Expression of the proto-oncogene fos (c-fos) by preimplantation blastocysts of the pig

The cellular proliferation and regulation which occur during the process of mammalian implantation have led to studies investigating the expression of proto-oncogenes during this period (Adamson, 1987a; Stewart et al. 1988). The expression of N-myc, an oncogene associated with neuroblastomas, is enhanced during murine embryogenesis, and decreases towards term (Jakobovits et al. 1985). C-myc, sis and ras have also been reported in human first-trimester villous trophoblast, but disappear later in gestation (Sarkar et al. 1986) and may be related to growth factor receptor activity (Goustin et al. 1985). The preimplantation period in the pig is a time of extensive cellular remodelling during which a blastocyst of about 1 cm in diameter extends to a filamentous thread up to 1·5 m in length within about 48 h (Geisert et al. 1982). Recent work has shown that transcriptional activation of the c-fos gene is linked to membrane depolarization and indicate that c-fos may play a central role in signal transduction by coupling early events associated with receptor occupancy to activation of other cellular genes (Sassone- Corsi et al. 1988). Activation of the c-fos gene may therefore have a critical function in porcine implantation and we decided to examine the expression of this proto-oncogene and the regulation of the fos protein around the time of implantation in this species.

Large white Devana gilts and sows were used at known gestational dates (day following service = day 1 of pregnancy) and the reproductive tract was removed at the abattoir. Animals were used at days 10(2), 11(2), 12(1), 13(4), 14(2), 15(4), 16(4), 17(6), 18(2), 19(2), 20(2) and 33(1) of gestation, the figures in brackets representing the number of animals used. Tissue was transported to the laboratory on ice and immediately flushed towards the cervical end with Hank’s Balanced Salts Solution (HBSS) (see Whyte et al. 1984 for details). Tissue was either snap-frozen in isopentane cooled in liquid nitrogen, fixed in paraformaldehyde/glutaraldehyde (see immunohistology), used for RNA preparation (see Northern blot), or cultured according to previous methods (Whyte et al. 1987).

For the construction of the fos oncogene probe, a Ikb fragment was isolated from proviral FBI murine osteosarcoma fos DNA (1·3 kb) (Curran et al. 1982) and subcloned into the Pstl site of the expression vector Bluescribe M13⁺ (Vector Cloning Systems, San Diego, USA). The orientation of the v-fos insert in Bluescribe was determined by restriction mapping with Sall and PvuII, thus enabling the synthesis of sense and antisense RNA transcripts from the T3 and T7 promoters, respectively. The structure of the /os/Bluescribe plasmid with restriction and promoter sites is shown in Fig. 1. The transcripts were labelled with α[³²P]UTP (uridine triphosphate, 800 Ci mmol⁻¹, Amersham) and unincorporated label was removed by ammonium acetate/ethanol precipitation. The specific activity of the resulting probes was between 1–5×10⁸ disints min⁻¹μg⁻¹.

Total RNA was prepared from pig blastocysts immediately after flushing from the uterus with ice-cold HBSS by the method of Chirgwin et al. (1979).

Samples of total RNA (usually 20 μg per track) were separated on 1% agarose gels containing MOPS buffer (0·02 m-3-(N-morpholino)propane sulphonic acid, 0-005 m-sodium acetate pH7·0, and 0·001 m-sodium EDTA), and 6% formaldehyde. The separated RNA was transferred to Hybond-N membranes (Amersham) by blotting overnight in 3 m-sodium chloride and 0 3 M-trisodium citrate (20×SSC). Hybridization was carried out at 65°C overnight in 50% deionized formamide, 0-5% sodium dodecyl sulphate (SDS), 6 × SSC, 5 × Denhardt’s (100 × Denhardt’s = 2% BSA, 2% Ficoll 400, 2% polyvinylpyrrolidine, PVP), 100 μg ml⁻¹ denatured salmon sperm deoxyribonucleic acid (DNA), and antisense fos probe at >5 ng ml⁻¹. Washing was at 65°C to a stringency of 0·1 × SSC, 0·1% SDS. The final wash was in 0·5 × SSC and 1 μg ml⁻¹ ribonuclease (RNase) A for 15 min at 37°C.

Total RNA (50 μg per sample) was denatured, diluted 1:1 with 20 × SSC, and applied to a nitrocellulose membrane equilibrated with 10 × SSC. The nitrocellulose was air-dried and fixed at 80°C under vacuum. The purified Ikb fragment of v-fos was ³²P-labelled by random priming with Klenow extension (Feinberg & Vogelstein, 1983). Hybridization was by a method modified from Taylor et al. (1984) by the exclusion of formamide from the hybridization and the use of 6% polyethylene glycol instead of dextran sulphate. Hybridization was carried out at 65 °C. Washing was at 65 °C to a stringency of 2 x SSC, 0-1% SDS. The blot was reprobed with an actin cDNA to determine the relative amounts of total RNA loaded (this assumed constant levels of actin mRNA).

Cryostat sections of tissues were prepared at 4–6 /rm thickness or tissue was fixed in 5% paraformaldehyde/1% glutaraldehyde/0·1 m-phosphate buffer pH 7·2 for 30 min at +4°C. The latter tissue samples were embedded in low- melting-point wax and sectioned at 5 μm. Preliminary experiments showed that fixation and histological processing gave hybridization signals equally as effective as those using cryostat material and because of its superior histological preservation, the former method was routinely adopted. The fixed tissue samples were cleared, after sectioning, through xylene and dehydrated through propan-2-ol into cold (4°C) PBS. Various modifications were attempted in order to optimize the results obtained from in situ hybridization. Proteinase K treatment (1 μg ml⁻¹, Sigma, 0·1 m-Tris-HCl buffer pH8·0, 50mm-ethylenediaminetetracetic acid, EDTA, at 37°C for 30 min) did not appear to affect subsequent accessibility of the hybridization probe and was therefore discontinued. Two treatments, however, did appear to affect the quality of results obtained from in situ hybridization. To reduce background in the autoradiographs, sections were immersed in freshly prepared 0·25% acetic anhydride in triethanolamine (0·1 m, pH8·0) for 10min. Prehybridization was performed in 50% formamide/2 × SSC for 15 min at 37 °C. Prehydridization in a different medium to this (2h in buffer containing 50% formamide, 0·1% bovine serum albumin (BSA), 0·1% Ficoll 400, 0·1% PVP 360, 50mm- sodium phosphate, pH 7·0, 5 × SSC, 0·5% SDS, 5 mm-EDTA and 1 mg ml⁻¹ denatured salmon sperm DNA) did not appear to offer any qualitative advantages and was omitted from further hybridizations.

To each section 10 μl of preheated (37 °C) hybridization buffer (50% formamide, 2 × SSC, 10% dextran sulphate, 0·25% BSA, 0·25% Ficoll 400, 0·25% PVP 360, 250 him- Tris-HCl pH 7·5, 0·5% sodium pyrophosphate, 0·5% SDS, 250 μg ml⁻¹ denatured salmon sperm DNA) and the antisense or sense RNA probe (60 ng ml⁻¹ at about 5×10⁷disints min⁻¹μg⁻¹ per 10 μl buffer) was added. Tissue sections were covered with dimethyldichlorosilane (Sigma)-coated coverslips and incubated in a moist chamber (as hybridization buffer but without DNA) at 43°C for 16h. Coverslips were removed in wash buffer (4 × SSC, 50% formamide) and then given a wash in this medium at 50°C for one hour followed by two washes at 37°C for 1h in 4 × SSC alone. The sections were then treated with 20 μg ml⁻¹ RNase A (Sigma, boiled prior to use) in 0·5 NTE (0·5 m-NaCl, 10mm-Tris-HCl pH8·0, Iitim-EDTA) for 30 min at 37 °C to remove unhybridized single-stranded RNA. Sections were washed in 2 × SSC (30 min at 37 °C) followed by a wash in 01 x SSC for the same time. Sections were dehydrated through an alcohol series containing 0-3M-ammonium acetate (pH8-0). Sections were dried, counterstained with haematoxylin/eosin, dipped into emulsion (Kodak NTB-2) and exposed for 5–10 days at 4°C before developing and mounting. Sense fos RNA probes were used throughout as controls and did not hybridize to any of the tissue sections.

A peptide (CKEKEKLEFIL) was synthesized on a Biosearch 9500 peptide synthesizer (New Brunswick Scientific, Herts, UK) using solid-phase t-Boc chemistry. This represents a conserved sequence of c-fos and y-fos (amino acids 198 to 207) with a terminal cysteine to act as linker. The peptide was conjugated to PPD (Statens Seruminstitut, Copenhagen) using Sulpho-SMCC (Pierce Chemical Co., USA). The conjugate solution contained l mg ml⁻¹ PPD and l mg ml⁻¹ of the peptide in lOmm-sodium phosphate buffer (pH 7·0).

Dutch rabbits (approx 2·5 kg) were bled to provide pre- immune serum before being injected with 100 μl BCG i.d. (Statens Seruminstitut) in each of two sites. One month later the rabbits were immunized with PPD-peptide in IFA i.m. ands.c. as detailed by Lachmann et al. (1986).

For identification of /os-coded oncoprotein, tissue was used either as cryostat sections or after cold methanol fixation (see Whyte & Allen, 1985). The primary antiserum raised in rabbits to the synthetic peptide sequence of c-fos was used at dilutions of 1·5, 1·0, 0·5 and 0·1 μg ml⁻¹ PBS. The second- stage antibody was either fluorescein isothiocyanate-(FTTC) or rhodamine (TRITC)-conjugated anti-rabbit immunoglobulin raised in pig (Nordic, used at 1:50 and 1:100 dilutions in PBS). Simultaneous double-label fluorescence of troph- ectoderm with mouse monoclonal SN1/38 antibody (Whyte et al. 1984) was used with FITC and TRITC second-stage antibodies in order to identify trophectoderm in conjunction with fos oncoprotein.

The rabbit anti-/os peptide antibody raised to the synthetic peptide was tested by its ability to immunoprecipitate the fos oncoprotein. Pig blastocysts (14 days post coitum) were pulse- labelled with ‘⁵S-labelled methionine (1-255 Ci mmol⁻¹, Amersham) at 0·3 to 1·0 mCi ml⁻¹ for 15 min. The methods of lysis and immunoprecipitation followed those of Curran et al. (1984). The conditions of electrophoresis in sodium dodecyl sulphate and subsequent autoradiography are detailed in Whyte et al. (1985).

Cells derived from pig blastocysts were cultured according to previously published methods (Whyte et al. 1987). Conditioned medium from 2-hour-old cultures were added at 30% v/v to cells incubated for various time periods (Fig. 4) according to the methods detailed in Müller et al. (1986). Endometrial tissue was also obtained from the pigs when the uteri were opened, and conditioned medium from cultures of these was used at 30% v/v as a control for the blastocyst- conditioned medium. Neither this, nor the addition of nonconditioned medium, caused any prolongation of the c-fos protein indicating the specificity of blastocyst conditioned medium in maintaining fos product expression.

Major proteins were found at of 55 000 (fos) and 40000 (Fig. 2A) which is in agreement with published data for the fos protein and its associated molecules (Franza et al. 1987).

In agreement with data obtained for mouse and human extraembryonic tissues (Millier et al. 1983), the size of the fos transcript from porcine extraembryonic membranes was found to be 2·3kb (Fig. 2B). The levels of mRNA for the c-fos proto-oncogene were determined relative to actin mRNA levels from slot blots (Fig. 2C). The amount of mRNA appeared to decrease progressively to day 16 of gestation, but increased again by day 17. However, the majority of the fos message was associated with allantois by this stage of pregnancy. Tissues from four separate animals were analysed by slot blot.

The results of in situ hybridization of antisense fos probe to trophectoderm at day 14 of gestation showed that the fos message occurred over the majority of trophectodermal cells at this stage of pregnancy, but by day 19 only isolated trophectodermal cells showed hybridization (Fig. 3C). In contrast, the allantois continued to hybridize the fos probe over its entire epithelium right up to day 19 of pregnancy (Fig. 3D).

Indirect immunofluorescence confirmed the distribution of fos as revealed using in situ hybridization. Thus the trophectoderm was reactive over its entire epithelium up to about the 14th day of gestation (Fig. 3A), but only a proportion of trophectodermal cells remained reactive by day 19 of pregnancy (Fig. 3B). The epithelia used in the immunolocalization studies and in the in situ hybridizations were identified as trophectodermal in origin by use of the mAb SN1/38 (Whyte et al. 1984) and this was confirmed by doubleantibody immunofluorescence studies (Fig. 3A). Allantoic epithelium retained reactivity with the anti-fos antisera in all its cells at least up to day 19 of pregnancy, confirming the results obtained by in situ hybridization with the fos RNA probe (Fig. 3D).

When cells derived from pig blastocysts were main-tained in short-term culture (Whyte et al. 1987), their immunoreactivity with the anti-fos antisera disappeared over a period of time so that the cells were virtually nonreactive after 7h in culture (Fig. 4A,B) and completely unreactive after 24 h in culture (Fig. 4C). If conditioned medium from 2-hour cultures of pig blastocysts was added to a total of 30% v/v of the incubation medium, however, then c-fos protein could still be detected by indirect immunofluorescence 24 h after initial seeding (Fig. 4D,E and F). Double-label studies using mAb SN1/38 confirmed that the cells reactive with anti-fos antisera in culture were predominantly derived from the trophectoderm.

Preabsorption of the rabbit antiserum with the synthetic fos peptide (CKEKEKLEFIL) completely blocked immunolocalization of the product whereas preabsorption with an unrelated proto-oncogenic peptide (CTPSTQLHTGGLAVA, the C-terminal region of the int-2 peptide) had no effect on the anti-fos- antiserum binding. In addition, the preimmune sera also gave negative results, indicating the specificity of the anti-fos antiserum for the synthesized peptide.

Preliminary results using antisera constructed to synthetic peptide sequences of c-myc revealed that, although this oncoprotein was present in pig trophectoderm, its pattern of expression was variable and remained unaffected by addition of conditioned medium from 2 h cultures (data not shown).

A feature of c-fos expression is that although transcripts are present at low levels in embryos and fetuses, the extraembryonic membranes have high levels of this proto-oncogene. The pig is similar, therefore, to the other species in which this has been described, namely mouse and man (Müller et al. 1983; Adamson, 1987a). The expression of the fos gene is regulated by at least three intracellular messenger systems: the Ca²⁺/ phospholipid-dependent protein kinase C; c-AMP; and Ca²⁺/calmodulin; and published data by several groups suggest a general role of fos protein in the mechanisms of action of hormones and neurotransmitters (Morgan et al. 1987).

The results presented indicate that pig blastocysts produce fos protein of the characteristic 55 000 M_r type (Fig. 2A), and that this is associated with other cellular products, particularly p39 (Adamson et al. 1985). The fos product, although expressed initially at high levels in trophectoderm, appears to be confined mainly to allantois at later stages of gestation. The dramatic increase in c-fos in murine extraembryonic membranes has been shown to be largely confined to the amnion (Mason et al. 1985). The fos gene is known to be activated in growth and differentiation processes (Adamson, 1987b) and preimplantation blastocyst development in the pig shows extensive growth and differentiation, developing from a spherical blastocyst about 1cm in diameter on day 10 to a filamentous thread over one metre long by day 14 (Geisert et al. 1982). Induction of cellular proliferation of a variety of cell types with EGF, PDGF or FGF (Adamson, 1987b) is accompanied by the rapid induction of the fos gene and its expression in amnion cells in vitro can be prolonged by addition of placenta-conditioned medium (Müller et al. 1986). We have also been able to extend the expression of the c-fos gene product in pig trophectodermal cells in vitro by addition of 30% v/v conditioned medium from short-term blastocyst cultures known to contain trophectodermal cells (Fig. 4). The addition of conditioned medium from other cell cultures (pig endometrial cells) did not prolong c-fos oncoprotein expression in cultures of pig blastocyst material. Most inducers of c-fos reported so far are related to growth factors or mitogens, and our results would suggest autocrine regulation of the expression of this proto-oncogene in pig extraembryonic tissues such as has been reported for sis and myc proto-oncogenes in human trophoblast (Goustin et al. 1985).

Preliminary results indicate that uterine flushings from pregnant pigs can have a similar, though less marked, effect suggesting the presence of a growth factor or factors in the uterus of the pig. Flushings from non-pregnant animals did not prolong c-fos expression. Factors related to the acidic and basic forms of fibroblast growth factor have been identified in uterine tissue and uterine flushings of pigs during the preimplantation period (D. R. Brigstock and K. D. Brown, personal communication). Transient transcriptional activation of the c-fos gene following serum stimulation of susceptible cell types requires a conserved DNA element located 300 bp 5’ to the mRNA cap site (Treisman, 1986). This serum-induced transcription of fos protooncogene has recently been shown to be under negative feedback regulation (trans-repression) by the fos protein itself, and that the serum-responsive promoter element is associated with a nucleoprotein complex that contains the product of c-fos and the product of another proto-oncogene c-jun, the transcriptional factor AP-1 (for a review see Sassone-Corsi et al. 1988).

The c-fos proto-oncogene may also be related to the expression of MHC determinants by the preimplantation blastocyst. Such antigens are known to appear on the surface of the mouse blastocyst but disappear from the trophectodermal surface when the blastocyst is hormonally activated for implantation (Billington & Bell, 1983). We do not know for certain whether the preimplantation blastocyst of the pig expresses MHC determinants, but there may an association between our observation that the level of fos expression declines during trophectodermal development in the pig and the recent work of Kushtai et al. (1988) showing that activation of c-fos causes expression of MHC antigens in murine Lewis lung carcinoma clones. The c-fos proto-oncogene may not only play a role in the growth and differentiation of the pig blastocyst, but may also be important in immunological exchange between mother and embryo.

We thank Mr N. Huskisson for synthesizing the peptide, Mr I. King for preparing the cryostat sections used in this study, Dr S. J. Humphries for the actin cDNA probe, Drs R. B. Heap and J. C. Pascall for helpful comments, and Mrs J. Hood for typing the manuscript.

 
• BCG

  Bacillus Calmette-Guérin

 
• BSA

  bovine serum albumin

 
• DNA

  deoxyribonucleic acid

 
• EDTA

  ethylenediaminetetracetic acid

 
• EGF

  epidermal growth factor

 
• FGF

  fibroblast growth factor

 
• Fl’l’C

  fluorescein isothiocyanate

 
• HBSS

  Hank’s balanced salts solution

 
• i.d.

  intradermal

 
• kb

  kilobases

 
• MHC

  major histocompatibility complex

 
• IFA

  incomplete Freund’s adjuvant

 
• mAb

  monoclonal antibody

 
• MOPS

  3-(7V- morpholinojpropane sulphonic acid

 
• M_r

  relative molecular mass

 
• NTE

  Na/Tris/EDTA buffer

 
• PPD

  purified protein derivative

 
• PDGF

  platelet-derived growth factor

 
• PVP

  polyvinylpyrrolidine

 
• RNase

  ribonuclease

 
• RNA

  ribonucleic acid

 
• s.c.

  subcutaneous

 
• SDS

  sodium dodecyl sulphate

 
• SSC

  sodium chloride/trisodium citrate

 
• Sulpho-SMCC

  sulphosuccinimidyl–4-(N-maleimidomethyl)cyclohexane-l-carboxylate

 
• TRITC

  tetramethylrhodamine isothiocyanate

 
• UTP

  uridine triphosphate.