY, USA). They were transfected, using the DOTAP reagent transfection kit (Boehringer Mannheim, Germany) according to supplier's instructions, with 1 μg of DNA per 1×105 cells. Selection for stable transformants was performed during 3 weeks in medium containing 1 mg/ml Geneticin (GIBCO Laboratories, Grand Island, NY, USA). The emerging resistant clones were cloned either by end point dilution or with cloning plastic rings. Neomycin resistant clones were further grown in DMEM supplemented with 10% fetal calf serum and with 0.54 mg/ml Geneticin. Recombinant human inhibin was obtained from the National Hormone and Pituitary Program, Harbor-UCLA Medical Centre, Torrance, CA, USA. Activin was a generous gift of C Boitani (Dipartimento di Sanita Publica e Biologia Cellulare, Universita di Roma, Italy).

Primary Sertoli cell cultures were established from 8-day-old males according to published methods (Rich et al., 1983). Their degree of purity after 1 week of culture was estimated to be greater than 95% on the basis of the following criteria: cell morphology after vimentin immunostaining, presence of lipid inclusions evidenced by red oil staining, active phagocytosis measured by internalization of latex beads (Grandjean et al., 1997), synthesis of the inhibin α subunit, and the absence of Leydig cells identified by in situ determination of 3β-hydroxysteroid dehydrogenase activity.

In vivo promoter activity assay

Seventy-two hours after transfection with the DOTAP reagent transfection Kit, β-galactosidase activity was measured using the Galacton kit (TROPIX) according to supplier instructions.

Plasmids and DNA preparation

Plasmid DNA was extracted from bacteria using the Maxiprep Qiagen kit (Qiagen) according to supplier instructions. Each construct was checked by sequencing using the T7 sequencing kit (Pharmacia Biotech) or the Amplicycle PCR sequencing kit (Perkin Elmer). PGEMT-inh contained a full length α-inhibin cDNA inserted at the EcoRI site of vector pGEM-Teasy (Promega), pSKKL1 contained 1.1 kb of KL1 cDNA between HindIII and XbaI sites of vector pSK (Stratagene) and HSS26 contained the human cDNA ribosomal S26 protein cDNA insert (Vincent et al., 1993). Plasmid pRC/CMV WT1 (−/+) used for synthesis of WT1 antisense RNA probes contains 2060 bp of the WT1 cDNA at the XbaI site of vector pRC/CMV (Larsson et al., 1995). PSKinh used for synthesis of Inha antisense RNA probe contains a 256 bp fragment of the cDNA obtained by RT – PCR using oligonucleotides A and B (Su and Hsueh, 1992), cloned at the EcoRV site of vector pSK (Stratagene). The α-inhibin reporter plasmid, pInh-lacZ was obtained by cloning a 452 bp PCR fragment of the α-inhibin promoter (Su and Hsueh, 1992) at the EcoRI site of pNASSβ (CLONTECH). PCR amplification was performed on genomic DNA using primers C and D (see below for primer sequences and polymerization conditions). pcDNA I NEOinh was obtained by cloning the α-inhibin entire coding region (1216 bp) obtained by RT – PCR using primers E and F in vector pcDNA I NEO (Invitrogen).

RNA purification

Total RNAs were prepared from decapsulated normal testes, primary Sertoli cell cultures and Sertoli cell lines derived from tumor testis by the acid guanidinium thiocyanate method (Chomczynski and Sacchi, 1987). RNA concentrations were measured spectrophotometrically and the quality of the purified RNA preparations was checked by UV examination after agarose electrophoresis and ethidium bromide staining.

RNA analysis

Northern blot

 Twenty μg of total RNA from testes or from cells in culture were electrophoresed in 1% agarose gels containing 1×TAE and 0.17 volume of 37% formaldehyde. RNAs were transferred to Hybond N+ (Amersham) membrane by capillary blotting with 20×SSC. The filters were rinsed with 2×SSC and RNAs were fixed on membrane by UV irradiation (Ultraviolet crosslinker Amersham). The filters were prehybridized in 5×SSC, 50% formamide, 50 mM HEPES (pH 6.8), 2 mM EDTA, 5×Denhart, 0.1% SDS, for 4 h at 65°C and hybridized overnight at 65°C in 10% Dextran Sulfate, 2×Denhart, 5×SSC, 50 mM EDTA, 50 mM HEPES (pH 6.8), 40% formamide, 0.2% SDS, 50 μg/ml thymus sheared DNA, with a cRNA probe or with a random primed α32P-dCTP-labeled fragment. Filters were subsequently washed at 20°C in 2×SSPE, 0.1% SDS and at 65°C in 1×SSPE, 0.1% SDS, and exposed to Kodak film for 2 – 5 days at −70°C with an intensifying screen.

RNase protection assasy

 Labeled cRNA transcripts were produced from PSKinh and pRC/CMV WT1 (−/+) in a final volume of 10 μl containing 1×transcription buffer (Boehringer), 20 units of RNase inhibitor (Boehringer), 0.5 mM each of ATP, GTP, UTP, 0.1 mM CTP, 50 μCi of [α32P]CTP (ICN), 0.5 μg of linearized DNA template and 40 units of either T7 or SP6 RNA polymerase (Boehringer). Reactions were performed for 90 min at 37°C. Ten units of DNase I (Boehringer) were then added, and the mixtures incubated for a further 15 min. For each RNase protection assay, 60 μg of total RNA from either testis or cell lines were annealed to labeled cRNA (1×106 d.p.m.), first for 5 min at 85°C, then overnight at 55°C, in a final volume of 30 μl of hybridization buffer (40 mM PIPES pH 6.4, 1 mM EDTA pH 8, 0.4 mM NaCl, 80% formamide). The hybridization mixture was then incubated in 300 μl of digestion solution containing 1×RNase ONE digestion buffer and 24 units of RNase ONE (Promega) at 20°C for 30 min and at 37°C for a further 30 min. Digestion was terminated by the addition of 5 μl of a solution containing 10% SDS, 4 μg/μl tRNA and kept 5 min in ice. Samples were ethanol precipitated. Protected RNAs were fractionated on 7 M urea 6% polyacrylamide gel. The gel was exposed to Kodak BMR 1 film with an intensifying screen for 1 – 6 days.

Semi quantitative RT – PCR and PCR analysis

 One μg of DNase I treated total RNA from cell lines was converted to single stranded cDNA and amplified by PCR by using the TITAN kit (Boehringer), according to supplier's instruction. Programs were 30 min at 50°C for Inha (5 min for Hprt) and then 10 cycles at 94°C for 30 s, 58°C for 30 s and 68°C for 30 s, following by 25 cycles 94°C for 30 s, 65°C for 30 s and 68°C for 30 s, incremented for 5 s per cycle. The PCR was terminated by one cycle at 68°C for 7 min. Primers for these reactions were A and B for inhibin, and G and H for HPRT. Amplified cDNAs were separated by electrophoresis on 2% agarose gels.

RT – PCR and PCR were carried out in GeneAmp PCR System 2400 (Perkin Elmer). PCR assays were performed on 0.2 μg of genomic DNA using standard condition. Sequences of oligonucleotide primers and their positions in the reference α-inhibin sequences are as follows, numbered from the initiator ATG (Su and Hsueh, 1992), with italics corresponding to the restriction sites added for cloning the PCR products: A: 5′-230GATCCTGGAATAAGGCG247-3′, B: 5′-486CTAGGCCTGTGTGGAAC470-3′, C: 5′-CGGAATT-CC74CATAGTTCACTTGCCCGT57-3′, D: 5′-CGGGAT-CC338TCAACCTTAAGCACCCAG321-3′, E: 5′-ACGTACGTGAAT12TCCTAGACAGAAAGGGCACAG32-3′; F:5′-ACGTACGTGAATT1230CCTGGGGTGGTGAATTGACT1211-3′.


The annealing temperatures and number of cycles are as follows. A and B: 50°C, 35 cycles; C and D: 72°C, 35 cycles; E and F, G and H: 58°C, 7 cycles and 65°C, 30 cycles.

Statistical analysis

Data are expressed as mean±s.e.m. Number of measurements is indicated for each experiment. Statistical comparisons refer to Student's t-test for paired data.


  1. Barut B, Chauhan D, Uchiyama H and Anderson KC. . 1993 J. Clin. Invest. 92: 2346–2352.

  2. Boitani C, Stefanini M, Fragale A and Morena AR. . 1995 Endocrinology 136: 5438–5444.

  3. Chen C-LC. . 1993 Endocrinology 132: 4–5.

  4. Chen C-LC, Feng Z-M, Morris PL and Bardin CW . (1992). In: Molecular Biology of Reproduction, Serono Symposia. Spera G, Fabbrini A, Gnessi L and Bardin CW. (eds).. Raven Press: New York pp. 97–108.

  5. Chomczynski P and Sacchi S. . 1987 Anal. Biochem. 162: 156–159.

  6. Cowie A and Kamen R. . 1984 J. Virol. 52: 750–760.

  7. DeLucia AL, Lewton BA, Tjian R and Tegtmeyer P. . 1983 J. Virol. 46: 143–150.

  8. Dunbar CE, Browder TM, Abrams JS and Nienhuis AW. . 1989 Science 245: 1493–1496.

  9. Dyson N, Bernards R, Friend SH, Gooding LR, Hassell JA, Major EO, Pipas JM, Vandyke T and Harlow E. . 1990 J. Virol. 64: 1353–1356.

  10. Gaudray P, Tyndall C, Kamen R and Cuzin F. . 1981 Nucl. Acids Res. 9: 5697–5710.

  11. Grandjean V, Vincent S, Martin L, Rassoulzadegan M and Cuzin F. . 1997 Biol. Reprod. 57: 1115–1122.

  12. Grimaldi P, Geremia R, Albanesi C and Rossi P. . 1996 Mol. Cell. Endocrinol. 117: 167–173.

  13. Guerra GR, Franco J and Gonzales GF. . 1989 Arch. Androl. 22: 35–40.

  14. Hakovirta H, Kaipia A, Soder O and Parvinen M. . 1993 Endocrinology 133: 1664–1668.

  15. Hangoc G, Carow CE, Schwall R, Mason AJ and Broxmeyer HE. . 1992 Exp. Hematol. 20: 1243–1246.

  16. Huang HJ, Yee JK, Shew JY, Chen PL, Bookstein R, Friedmann T, Lee EY and Lee WH. . 1988 Science 242: 1563–1566.

  17. Jakubowiak A, Janecki A, Tong D, Sanborn BM and Steinberger A. . 1991 Mol. Cell. Endocrinol. 82: 265–273.

  18. Keating MT and Williams LT. . 1988 Science 239: 914–916.

  19. Larsson S, Charlieu J, Miyagawa K, Engelkamp D, Rassoulzadegan M, Ross A, Cuzin F, van Heyningen V and Hastie N. . 1995 Cell 81: 391–401.

  20. Matzuk MM, Finegold MJ, Su J-GJ, Hsueh AJW and Bradley A. . 1992 Nature 360: 313–319.

  21. Pal D, Collins TJ and Parkening TA. . 1991 Biol. Reprod. 45: 869–875.

  22. Paquis-Flucklinger V, Michiels J, Vidal F, Alquier C, Pointis G, Bourdon V, Cuzin F and Rassoulzadegan M. . 1993 Oncogene 8: 2087–2094.

  23. Pech N, Hermine O and Goldwasser E. . 1993 Blood 82: 1502–1506.

  24. Pelletier J, Schalling M, Buckler AJ, Rogers A, Haber DA and Housman D. . 1991 Genes Dev. 5: 1345–1356.

  25. Rassoulzadegan M, Gaudray P, Canning M, Trejo-Avila L and Cuzin F. . 1981 Virology 114: 489–500.

  26. Rich KA, Bardin CW, Gunsalus GL and Mather JP. . 1983 Endocrinology 113: 2284–2293.

  27. Skinner MK. . 1993 The Sertoli Cell. Russell LD, Griswold MD (eds). Cache River Press: Clearwater FL pp. 237–247.

    Google Scholar 

  28. Su JG and Hsueh AJ. . 1992 Biochem. Biophys. Res. Commun. 186: 293–300.

  29. Tjian R. . 1978 Cell 13: 165–179.

  30. van Dissel-Emiliani FM, Grootenhuis AJ, de Jong FH and de Rooij DG. . 1989 Endocrinology 125: 1899–1903.

  31. Vincent S, Marty L and Fort P. . 1993 Nucl. Acids Res. 21: 1498.

  32. Vincent S, Segretain D, Nishikawa S, Nishikawa S, Sage J, Cuzin F and Rassoulzadegan M. . 1998 Development 125: 4585–4593.

  33. Zhu Z, Veldman GM, Cowie A, Carr A, Schaffhausen B and Kamen R. . 1984 J. Virol. 51: 170–180.

Download references


We thank C Boitani for the generous gift of recombinant activin, F Ranc for help in cell culture experiments and L Martin for help in setting up the RNase protection assay. The expert technical assistance of M Cutajar and Y Fantéi is gratefully acknowledged. Recombinant human hormones were kindly provided by the National Hormone and Pituitary Program of the National Institute of Diabetes and Digestive and Kidney Diseases. This work was supported by grants from Association pour la Recherche sur le Cancer (France).

Author information



Corresponding author

Correspondence to François Cuzin.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Lopez, P., Vidal, F., Rassoulzadegan, M. et al. A role of inhibin as a tumor suppressor in Sertoli cells: down-regulation upon aging and repression by a viral oncogene. Oncogene 18, 7303–7309 (1999).

Download citation


  • testis
  • gonadal tumor
  • inhibin-activin
  • polyoma virus
  • T antigen
  • gene regulation

Further reading


Quick links