Mice are the envy of the genetics world thanks to the ease with which these animals can be genetically altered by tinkering with cultured embryonic stem (ES) cells. When it comes to making straight transgenics, however, the mouse no longer stands apart: for this and most other species, DNA needs to be injected into embryos or germ cells, which leads to a low but nevertheless unwelcome degree of mosaicism. Kayoko Kurita and colleagues have discovered how to do away with such inefficiencies: they have successfully created transgenic sperm by genetically modifying zebrafish sperm precursors that are grown in vitro.

Hundreds of developmental zebrafish mutants have been created, and the genome sequencing project has highlighted many more genes that could be modified or inactivated. The available options for fish transgenesis, however, are currently inadequate: for example, when DNA is injected into oocytes, embryos or the male pronucleus of a zygote, the resulting transgenic organism is usually mosaic for the transgene, meaning that germline transmission cannot be ensured until the following generation. The most straightforward way of avoiding this inconvenience would be to genetically modify sperm before fertilization, so that all cells of the F1 progeny contain the transgene; however, this approach has not been met with much success, in vivo or in vitro, as mature sperm are refractory to transgene insertion.

To avoid the problem, Kurita and co-workers set out to introduce the transgene into sperm precursors. Primary cultures of zebrafish male germ cells were infected with retro-viral vectors derived from the Moloney murine leukaemia virus. The in vitro matured sperm were then used to produce transgenic offspring. Of the 89 successfully fertilized eggs that developed into adult fish (out of 104), 5 carried the transgene. Importantly, the transgene was transmitted to offspring in a Mendelian fashion, thereby proving that the parent fish was not mosaic.

Five transgenic fish out of 89 might not seem like a vastly high rate of insertion. However, this is comparable to the rate of success of current transgenic approaches, which lead to mosaic progeny. The authors speculate that their protocol — with the appropriate tweaks — could be applied to rapidly alter the sperm genomes of other animals, including humans.