Search

In vertebrates, the engrailed genes are expressed at early neurula stage in a narrow stripe encompassing the midbrain-hindbrain boundary (MHB), a region from which a peculiar structure, the isthmus, is formed. Knock-out experiments in mice demonstrated that these genes are essential for the development of this structure and of its derivatives. In contrast, little is known about the effect of an overexpression of engrailed genes in vertebrate development. Here we report the isolation of Ol-eng2, a medaka fish (Oryzias latipes) engrailed gene. We have monitored the effects of its widespread expression following mRNA injections in 1- and 2-cell medaka and Xenopus embryos. We found that the ectopic expression of Ol-eng2 predominantly results in an altered development of the anterior brain, including an inhibition of optic vesicle formation. No change in the patterns of mesencephalic and telencephalic markers were observed. In contrast, expressions of markers of the diencephalon were strongly repressed in injected embryos. Furthermore, the endogenous Ol-eng2, Pax2, Wnt1 and Fgf8, which are essential components of the MHB genetic cascade, were ectopically expressed in this region. Therefore, we propose that Ol-eng2 induces de novo formation of an isthmus-like structure, which correlates with the development of ectopic midbrain structures, including optic tectum. A competence of the diencephalon to change to a midbrain fate has been demonstrated in isthmic graft experiments. Our data demonstrate that this change can be mimicked by ectopic engrailed expression alone.

Gynogenesis could be an efficient way for producing inbred lines in commercial fish species. Gamma-irradiation of sperm gives haploid embryos that all die without hatching; in the present study, we optimized heat treatment of the eggs, in order to produce high rates of diploid gynogenetics. When the eggs are heated to 26 °C for 20 min after 25 min of development, 80% of the embryos hatch, and all the resulting fry are diploid; nevertheless, high mortalities are recorded until feeding start. The monosex female constitution of gynogenetic offspring confirms the female homogamety in the rainbow trout.

Gene transfer into fish embryo is being performed in several species (trout, salmon, carps, tilapia, medaka, goldfish, zebrafish, loach, catfish, etc.). In most cases, pronuclei are not visible and microinjection must be done into the cytoplasm of early embryos. Several million copies of the gene are generally injected. In medaka, transgenesis was attempted by injection of the foreign gene into the nucleus of oocyte. Several reports indicate that the injected DNA was rapidly replicated in the early phase of embryo development, regardless of the origin and the sequence of the foreign DNA. The survival of the injected embryos was reasonably good and a large number reached maturity. The proportion of transgenic animals ranged from 1 to 50% or more, according to species and to experimentators. The reasons for this discrepancy have not been elucidated. In all species, the transgenic animals were mosaic. The copy number of the foreign DNA was different in the various tissues of an animal and a proportion lower than 50% of F1 offsprings received the gene from their parents. This suggests that the foreign DNA was integrated into the fish genome at the two cells stage or later. An examination of the integrated DNA in different cell types of an animal revealed that integration occurred mainly during early development. The transgene was found essentially unrearranged in the fish genome of the founders and offsprings. The transgenes were therefore stably transmitted to progeny in a Mendelian fashion. Southern blot analysis revealed the presence of possible junction fragments and also of minor bands which may result from a rearrangement of the injected DNA. In all species, the integrated DNA appeared mainly as random end-to-end concatemers. In adult trout blood cells, a small proportion of the foreign DNA was maintained in the form of non-integrated concatemers, as judged by the existence of end fragments. The transgenes were generally only poorly expressed. The majority of the injected gene constructs contained essentially mammalian or higher vertebrates sequences. The comparison of the expression efficiency of these constructs in transfected fish and mammalian cells indicates that some of the mammalian DNA sequences are most efficiently understood by the fish cell machinery. Chloramphenicol acetyl transferase gene under the control of promoters from Rous sarcoma virus, and human cytomegalovirus, was expressed in several tissues of transgenic fish. Chicken δ-crystallin gene was expressed in several tissues of transgenic fish. Rainbow trout growth hormone cDNA driven by the Rous sarcoma virus promoter was expressed in transgenic carps leading to a faster growth of these animals. The antifreeze protein gene from flounder was expressed in transgenic salmon. These data indicate that transgenesis in fish is relatively easy but that fish gene sequences must be preferably used to obtain a good expression of the transgenes. Fish is a good biological model, specially for developmental studies and it is an increasing part of human food. For these reasons, transgenesis in fish is most likely to be more and more practised in the coming years.

Dietary administration of various estrogens for three months from swim-up stage resulted in excess of females, the remainder of the treated groups consisting of males and hermaphrodites. Mature hermaphrodites were self-fertilized or mated with standard males and females. These hermaphrodites and some of the estrogen-treated females proved to be genetic males; the frequencies of males obtained from their ova averaged 76.6%, suggesting viability of the YY genotype. Four of nine tested males of those progenies provided all male offspring when mated with standard females.

Screening of a medaka (Oryzias latipes) adult brain cDNA library, with a degenerated probe corresponding to the most conserved region of helix III of the homeodomain, led to the isolation of a gene homologous to a murine orphan Hoxgene, named Gsh-1. We have called this gene Ol-Gsh 1(Oryzias latipes-Gsh 1). Molecular analysis of the Ol-Gsh 1 putative protein points to potential functional domains which are highly conserved between fish and mouse genes. Whole-mount in situ hybridization shows that Ol-Gsh 1is expressed in several waves during embryonic development. Transcripts are found in many regions of the central nervous system: the spinal cord, dorsal rhombencephalon, optic tectum, dorsal diencephalon, hypothalamus anlagen and rostral telencephalon. This multimodal expression pattern, strikingly conserved between fish and mammals, is reminiscent of both clustered and orphan homeobox genes. In addition, each expression wave is initiated in the fish embryo earlier than in the mammalian embryo, relative to the time scale defined by somitogenesis. We propose that Ol-Gsh 1may be involved in conserved developmental pathways and in particular may be linked to proliferation events. Mouse Gsh-1was shown to participate in neuro-endocrine functions of the hypothalamus. From late developmental stages onwards, Ol-Gsh 1expression is also restricted to the hypothalamus. The expression pattern in this structure raises interesting questions concerning a fully or partially conserved function for these genes.

Nine induced tetraploid females were artificially inseminated by UV-irradiated sperm collected from diploid males, in order to induce the gynogenetic development of their ova. Most of the resulting embryos were diploid (or minor aneuploids). Several gynogenetic tetraploids, likely to issue from unreduced ova, were also detected in these progenies. The same females fertilized by normal sperm of diploid males gave a majority of triploids and several pentaploids, while the fertilization by normal sperm of tetraploid males gave rise to a majority of tetraploids and one hexaploid. The same crosses, after the eggs had been heat-shocked to double the maternal genetic contribution, yielded about three-quarters pentaploids and one quarter haploids (normal sperm of diploids), or three-quarters hexaploids and one quarter diploids (normal sperm of tetraploids). These haploids and diploids are likely to result from androgenesis.

Experiments, supported by extensive karyology, were carried out to evaluate the different techniques used for chromosome manipulation in rainbow trout. Eggs, when subjected to early heat shocks, changed from haploidy to diploidy and from diploidy to triploidy. In this respect heat shocks differ from pressure shocks which induce gradual transitions between successive ploidy levels. Sperm treatment with dimethylsulphate yields haploid embryos containing residual sperm chromatin fragments, in contrast to treatment with ultraviolet rays.

Experiments were conducted to estimate the viability of diploid and induced triploid hybrids between the rainbow trout female and the grayling male. Both are unable to hatch, even so the triploid dies later than the diploid. Insemination of rainbow trout eggs by UV irradiated sperm of grayling results in gynogenetic rainbow trouts when viability is restored by heat shocks inhibiting the second division of egg meiosis.

First generation tetraploids were produced by hydrostatic pressure treatment before the first cleavage and raised until the adult stage. Their survival and growth were severely depressed when compared to the diploid control: after two years, no ovulated females were found although males produced sperm at 1 and 2 years of age and were mated individually with diploid females. The progenies were consistently normal with high survival rates. They were found to be almost all triploids by karyology, which failed to detect a significant rate of aneuploidies. However, the fertilizing ability of tetraploid males was always low (0 to 97% of the control; average 40%). Several arguments presented here support the hypothesis that diploid spermatozoas, which are wider than haploid ones, would be frequently blocked during their penetration through the micropyle canal. Second generation tetraploids were produced after such matings by heat shocks, causing the retention of the second polar body. Their survival and growth were much more satisfactory than in the first generation, although still lower than in diploid and triploid controls issuing from diploid parents. Performances of second generation triploids were comparable to those of diploids, and slightly better than those of conventional triploids issuing from diploid parents. 94.5% of the second generation tetraploids were male.