The angle of the dorsoventral axis with respect to the anteroposterior axis in the Drosophila embryo is controlled by the distribution of gurken mRNA in the oocyte.

Like most organisms, Drosophila embryos develop in relation to two orthogonal axes, the anteroposterior and dorsoventral. These two axes are established by four independent systems of maternal information. Mutations in any system disrupt either the anteroposterior or the dorsoventral patterning of the embryo but never affect the orthogonal orientation of the axes relative to each other. Here we show by analyzing their embryonic fate map, that K10 embryos still possess a dorsoventral polarity. However, instead of forming a 90 degrees angle, the dorsoventral and the anteroposterior axes lie parallel to each other. This axis misorientation was partially corrected by decreasing the wild-type grk gene copy number such that the embryos issued from K10/K10; grk/+ females showed a variability in their fate map which could be interpreted as a progressive rotation of dorsoventral axis relative to the unmodified anteroposterior axis. This rotation is maximal in the K10 embryos where it reaches 90 degrees and results in the congruence of the two axes. The alteration of the embryonic fate map could be traced back to oogenesis where it was shown to correlate with the mislocalization of the grk transcripts.

A regulatory function for K10 in the establishment of dorsoventral polarity in the Drosophila egg and embryo.

Several lines of evidence suggest that the origin of pattern formation of Drosophila embryos must be traced back to oogenesis, to the polarity of the egg chamber. A few early-acting genes, K10, top, grk and cni, have been identified which are assumed to function in a signal transduction process between the germline oocyte and the somatic follicle cells, during which the egg chamber acquires a dorsoventral polarity. K10 has been cloned and was shown to encode a putative transcription factor specifically acting in the oocyte nucleus. In order to characterize further the function of K10, we have analyzed its genetic interactions with grk, top and cni. We show that grk behaves as a dominant partial suppressor of K10. Analysis of the rescuing process of the K10 phenotype by grk shows that: (1) K10 is not indispensable for the establishment of dorsoventral polarity of the egg chamber, since its lack of function can be compensated for by reducing the grk wild-type copy number; (2) grk function is highly dose-sensitive; (3) the rescue process shows an anteroposterior effect suggesting that K10 may also interact with genes involved in anteroposterior pattern formation. These results are compatible with a model in which grk is a dorsalizing signal emanating from the oocyte nucleus, whose level of expression is regulated negatively by the K10 product.

[Determination of the dorso-ventral polarity of the Drosophila embryo]. / Déterminisme de la polarité dorso-ventrale de l'embryon de drosophile.

Embryonic pattern formation has been studied extensively in many organisms. In Drosophila, the powerful combination of genetics cytoplasm transplantation experiments, as well as recent molecular data, have helped to elucidate the mechanisms responsible for the establishment of embryonic polarity. A small number of genes, most of them maternally expressed, are involved in this process and participate in four independent systems--three for the antero-posterior axis (A/P) and one for dorsoventral axis (D/V)--which define various embryonic territories by specifically localized cues. This review concerns the definition of the dorsoventral polarity responsible for the establishment of the germ layers of the embryo. Dorsoventral development is regulated by a single group of maternally expressed genes: the "dorsal group" of genes. It includes 11 genes, the loss of function of any of which results in a dorsalized development, whereas mutation of the 12th gene, cactus, results in a ventralized development. These genes are arranged according to a functional hierarchy, and have been shown to cooperate in the formation of a graded nuclear concentration of the dorsal gene product. The dorsal product corresponds to the dorsoventral morphogen and is homologous to the transcription factor NF-kappa B. Among the 11 genes of the dorsal group, 3 are required in the somatic line. This suggests the existence of inductive signals originating during oogenesis from the follicle cells that surround the developing oocyte. This somatically expressed spatial information probably controls dorsoventral development by defining the polarity of a signal transducing pathway that specifically activates the nuclear uptake of the dorsal product. This model, highlights the importance of the polarity of the egg chamber, and suggests that it is the oocyte nucleus due to its asymmetrical localization, that determines the dorsoventral pattern formation of the embryo.

Role of the oocyte nucleus in determination of the dorsoventral polarity of Drosophila as revealed by molecular analysis of the K10 gene.

In Drosophila, the establishment of dorsoventral polarity of the developing embryo depends on the expression of at least 11 maternally acting genes. Mutant females that lack any of these gene activities produce normally shaped eggs that develop into dorsalized embryos. The female sterile K10 mutation differs from these mutants, because in addition to the dorsalized development of the embryo, it causes a dorsalization of the egg shape. During oogenesis, the K10 gene is specifically expressed in the oocyte. Antibodies raised against a beta-galactosidase-K10 fusion protein were used to visualize the K10 product in ovaries by indirect immunofluorescence. The protein, which contains a putative DNA recognition helix, accumulates in the nucleus of the oocyte, where it is assumed to have a regulatory function. Our results thus indicate that the controlled expression of some of the genes of the oocyte nucleus is essential for the determination of the dorsoventral polarity of the oocyte and possibility of the developing embryo.

DNA sequences homologous to the Drosophila opa repeat are present in murine mRNAs that are differentially expressed in fetuses and adult tissues.

A mouse embryonic cDNA containing two opa-like (CAX)n repeats was isolated on the basis of its cross-hybridization with a Drosophila K10 cDNA. Such repeated sequences were present in different murine mRNAs, some of which were specifically expressed during fetal life or in different adult tissues. This suggests that, as already described for Drosophila, opa-like sequences are parts of proteins involved in ontogenic or cell-type-specific functions in vertebrates. However, unlike Drosophila, such repeated sequences were not found within the murine homeo-boxes containing genes of the Hox-1 complex.

Oocyte-specific transcription of fs(1)K10: a Drosophila gene affecting dorsal-ventral developmental polarity.

The expression of the fs(1)K10 gene is required in early oogenesis for the establishment of the dorsal-ventral polarity of the oocyte, and later in the embryo. P-element-mediated transformation shows that the K10 function is located within a fragment of DNA of 5 kb, which encodes four RNA species. A major transcript of 3.1 kb is likely to be responsible for the K10 function. It is abundant in ovaries and in early developing embryos. Thus its expression profile corresponds closely to that which could be anticipated from the biological characteristics of the mutation. In situ hybridization on ovary sections shows that the gene is not only specifically transcribed in the germ line (which is consistent with the germ-line dependence of the mutation), but that its expression is also cell-specific since it is apparently restricted to the oocyte.

A 43 kilobase cosmid P transposon rescues the fs(1)K10 morphogenetic locus and three adjacent Drosophila developmental mutants.

The K10 female sterility locus involved in establishment of the embryonic dorsoventral axis maps genetically to the 2E2-2F1 interval of the Drosophila X chromosome. We microdissected the 2E2-2F3 region from salivary gland chromosomes and used clones obtained from the microdissected fragments to establish a chromosomal walk covering more than 200 kb. To identify the K10 gene we used P-mediated transformation with cosmid clones constructed in cos-P, a cosmid vector incorporating the terminal repeats of the P element. Clone cos9, containing a 43 kb insert, transformed the germ line of homozygous K10 females and allowed production of normal progeny. It also rescued three genes, crooked neck, pecanex, and kurz, which map genetically near K10. Transformation experiments using smaller fragments of cos9 localize the K10+ function within 11 kb. Northern blots hybridized with probes from this region indicate the presence of several mRNA species. Each transcript has been assigned to a complementation group.

Nucleotide sequence at the 5' extremity of tobacco-mosaic-virus RNA. 2. The coding region (nucleotides 69-236).

In the preceding paper it was shown that the first A-U-G codon in tobacco mosaic virus RNA is separated from the 5' terminus by a sequence of 68 nucleotides devoid of internal guanosine residues. In this paper we present the sequence of 165 residues immediately following the first potential initiation codon. The characterized sequence contains four nonsense codons but none are in phase with the prospective initiation codon. Several lines of evidence, including direct characterization of the portion of the RNA molecule which binds to and is protected by the ribosome in the course of initiation, all support the idea that the A-U-G at position 69-71 is a functional initiation signal for viral protein synthesis.

Monocistronic translation of alfalfa mosaic virus RNAs.

The four alfalfa mosaic virus RNAs (respectively 24 S, 20 S, 17 S and 12 S) have been used separately as messengers in two in vitro protein synthesizing systems: wheat germ and rabbit reticulocyte lysate. In both systems a polypeptide corresponding to the translation of the entire length of the RNA can be found for RNAs 24 S, 20 S and 12 S, but not for 17 S RNA, the translation product of which is only 35,000 daltons. The number of initiation sites has been determined for each RNA by analyzing the initiation peptides synthesized in the presence of spasomycin and show that there is only one initiation or binding site perRNA. We thus conclude that each AMV RNA behaves as a monocistronic messenger in in vitro translating systems.