| Abstract |
With the goal of developing techniques for DNA insertional mutagenesis in
zebrafish, we established procedures for rapidly obtaining and injecting
large numbers of fertilized eggs. Using either of two plasmid constructs,
we injected uncut DNA into fertilized eggs at the one-or two-cell stage.
Injected eggs were raised to sexual maturity and the frequency of
transgenic founder fish determined by pair mating them and testing DNA
extracted from pools of 16-hour embryos using PCR and subsequently
Southern analysis. Eggs injected with one of two different plasmids
yielded no transgenic fish, but 7-25% (19/115 overall) of the eggs
injected with the other transmitted the injected sequences to their
offspring (F1). Of nineteen lines studied further, all were able to pass
the foreign DNA sequences to the next (F2) generation. Inheritance in the
F2 was Mendelian in the seventeen lines tested. PCR and Southern analysis
indicated that the plasmid sequences were amplified. Three founder fish
possessed more than one integration event, and multiple integrations were
found to occur both in the same germline precursor cell, as well as in
different precursors. Although the plasmids injected into the embryos
contained a functional lacZ gene, of twelve lines tested, none of the
transgenic offspring were found to be XGAL-positive. Of thirteen
transgenic lines tested, in only one line was RNA detected in the
offspring. The ability to obtain and inject large numbers of zebrafish
eggs combined with a high frequency of germline integration may be steps
towards the goal of being able to perform insertional mutagenesis with
this organism.
In a screen for mutant phenotypes caused by the insertion of transgenes
into the zebrafish genome, nineteen transgenic lines were bred to
homozygosity. None of the nineteen lines displayed a discernible phenotype
in the homozygous state. Homozygous adult fish were identified in each of
the nineteen transgenic lines, and all were fertile. Seventeen lines were
screened for mutations exhibiting a strict maternal phenotype, and no
mutants were identified. Homozygous lines were generated from each of
twelve transgenic lines, and all were fertile. Cloning the sequences
flanking the insertions of two transgenic lines revealed that these
transgenes had integrated into single-copy sequences of the zebrafish
genome. Analysis of one transgenic line studied further revealed that the
insertion of the transgene had occurred with very little disruption to the
host chromosome: 12 bp had been deleted from the host chromosome, while 23
bp had been deleted from the plasmid.
In an effort to determine whether the insertion of DNA randomly into the
genome of the zebrafish is potentially mutagenic, gene trap constructs
were introduced into cultured zebrafish cells. This cell line, PAC2, is
highly transfectable and demonstrates a high efficiency of cloning. A
second cell line which was also established, HetA, is also transfectable,
but these cells are not clonable. The introduction of lacZ gene trap
constructs into PAC2 cells resulted in a very low frequency of
XGAL-positive colonies. It was found that only by using a very strong
promoter or by concentrating the protein in the nucleus can a level of lacZ
expression be attained for efficient detection by the XGAL test. In
contrast, PAC2 cells are capable of activating a promoterless neomycin
gene at a frequency similar to that observed in cultured mouse cells.
Analysis of the fusion transcript of one of the neomycin resistant PAC2
clones demonstrated that the neomycin resistance gene was activated by
fusion with a gene normally expressed in PAC2 cells. Thus, DNA is capable
of integrating into expressed zebrafish genes, indicating that random DNA
integration is a feasible approach for insertional mutagenesis in the
zebrafish.
|
|---|
