GENETIC ANALYSIS OF ZEBRAFISH EMBRYO DEVELOPMENT

Human Genome Research

Investigators

Linked publications & trials

Abstract

Zebrafish (Danio rerio) has emerged as ?Drosophila of the vertebrates? to study genetic controls of development and disease. It has several advantages over mice due to its small size, large progeny, and short reproduction cycle. Transparent embryos and ex vivo development allows noninvasive observations of embryonic development. Furthermore, the ability to do saturation mutagenesis using ENU (N-ethyl-N-nitrosourea) followed by forward and reverse genetic technologies to isolate mutants has a great potential in functional genomics. In order to identify novel genes required for embryonic hematopoiesis, particularly those affecting early stages as well as myelopoiesis, we conducted a whole mount RNA in situ hybridization screen of ENU-mutagenized F2 haploid embryos with antisense probes for cbfb and l-plastin, genes expressed in the hematopoietic stem cells and myeloid lineage cells, respectively. One mutant line, mummy, was identified from the screen that showed reduced cbfb expression and total lack of l-plastin expression. In addition, mummy mutants have reduced scl, gata1 and hemoglobin expression. These results suggested that the defect in mummy is at or above HSC level. The mummy mutant embryos also suffer from widespread cell death and die around 2 days post fertilization. Genetic mapping and positional cloning identified the mutated gene as dhx8, which encodes an RNA helicase. The yeast homolog of dhx8, prp22, functions during mRNA splicing. Morpholino injection has confirmed that reduced expression of dhx8 can produce a similar phenotype. The wild-type full-length zebrafish dhx8 cDNA but not the mutated dhx8 cDNA identified in mummy was able to rescue the mummy defect. Lack of dhx8 function in the mummy mutant embryos specifically affected the production of spliced transcripts of certain genes but not others, which may account for the tissue-specific phenotype in this mutant. Our results demonstrate that mutations in an RNA helicase involved in splicing may show tissue specific developmental defects. A manuscript reporting this finding is currently under preparation. We have also undertaken an effort to generate a mutant library archived as a sperm bank and a DNA bank from ENU-mutagenized F1 males, in order to screen for mutations in specific genes of interest. DNA was used to screen for mutations by sequencing or other heteroduplex screening methods. The corresponding sperm samples were used to revive the mutant fish by in vitro fertilization. In the last year we performed two more rounds of ENU mutagenesis and archived sperm and DNA samples from 1800 F1 male fish. We have now a DNA/sperm library from 3100 F1 male fish. Our goal is to generate ~5,000 mutants to ensure efficient recovery of multiple mutant alleles for any gene screened. Since success of resequencing projects depends on the ability to recover lines from the frozen sperm, we evaluated several published protocols on sperm cryopreservation and optimized the conditions for testes extraction and sperm cryopreservation to ensure maximum recovery of all mutations. We have continued our effort to screen for mutations in selected genes. Analysis of resequencing data has proven to be a challenge. Therefore, we analyzed the resequencing data using multiple parameters and programs to maximize the mutation detection with minimal false positive detection. A new program, developed by our collaborator in NCI, Dr. Jinghui Zhang, was very sensitive and accurate in recognizing heterozygous mutations from the resequencing data. A manuscript describing this new program has been accepted for publication. Since sequencing is expensive and analysis of such large number of samples is a challenge, we have explored alternative ways of screening for rare mutations. TILLING (Targeting Induced Local Lesions IN Genomes) is a method that combines heteroduplex screening with fragment cleavage (Till et al, Genome Res. 13:524, 2003), thereby giving unique fragments for mutations despite the presence of common polymorphisms, estimated at ~1/200bp in the zebrafish genome. Using a known mutation in gata1, we optimized conditions for TILLING using capillary electrophoresis, allowing the use of existing sequencing machines in the Institute. Screening of two exons from lft1 and lft2 in 1235 F1 males using the optimized protocol led to the identification of three new mutations. Screening was very efficient since mutations could be detected even after four-fold pooling of PCR products prior to heteroduplex formation. TILLING, therefore, provides a cost-effective and faster alternative to sequencing. Amplicons with fewer common SNPs are ideal for TILLING whereas amplicons with several SNPs and in/del polymorphisms are best suited to sequencing.