Biochimie 70 (1988) 895- 899 (~) Soci6t6 de Chimie biologique / Elsevier, Paris

895

Proto-oncogenes and embryonic development

Marcel MI~CHALI, Michel GUSSE, Sophie VRIZ, Michael TAYLOR*, Yannick ANDI~OL, Jacques MOREAU, Jacques HOURDRY, Michel LEIBOVICI, Annie BRULFERT, Genevieve ALMOUZNI and Sylvie MOUSSERON-GRALL

Institut Jacques Monod, Tour 43, 2, place Jussieu, 75251 Paris Cedex 05, France

(Received 11-3-1988, accepted 24-3-1988)

Summary - - The role of proto-oncogenes in embryonic development was investigated using one of the most characterized vertebrates, the amphibian Xenopus laevis. Genes which belong to the major proto-oncogene families have been detected in Xenopus genome. The developmental control of the myc gene was assayed using a characterized Xenopus myc probe and specific antibodies. The myc gene is highly expressed as a stable maternal mRNA in oocyte, and an unfertilized egg contains 5 × !05-fold the myc RNA content of a proliferative somatic cell. The myc RNA store is evenly distributed in the oocyte and the egg. Fertilization triggers a post-transcriptional control of the gene and the RNA store is progressively degraded to a constitutive value of 10 to 30 myc RNA copies registered per gastrula embryonic cell. The 62K myc protein is accumulated late in oogenesis. This uncoupling of myc expres- sion and cell proliferation appears as a specific developmental regulation of the myc gene, adapted to the series of rapid cell cleavages occurring after fertilization.

proto-oncogenes / myc / Xenopus laevis / development

Introduction

Genes involved in cancerogenesis have been identified these last years in the vertebrate genome. They belong to the oncogene family, by analogy with their essential role in the neoplastic process [1-3]. This purely experimental defini- tion can be confusing as it relies on an uncon- trolled or abnormal expression of these genes. Their role in normal cells is still largely unknown but, by analogy with their functions in a cancer cell, they could be related to cell proliferation and differentiation. A developing embryo is char- acterized by three overlapping programs: pro- liferation, migration and differentiation. These three programs are clearly expressed, but in an abnormal fashion in a neoplastic cell. Simil- arities have often been observed between a neo- plastic cell and an embryonic cell at both the cellul- ar and the molecular level, and could reflect common gene functions used in these two cell

types. Thus, involvement of p. roto-oncogenes in embryonic development Is a reasonable hypothesis and we are testing it through the development of one of the most characterized vertebrates, the amphibian Xenopus laevis. During early development the fertilized egg cleaves to form many daughter cells. This is the cleavage stage, common to all vertebrate deve- lopmen,. After fertilization the egg starts a series of twelve synchronous cell divisions every 35 min [4]. Thus, during that period, Xenopus embryonic cells divide more rapidly than a bac- terium. Interestingly, these rapid cell cycles, although highly dependent on protein synthesis, do not require gene expression [4, 5]. Gene func- tions required for early development are ex- pressed during the formation'of the egg, and the mRNA encoding for these functions belong to the Class of maternal mRNAs [6].

Uncontrolled cell division is probably the most dramatic event linked to the neoplastic

*Present address: C.R.C. Molecular Embryology Group, Department of Zoologie, Downing Street, Cambridge, CB & El, U.K.

896 M. MEchali et al.

phenotype. Gene functions involved in the un- limited growth potential of a cancer cell are largely unknown but immortalizing proto-oncogenes are good candidates for such a role [1-3]. We postulated that such gene products should be highly active during early embryonic develop- ment and therefore should be highly expressed during oogenesis, before fertilization. We show that the myc proto-oncogene belongs to these early expressed genes.

Materials and methods

Xenopus oocytes and embryos Xenopus oocytes were removed from anesthetized

females and classified under the microscope accord- ing to the 6 stages described by Dumont [7]. Embryos were harvested at different developmental stages and processed as previously described for RNA analysis [81.

In situ hybridizations Oocytes and embryos were fixed in 4% paraformal- dehyde and embedded in ester-wax following slight modifications of a method previously described by Petavy [9]. 7-/zm cut sections were hybridized with randomly primed 355 probes (3 x 10 s cpm//zg) according to Feinberg and Vogelstein [10]. A HincII- SacI fragment of the myc RNA coding sequence [8] was used as the template probe.

Northern blot analysis RNA from oocytes and embryos was extracted and

10"/

106

l°S t a .

~ 10 4 g f,n C q L_

u 10 3 E

102

10

Oogenesis Embryonic Development

' I I 6 12 18

Months

~(>. -~Fei r t i I izat. ion

-I '1 1

I i-

(

,- oMBT N

', H S T F T : ~ o o-4 I.,-o

:" G . . . . . . - ~ , ~ - ~ - . - i , - ~ I I I I I I I ! - - / ' ' -

24 0 i0 20 30 40 50 60 70 288 Hours p o s t - f e r t i l i z a t i o n

Fig. 1. Variation of the amount of myc RNA molecules per cell during oogenesis and embryonic development. Total RNA from each developmental stage was analyzed by Northern blot hybridization. The experiment was internally controlled both by the total RNA extracted and the level of 18S and 28S y tbosomal RNA, and the 5S RNA exhibit published values. The quantitation was also done using a calibrated dot-blot hybridization in a separate experiment. Finally the values found were in agreement with an independent estimate of myc recombinants in the oocyte library ([8] for details). Ttne mean number of myc RNA mole- cules per cell is indicated as a function of the oocyte growth or embryonic development.

Proto-oncogenes and embryonic development 897

analysed by Northern blot as previously described [8].

Results

All classes o f oncogenes are present in the Xenopus genome

By using controlled hybridization conditions and known oncogene probes we have been able to detect different proto-oncogenes in the genome of Xenopus (Moreau et al., submitted). In most cases these genes were found in duplicate, in accordance with the hypothesis of total genome duplic~,tion which occurred thirty million years ago in Xenopus [11]. Members of the proto- oncogene kinase family, nuclear oncogene family, receptor-like family have been detected. Oncogenes encoding for guanine nucleotide binding-like proteins (ras) appear the most conserved. This study confirmed the already reported high level of conservation of proto- oncogenes during evolution, although-limited here to the vertebrate phylum.

2°7-

1.8-

! ÷

~ e m

• i

The immortalizing proto-oncogene c-myc is expressed during oogenesis

By using Northern blot hybridizations and a heterologous human c-myc probe, we have been able to detect the c-myc gene in the poiy(A) ÷ fraction of oocyte RNA. The Xenopus cDNA has been characterized and has been found to be 65% homologous with the human c-myc coding sequence. An extensive analysis of its expression has been carried out during both early oogenesis and embryonic development [8], and the overall results are summarized in Fig. 1. The myc gene RNA is accumulated at a very high level as soon as in early oogenesis, and reaches a maximum level during mid-oogenesis. This high level is maintained constant during the last 12 months of oogenesis. A mature oocyte or an unfertilized egg contains 5 x 106 myc RNA molecules, most of them present as poly(A) + RNA (Fig. 2). We compared this value with other proliferative cell types and found that a somatic growing cell in culture or a preleukemic cell contains 10 to 30 myc RNA copies ([8] and our unpublished results). Therefore an egg contains 5 x 105 fold the amount of myc RNA found in a proliferative somatic cell. After fertilization this exceptional myc RNA level decreases until gastrula, and the myc RNA level per cell is then similar to the

Fig. 2. The oocyte myc RNA population is polyadenylated. Total oocyte RNA was separated in a poly(A) • and a poly(A)- RNA fraction by chromatography on poly(U)- Sepharose. 10/~g RNA was fractionated on formaldehyde agarose gel and the Northern blot was probed with a Xeno- pus myc probe as described [8].

level found in a proliferative cell. This decrease is due to two additional causes: first, fertilization triggers a post-transcriptional regulation of the myc gene so that 90% of the RNA store of the egg becomes progressively degraded in 9 h. Second, during the same stage the remaining pool of myc RNA molecule is distributed among the daughter cells of the cleaving embryo [8]. From the gastrula-neurula stage the myc RNA level per embryonic cell reaches the constitutive value found in proliferative somatic cells and this value is maintained during further embryonic development probably by riew transcription starting at the neurula stage.

The myc protein is stored during late ooge- nesis

An analysis of the myc protein level was carried out using c-myc antibodies provided by G. Evan

898 M. M6chali et al.

1 D

total RNA population as the vegetal pole of the oocyte contains a high proportion of yolk. A similar result was observed with the histone H4 mRNA, a known unlocalized mRNA [13].

- 9 7 K d

-68Kd

- 5 5 K d

Fig. 3. The nlyc protein is synthesized in oocytes. Lane 1: Oocvtes were labeled 2(1 h with 2 mCi /ml 35S met and the °

extracted proteins were immunoprecipitated with a c-myc antibody [19 l. Lane 2: The antisera were first preincubated with the myc peptide used as an antigen• Immunoprecipitat- ed proteins were analyzed on an SDS 10% polyacrylamide gel.

(Ludwig Institute, MRC Cambridge). All anti- bodies used recognized a 62K protein (Fig. 3) which is the one encoded by the Xenopus myc RNA, as found by an ybrid selection exper- iment. This protein is accumulated only in late oogenesis and its amount reaches a maximum level in mature oocytes ([8] and Gusse et al., sub- mitted). Thus the rnyc gene is expressed both at the RNA level ant~ a~ the protein level during oogenesis.

The myc R N A is not a locafized determinant

The localization of maternal determin.~nts in the unfertilized eggs has been suggested as an attrac- tive explanation to the future determinations of the early embryo ([6] for a review). As myc RNA clearly belongs to the maternal mRNA family we assayed for its localization in the oocyte. In situ hybridizations were performed on oocyte cut sections and showed that myc RNA was evenly distributed in the cytoplasm of oocy- tes (Fig. 4A). Fig. 4B shows a Northern blot ana- lysis of animal and vegetal parts of the oocyte separated by cutting the oocyte in two parts [ 12]. The myc RNA appears less concentrated in the vegetal pole. However this result does not reflect a specific localization of myc RNA among the

Discussion

Members of the different oncogene families have been detected in the amphibian Xenopus genome. We assayed for the developmental con- trol of the c-myc proto-oncogene. C-myc has been the focus of numerous studies and its expression was found to be modified in the major mechanisms leading to cell transforma- tion, including proviral insertion, chromosomal translocation, and gene amplification [14]. Two of its most characteristic features are the tight coupling between the level of its expression and the rate of cell proliferation, as well as the short half-life of its mRNA product [15, 16]. As an increase in the level of c-myc RNA was observed in nearly all cell types entering into the cell cycle, it has been suggested that c-myc expression was necessary for cell proliferaticm to occur. Our study shows an unusually high level of c.myc

B 1 2

- , - . ~ . ~ . . . . ~ .

¢f,r _<

~ - 2 . 7 K b

Fig. 4. The oocyte myc RNA is not a localized maternal determinant. A) An in situ hybridization was performed on an oocyte stage 1 [7], using a randomly primed myc DNA probe (Materials and methods). B) Oocytes were frozen in dry ice and sectioned to cut off the vegetal (lane 1) or animal part (lane 2) as described by Rebagliati et al. [12].

Proto-oncogenes and embryonic development 899

expression during oogenesis, therefore in non- dividing cells. An unfertilized egg contains 5 x 106 myc RNA molecules, thus 105 times the amount of myc RNA found in a proliferative somatic cell either normal or transformed. A similar level of activity was also found by King etal. [17] for the myc gene and Soussi et al. [18] for another immortalizing oncogene, p53. Assuming a maximal rate of RNA synthesis and a maximum loading of RNA polymerases on the c-myc transcription units, it can be calculated that at least 280 days are necessary to reach this level [8]. Thus this store of myc mRNA and the maintenance of this store during oogenesis imply an unusual stability of the myc RNA transcribed during oogenesis. Overall this study shows an uncoupling of myc RNA synthesis and cell proli- feration during the formation of the egg. This uncoupling is also detected at the level of protein synthesis as the myc protein is synthesized during oogenesis. After fertilization a post- transcriptional regulation of the gene occurs, with a progressive degradation of 90% of the myc RNA store during the cleavage stage of the embryo. The remaining pool of rnyc RNA is dis- tributed among the daughter cells of the embryos such that at gastrula a constitutive value of 10 to 30 copies of myc RNA per embryonic cell is registered. The Xenopus early embryonic development is characterized by an active cell proliferation producing the 8000 cells of a blastula in 9 h, without transcription of new mRNAs [4]. According to our hypothesis the egg accumulates an unusual amount of myc RNA to support the active cell proliferation which will be induced by fertilization. This uncoupling of myc gene expression and cell pro- liferation appears therefore as a specific deve- lopmental control of the myc gene, adapted to the early embryonic development. This study also shows that myc expressiop per se is not suf- ficient to trigger cell proliteration but may be considered as a sign of a cell committed to a pro- liferative pathway.

Acknowledgments

This work was supported by the Association pour la Recherche sur le Cancer and CNRS Grants ATP No. 960110 aild 960097. M.V.T. was supported by an EMBO Long Term Fellowship.

References

1 Bishop J. M. (1986) Trends Genet. 1,245-249 2 Bishop J. M. (1987) in: The Molecular Genetics

o f Cancer. Sciences 235,305- 311 3 Varmus H. E. (1984) Annu. Rev. Genet. 18,

553-612 4 Newport J. & Kirschner M. (1982) Cell 30,

675- 685 5 Brachet T. J. & Denis H. (1963) Nature 198,

205-206 6 Davidson E. H. (1987) in: Gene Activity in Early

Development. Academic Press, N.Y. 7 Dumont J.N. (1972) J. Morphol 136,

1653-1654 8 Taylor M. V., Gusse M., Evan G. I., Dathan N.

& Mechali M. (1986) EMBO J. 5, 3563-3570 9 Petavy G. (1985) Stain Technol. 60,321-330

10 Feinberg A. P. & Vogelstein B. (!983) Anal. Biochem. 132, 6-13

11 Bisbee C.A. , Baker M.A., Wilson A.C. , Hadji-Azimi I. & Fishberg M. (1977) Science 195, 785-787

12 Rebagliati M. R., Weeks D. L., Harvey R. P. & Melton D. A. (1985) Cell 42, 769-777

13 Weeks O. L. & Melton D. A. (1987) Cell 51, 861-867

14 Cole M.D. (1986) Annu. Roy. Genet. 20, 361-384

15 Kelly K., Cochran B. H., Stiles C. & Leder P. (1983) Cell 35,603-610

16 Dani ~'., Blanchard J. M., Piechaczyk M., El Sabouty S., Marty L. & Jeanteur P. (1984) Proc. Natl. Acad. Sci. USA 81, 7046-7050

17 King M. W., Roberts J. M. & Eisenmman R. N. (1986) Mol. Cell. Biol. 6, 4499-4508

18 Soussi T., Caron de Fromentel C., M~chali M., May P. & Kress M. (1987) Oncogene 1, 71-78

19 Evan G. I. & Hancock D. C. (1985) Cell 43, 253-261