|Home | About | Journals | Submit | Contact Us | Français|
Several technologies for studying gene function in zebrafish embryos, larvae, and adults were discussed at the Targeted and Conditional Gene Expression workshop. The workshop included six oral presentations followed by a community discussion of current approaches, their limitations, and possible solutions. The research talks focused on new methods for controlling the activity of endogenous or exogenous genes through genetic and/or chemical strategies. Approaches for conditionally regulating endogenous gene function included short-hairpin RNA (shRNA) expression, reversible transposon-induced mutations, and caged morpholinos. Technologies for controlling exogenous gene activity included a chemically inducible progesterone receptor-based transactivator, the Cre-LoxP system, and estrogen receptor fusion proteins. General considerations for using the Gal4/UAS system in zebrafish were also discussed.
Adam Hurlstone kicked off the workshop by reporting his laboratory's studies of shRNA-mediated gene silencing in zebrafish embryos. In contrast to short-interfering RNAs, expressed shRNAs do not appear to cause nonspecific developmental defects, and the Hurlstone laboratory reported that they have successfully blocked enhanced green fluorescent protein (EGFP) expression in transgenic embryos by injecting them with shRNA expression constructs. They have also transiently inhibited endogenous tyrosinase expression and pigment formation using an shRNA construct driven by the Fugu tyrp1 promoter. Efforts to generate transgenic zebrafish that stably express shRNAs in a tissue-specific and potentially reversible manner are underway. Thus far, their preliminary studies suggest that shRNA expression levels must reach a certain threshold to convey efficient endogenous gene silencing, perhaps necessitating amplification strategies such as the Gal4/UAS system or multiple, tandem shRNAs against a given target.
Steve Ekker described a complementary genetic approach for reversibly blocking endogenous gene function in zebrafish. His laboratory has created “gene-break” transposon-induced mutations, in which splice acceptor and donor sites flanked by LoxP sites are inserted into genomic DNA through transposon-mediated integration. The insertional constructs also contain red fluorescent protein (RFP) and EGFP coding sequences that generate an RFP-containing truncated gene product and EGFP reporter when the heterologous splice sites are used during RNA maturation. Using this approach, the Ekker laboratory has initiated a screen for zebrafish mutants that exhibit abnormal nicotine sensitivity. They have also demonstrated the reversibility of “gene-break” mutations by either Cre expression or injection of a morpholino targeting the splice acceptor.
James Chen then concluded the discussion of endogenous gene control by presenting his laboratory's work on caged morpholinos. In their approach, conventional 25-base morpholinos are conjugated to a complementary oligomer, typically 10 bases in length, through a photocleavable linker. The ability of the caged morpholino to anneal to its RNA target can then be actuated by light, exploiting the optical transparency of zebrafish embryos. Using the no tail gene as a proof of principle, the Chen laboratory demonstrated the ability of caged morpholinos to convey both temporal and spatial control of gene expression. Cells that have been selectively irradiated to uncage the morpholino can be simultaneously labeled using the Kaede protein, which undergoes a green-to-red photoconversion. Phenotypes observed for caged morpholinos against etsrp and sox32 were then discussed, illustrating the generality of this approach. Chen also described his laboratory's synthesis of a caged morpholino with a two-photon-sensitive linker, as well as alternative morpholino caging strategies developed by Jeffrey Yoder, Alexander Deiters, and Alan Mayer.
The workshop's discussion of technologies for exogenous gene expression was initiated by Sergey Parinov's presentation about a new chemically inducible expression system. Using a chimeric transactivator composed of the LexA DNA-binding domain, the p65/NF-κB activation domain, and the progesterone receptor ligand-binding domain, his laboratory was able to achieve RU486-dependent gene expression in stable transgenic zebrafish. The system works as a two-component system, in which the driver and effector constructs can be combined through genetic crosses, and Parinov group was also able to create various driver lines with EGFP reporters using the Ac/Ds transposon system from maize. RU486-dependent transcription is detectable within 1h of agonist treatment, and maternal expression can be induced by treating female adults prior to mating. The off-kinetics of the system upon RU486 removal is significantly slower, perhaps reflecting the accumulation of this drug in zebrafish embryos and adults. Using this approach, one can create transgenic lines that carry inducible, deleterious genes, as demonstrated by the Parinov laboratory's generation of a Tg(krt8:EGFP-K-rasV12) cancer model.
Alejandro Gutierrez, a postdoctoral fellow in Thomas Look's laboratory, next presented his work on tumor models that involve conditional activation of the Myc oncogene. In one approach, this group has developed a zebrafish model of T-cell acute lymphoblastic leukemia (T-ALL) by creating a Tg(rag2:LoxP-DsRed2-LoxP-EGFP-Myc) line. In their uninduced state, these fish express DsRed2 in their T-cells and do not express Myc or develop tumors. When crossed with an Tg(hsp70:Cre) line, the resulting double transgenic progeny can undergo heat shock–induced Cre/LoxP excision, to place Myc expression under control of the rag2 promoter. This excision process occurs with variable efficiency; however, the heat-shocked double transgenics develop T-ALL with 80–90% penetrance. As an alternative approach, the Look laboratory also created a Tg(rag2:Myc-ER) line, which selectively expresses a Myc-estrogen receptor fusion protein in T-cells. Upon treatment of these fish with 4-hydroxytamoxifen (4-HT), 100% develop tumors within 8–12 weeks, and most of these tumors undergo remission upon 4-HT removal. Both inducible strategies exhibit varying degrees of basal activation, however. Cre expression under control of the hsp70 promoter is leaky, and the Myc-ER fusion protein exhibits some activity in the absence of 4-HT.
Kazuhide Asakawa, a postdoctoral fellow in Koichi Kawakami's group, then rounded off the workshop's oral presentations by providing pointers about the Gal4/UAS system in zebrafish that they have learned through studies of various Gal4 driver lines and UAS-dependent effector lines. For example, they have observed that Gal4-VP16 transactivator toxicity can be mitigated by using a minimal activation domain from the VP16 protein (Gal4FF). In their studies of enhancer traps utilizing Tol2 transposable elements and hsp70:Gal4FF, they found that the hsp70 promoter caused background Gal4 expression in heart and skeletal muscle. Their investigations of gene traps using the Tol2 system also revealed that the splice acceptor sequence is critical for maximizing tissue specificity, with the gata6 splice acceptor minimizing background Gal4 expression. Finally, Asakawa reported that their studies of various 5xUAS:tetanus toxin light chain (TeTxLC) lines suggest that UAS:transgenes are subject to chromosomal positional effects and that UAS-dependent reporters can persist after Gal4 expression is lost. Careful screening of multiple UAS:transgene insertions and direct confirmation of transgene expression by epitope tagging or immunostaining are therefore required to correctly interpret one's experimental results. These observations complement findings reported at the meeting by Mary Goll and Marnie Halpern that 14×UAS-based promoters can undergo methylation-dependent silencing, at least in some contexts, and it was suggested that limiting the number of UAS repeats might minimize this problem.
Upon the conclusion of these six research talks, the workshop participants discussed a number of additional items. Since a couple of the talks described gene-targeting approaches that rely upon the Cre/LoxP system, the generality and potential limitations of the Cre recombinase in zebrafish were considered. The point was made that Cre is adapted to function at 37°C, and that flipase/flipase recombination target (FLP/FRT) from Saccharomyces cerevisiae, which is widely used in Drosophila, might be preferable for an organism raised at 28°C. Among the audience, there was no experience of using the FLP/FRT system in zebrafish, nor is there any mention of this application in the published literature. Moreover, the workshop attendees had positive experiences with the Cre/LoxP system, and it is anticipated that numerous transgenic zebrafish lines with spatiotemporally regulated Cre activity will become available to the community in the coming years. This would allow even greater facility for regulating transgenes in zebrafish models.
The description by Asakawa about the use of TeTxLC to selectively remove neurons led to a follow-up discussion of cell-ablation methods that are applicable to other cell types. Feedback from the audience supported the use of bacterial nitroreductase as a suicide gene in combination with the prodrug metronidazole, which has been used to ablate a wide spectrum of cell types in both larval and adult zebrafish with limited effects on surrounding tissues.
Finally, there was a lively exchange about how best to catalog and distribute the enhancer and gene trap lines that are being generated by several laboratories. While these lines are freely available to the wider community through the host laboratory, they are seldom distributed by the Zebrafish International Resource Center (ZIRC). How ZIRC might facilitate the deposition of these lines by individual laboratories, community access, and maintenance of a user-friendly database were discussed. Unfortunately, a ZIRC representative was not available at the workshop for comments or suggestions.