1. Cloning of the Ovum mutant mutation (Om) which induces embryonic death around the blastocyst stage and entails a parental effect (P. Baldacci, M. Cohen-Tannoudji)
The DDK inbred strain of mice carries a conditional letal mutation which is manifested in outcrosses and depends on the direction of the cross. We had mapped the Om locus and have now established its physical map by isolating a contig of BACs which encompasses the genetic region containing Om. Exon trapping, cDNA selection and sequencing (in collaboration with the Génoscope, Evry, France) has allowed us to identify ca. fifty genes. Among those, two of them, which are expressed both in the oocyte and the testis, exhibit a difference in their open reading frame which is unique to DDK. We now develop functional approaches which should allow us to demonstrate that these mutations are indeed directly involved in the letality of the (DDK x non DDK)F1 embryos.
2. Targeted mutagenesis to analyse gene function in vivo and search for developmental genes by a gene-trap approach
Null mutations in genes coding for intermediate filament (IF) (E. Colucci-Guyon)
We have generated knock-out (KO) mice for the vimentin gene. This year, we have addressed the hypothesis of an implication of this IF in certain types of tumors. To that end, we have used the teratocarcinoma model; Vim -/- and Vim +/+ ES cells were injected in ectopic sites either in Vim -/- or in Vim +/+ mice. We found that teratocarcinoma formation was obtained in all four combinations with the same efficiency, thus dismissing a role of vimentin in tumor formation at least in this system. We have also created knock-out mice for the cytokeratin K6 and demonstrated in collaboration with the group of P. Coulombe (John Hopkins University, Baltimore), that this cytokeratin plays an essential structural role in oral mucosa.
Knock-out mutations in the SNF5/INI1 and vHNF1 genes (J. Barra in collaboration with the group of Dr. M. Yaniv, Pasteur Institute)
We have analysed the phenotype induced by null mutations in these genes. In the case of SNF5/INI1, we demonstrated that the embryos homozygous for the null mutation die around implantation, and therefore that SNF5/INI1 gene product is essential for completion of preimplantation development. Furthermore, mice heterozygous for the mutation developped nervous system and soft tissue sarcomas, in which wild type allele was lost, thus strongly suggesting that SNF5/INI1 functions as a tumor suppressor gene.
We had shown that embryos homozygous for a null mutation in vHNF1 gene die around day 7 of gestation, due to a defect in visceral endoderm differentiation. This year, we have begun to address the role of vHNF1 in organogenesis, using a strategy of conditional mutagenesis. Preliminary results suggest a very early defect in kidney formation, as well as anomalies in pancreas and liver development.
Gene trap in ES cells (J. Barra)
This approach is aimed at identifying developmentally important genes, via the creation of insertion mutations. The latter are generated via the random insertion in ES cells of a "trap" vector comprising a reporter gene (coding for a fusion protein between b-galactosidase and a protein confering resistance to an antibiotics), without promoter/regulatory sequences. Thus the reporter gene is expressed only if inserted functionally in a gene active in ES cells. We have pursued this year the analysis of two insertional mutations, one in the SSeCKS gene, the other in the TRAPa gene. In the first case, molecular analysis of the insertion revealed that SSeCKS encodes in fact two proteins one of which is a truncated form of the other ; interestingly these two isoforms are differentially expressed, one (the truncated isoform) being expressed exclusively in the male germline, the other exhibiting wide expression in particular in the central nervous system ; furthermore only the synthesis of the latter is disrupted by the insertion. Surprisingly, mice homozygous for the insertion develop without an obvious phenotype. However given the postulated importance of the signal transduction pathway mediated by PKA in learning and memory, it may very well be that these mice have defects in these processes.
The insertion in TRAPa is embryonic letal around day 14 of gestation and preliminary observations suggest that mutant embryos are defective in heart morphogenesis.
3. An approach to improve the efficiency of gene targeting (M. Cohen-Tannoudji)
We have recently developped a strategy for mouse germline modifications based on the stimulation of endogenous DNA repair processes. This approach allows to enhance considerably the targeting efficiency at a given locus. Our aim is now to : i) to improve the targeting methods by introducing targeted modifications directly into the zygote, thus avoiding the use of ES cells. ii) Create genetic modifications in somatic cells. iii) Identify permissive sites for the targeted expression of genes involved in human pathologies or of therapeutic agents.
Recently we have generated a reporter gene comprising the coding sequences of LacZ separated by an I-SceI site. The reporter gene is not functional, due to the presence of duplicated regions surrounding the I-SceI. An introchromosomal recombination will be induced by I-SceI, thus permitting to restaure LacZ activity which can be easily monitored. We have placed this reporter gene under the control of different promoters and derived the corresponding transgenic mice, which will allow us to test the feasability of homologous recombination in ovo.
A core facility for the generation of transgenic mice (V. Guyot)
Our laboratory is in charge of a transgenic facility which generates transgenic mice for the groups of the Pasteur. Institute.