Schematics of the new RNAi backbone for zebrafish gene silencing. (A) Tol2kit compatible p3E presenting an optimized empty synthetic pri-miR (or RNAi cassette) embedded into a β-globin intronic sequence. This presented pri-miR is designed to allow rapid directional insertion of synthetic pre-miR of choice. Synthetic pre-miR(s) of choice are generated by annealing specific top and bottom RNAi oligos designed to release a mature miRNA directed against the 3′UTR of a gene-of-interest [see ‘Methods’ section and previous material (5)]. The RNAi cassette is flanked by restriction sites allowing subsequent and repetitive chaining [see ‘Methods’ and Supplementary Figure S2) for generating p3E-RNAi plasmid with multiple pri-miR for either increasing the potency of the desired knockdown or targeting multiple genes at the same time. (B) Without intron, the pre-miR/RNAi cassette is transcribed along with a co-marker on the same RNA, which is cut by Drosha in the nucleus for releasing the associated pre-miR. This cut leaves the mRNA without polyA tail and leads to its rapid degradation. The associated fluorescence is not a good indicator of the activity and amount of synthetic miRNA produced. (C) The presence of the intronic sequence is designed to rescue the co-expression of the marker (see also Supplementary Figure S1).

Validating effect of intron based RNAi approach on endogenous miRNA processing. (A–C), (E–G) and (I–K) Ubiquitously expressing mCherry fluorescent embryos from Tg(ubi:mCherry:SPmiR137) incrosses were injected with 25 pg of Tol2 transposase combined with 50 pg of control transgenes 157-control or anti-miR137 expressing constructs—153 (without intron), 155 (with intron). All injected plasmids led to mosaic eGFP expression. Plasmids 153 and 155, but not 157-control, led to a reduction of red fluorescence, hence knockdown of the integrated mCherry sensor transgene. A minimum of six larvae were analyzed per condition to estimate knockdown efficiency. Graphs D, H and L are representations of pixel intensities (each dot presents a grey value) of green versus red fluorescence (presented as a percentage) from z-stacks imaged with a confocal microscope at 4 day-post-fertilization (dpf). Scale bar is 100 μm.

Validation of intron based RNAi approach on rescuing co-expression of fluorescent markers. (A–D), (E–H) and (I–L) Heterozygous Tg(ubi:mCherry:SPmiR137) embryos with ubiquitous mCherry expression were injected with Tol2 transposase and 50 pg of 4× anti-miR137 miRNAs construct without intron (plasmid 154) and with intron (plasmid 156), including a control (plasmid 157). A minimum of five larvae were imaged at 4 dpf using a confocal microscope to correlate the presence of 4× anti-miR137 expressing eGFP (with and without intron) to decrease/absence of mCherry fluorescence. Pixel intensities (grey values) of green versus red fluorescence (presented as a percentage) plotted in graphs H demonstrate no correlation of fluorescence in embryos injected with plasmid 154, similar to control 157. Comparatively, a linear correlation of green versus red fluorescence (L) is evidenced in embryos injected with plasmid with intron (plasmid 156).

Conditional Cre/Lox RNAi genetic system. (A) Schematic representation of the driver transgene, with here a ubiquitous promoter, named ubi:iCre. Tissue-specific promoters can obviously be used. (B) Schematic representation of the silent/conditional gene-silencing responder transgene. In the absence of iCRE, only BFP will be expressed and the animal will remain unaffected by the RNAi cassette, thereby could be maintained without special care. In the presence of iCre, genetic recombination will lead to excision of the BFP cassette and expression of mKate along with the miRNAs of choice; with here 4× anti-smn1 miRNAs, resulting in potent zebrafish smn1 knockdown. This recombination is irreversible.

Representative snapshots and phenotypic analysis of 50 hpf zebrafish larvae with unfloxed (A–D) or floxed (E–H) integrated RNAi transgene. (A–D) In the absence of iCRE, the transgene remains unfloxed triggering BFP expression and not inducing any significant developmental defect in the analyzed embryos. (E–H) In the presence of iCRE, BFP is excised and mKate expressed along with anti-smn1 miRNAs, triggering SMA-hallmarks such as motoneuron defects. Images represent standard deviation projections from confocal z-stack acquisitions using three channels (merged) to detect eGFP (green, motoneurons), tagBFP (blue, ubiquitous expression) and mKate (red, ubiquitous expression). White arrows indicate abnormal CaP motoneurons. (I) Number of ventral motoneuron (CaP) abnormalities observed per side of 50 hpf larvae. ubi-floxed tg(loxSMN) animals were also injected with hsa-SMN1 mRNA to test the specificity of the observed phenotype. (J) Animal size comparison at 52 hpf. (K) Survival assay (Kaplan Meier) for the different lines and conditions tested. Means of 20 larvae ± SEM.

Cell-specific miR-delivery/RNAi in zebrafish does not require transposase. We tested the ability of the presented miR-delivery system to enable rapid cell-specific experiments. The approach is similar to the method conducted in Figures 2 and 3 but aimed at evaluating the effect of different injection mix. (A) Schematic example of the expression obtained following the injection of a motoneuron-specific miR-delivery construct (83_HB9_mKate_smn4141) within the cell of one-cell stage tg(MN:GFP) animals. Arrowhead indicating mKate positive motoneurons with positive transgene expression. (B) Titration/Toxicity test of Tol2-construct injections (83_HB9_mKate_smn4141) for triggering cell-specific expression. (C) Number (percentage) of animals presenting at least 1× mKate positive motoneuron. (D) Average number of mKate-positive CaP motoneurons per positive animal. Eighty embryos per condition have been analyzed (means ± SEM). These results suggest that a dose of 50 pg without transposase may be ideal for setting up cell-specific approaches and studies.