Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Methods. Author manuscript; available in PMC 2010 November 1.
Published in final edited form as:
PMCID: PMC2764541

The Tol2-mediated Gal4-UAS method for gene and enhancer trapping in zebrafish


The Gal4-UAS system provides powerful tools to analyze the function of genes and cells in vivo and has been extensively employed in Drosophila. The usefulness of this approach relies on the P element-mediated Gal4 enhancer trapping, which can efficiently generate transgenic fly lines expressing Gal4 in specific cells. Similar approaches, however, had not been developed in vertebrate systems due to the lack of an efficient transgenesis method. We have been developing transposon techniques by using the madaka fish Tol2 element. Taking advantage of its ability to generate genome-wide insertions, we developed the Gal4 gene trap and enhancer trap methods in zebrafish that enabled us to create various transgenic fish expressing Gal4 in specific cells. The Gal4-expressing cells can be visualized and manipulated in vivo by crossing the transgenic lines with transgenic fish lines carrying various reporter and effector genes downstream of UAS (upstream activating sequence). Thus, the Gal4 gene trap and enhancer trap methods together with UAS lines now make detailed analyses of genes and cells in zebrafish feasible. Here, we describe the protocols to perform Gal4 gene trap and enhancer trap screens in zebrafish and their application to the studies of vertebrate neural circuits.

1. Introduction

The yeast transcriptional activator Gal4 has a modular structure consisting of the DNA-binding domain and the transcriptional activation domain (1, 2). Gal4 binds to its specific recognition sequence UAS (for upstream activating sequence) and activates transcription of target genes (3). Since Gal4 expressed in particular tissues stimulates expression of a gene linked to UAS in a tissue-specific manner, the Gal4-UAS system has been employed to analyze gene functions in vivo in Drosophila. Namely, the P element-mediated enhancer trapping efficiently creates a number of fly lines expressing Gal4 in specific cells, and genes of interest are expressed in a spatially and temporally regulated fashion in the Gal4-expressing cells (4).

The zebrafish is a useful model for genetic studies of vertebrate systems. Several hundreds of fertilized eggs can be obtained from a single mating and a large number of adult fish can be maintained in a limited laboratory space. Due to these advantages, a variety of genetic approaches for investigating gene function have been carried out; i.e., chemical mutagenesis (5, 6), retroviral insertional mutagenesis (79), target-selected mutagenesis (10, 11), zinc-finger nuclease-based mutagenesis (12, 13) and morpholino knock-down (14). In addition to these, targeted gene expression in specific tissues by using the Gal4-UAS system was described in zebrafish (15). However, the usefulness of the Gal4-UAS method had been limited since construction of transgenic lines expressing Gal4 in various tissues and cells had been laborious and time-consuming, mainly because of the lack of an efficient transgenesis method.

We have been developing transposon techniques by using the Tol2 transposable element (16). We cloned a cDNA encoding the transposase protein from the medaka fish Tol2 element and developed a two-component transposition system (17, 18). When a plasmid carrying the Tol2 element is injected to zebrafish embryos with the Tol2 transposase mRNA synthesized in vitro, the Tol2 element is excised from the plasmid and integrated into the genome by the activity of the transposase (19). Because of the high transposition efficiency in the germ line and the capacity to carry a large DNA fragment (19, 20), the Tol2-mediated transgenesis has become a popular method to create transgenic zebrafish and has been applied to gene trap and enhancer trap approaches (2123)

Recently, the Tol2 transposon system was successfully applied to the Gal4 gene trap and enhancer trap methods in zebrafish (2427). First, we constructed a novel Gal4 variant Gal4FF, which we think suitable for transcription activation in zebrafish. Second, we developed gene trap and enhancer trap constructs by using the Tol2 transposon vector and Gal4FF. Third, we constructed transgenic fish carrying fluorescent reporter genes downstream of UAS and performed screens to identify transgenic fish expressing Gal4 in specific tissues and cells. Finally, we constructed transgenic fish carrying an effector gene, in this case a gene for tetanus toxin, downstream of UAS, and demonstrated that our system can inhibit neural functions. Here we describe the protocols to perform the Gal4 gene trap and enhancer trap screens in zebrafish and to apply the Gal4 transgenic fish to the studies of vertebrate neural functions.

2. Methods and materials

2.1. A novel transcriptional activator Gal4FF

We developed a novel transcriptional activator Gal4FF that contains the 147 amino acids DNA-binding domain from Gal4 and two tandem repeats of a 13 amino acids transcription activation module (PADALDDFDLDML) from VP16 (Fig. 1) (28, 29). In previous studies (15, 24, 25, 3032), the full-length Gal4 or Gal4-VP16 that contained the DNA binding domain and the VP16 activation domain was used for the Gal4-UAS system in zebrafish. We decided to employ Gal4FF since it is less toxic than Gal4-VP16 (see below).

Figure 1
The structures of full-length Gal4, Gal4-VP16 and Gal4FF

2.2. Gal4FF enhancer trap and gene trap constructs

T2KhspGFF is the enhancer trap construct that contains 638-bp DNA of the hsp70 promoter region, the Gal4FF gene and the SV40 polyA signal between essential cis-sequences of Tol2 (Fig. 2A, B). T2KSAGFF is the gene trap construct that contains a splice acceptor from the rabbit β-globin gene, the Gal4FF gene and the SV40 polyA signal between essential cis-sequences of Tol2 (Fig. 2B) (26).

Figure 2
The Tol2 transposable element and constructs

2.3. The UAS plasmid

To facilitate development of UAS reporter and effector fish, we constructed T2MUASMCS that contained five tandem repeats of the Gal4-recognition sequence (5×UAS), a TATA sequence, a multi-cloning site (MCS) and SV40 ployA between essential cis-sequences of Tol2 (Fig. 2C). Any genes cloned at the MCS are expected to be expressed in the presence of Gal4 (26).

2.4. The UAS:GFP reporter line

T2KUASGFP is a reporter construct that contains the EGFP gene downstream of 5×UAS between essential cis-sequences of Tol2 (Fig. 2C). The plasmid carrying the T2KUASGFP construct was injected into fertilized eggs with the Tol2 transposase mRNA. Eight injected fish were crossed with wild-type fish and the resulting F1 embryos were injected with a Gal4FF expression plasmid DNA. The offspring from five injected fish showed GFP expression. These GFP-positive fish were raised, analyzed by Southern blot analysis, and fish carrying single T2KUASGFP insertions were identified. One of such lines showed ubiquitous GFP expression when crossed with a line expressing Gal4FF ubiquitously and was established as the UAS:GFP reporter line. Inverse PCR analysis revealed that the T2KUASGFP insertion was located within the Nedd4-binging protein 1 gene (NM_199787) (26).

2.5. The UAS:RFP reporter line

T2ZUASRFP was constructed by cloning the mRFP1 gene (33) at the XhoI site of the T2MUASMCS plasmid (Fig. 2C). The plasmid carrying T2ZUASRFP was injected to fertilized eggs with the Tol2 transposase mRNA. Nine injected fish were crossed with the enhancer trap line hspGGFF15, which expressed Gal4FF in the central nervous system, and the resulting F1 embryos were analyzed for RFP fluorescence. The F1 embryos from seven injected fish showed RFP expression. These RFP-positive fish were raised, analyzed by Southern blot analysis, and fish lines carrying single T2ZUASRFP insertions were identified. One of such lines that showed the strongest RFP expression when crossed with the hspGGFF15 fish was established as the UAS:RFP reporter line. Inverse PCR analysis revealed that the T2ZUASRFP insertion was located in a solute carrier protein gene (Slc12a8) (26).

2.6. The UAS-tetanus toxin light chain effector lines

Tetanus toxin light chain (TeTxLC) cleaves a vesicle membrane protein Synaptobrevin and thereby blocks neurotransmitter release from synaptic vesicles (34). T2SUASTeTxLCCFP was constructed by cloning a gene encoding the CFP-tagged TeTxLC (35) between the EcoRI and XhoI sites of the T2MUASMCS plasmid (Fig. 2C). The plasmid DNA carrying T2SUASTeTxLCCFP was injected to fertilized eggs with the Tol2 transposase mRNA. The injected fish were raised to adulthood and crossed with the SAGFF73A fish, which expressed Gal4FF in the whole body (Fig. 4A). Various levels of CFP expression were detected in the F1 offspring. The founder fish whose F1 embryos expressed CFP at high levels and were immotile due to strong expression of the TeTxLC:CFP fusion gene was mated again with wild type fish, and their progeny were raised to adulthood. Genomic DNA from caudal fins of these F1 fish were analyzed by PCR and Southern blot hybridization, and the UAS:TeTxLC:CFP effector fish line that carried a single T2SUASTeTxLCCFP insertion at the CSPP1 gene (XM_693371) was established. In this manuscript, the UAS:TeTxLC:CFP fish are crossed with the enhancer trap (hspGGFF27A) and gene trap (SAGFF31B and SAGFF36B) lines (26) (Fig. 5), and behaviors of double transgenic offspring are analyzed.

Figure 4
Molecular characterization of the SAGFF73A insertion
Figure 5
Targeted inhibition and visualization of neural circuits by the UAS:TeTxLC:CFP fish

2.7. Tol2-mediated transgenesis

Synthesis of the Tol2 transposase mRNA in vitro is carried out as described previously (36). ~1 nl of a DNA/RNA mixture containing 25 ng/µl of circular plasmid DNA carrying a Tol2 construct and 25 ng/µl of the transposase mRNA are injected into fertilized eggs. To confirm that transposition occurred in the injected embryos, DNA is extracted from the embryos and the excision assay is performed as described (36). PCR condition for the excision assay is: 35 cycles of 94 °C for 20s; 55 °C for 20s; 72 °C for 20s. Primers used for this assay hybridize with plasmid sequences adjacent to the Tol2 construct. For T2KhspGFF or T2KSAGFF, TYR1 (5’-AAG GCT CTT GGA TAC GAG TAC GCC-3’) and BS1 (5’-AAC AAA AGC TGG AGC TCC ACC G-3’) are used to amplify PCR products of ~250 bp. For T2MUASMCS and its derivatives, excL1 (5’-ATT TCA CAC AGG AAA CAG CTA TGA-3’) and excR2 (5’-CAC GGA AAT GTT GAA TAC TCA TAC TC-3’) are used to amplify a PCR product of ~200 bp. The Tol2 construct excised from the injected plasmid is integrated in the genome in germ cells during embryogenesis. The injected fish are raised to adulthood and crossed with non-injected fish. Transgenic fish are identified in the progeny from 50 to 70% of the injected fish (21).

2.8. Southern blot hybridization analysis

Primers GAL4FF-f2 (5’-ATG AAG CTA CTG TCT TCT-3’) and GAL4FF-r2 (5’-TCT AGA TTA GTT ACC CGG-3’) are used to amplify the open reading frame of Gal4FF. The PCR product is labeled with [α-32P]dCTP and used to hybridize with the T2KhspGFF and T2KSAGFF insertions. The 32P-labeled EGFP probe (36) is used to detect the T2KUASGFP and T2SUASTeTxLCCFP insertions since it hybridizes with the CFP sequence. Primers mRFP1-f1 (5’-CCG CTC GAG ATG GCC TCC TCC GAG GAC GT-3’) and mRFP1-r1 (5’-ACG CGT CGA CTT AGG CGC CGG TGG AGT GGC-3’) are used to amplify the open reading frame of mRFP1. The 32P-labeled RFP probe is used to detect the T2ZUASRFP insertion. ~5 µg of Genomic DNA from tail fins of transgenic fish are digested with BglII, separated on 1% agarose gel and transferred to a Hybond-XL (Amersham) membrane. The membrane is hybridized with the 32P-labeled probes as described previously (36). When multiple bands are identified in transgenic fish by Southern blot analysis, the fish should be crossed with non-injected fish and fish with fewer insertions or single insertions can be identified in the next generation.

2.9. Identification of transposon integration site

To analyze the integration sites of Tol2, inverse PCR is performed (20, 36). In this study, GFP-positive Gal4FF transgenic fish always contain the UAS:GFP transgene also. Therefore, we first identify fish with single Gal4FF insertions by Southern blot hybridization analysis to avoid any complications. Then, the Gal4FF;UAS:GFP double transgenic fish are crossed with wild-type fish, and genomic DNA samples from their embryos are analyzed by PCR to detect DNA samples that contain the Gal4FF insertions only. To amplify junction fragments containing the Tol2 right end and genomic DNA, DNA samples are digested with MboI and self-ligated, and PCR is performed by using primers RF1 (5’-CCT CGA GTT TTC TTT CTT GCT TT-3’) and RR1 (5’-GCT AAC CAT GTT CAT GCC TTC T-3’). If PCR product is not amplified in the first PCR, a second PCR should be performed by using primers RF2 (5’-TTT TAC TCA AGT AAG ATT CTA GCC A-3’) and RR2 (5’-GTT CAT GCC TTC TTC TCT TTC CTA-3’). Adaptor-ligation PCR can also be used to analyze the insertions (20). The PCR products amplified by inverse PCR or adaptor-ligation PCR is gel-extracted and sequenced. The genomic DNA surrounding the Tol2 insertions are analyzed by using the ensembl database (

2.10. Characterization of SAGFF73A, an insertion of the exon trap

When a candidate gene trapped by the gene trap construct is identified, RT-PCR can be carried out to identify a fusion transcript. For instance, in the case of the SAGFF73A insertion, inverse PCR and the database analysis identified that the insertion is located within the zfand5b gene encoding a zinc-finger protein. Total RNA was prepared from 38 SAGFF73A;UAS:GFP embryos at 24 hours post fertilization (hpf), and used for cDNA synthesis with an oligo-dT primer and SuperScript First-Strand Synthesis System for RT-PCR (Invitrogen). Using the cDNA as a template, 30 cycles of a PCR (94°C, 30 sec; 55°C, 30 sec; 72°C, 2 min) was carried out using a forward primer in the first exon of zfand5b (zfand5b-f1: 5’-ATA GTA CAC ACC GAA ACG GAC AC-3’) and a reverse primer in the Gal4FF coding sequence GAL4FF-r1 (5’-CTG TGA CGG CAT CTT TAT TCA C-3’).

2.11. Behavioral analysis

Gal4FF;UAS:GFP double transgenic fish that express GFP in neurons in the embryonic stages are crossed with the UAS:TeTxLC:CFP fish. The CFP-positive embryos are selected by observing under a fluorescence stereomicroscope (MZ16FA, Leica). A gentle tactile stimulus is applied to the tail of each embryo at 48–80 dpf with a needle (NN-2725R, TERUMO). The behavior of embryos is recorded by using a dissecting microscope (Stemi 2000-C, Zeiss) equipped with a high-speed digital video camera (FASTCAM-512PCI, Photoron).

2.12. Whole mount immunofluorescence

To analyze TeTxLC:CFP expression, the rabbit polyclonal anti-GFP antibody (1:500 dilution; Invitrogen A6455) is used as a primary antibody. For immunostaining of motor neurons, the mouse monoclonal anti-Hb9 antibody (1:25 dilution; Developmental Studies Hybridoma Bank) is used as a primary antibody. For immunostaining of Rohon-Beard sensory neurons, the mouse monoclonal zn-12 antibody (1:250 dilution; obtained from the Zebrafish International Resource Center) is used as a primary antibody. Secondary antibodies used are goat anti-rabbit Alexa Fluor 488 (1:1000 dilution) and goat anti-mouse Alexa Fluor 633 (1:1000 dilution) (Molecular Probes). The CFP-positive Gal4FF;TeTxLC:CFP transgenic embryos at 30–48 hpf were transferred to 1.5 ml microtubes and fixed for 2–3 h in 1 ml of PBS containing 4% paraformaldehyde. The samples were washed with 400 µl of PBS-TX (PBS containing 0.5% Triton X-100) for 20 min 4 times, soaked in 1 ml of PBS-TX containing 10% BSA for 1 h, and incubated in the primary antibody diluted in PBS-TX/1% BSA overnight at 4°C. The samples were washed with 400 µl of PBS-TX for 20 min 4 times to remove the primary antibody, soaked in 1 ml of PBS-TX containing 10% BSA for 30 min and incubated in the secondary antibody diluted in PBS-TX/1% BSA for 2 h at room temperatures. The secondary antibody was removed by washing samples with 400 µl of PBS-TX for 20 min 4 times.

2.13. Microscopic analysis

A fluorescence stereomicroscope (MZ16FA, Leica) equipped with a CCD camera (DFC300FX, Leica) is used to observe and take images of GFP, RFP and CFP expressing embryos. Embryos immunostained with antibodies are soaked in 30 % glycerol, mounted on a slide glass (S8111, MATSUNAMI) with a cover glass (22 × 22 mm MICRO COVER GLASS, MATSUNAMI) and subject to confocal microscopy using a Zeiss LSM510META laser confocal microscope. Images of embryo were acquired as serial sections along the z-axis at 1.0 µm-interval and processed by using Zeiss LSM Image Browser and Adobe Photoshop CS2.

3. Results

3.1. Gal4FF gene trap and enhancer trap screens

A scheme for the Gal4FF gene trap and enhancer trap screening is shown in Fig. 3A. A plasmid DNA carrying T2KSAGFF or T2KhspGFF was injected into fertilized eggs with Tol2 transposase mRNA synthesized in vitro. In each microinjection experiment, 8–16 injected embryos were analyzed by the excision assay to confirm that transposition occurred in the injected embryos (Fig. 3A). When the excision product is detected in less than 70–80% of injected embryos, this indicates that transposase did not work properly in the injected embryos and they should be discarded to avoid wasting the labor of raising and mating unintegrated founder fish. In such cases, we recommend performing microinjection experiments again.

Figure 3
The Gal4FF gene trap and enhancer trap screens

The injected fish were raised to adulthood and crossed with homozygous UAS:GFP reporter fish. It is important to maintain sufficient numbers of UAS:GFP reporter fish of nearly the same age and body size as the injected fish to facilitate mating and screening. For each cross, more than forty F1 embryos were analyzed for GFP expression on 1 day and 5 days postfertilization (Fig 3B). Observations may be performed at any stage of interest. In some cases, F1 fish from the same founder fish showed multiple GFP expression patterns, indicating that more than one insertion causing Gal4 expression is transmitted to the F1 fish. The F1 fish that showed different GFP expression patterns should be kept separately in different fish tanks and analyzed by Southern blot hybridization.

The 213AGal4FF gene trap line expressed GFP in the central nervous system when crossed with the UAS:GFP fish (Fig. 3C). The 213AGal4FF;UAS:GFP double transgenic fish was crossed with the UAS:RFP reporter fish. The resulting 213AGal4FF;UAS:RFP double transgenic embryos expressed RFP in the same expression pattern as GFP, indicating that the Gal4FF transgenic fish can express different genes downstream of UAS in the same pattern (Fig. 3C).

3.2. Molecular characterization of Gal4FF trap lines

First, the GFP-positive F1 fish are analyzed by Southern blot hybridization to identify fish with single insertions that are responsible for the observed expression patterns. When only fish with multiple insertions are identified, F1 fish with the smallest number of insertions are outcrossed to obtain fish with single or fewer insertions in the F2 generation. Second, integration sites of the Tol2 construct are analyzed by inverse PCR and adaptor-ligation PCR using genomic DNA from fish with single insertions. The DNA sequences surrounding the Tol2 insertion are analyzed by using the ensembl database. Third, based on the genomic information obtained from the database analysis, genes (enhancers) that cause the observed Gal4FF expression patterns may be identified. In the case of gene trapping, a transcript trapped by the gene trap insertion is identified by analyzing mRNA. We describe the characterization of the SAGFF73A insertion as an example.

Fish injected with the T2KSAGFF plasmid and the transposase mRNA were crossed with the UAS:GFP reporter fish (Fig. 3A) and F1 embryos that expressed GFP in the whole body at 1 day postfertilization were identified and named SAGFF73A (Fig. 4A). The F1 fish carried six insertions (Fig. 4B). The F1 fish was outcrossed and F2 fish with three insertions were identified (Fig. 4C). These F2 fish had one insertion in common, suggesting that the insertion was responsible for the GFP expression. The F2 fish was further outcrossed and F3 fish with single insertions were identified (Fig. 4D). When these F3 fish were crossed with wild type fish, the GFP expression pattern was transmitted to the next generation in a Mendelian fashion.

The SAGFF73A;UAS:GFP fish was crossed with wild type fish and DNA samples prepared from 8 embryos were analyzed by PCR. Then, the DNA sample that contained the SAGFF73A insertion but not the UAS:GFP reporter gene was used for inverse PCR analysis. Single PCR products containing the Tol2-end sequences and flanking genomic sequences were amplified from both ends of the insertion (Fig. 4E). An 8-bp target site duplication was observed (Fig. 4F), confirming that the integration was created through transposition. The genomic DNA sequence was analyzed by using the ensemble database. The genomic sequence was mapped within the first intron of the zfand5b gene on chromosome 10 (Fig. 4G). The orientation of the Gal4FF gene was the same as that of the zfand5b gene (Fig. 4H). The expression pattern of zfand5b was consistent with the whole body GFP expression seen in the SAGFF73A;UAS:GFP embryos (ZFIN: Currently, in ~50% of the cases, the sequence can be mapped on the genome, and in the rest, the insertions can not be mapped because BLAST (BLAT) searches do not match any sequences or match more than one loci.

To determine whether the splice acceptor on the SAGFF73A insertion trapped the zfand5b transcript, we prepared total RNA from the SAGFF73A;UAS:GFP heterozygous embryos at 24 hpf and performed RT-PCR by using a primer in the 5’ non-coding exon of the zfand5b gene and a primer in the Gal4FF coding sequence (Fig. 4H). A fusion transcript was detected (Fig. 4I), indicating that the splice acceptor on the insertion indeed trapped the zfand5b transcript. Also, we performed RT-PCR using the same 5’ primer and a primer in the 3’ UTR (Fig. 4H). The full-length transcript of the zfand5b gene was reduced in the heterozygous embryos (Fig. 4I), suggesting that the insertion of the gene trap construct interfered with normal splicing.

3.3 Inhibition of neural functions by using the UAS:TeTxLC:CFP effector line

We performed the gene trap and enhancer trap screens and identified three trap lines, hspGGFF27A, SAGFF31B and SAGFF36B, that directed UAS:GFP reporter expression in different types of neurons (26). These Gal4FF;UAS:GFP double transgenic fish were crossed with the UAS:TeTxLC:CFP effector fish. Embryos that showed CFP expression, namely double transgenic for Gal4FF and UAS:TeTxLC:CFP, were collected and used for further analysis (Fig. 5A–D).

These embryos were analyzed by the touch response assay. When a tactile stimulus is applied to a wild type embryo with a needle, the embryo showed a rapid escape straight away from the stimulus (Fig. 5H). On the other hand, the hspGGFF27A;UAS:TeTxLC:CFP double transgenic embryo did not show the touch response. The embryos did not show spontaneous movement either, suggesting that motility was abolished in the embryo. The SAGFF31B;UAS:TeTxLC:CFP double transgenic embryo could respond to the touch but showed an abnormal escape behavior, suggesting that coordinated movement was defective in the embryo. The SAGFF36B;UAS:TeTxLC:CFP double transgenic embryo did not respond to the touch. However, the embryo could move spontaneously, suggesting that the embryo was defective in sensing the touch.

To identify neural circuits affected or inhibited by the TeTxLC:CFP expression in these embryos, we performed immunostaining of the TeTxLC:CFP fusion protein by using the anti-GFP antibody (Fig. 5E–G). These embryos were also treated with the anti-Hb9 antibody that stained motor neurons or the zn-12 antibody that stained Rohon-Beard sensory neurons. In the hspGGFF27A;UAS:TeTxLC:CFP embryo, strong TeTxLC:CFP expression in hindbrain interneurons were detected (Fig. 5E). In the SAGFF31B;UAS:TeTxLC:CFP embryo, the fusion protein was expressed in a number of spinal interneurons and some motor neurons (Fig. 5F). In the SAGFF36B;UAS:TeTxLC:CFP embryo, the fusion protein was predominantly expressed in Rohon Beard sensory neurons (Fig. 5G). These suggested that the observed abnormalities in the touch response behavior were caused by inhibition of, at least partly, the functions of these neural circuits by expression of the TeTxLC:CFP fusion protein.

4. Discussion and perspectives

4.1. The choice of Gal4FF

We employed Gal4FF to develop the Gal4-UAS transactivation system. Scheer and Campos-Ortega created transgenic fish by using the full-length Gal4 (15), and Koster and Fraser performed the transient gene expression analysis by using Gal4-VP16 since they thought that the full-length Gal4 was not strong enough in zebrafish (30). Although we have not conducted an exact comparison between Gal4FF and other Gal4 variants from the view of the transactivation activity, we think the Gal4FF-UAS system is highly efficient since GFP expression in the Gal4FF gene trap and enhancer trap lines described in this study is in general much stronger than those in trap lines created by using GFP in our previous studies (21, 23). In the previous studies that employed Gal4-VP16, toxic effects associated with the Gal4-VP16 expression have been observed in both transient and stable transgenesis (24, 30, 37, 38). We have not observed such toxicity in the embryos injected with 1 pg of Gal4FF mRNA while injection of the same amount of Gal4-VP16 mRNA caused severe developmental defects (26). Also, toxicity associated with the Gal4FF expression has not been observed in our gene trap and enhancer trap lines. We think that this difference is due to the transactivation domains. It was hypothesized that a phenomenon called “squelching”, i.e., titration of a host transcription machinery by VP16, is the cause of the toxicity (39). The FF activation domain may not cause “squelching”.

4.2. Background expression of the GFP reporter gene

In most of the hspGFF;UAS:GFP enhancer trap lines, GFP expression is observed in the heart at day 1 and in the heart and skeletal muscle at day 5 (26). Similar expression patterns were observed in other enhancer trap screens that use the hsp70 promoter and Gal4-VP16 (24) or the hsp70 promoter and GFP (23). Therefore, we think that these are likely to be caused by a basal activity of the hsp70 promoter. It may be possible to reduce the background expression by dissecting and making a shorter hsp70 promoter. The study along this line is in progress in our lab. Also, in ~20% of the SAGFF;UAS:GFP gene trap lines, GFP expression is observed in non-neuronal cells in the spinal cord weakly at days 3–5 in addition to specific GFP expression patterns (26). We think that the GFP expression is probably associated with the rabbit β-globin gene sequence on the T2KSAGFF construct since we do not detect such expression when the splice acceptor was changed (unpublished observations). An improved version of the gene trap construct is currently under construction. It should be noted that the background spinal expression did not affect the study of spinal neurons by using the tetanus toxin transgenic line (26).

4.3. Variegated (Mosaic) expression of UAS-transgene

The GFP reporter expression in our Gal4FF;UAS:GFP fish is reproducible and persists from generation to generation (at present, up to the F7 generation). Variegated (or mosaic) expression of UAS-reporter genes has been observed in transgenic lines expressing Gal4-VP16 under the control of specific enhancer/promoters that are created by the meganuclease (I-SceI)-mediated transgenesis method or by microinjection of the plasmid DNA (24, 31). The variegated expression has also been reported in the gene trap lines expressing Gal4-VP16 that were created by using Tol2-mediated transgenesis (25). In contrast, variegated expression of the UAS:GFP reporter was rarely observed in our enhancer trap and gene trap lines. The exact cause of variegated expression is not known. This may involve “squelching” described above or gene silencing which is often associated with transgenes integrated as concatemers. The use of Gal4FF and Tol2-mediated transgenesis may avoid these problems. Or alternatively, we utilized five tandem repeats of the Gal4 recognition sequence (5×UAS) while in most of other studies 14 (or more) repeats of UAS were utilized which may increase the chance for suffering methylation since the UAS sequence itself is highly G:C rich (40). It should be noted that in our hspGFF;UAS:GFP enhancer trap lines the background GFP expression in the skeletal muscle is variegated unlike their specific expression patterns (data not shown). It is important to elucidate the cause of variegated expression to achieve targeted expression of effector genes with high accuracy.

4.4. Position effects on UAS-transgene expression

We have observed different levels of reporter and effector gene expression when the same Gal4FF fish were crossed with fish carrying the same UAS transgene construct at different chromosomal loci, suggesting that expression from UAS construct is sensitive to position effects. Therefore, it is practically important to create tens of different UAS transgenic fish and select one that gives rise to a high level of expression when crossed with a Gal4FF tester line. In fact, the UAS:TeTxLC:CFP fish was established by testing ~75 different insertions for their abilities to inhibit neural functions. It is interesting to note that UAS-reporter and UAS-effector insertions thus established are often located within regions that are thought to be transcriptionally competent or active.

4.5. Applications of the Gal4 gene trap and enhancer trap methods

The Gal4 gene trap and enhancer trap methods can be applied to manipulate the function of specific cell types. We demonstrated that neuronal functions can be inhibited by the UAS:TeTxLC:CFP effector fish (26). The results described here suggest that the Gal4-expressing neural circuits (hindbrain interneurons, spinal interneurons and sensory neurons) regulate distinctive behaviors (motility, coordinated movements and sensing the touch). In vertebrates, such analyses have been carried out mostly by electrophysiological approaches that use fixed or spinalized animals (41). Our system enables the study of the neuronal functions by using intact vertebrate embryos. Davison et al. developed transgenic fish that carried the nitroreductase gene downstream of UAS. By crossing the effector fish with Gal4 lines, cells in the floorplate and the notochord were successfully eliminated (25). In the future, development of various UAS effector lines that inhibit or modify cellular functions should increase the usefulness of the Gal4 expressing lines in the study of the vertebrate system.


This work was supported by a postdoctoral fellowship from the Japan Society for the Promotion of Science (to K.A.), a Sasakawa Scientific Research Grant from the Japan Science Society (to K.A.), NIH/NIGMS R01 GM069382, the National BioResource project and grants from the Ministry of Education, Culture, Sports, Science and Technology of Japan.

The Tol2 constructs and fish lines described here are available upon request.


Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.


1. Keegan L, Gill G, Ptashne M. Science. 1986;231:699–704. [PubMed]
2. Ma J, Ptashne M. Cell. 1987;48:847–853. [PubMed]
3. Fischer JA, Giniger E, Maniatis T, Ptashne M. Nature. 1988;332:853–856. [PubMed]
4. Brand AH, Perrimon N. Development. 1993;118:401–415. [PubMed]
5. Driever W, Solnica-Krezel L, Schier AF, Neuhauss SC, Malicki J, Stemple DL, Stainier DY, Zwartkruis F, Abdelilah S, Rangini Z, Belak J, Boggs C. Development. 1996;123:37–46. [PubMed]
6. Haffter P, Granato M, Brand M, Mullins MC, Hammerschmidt M, Kane DA, Odenthal J, van Eeden FJ, Jiang YJ, Heisenberg CP, Kelsh RN, Furutani-Seiki M, Vogelsang E, Beuchle D, Schach U, Fabian C, Nusslein-Volhard C. Development. 1996;123:1–36. [PubMed]
7. Amsterdam A, Burgess S, Golling G, Chen W, Sun Z, Townsend K, Farrington S, Haldi M, Hopkins N. Genes Dev. 1999;13:2713–2724. [PubMed]
8. Amsterdam A, Nissen RM, Sun Z, Swindell EC, Farrington S, Hopkins N. Proc Natl Acad Sci U S A. 2004;101:12792–12797. [PubMed]
9. Golling G, Amsterdam A, Sun Z, Antonelli M, Maldonado E, Chen W, Burgess S, Haldi M, Artzt K, Farrington S, Lin SY, Nissen RM, Hopkins N. Nat Genet. 2002;31:135–140. [PubMed]
10. Wienholds E, Schulte-Merker S, Walderich B, Plasterk RH. Science. 2002;297:99–102. [PubMed]
11. Wienholds E, van Eeden F, Kosters M, Mudde J, Plasterk RH, Cuppen E. Genome Res. 2003;13:2700–2707. [PubMed]
12. Meng X, Noyes MB, Zhu LJ, Lawson ND, Wolfe SA. Nat Biotechnol. 2008;26:695–701. [PMC free article] [PubMed]
13. Doyon Y, McCammon JM, Miller JC, Faraji F, Ngo C, Katibah GE, Amora R, Hocking TD, Zhang L, Rebar EJ, Gregory PD, Urnov FD, Amacher SL. Nat Biotechnol. 2008;26:702–708. [PMC free article] [PubMed]
14. Nasevicius A, Ekker SC. Nat Genet. 2000;26:216–220. [PubMed]
15. Scheer N, Campos-Ortega JA. Mech Dev. 1999;80:153–158. [PubMed]
16. Kawakami K. Genome Biol. 2007;8 Suppl 1:S7. [PMC free article] [PubMed]
17. Kawakami K, Koga A, Hori H, Shima A. Gene. 1998;225:17–22. [PubMed]
18. Kawakami K, Shima A. Gene. 1999;240:239–244. [PubMed]
19. Kawakami K, Shima A, Kawakami N. Proc Natl Acad Sci U S A. 2000;97:11403–11408. [PubMed]
20. Urasaki A, Morvan G, Kawakami K. Genetics. 2006;174:639–649. [PubMed]
21. Kawakami K, Takeda H, Kawakami N, Kobayashi M, Matsuda N, Mishina M. Dev Cell. 2004;7:133–144. [PubMed]
22. Parinov S, Kondrichin I, Korzh V, Emelyanov A. Dev Dyn. 2004;231:449–459. [PubMed]
23. Nagayoshi S, Hayashi E, Abe G, Osato N, Asakawa K, Urasaki A, Horikawa K, Ikeo K, Takeda H, Kawakami K. Development. 2008;135:159–169. [PubMed]
24. Scott EK, Mason L, Arrenberg AB, Ziv L, Gosse NJ, Xiao T, Chi NC, Asakawa K, Kawakami K, Baier H. Nat Methods. 2007;4:323–326. [PubMed]
25. Davison JM, Akitake CM, Goll MG, Rhee JM, Gosse N, Baier H, Halpern ME, Leach SD, Parsons MJ. Dev Biol. 2007;304:811–824. [PMC free article] [PubMed]
26. Asakawa K, Suster ML, Mizusawa K, Nagayoshi S, Kotani T, Urasaki A, Kishimoto Y, Hibi M, Kawakami K. Proc Natl Acad Sci U S A. 2008;105:1255–1260. [PubMed]
27. Asakawa K, Kawakami K. Dev Growth Differ. 2008;50:391–399. [PubMed]
28. Sadowski I, Ma J, Triezenberg S, Ptashne M. Nature. 1988;335:563–564. [PubMed]
29. Baron U, Gossen M, Bujard H. Nucleic Acids Res. 1997;25:2723–2729. [PMC free article] [PubMed]
30. Koster RW, Fraser SE. Dev Biol. 2001;233:329–346. [PubMed]
31. Sagasti A, Guido MR, Raible DW, Schier AF. Curr Biol. 2005;15:804–814. [PubMed]
32. Inbal A, Topczewski J, Solnica-Krezel L. Genesis. 2006;44:584–588. [PubMed]
33. Campbell RE, Tour O, Palmer AE, Steinbach PA, Baird GS, Zacharias DA, Tsien RY. Proc Natl Acad Sci U S A. 2002;99:7877–7882. [PubMed]
34. Schiavo G, Benfenati F, Poulain B, Rossetto O, Polverino de Laureto P, DasGupta BR, Montecucco C. Nature. 1992;359:832–835. [PubMed]
35. Harms KJ, Tovar KR, Craig AM. J Neurosci. 2005;25:6379–6388. [PubMed]
36. Kotani T, Nagayoshi S, Urasaki A, Kawakami K. Methods. 2006;39:199–206. [PubMed]
37. Argenton F, Arava Y, Aronheim A, Walker MD. Mol Cell Biol. 1996;16:1714–1721. [PMC free article] [PubMed]
38. Emelyanov A, Parinov S. Dev Biol. 2008;320:113–121. [PubMed]
39. Gill G, Ptashne M. Nature. 1988;334:721–724. [PubMed]
40. Halpern ME, Rhee J, Goll MG, Akitake CM, Parsons M, Leach SD. Zebrafish. 2008;5:97–110. [PubMed]
41. Sillar KT, Roberts A. Nature. 1988;331:262–265. [PubMed]