Vertebrate Genome Regulation using Zebrafish
(January 1999 - December 2008)
Zebrafish have a generation time of only three months and are bred in the laboratory in large numbers, making them suitable for transgene- based genomic analysis. Embryonic development from a fertilized egg to a swimming fish with a functional visual system takes only five days, and development of the entire body can be studied in great detail in the living animal using transgenic fish where gene expression is visible under the microscope. As the zebrafish genome nears completion it becomes evident that this small vertebrate has a developmental gene repertory that is near identical to that of humans. The way the genome shapes a vertebrate embryo is studied in zebrafish in forward genetic screens that aim either at recovering mutations that cause specific defects during embryonic development as well as through examining gene expression. Mutations that affect the activity of crucial genes can also give important clues about human heritable diseases. For instance all major vertebrate organs are found in the young fish and their development can be observed. Our own interest is the development of the embryonic forebrain, specifically the retina, but we have recently started to work on many more organs, based on the finding that developmental regulatory genes affect many different structures. For example the fish below shows expression of GFP under the control of regulatory elements of the inhibitor of differentiation 1 gene (id1), which is expressed in the bones, skin, retina, pineal, optic tectum, and cerebellum.
Photo by Mary Laplante
All teleosts genomes have undergone an additional duplication and as a result a number of genes are found in duplicate in the zebrafish genome when compared to tetrapods. The duplicated genes share the function of the ancestral single gene. An example is shown below, where the two loci of the pax6.1 and pax6.2 genes are seen to have complementary expression patterns in the retina, diencephalon, hindbrain and spinal cord, where the mammalian single gene is highly expressed. We have found that, as postulated by the duplication, degeneration, complementation (DDC) model, this ‘subfunctionalization’ is based upon the loss of specific regulatory elements from one or the other locus when compared to the human chromosomal locus.
Pax6.1 (top) and pax6.2 (bottom) enhancer detection lines show complementary patterns of expression. A comparison of the chromosomal loci shows that both loci in fish contain fewer regulatory elements (green ovals) than the single locus in the human genome. PAX6 (in red) has been kept in both copies, whereas the neighboring bystander genes are conserved only in one or the other copy in fish. For more information see Kikuta et al. (2007).
Our laboratory uses the zebrafish to characterize vertebrate transcriptional regulation at the genome level. In an enhancer detection screen where we generated and screened around 15000 retroviral insertions in the zebrafish germ line, we kept and characterized about 350 transgenic lines with yellow fluorescent protein (YFP) expression in the central or peripheral nervous system and mapped the insertions to the zebrafish genome sequence.
Surprisingly, a number of gene expression patterns were found more than once, and upon mapping we discovered that they demarcate the same large chromosomal segments of several hundred kb around developmental regulatory genes, suggesting that these genes are regulated by long-range regulatory elements that are situated at a long distance from the coding sequence. In other cases we found that, while the insertion was near or even inside a gene, the expression pattern was that of a gene further away, which in every case were also regulatory genes (encoding transcription factors, growth factors or micro RNAs, to name a few).
Some of these chromosomal segments that are now termed genomic regulatory blocks were also independently discovered through bioinformatic methods by way of association of large numbers of highly conserved non coding elements (HCNEs) around regulatory genes (see Lenhard lab web site). Since then, many of these HCNEs have been shown to act as positive regulatory elements, or enhancers, and they are indeed found not only close to the gene, but far up- or downstream of the gene they regulate, as well as in introns of neighboring genes, in some cases (in the human genome) a Mb or more away. Our main projects at present are the characterization and annotation of entire genomic regulatory blocks for their regulatory content, especially those that have likely connection to human genetic diseases.