A novel inducible mutagenesis screen enables to isolate and clone both embryonic and adult zebrafish mutants

Conventional genetic screens for recessive mutants are inadequate for studying biological processes in the adult vertebrate due to embryonic lethality. Here, we report that a novel inducible mutagenesis system enables to study gene function in both embryonic and adult zebrafish. This system yields genetic mutants with conditional ectopic over- or under-expression of genes in F1 heterozygotes by utilizing inducible Tet-On transcriptional activation of sense or anti-sense transcripts from entrapped genes by Tol2 transposase-meditated transgenesis. Pilot screens identified 37 phenotypic mutants displaying embryonic defects (34 lines), adult fin regeneration defects (7 lines), or defects at both stages (4 lines). Combination of various techniques (such as: generating a new mutant allele, injecting gene specific morpholino or mRNA etc) confirms that Dox-induced embryonic abnormalities in 10 mutants are due to dysfunction of entrapped genes; and that Dox-induced under-expression of 6 genes causes abnormal adult fin regeneration. Together, this work presents a powerful mutagenesis system for genetic analysis from zebrafish embryos to adults in particular and other model organisms in general.


Construction of the first generation pIDM vector:
The first step: The rtTA DNA fragment was amplified with the primer pair rtTA-BamHI-F and -R from the pT2HB-PUHrT62 plasmid. The amplified rtTA DNA fragment was then digested with BamHI and cloned into a pIRES 2 -EGFP plasmid (Clontech) to generate the prtTA-IRES2-EGFP construct. The prtTA-IRES2-EGFP construct was then used to amplify the rtTA-IRES-Egfp-polyA fragment with the primer pair rtTA-IRES-Egfp-SV40-NcoI-F and -R.
The third step: The NcoI digested rtTA-IRES-Egfp-polyA fragment from the first step was ligated into the pT2-β-actin/elf1a-TRE3G vector from the second step. The TRE3G-β-actin-rtTA-IRES-Egfp-polyA fragment was then amplified with TRE3G-SwaI-F and polyA-HindIII-R primers and digested with SwaI and HindIII. The restriction enzyme digested fragment was cloned into a pTol2mini plasmid to generate the first generation pIDM vector.

Construction of the second generation pIDM vector:
A homologous recombination ligation method was used in the construction of the second generation pIDM vector.

Linker-mediated PCR for identification of pIDM insertion sites
Linker annealing was performed according to the protocol described in a previous study 2 . A cocktail containing 50 μl of 25 μM AluI or BfaI linker+, 50 μl of 25 μM AluI or BfaI linkerand 2 μl of 5 M NaCl was incubated at 95 °C for 5 minutes and then slowly cooled down to room temperature to allow the linkers to anneal. Annealed linkers were stored at -20°C.
Linker-mediated PCR was performed as described previously 3 . Briefly, 3 μg genomic DNA was digested with AluI or BfaI overnight. The digested DNA was purified and dissolved in 30 μl H 2 O. The Linker DNA together with the digested genomic DNA fragments was ligated with T4 DNA ligase (NEB). The ligation solution was used as the template for the first-round PCR with a pair of Linker primer and Tol2 5' N1 primer. Then the first-round PCR product was used as the template for the second-round PCR with another pair of primers: Linker AluI (BfaI) nested primer and Tol2 5' N2 primer.
After the second-round of PCR, the PCR products were purified using a PCR clean kit (Axygen) according the manufacturer's instructions and then ligated to a pGEMT-Easy vector (Promega). The Tol2 5' S1 primer was used to sequence the PCR products. The sequence of the PCR product was blasted with the database GRCz10. The sequence of genomic DNA with a high identity (>95%) to the PCR product was considered to be the insertion site. If the similarity between the PCR product and genomic sequences was less than 95%, the insertion site was considered as non-identified (N.D).

Segregation of mutation responsible for fin regeneration defect from multiple-inserted lines
From line pIDM-A28, LM-PCR amplified two insertion sites. The first insertion located in the 1 st intron of galnt2 gene with the Tet-On promoter oriented in the opposite direction (Fig.   6A). The second insertion was unable to be identified due to that the sequence of second LM-PCR amplified DNA fragment has high similarity to a number of genomic DNAs at different locations. We crossed fin regeneration defective F 1 mutants with WT to set up F 2 population. About 50% EGFP positive embryos was found in the offspring of one of these crosses, which suggested that this F 1 mutant fish may contain single insertion. Then we raised the EGFP-positive embryos from this cross to generate F 2 generation. Using a pair of galnt2 and insertion specific primers, the insertion within the galnt2 gene was identified in all of the F2 fish (Fig. S17A). Southern blot showed that only one insertion site was detected in the F 2 offspring from this F 1 mutant as well as the F 3 offspring from one of the F 2 mutant adult fish (Fig. S17A). The results demonstrated that all of these F 2 mutants had only one insertion within the galnt2 gene.
There were three insertion sites in both pIDM-A3 and E7. Single-inserted mutant was segregated from F2 individuals with fin regeneration defects in both lines and the EGFP positive embryos from the F2 individuals were used to generate F3 population. The single insertion was confirmed by southern blot with the F3 offspring from the F2 mutant as well as the F4 offspring from the F3 population in both lines (Fig. S17C,D). The insertion in segregated pIDM-A3 and E7 located in the 4th intron of the γ-glutamyltransferase7-like (ggt7l) gene and the 1st intron of cry61 gene with the Tet-on promoter oriented in the opposite direction (Fig. 6A), respectively. Using a pair of gene and insertion specific primers, the insertion site was identified in all of the F3 mutant fish from both of lines (Fig. S17C,D).
Dox-dependent fin regeneration defective phenotype was also observed in the F3 population

sema5ba-R TGCGTGTTCGTTGGTAGTGT
Primers for qRT-PCR to analyze the relative levels of sense mRNA