Genetic analysis of T-cell development in zebrafish
In the last fifteen years, we have conducted and now essentially concluded two genetic screens in zebrafish to identify recessive mutations affecting T lymphocyte development in early embryogenesis and identified the responsible genes in about 40 complementation groups by positional cloning. In most cases, mis-sense mutations were identified.
The fact that several genes were represented by two distinct alleles each suggests that we have captured a substantial fraction of genes or at least representatives of pathways whose mutations affect early T-cell development in zebrafish. The affected genes fall into several functional categories according to their known functions. The first group of mutants comprises key regulators of haematopoiesis and lymphopoiesis (for instance, cmyb, ikzf1/ikaros, il7r, jak3, etc.). A second group comprises genes whose products are known to play a role in DNA replication and repair and the regulation of the cell cycle (for instance, pole1, mcm10, dnmt1, etc.). Other groups comprise genes involved in pre-mRNA processing (for instance, lsm8, gemin5, tnpo3, cstf3, etc.). Interestingly, despite the ubiquitous expression pattern of many of these genes, it appears that T-cell development is particularly sensitive to the phenotypic effects of these mutations. The detailed analysis of a recessive missense mutation in the dnmt1 gene encoding the zebrafish DNA maintenance methyltransferase indicated that this mutation caused a selective loss of the lymphoid lineage; otherwise the fish appear normal. When dnmt1m/mmales are crossed with wild-type females, the resulting dnmt1+/m off-spring exhibit a specific loss of larval T cell development, unlike heterozygous animals arising from a cross of two heterozygous parents. This unique haematopoietic phenotype re-appears in the following generations, and, in later generations, is transmitted even by genotypically wild-type males and females; in follow-up studies, this unequivocal case of transgenerational inheritance was traced to specific DNA methylation marks and correlated to aberrant transcription levels of key regulators of T cell development. Moreover, we have generated mouse models of Dnmt1 missense mutations, and, using structure-guided mutagenesis of the target-recognition domain of DNMT1, generated a viable mouse model with selective lymphoid deficiency. The identification of a mis-sense mutation in pole1 led us to explore the function of non-enzymatic accessory units of the POLE holoenzyme. To this end, we generated a series of mutations in the Pole3 gene using the CRISPR/Cas9 technology. These studies unexpectedly revealed that the replacement of acidic residues in the C-terminus of POLE3 caused a selective loss of T and B cells, while sparing other haematopoietic lineages. These results support the emerging notion of cell type-specific functions of general purpose factors.
Our genetic screen uncovered several components of the IL-7 signalling pathway, but did not identify the il7 gene itself. Using comparative genomics, we isolated this cytokine gene, generated a null allele and found that the phenotypic consequences are much milder than those seen with the il7r mutant. This discrepancy led to the discovery of a degenerate network of cytokines supporting T cell development in zebrafish, consisting of il7, il2, il15, and an as yet presumptive tslp homolog. This is in stark contrast to the situation in mice, where T cell development entirely depends on IL-7. We are now following up these observations by generating genetic interaction maps aiming at identifying synthetic lethality as a means to interfere with the development of T cells and their malignant derivatives.