INVESTIGATION

Stable Silencing in Zebrafish with Spatiotemporally Targetable RNA Interference

Zhiqiang Dong,1 Jisong Peng,1 and Su Guo2 Department of Bioengineering and Therapeutic Sciences, Programs in Human Genetics and Biological Sciences, University of California, San Francisco, California 94143-2811

ABSTRACT The ability to regulate gene activity in a spatiotemporally controllable manner is vital for biological discovery that will impact disease diagnosis and treatment. While conditional is possible in other genetic model , this technology is largely unavailable in zebrafish, an important vertebrate model for functional gene discovery. Here, using short hairpin RNAs (shRNAs) designed in the microRNA-30 backbone, which have been shown to mimic natural microRNA primary transcripts and be more effective than simple shRNAs, we report stable RNA interference-mediated gene silencing in zebrafish employing the yeast Gal4-UAS system. Using this approach, we reveal at single-cell resolution the role of atypical protein kinase Cl (aPKCl) in regulating neural progenitor/stem cell division. We also show effective silencing of the one-eyed-pinhead and no-tail/brachyury . Further- more, we demonstrate stable integration and germ-line transmission of the UAS-miR-shRNAs for aPKCl, the expressivity of which is controllable by the strength and expression of Gal4. This technology shall significantly advance the utility of zebrafish for understand- ing fundamental vertebrate biology and for the identification and evaluation of important therapeutic targets.

TUDIES of genetic model organisms contribute tremen- RNA interference (RNAi) is a powerful approach to Sdously to our understanding of diverse biological pro- dissect gene function (Fire 2007; Mello 2007). It was orig- cesses and human disorders at molecular and cellular levels inally discovered in , where the obser- (Davis 2004; Aitman et al. 2011). As a recently established vation of gene inactivation by both sense and antisense animal model, zebrafish has salient features that promise to RNAs (Guo and Kemphues 1995) led to the finding of double- provide new insights into biology and medicine. These in- stranded RNA (dsRNA)-mediated gene silencing (Fire et al. clude rapid external development, transparent embryonic 1998). In vertebrates, the effectiveness of RNAi was hindered and larval stages, simpler vertebrate organ systems, and by the dsRNA-induced interferon response causing nonspecific amenability for genetic and chemical screening. The utility effects (Manche et al. 1992; Stark et al. 1998) until the finding of zebrafish, however, has been hampered by the inability of efficacious small interfering RNAs (siRNAs) (Elbashir et al. to silence genes in a spatiotemporally controllable manner. 2001). However, their delivery into zebrafish remains either Although valuable reverse genetic methods are available ineffective or has nonspecific effects (Gruber et al. 2005; (Nasevicius and Ekker 2000; Wienholds et al. 2002; Doyon Zhao et al. 2008). Likewise, simple short hairpin RNAs et al. 2008; Meng et al. 2008; Huang et al. 2011; Sander (shRNAs) driven by RNA polymerase III promoters (Paddi- et al. 2011; Bedell et al. 2012), none offer gene silencing at son et al. 2002) also appears ineffective in zebrafish (Wang any desired stages of the life cycle and in any cell types of et al. 2010). interest. Micro-RNAs (miRNAs) are endogenous 21- to 23-nucleotide RNAs that can regulate . Originally discov- ered in C. elegans (Lee et al. 1993; Reinhart et al. 2000), they

Copyright © 2013 by the Genetics Society of America were later found conserved across animals and plants (Bartel doi: 10.1534/genetics.112.147892 2009; Ebert and Sharp 2012). The observation that de- Manuscript received November 21, 2012; accepted for publication January 12, 2013 signed miRNAs can inhibit cognate mRNA expression in Supporting information is available online at http://www.genetics.org/lookup/suppl/ doi:10.1534/genetics.112.147892/-/DC1. human cells (Zeng et al. 2002) has paved the way for recent 1These authors contributed equally to this work. developments of shRNAs employing the primary miR-30 2Corresponding author: Department of Bioengineering and Therapeutic Sciences, Programs in Human Genetics and Biological Sciences, University of California, San backbones (miR-shRNAs) (Dickins et al. 2005; Silva et al. Francisco, CA 94143-2811. E-mail: [email protected] 2005). This natural configuration is reported 12 times more

Genetics, Vol. 193, 1065–1071 April 2013 1065 efficient than simple hairpin designs. MiR-shRNAs are re- DNA and RNA injection cently shown to be effective in vivo in mice (Dickins et al. DNA and RNA injections were performed at the one- 2005; Premsrirut et al. 2011; Zuber et al. 2011) and zebra- cell stage. For genetic mosaic analyses, RNAi constructs fish (Dong et al. 2009; De Rienzo et al. 2012). However, the (UAS-TdTomato-shRNA1apkcl-miniTol2 or UAS-TdTomato- capability of germ-line transmission is only reported by one shRNA5apkcl-miniTol2; 75 pg per ) were co-injected study in zebrafish (Dong et al. 2009), where functional ef- with EF1a-GFF-PT2KXIG (elongation factor 1a regulatory ficacy and specificity of silencing were not quantitatively element-driven Gal4) (50 pg per embryo) and Tol2 tranpo- documented. Therefore, the potential of RNAi for stable sase (10 pg per embryo). Sense strand-capped RNA was and conditional gene silencing in zebrafish remains synthesized by SP6 from NotI-linearized plas- uncertain. mid using the mMESSAGE mMACHINE system (Ambion). With the goal of achieving spatiotemporal and dosage RNAi-resistant sense RNAs for apkcl, ntla and οepa, as well control of gene silencing, we have exploited the miR-shRNA as RNAs encoding the oep-shRNA and ntla-shRNA, were technology in combination with the bipartite Gal4/upstream injected at 400 pg per embryo. For time-lapse in vivo imag- activating sequence (UAS) system (Ptashne 1988), which ing, 3NLS:EGFP sense RNA (200 pg per embryo) was co- offers excellent versatility for spatiotemporal control to- injected with EF1a-GFF-PT2KXIG DNA (50 pg per embryo) gether with amplification of gene expression level. Using and 5xUAS-tdTomato-shRNAapkcl plasmid DNA (75 pg per atypical protein kinase Cl (apkcl), one-eyed pinhead (oep), embryo). and no tail/brachyury (ntl) as test loci, we show that RNAi in fi zebrafish is effective, rapid, dosage controllable, stable with Heat shock of zebra sh conditional capability, and scalable. Heat shock of zebrafish embryos was performed at the stage of 8 hours postfertilization (hpf) (75% epiboly). Fertilized embryos were collected from the cross between the UAS- Materials and Methods TdTom-miR-shRNA1apkcl founder A and the hsp70-Gal4 transgenic animal, and were raised at 28.5 before heat Zebrafish strains shock. A 500-ml glass beaker filled with 100 ml egg water Wild-type embryos were obtained from natural spawning of was preheated to 37 in a water bath right before heat AB adults. The transgenic line Tg[ubi-GFF] (Asakawa and shock. Around 100 embryos were transferred to the pre- Kawakami 2010) was a gift from K. Kawakami. The apkcl heated egg water at 8 hpf and incubated in a 37 water bath mutant heart and soul was kindly provided by D. Stainier. for 1 hr. The embryos were then transferred to 28.5 egg water immediately and raised in a 28.5 incubator until ShRNA design analysis. ShRNA design was performed using the siRNA design tool Phenotypic analyses from the following website: http://www.genscript.com/ design_center.html, which ranks target candidates based Phenotypic analyses were performed manually under a bright on a proprietary algorithm using DE parameters. We further field dissection microscope. For apkcl RNAi, four morphologi- refined the list by removing the candidates that have more cal phenotypes were analyzed, which are typically observed than three G/C in 6–11 nt of the target sequences, based on in the apkcl mutant (heart and soul) (Horne-Badovinac et al. a thermodynamic study of siRNA (Ui-Tei et al. 2008). 2001): (1) defect in heart tube assembly, (2) patchy retinal pigmented epithelium, (3) failure of the brain ventricle to Vector construction inflate, and (4) abnormal body curve. The number of em- The shRNA precursor structure was designed based on the bryos that are positive for each phenotype as well as the total miR-30e precursor in zebrafish. The miR-30e target and number of all embryos was counted for both RNAi groups and guide sequences were replaced with the predicted shRNA control groups. Statistic analyses were performed based on target and guide sequences. To best mimic the miR-30e positive percentage for each phenotype. For ntla and oep precursor structure, we kept the original loop region and the RNAi, the typical no-tail phenotype and one-eyed pinhead flanking sequences of the miR-30e precursor and introduced phenotype were analyzed. two mismatched “TT” into the shRNA guide sequence be- Time-lapse in vivo imaging tween nt 12 and nt 13. We synthesized a pair of comple- mentary oligos that include all the designed shRNA To observe the division orientation of radial glia progenitor structures (guide sequence, loop sequence, target sequence, cells, time-lapse in vivo imaging was performed on embryos and the flanking sequences) and the annealed oligos were injected with 3NLS:EGFP sense RNA, EF1a-GFF-PT2KXIG inserted into zebrafish miR-30e backbone in expression vec- DNA, and 5xUAS-tdTomato-shRNAapkcl-miniTol2 DNA (see tors kindly provided by the late Ting Xi Liu (Dong et al. DNA and RNA injection for detail) at 24 hpf. The imaging 2009). The following 5xUAS sequence was used: cggagtactg was performed as previously described (Dong et al. 2012). tcctccgagcggagtactgtcctccgagcggagtactgtcctccgagcggagtactgtcct In brief, embryos were mounted in low melting point aga- ccgagcggagtactgtcctccg. rose and properly placed on a temperature-controlled stage

1066 Z. Dong, J. Peng, and S. Guo of the confocal microscope. We used a Nikon C1 spectral HEX-modified gapdhs probe (59-HEX-TTCCAGTGCATGAAGCC confocal microscope with upright objectives. Fluorescently TGCTGAGAT-39) were used. Data were analyzed using Quanta- labeled individual neural progenitor cells were imaged for soft version 1.2.10.0. 2hrwithafixed interval of 90 sec under a ·40 water-dipping Stable transgenesis objective. The parameters of confocal imaging were deter- mined as sufficient to capture the orientation of cell division, Stable UAS-miR-shRNAapckcl transgenic lines were generated while reducing photobleaching during the imaging period. using the Tol2 transposon system as described previously (Suster et al. 2009). RNAi constructs (UAS-TdTomato-MiR- Immunohistochemistry shRNA1apkcl-miniTol2 or UAS-TdTomato-miR-shRNA5apkcl- Immunohistochemistry was performed on whole mount miniTol2; 50 pg per embryo) together with Tol2 transposase embryos as described previously (Guo et al. 1999). Embryos (10 pg per embryo) were injected into one-cell-stage em- fi were fixed with 4% paraformaldehyde. Embryos older than bryos of AB wild-type sh. Injected embryos were raised 60 hpf were treated with proteinase K before incubation in to adulthood (G0) and screened for germ-line transmission fi primary antibodies (rabbit anti-aPKC 1:100, Santa Cruz sc- by crossing with Tg[Ubi-GFF] transgenic sh. The germ-line 216) at 4 overnight. After four washes with phosphate- transgenic founders would yield embryos with TdTomato buffered saline with 0.1% Tween, 0.5% Triton X-100, and expression, which could be easily distinguished under the fl 1% DMSO, the embryos were incubated with Alexa Fluor epi uorescent microscope. The embryos with ubiquitous secondary antibody (1:2000, Invitrogen) at 4 overnight. TdTomato expression were raised to adulthood (F1), which apckcl After wash and glycerol sequential treatment, the embryos were carry both UAS- RNAi and Ubi-GFF . Addi- mountedandimagedusingNikonC1confocalmicroscope. tionally, UAS-RNAi G0 germ-line founders were crossed with AB wild-type fish. All the progenies were raised to adult- – Quantitative real-time RT PCR analyses hood and genotyped by PCR using TdTomato-specificpri- For the analyses of knockdown of apkcl expression by tran- mers to screen for stable transgenic F1 fish. Stable transgenic l sient RNAi, the control or RNAi embryos were collected at F1 were obtained for UAS-TdTomato-shRNA1apckc and UAS- l 10 hpf and total RNA from pooled embryos (20 embryos per TdTomato-shRNA5apckc . sample for each group) was extracted using Trizol reagent (Invitrogen) followed by treatment with Turbo DNA-free Results and Discussion DNase (Ambion). First-strand cDNA was reverse transcribed Conditional miR-shRNA expression system using oligo-dT primers and Superscript reverse transcriptase (Invitrogen). Real-time PCR amplifications of apkcl, oep, We designed shRNAs employing the primary miR-30 back- ntla, and gapdhs were carried out with SYBR Green PCR bones (hence called miR-shRNAs) (Figure 1A). The fluores- Master Mix (Applied Biosystems). Primers used were: cent reporter TdTomato conveniently marked the cells that express the miR-shRNAs. Three genes were used as the test apkcl F: 59-CCCGCACCAAGTCCGGGTAA-39, set: (1) The atypical protein kinase C lambda (aPKCl), for apkcl R: 59-GCTGAGAAGAAACGGTGCACGGA-39; which a distinct loss-of-function phenotype has been docu- oep F: 59-ATGGACTTTTGCATTGCTTCCCACA-39, mented in zebrafish. We chose this gene also because its oep R: 59-AGTGTTCTGAGGGAGCCCGACC-39; disruption produces a highly specific cellular phenotype, ntla F: 59-CCCAGCCATTACTCCCACCGC-39, that is, an alteration of spindle rotation in dividing radial ntla R: 59-TGGGCCAGGGTTCCCATCCC-39; glia progenitor cells (Horne-Badovinac et al. 2001). (2) The gapdhs F: 59-ACTCCACTCATGGCCGTTAC-39, no tail a (ntla) and (3) one-eyed pinhead (oep) genes, for gapdhs R: 59-TGAGCTGAGGCCTTCTCAAT-39. which distinct loss-of-function phenotypes can be readily assessed (Schulte-Merker et al. 1994; Zhang et al. 1998). Droplet digital PCR analyses Using the Web-based shRNA design tool (www.genescript. com) together with the filtering criteria based on thermody- For the analyses of knockdown of apkcl expression by stable namic properties (Ui-Tei et al. 2008), six shRNAs targeting RNAi transgenesis, the sibling control or RNAi transgenic the apkcl gene (Supporting Information, Figure S1), four embryos were collected at 24 hpf and total RNA from single shRNAs each that targeted the ntla (Figure S2), and oep embryos was extracted using Trizol reagent (Invitrogen) (Figure S3) genes were selected. followed by treatment with Turbo DNA-free DNase (Ambion). Transient in vivo RNAi is highly potent in knocking First-strand cDNA was reverse transcribed using oligo-dT down gene activity primers and Superscript reverse transcriptase (Invitrogen). Droplet Digital PCR amplifications of apkcl and gapdhs were For functional validation of shRNAs, we employed a tran- carried out with primers and probes carrying fluorophore sient in vivo transgenesis method. Although the mosaic ex- modifications. The sequences of the primers are the same pression is an inevitable feature associated with transient as used in qRT–PCR analyses. Additionally, FAM-modified apkcl transgenesis, this can be alleviated by coexpression of the probe (59-FAM-ATGTGCTCCATGGACAATGACCAGCT-39)and Tol2 transposase (Li et al. 2010) and selection of high

RNAi in Zebrafish 1067 Figure 1 Functional validation of miR- shRNAs by transient in vivo transgenesis. (A) Diagram of the conditional miR-shRNA expression system. The guide stand (bot- tom) is highlighted in light blue. (B) Mosaic expression of miR-shRNA5apkcl (left), miR- shRNA4oep (top right), and miR-shRNA4ntla (bottom right) causes morphological defects similar to those observed in respective mutants, which can be rescued by delivery of shRNA-resistant wild-type mRNAs. The miR-30e vector-injected embryos serve as controls. (C) Quantitative RT–PCR analysis shows a gene-specific reduction of endoge- nous mRNA levels by shRNAs cognate to their target genes. The miR-30e vector- injected embryos serve as controls. (D) Fluo- rescent immunostaining of aPKCl (green) shows knockdown effects at the protein level by different miR-shRNAsapkcl. In a sin- gle miR-shRNAapkcl-expressing cell, as indi- cated by the tdTomato fluorescence (red), miR-shRNA1apkcl (middle panel) and miR- shRNA5apkcl (left panel) lead to significant knockdown of aPKCl, whereas the ineffec- tive miR-shRNA7apkcl does not show obvi- ous knockdown effect (right panel). (E) Mosaic expression of miR-shRNA5apkcl causes defects in mitotic division orientation of radial glia progenitors in the developing zebrafish forebrain. Left: examples of mitotic division orientation in the miR-30e vector- injected (control) and miR-shRNA5apkcl- expressing radial glia progenitors. Right: Quantification of division orientation in miR-30e vector control and miR-shRNA5apkcl- expressing radial glia progenitors (n =27for control, and n = 24 for miR-shRNA5apkcl). The nuclei of radial glia progenitors were labeled with 3NLS:EGFP (green) and the individual ra- dial glia progenitors expressing UAS-miR30e vector (control) or UAS-miR-shRNA5apkcl were highlighted by tdTomato (red).

expressers with the visible reporter TdTomato (Figure S4). led to gene-specific phenotypes (with variable penetrance) DNA plasmids carrying the Pef-1a-Gal4 (elongation factor 1a that could be rescued by delivery of shRNA-resistant wild- promoter-driven Gal4) and UAS-miR-shRNA together with type mRNAs for the respective genes (Figure 1B, Table S1, the transposase were co-injected into one-cell-stage zebra- Table S2, and Table S3). Quantitative RT–PCR analysis us- fish embryos. Three of six miR-shRNAs targeting apkcl, and ing total RNAs from 10 hpf shRNA-expressing embryos three of four miR-shRNAs targeting either oep or ntla genes showed a gene-specific reduction of endogenous mRNAs

1068 Z. Dong, J. Peng, and S. Guo cognate to the shRNA target genes, and codelivery of two To determine whether a greater impairment of apkcl shRNAs (1 + 5) targeting apkcl further decreased the en- gene activity might be achievable if the strength of the dogenous apkcl level (Figure 1C). Immunostaining of the Gal4 driver was increased, we used KalTA4, an optimized aPKC protein (recognizing both l and u isoforms) showed at version of Gal4 for zebrafish (Distel et al. 2009). The KalTA4 single-cell levels that the two most effective shRNAs, apkcl- mRNA was injected into one-cell stage-F2 embryos from Tg shRNA5 and apkcl-shRNA1, diminished the apically local- [UAS-TdTom-miR-shRNA1apkcl A; Ubi-GFF] fish. While the ized aPKC protein in radial glia progenitors, whereas the KalTA4 mRNA-injected, nontransgenic siblings were mostly ineffective apkcl-shRNA7 did not (Figure 1D). normal, KalTA4 mRNA-injected, TdTom-positive transgenic embryos showed an apkcl mutant-like morphology (Figure Transient in vivo RNAi is a powerful tool for genetic 2E, 4 and Table S6). Similarly, EF1a-GFF DNA-injected, mosaic analysis TdTom-positive transgenic embryos also showed an apkcl A useful application of transient RNAi is in vivo genetic mutant-like phenotype (Figure 2E, 5 and Table S7). To as- mosaic analysis at single-cell resolution. Using time-lapse sess the endogenous apkcl mRNA level, we employed the imaging, we further analyzed the miR-shRNA-expressing Droplet Digital (dd) PCR technology (Figure S5), which individual radial glia progenitor cells for the mitotic spindle measures target (e.g., apkcl mRNA) and control (e.g., gapdh rotation phenotype associated with the loss of apkcl gene mRNA) molecules in the same biological reaction for superb activity (Horne-Badovinac et al. 2001). While most control precision and sensitivity (Hua et al. 2010). The results progenitors (.90%) displayed a division angle between 60 showed a clear correlation between the phenotypic severity and 90 and none between 0 and 30 (Figure 1E and File and the apkcl mRNA knockdown level (Figure 2E). Together, S1), a significant portion of the apkcl-shRNA5-expressing these results uncover a dosage-dependent knockdown of progenitors had a division angle between either 0 and apkcl gene activity through stable UAS-miR-shRNA transgen- 30 or 30 and 60 (Figure 1E and File S2), a ratio that esis and controllable expression of Gal4. was similar to what has been observed in the apkcl mutant (heart and soul, has) retina (Horne-Badovinac et al. 2001). Conclusions Together, these results validate the efficacy and specificity of miR-shRNAs in vivo and demonstrate the saliency of tran- We show that RNAi employing the miR-shRNA and Gal4/ sient RNAi for in vivo genetic mosaic analysis at single-cell UAS system is a promising technology for conditional gene resolution. silencing in zebrafish. Compared to previous reports (Dong et al. 2009; De Rienzo et al. 2012), our bipartite Gal4/UAS- Stable gene silencing with UAS-miR-shRNAapkcl based design offers greater versatility and synergizes with transgenic lines the ongoing development of Gal4 enhancer trap lines in Next we established transgenic lines that carry the UAS- zebrafish. TdTom-miR-shRNA1apkcl or UAS-TdTom-miR-shRNA5apkcl Two platforms are established that can be used either using the Tol2 transposon system. G0 germ-line independently or in combination to serve different research transgenic founders were identified as those that produced needs. The first platform is the in vivo transient RNAi. We red fluorescent progeny when crossed with the Ubiquitous- demonstrate that this method leads to robust and specific Gal4 (Ubi-GFF) transgenic animal (Asakawa and Kawakami phenotypes associated with three different genes, apkcl, 2010). One of 40 UAS-TdTom-miR-shRNA1apkcl and 3 of 50 oep, and ntla. We also show that it is a powerful tool for UAS-TdTom-miR-shRNA5apkcl founders transmitted the trans- genetic mosaic analysis at single-cell resolution. The ability genes to germ line (Figure 2A, Table S4,andTable S5). We to detect significant and dosage-dependent decrease of tar- noted that among the F1’s derived from the UAS-TdTom-miR- get mRNA expression also makes the in vivo transient RNAi shRNA1apkcl founder (A) and Tg[Ubi-GFF] cross, 30% of the a high throughput screening tool for identifying effective TdTom-positive embryos displayed a heart defect with car- miR-shRNAs. The second platform is the in vivo stable RNAi. diac edema that was strictly correlated with the expression of We show that the miR-shRNA transgenes can be stably TdTom (Figure 2B, middle panels compared to top panels). inherited in subsequent generations and leads to significant This heart defect was rescued by the delivery of shRNA- gene knockdown that is temporally regulatable and depen- resistant apkcl mRNA (Figure 2, B, bottom panels, and C). dent on the strength and expression of the Gal4 driver. Furthermore, to test for the spatiotemporal effect of RNAi, we However, it is worth pointing out that the stable UAS- crossed the UAS-TdTom-miR-shRNA1apkcl founder (A) with Tg shRNAapkcl transgene, which is likely present at single or [hsp-gal4], and subjected the resulting embryos to a transient few copies in the genome, is insufficient to produce the heat shock. The data showed an shRNA-dependent cardiac severe mutant phenotype observable in transient RNAi, defect upon heat shock (Figure 2D). Together, these results when crossed with stable Gal4 lines including several tis- demonstrate the stable integration and germ-line trans- sue-specific Gal4 lines tested. This is likely to be a general mission of RNAi in zebrafish and uncover a cardiac impair- concern associated with stable RNAi, and is possibly due to ment by UAS-TdTom-miR-shRNA1apkcl that is temporally the difference in shRNA transgene copy numbers between regulatable. transient and stable RNAi. Increasing the UAS-shRNA copy

RNAi in Zebrafish 1069 Figure 2 Functional validation of miR- shRNAs by stable in vivo transgenesis. (A) Schematic shows the stable transgenesis of miR-shRNAsapkcl in zebrafish using the Tol2

transposon system. (B) F1 embryos derived from a cross between the UAS-TdTom-miR- shRNA1apkcl founder A and Ubi-GFF trans-

genic line. TdTom-negative F1 embryos are normal (top panels). In contrast, 30% of

the TdTom-positive F1 embryos display a heart defect with cardiac edema (middle panels). This heart defect is rescued by the delivery of shRNA-resistant apkcl mRNA (bottom panels). (C) Quantification of heart defect and rescue in B. (D) Embryos (60 hpf) derived from a cross between the UAS- TdTom-miR-shRNA1apkcl founder A and hsp70-Gal4 transgenic animal. Heat shock was performed at 8 hpf at 37 C for 1 hr. The TdTom-positive embryo (right) dis- played a heart defect with cardiac edema, whereas the heat-shocked control sibling (left) did not. (E) Droplet digital PCR analyses of endogenous apkcl mRNA levels in stable UAS-TdTom-miR-shRNAsapkcl transgenic lines at different conditions of Gal4 induction. The representative images of 48-hpf embryos for each condition were shown on top of the bar graph. *P , 0.05, **P , 0.01, ***P , 0.001, vs. control sibling.

numbers through homozygosing the transgene or crossing transgenic line; the Guo lab and B. Lu for helpful discussions two different UAS-shRNA lines that potentially quadruples and comments on the manuscript; Camille Troup and David the copy numbers did not produce an obvious difference in Merrium from the Bio-Rad Laboratories for discussion on the our hands, suggesting a great many more copy numbers are dd-PCR technology; and David Merrium for generously necessary to produce a transient RNAi-like effect in stable providing access to the equipment and for technical advice. lines. It may not be feasible to incorporate tens or hundreds This work was supported by National Institutes of Health of miR-shRNA copies in the transgene for concerns of cloning grants (NS042626 and DA023904). and transgene stability. Instead, the efficacy of stable RNAi transgenic systems may be further enhanced by incorporating the dual Tet-regulatory system (Knopf et al. 2010) recently Literature Cited developed in zebrafish for further amplifying the UAS trans- Aitman, T. J., C. Boone, G. A. Churchill, M. O. Hengartner, T. F. gene expressivity. Taken together, this study demonstrates the Mackay et al., 2011 The future of model organisms in human fi rst prototype of stable and conditional RNAi-mediated gene disease research. Nat. Rev. Genet. 12: 575–582. silencing employing the Gal4/UAS system in zebrafish that Asakawa, K., and K. Kawakami, 2010 A transgenic zebrafish for mon- has the potential to transform the utility of this vertebrate for itoring in vivo microtubule structures. Dev. Dyn. 239: 2695–2699. uncovering novel biology and modeling human diseases. Bartel, D. P., 2009 : target recognition and regulatory functions. Cell 136: 215–233. Bedell, V. M., Y. Wang, J. M. Campbell, T. L. Poshusta, C. G. Starker Acknowledgments et al., 2012 In vivo genome editing using a high-efficiency TALEN system. Nature 491: 114–118. We thank the late Tingxi Liu for the zebrafish miR-30e vector; Davis, R. H., 2004 The age of model organisms. Nat. Rev. Genet. Kazuhide Asakawa and Koichi Kawakami for the Ubi-GFF 5: 69–76.

1070 Z. Dong, J. Peng, and S. Guo De Rienzo, G., J. H. Gutzman, and H. Sive, 2012 Efficient shRNA- Mello, C. C., 2007 Return to the RNAi world: rethinking gene mediated inhibition of gene expression in zebrafish. Zebrafish 9: expression and evolution. Cell Death Differ. 14: 2013–2020. 97–107. Meng, X., M. B. Noyes, L. J. Zhu, N. D. Lawson, and S. A. Wolfe, Dickins, R. A., M. T. Hemann, J. T. Zilfou, D. R. Simpson, I. Ibarra 2008 Targeted gene inactivation in zebrafish using engineered et al., 2005 Probing tumor phenotypes using stable and regu- zinc-finger nucleases. Nat. Biotechnol. 26: 695–701. lated synthetic microRNA precursors. Nat. Genet. 37: 1289–1295. Nasevicius, A., and S. C. Ekker, 2000 Effective targeted gene Distel, M., M. F. Wullimann, and R. W. Köster, 2009 Optimized “knockdown” in zebrafish. Nat. Genet. 26: 216–220. Gal4 genetics for permanent gene expression mapping in zebra- Paddison,P.J.,A.A.Caudy,E.Bernstein,G.J.Hannon,andD.S. fish. Proc. Natl. Acad. Sci. USA 106: 13365–13370. Conklin, 2002 Short hairpin RNAs (shRNAs) induce se- Dong,M.,Y.F.Fu,T.T.Du,C.B.Jing,C.T.Fuet al., 2009 Heritable quence-specific silencing in mammalian cells. Genes Dev. and lineage-specific gene knockdown in zebrafish embryo. PLoS 16: 948–958. ONE 4: e6125. Premsrirut,P.K.,L.E.Dow,S.Y.Kim,M.Camiolo,C.D.Maloneet al., Dong, Z., N. Yang, S. Yeo, A. Chitnis, and S. Guo, 2012 Intra- 2011 A rapid and scalable system for studying gene function in lineage directional Notch signaling regulates self-renewal and mice using conditional RNA interference. Cell 145: 145–158. differentiation of asymmetrically dividing radial glia. Neuron Ptashne, M., 1988 How eukaryotic transcriptional activators 74: 65–78. work. Nature 335: 683–689. Doyon, Y., J. M. McCammon, J. C. Miller, F. Faraji, C. Ngo et al., Reinhart, B. J., F. J. Slack, M. Basson, A. E. Pasquinelli, J. C. Bet- 2008 Heritable targeted gene disruption in zebrafish using de- tinger et al., 2000 The 21-nucleotide let-7 RNA regulates de- signed zinc-finger nucleases. Nat. Biotechnol. 26: 702–708. velopmental timing in Caenorhabditis elegans. Nature 403: Ebert, M. S., and P. A. Sharp, 2012 Roles for microRNAs in 901–906. conferring robustness to biological processes. Cell 149: Sander, J. D., L. Cade, C. Khayter, D. Reyon, R. T. Peterson et al., 515–524. 2011 Targeted gene disruption in somatic zebrafish cells using Elbashir, S. M., J. Harborth, W. Lendeckel, A. Yalcin, K. Weber engineered TALENs. Nat. Biotechnol. 29: 697–698. et al., 2001 Duplexes of 21-nucleotide RNAs mediate RNA in- Schulte-Merker, S., F. J. van Eeden, M. E. Halpern, C. B. Kimmel, terference in cultured mammalian cells. Nature 411: 494–498. and C. Nüsslein-Volhard, 1994 no tail (ntl) is the zebrafish Fire, A., S. Xu, M. K. Montgomery, S. A. Kostas, S. E. Driver et al., homologue of the mouse T (Brachyury) gene. Development 1998 Potent and specific genetic interference by double- 120: 1009–1015. stranded RNA in Caenorhabditis elegans. Nature 391: 806–811. Silva, J. M., M. Z. Li, K. Chang, W. Ge, M. C. Golding et al., Fire, A. Z., 2007 Gene silencing by double-stranded RNA. Cell 2005 Second-generation shRNA libraries covering the mouse Death Differ. 14: 1998–2012. and human genomes. Nat. Genet. 37: 1281–1288. Gruber, J., H. Manninga, T. Tuschl, M. Osborn, and K. Weber, Stark, G. R., I. M. Kerr, B. R. Williams, R. H. Silverman, and R. D. 2005 Specific RNAi mediated gene knockdown in zebrafish Schreiber, 1998 How cells respond to interferons. Annu. Rev. cell lines. RNA Biol. 2: 101–105. Biochem. 67: 227–264. Guo, S., and K. J. Kemphues, 1995 par-1, a gene required for Suster, M. L., H. Kikuta, A. Urasaki, K. Asakawa, and K. Kawakami, establishing polarity in C. elegans embryos, encodes a putative 2009 Transgenesis in zebrafish with the tol2 transposon sys- ser/thr kinase that is asymmetrically distributed. Cell 81: tem. Methods Mol. Biol. 561: 41–63. 611–620. Ui-Tei, K., Y. Naito, K. Nishi, A. Juni, and K. Saigo, Guo, S., S. W. Wilson, S. Cooke, A. B. Chitnis, W. Driever et al., 2008 Thermodynamic stability and Watson-Crick base pairing 1999 Mutations in the zebrafish unmask shared regulatory in the seed duplex are major determinants of the efficiency of pathways controlling the development of catecholaminergic the siRNA-based off-target effect. Nucleic Acids Res. 36: 7100– neurons. Dev. Biol. 208: 473–487. 7109. Horne-Badovinac, S., D. Lin, S. Waldron, M. Schwarz, G. Mbamalu Wang,L.,J.Y.Zhou,J.H.Yao,D.R.Lu,X.J.Qiaoet al., et al., 2001 Positional cloning of heart and soul reveals multi- 2010 U6 promoter-driven siRNA injection has nonspecific ple roles for PKC lambda in zebrafish organogenesis. Curr. Biol. effects in zebrafish.Biochem.Biophys.Res.Commun.39: 11: 1492–1502. 1363–1368. Hua, Z., J. L. Rouse, A. E. Eckhardt, V. Srinivasan, V. K. Pamula Wienholds, E., S. Schulte-Merker, B. Walderich, and R. H. A. Plas- et al., 2010 Multiplexed real-time polymerase chain reaction terk, 2002 Target-selected inactivation of the zebrafish rag1 on a digital microfluidic platform. Anal. Chem. 82: 2310–2316. gene. Science 297: 99–102. Huang, P., A. Xiao, M. Zhou, Z. Zhu, S. Lin et al., 2011 Heritable Zeng, Y., E. J. Wagner, and B. R. Cullen, 2002 Both natural and in zebrafish using customized TALENs. Nat. Bio- designed micro RNAs can inhibit the expression of cognate technol. 29: 699–700. mRNAs when expressed in human cells. Mol. Cell 9: Knopf, F., K. Schnabel, C. Haase, K. Pfeifer, K. Anastassiadis et al., 1327–1333. 2010 Dually inducible TetON systems for tissue-specific con- Zhang, J., W. S. Talbot, and A. F. Schier, 1998 Positional clon- ditional gene expression in zebrafish. Proc. Natl. Acad. Sci. USA ing identifies zebrafish one-eyed pinhead as a permissive 107: 19933–19938. EGF-relatedligandrequiredduring gastrulation. Cell 92: Lee, R. C., R. L. Feinbaum, and V. Ambros, 1993 The C. elegans 241–251. heterochronic gene lin-4 encodes small RNAs with antisense Zhao, X. F., A. Fjose, N. Larsen, J. V. Helvik, and Ø. Drivenes, complementarity to lin-14. Cell 75: 843–854. 2008 Treatment with small interfering RNA affects the micro- Li, Q., D. Ritter, N. Yang, Z. Dong, H. Li et al., 2010 A systematic RNA pathway and causes unspecific defects in zebrafish em- approach to identify functional motifs within vertebrate devel- bryos. FEBS J. 275: 2177–2184. opmental enhancers. Dev. Biol. 337: 484–495. Zuber, J., K. McJunkin, C. Fellmann, L. E. Dow, M. J. Taylor et al., Manche, L., S. R. Green, C. Schmedt, and M. B. Mathews, 2011 Toolkit for evaluating genes required for proliferation and sur- 1992 Interactions between double-stranded RNA regula- vival using tetracycline-regulated RNAi. Nat. Biotechnol. 29: 79–83. tors and the protein kinase DAI. Mol. Cell. Biol. 12: 5238– 5248. Communicating editor: B. Goldstein

RNAi in Zebrafish 1071 GENETICS

Supporting Information http://www.genetics.org/lookup/suppl/doi:10.1534/genetics.112.147892/-/DC1

Stable Gene Silencing in Zebrafish with Spatiotemporally Targetable RNA Interference

Zhiqiang Dong, Jisong Peng, and Su Guo

Copyright © 2013 by the Genetics Society of America DOI: 10.1534/genetics.112.147892

Figure S1 The design of miR-shRNAs targeting apkcλ. (A) The sequences of six shRNAs targeting apkcλ. The guide strands are highlighted in pink. (B) The location of shRNA target sites in the apkcλ mRNA.

2 SI Z. Dong, J. Peng, and S. Guo

Figure S2 The design of miR-shRNAs targeting ntla. (A) The sequences of four shRNAs targeting ntla. The guide strands are highlighted in pink. (B) The location of shRNA target sites in the ntla mRNA.

Z. Dong, J. Peng, and S. Guo 3 SI

Figure S3 The design of miR-shRNAs targeting oep. (A) The sequences of four shRNAs targeting oep. The guide strands are highlighted in pink. (B) The location of shRNA target sites in the oep mRNA.

4 SI Z. Dong, J. Peng, and S. Guo

Figure S4 Representative images of TdTom-miR-shRNA expressing embryos in transient transgenesis. (A) The embryo with low expression mosaicism, which is used for our analysis. (B) The embryo with high expression mosaicism, which is not used for our analysis.

Z. Dong, J. Peng, and S. Guo 5 SI

Figure S5 Droplet Digital PCR analyses of endogenous apkcλ mRNA levels in stable UAS-TdTom-miR-shRNAapkcλ transgenic lines coupled with the KalTA4 driver. (A-B) FAM intensity plotted against HEX intensity. (A) Sibling control embryos injected with KalTA4 RNA. (B) UAS-TdTom-miR-shRNA1apkcλ-A transgenic embryos injected with KalTA4 RNA. In both (A) and (B), pink lines are thresholds used to assign individual droplets as either positive or negative, which divide the graph into four sections: top left (droplets positive for apkcλ), top right (droplets positive for both apkcλ and gapdhs), bottom right (droplets positive for gapdhs) and bottom left (negative droplets with quenched un-hydrolyzed probes). With similar amount of gapdhs positive droplets, (b) shows less apkcλ positive droplets compared with (A), indicating reduced apkcλ mRNA level. (C) Statistic analysis on Droplet Digital PCR results of (A) and (B). *: P < 0.05, **: P < 0.01, vs sibling.

6 SI Z. Dong, J. Peng, and S. Guo

File S1

Time-lapse of a single TdTom-miR-30e vector control -expressing radial glia progenitor during mitotic division. The division ends at an angle close to 90 degree. The interval between each frame is 90 seconds.

File S1 is available for download at http://www.genetics.org/lookup/suppl/doi:10.1534/genetics.112.147892/-/DC1.

screenshot from File S1

Z. Dong, J. Peng, and S. Guo 7 SI

File S2

Time-lapse of a single TdTom-miR-shRNA5apkcλ -expressing radial glia progenitor during mitotic division. The mitotic spindle undergoes active rotation and the division ends at an angle close to 0 degree. The interval between each frame is 90 seconds.

File S2 is available for download at http://www.genetics.org/lookup/suppl/doi:10.1534/genetics.112.147892/-/DC1.

screenshot from File S2

8 SI Z. Dong, J. Peng, and S. Guo

Table S1 The efficacy and specificity of different apkcλ-shRNAs based on morphological phenotypes. UAS-TdTom-miR- shRNAapkcλ -miniTol2 plasmid was co-injected with Pef1α-GFF-PT2KXIG and Tol2 transposase into wild-type embryos at 1-cell stage and phenotypes were analyzed at ~48 hpf. The shRNA-resistant apkcλ mRNA (coding region only) was used for rescue. The miR-30e vector –injected embryos served as controls. *: P < 0.05; **: P < 0.01; ***: P < 0.001 vs Ctrl., ##: P < 0.01; ###: P < 0.001 vs apkcλ-shRNA5, Z-test.

Z. Dong, J. Peng, and S. Guo 9 SI

Table S2 The efficacy and specificity of different ntla-shRNAs based on morphological phenotypes. miR-ntla-shRNA (in pCS2) was in vitro transcribed and injected into wild-type embryos at 1-cell stage and phenotypes were analyzed at ~24 hpf. The shRNA-resistant ntla mRNA (coding region only) was used for rescue. The miR-30e vector –injected embryos served as controls. **: P < 0.01; ***: P < 0.001 vs Ctrl., ###: P < 0.001 vs ntla-shRNA4, Z-test.

10 SI Z. Dong, J. Peng, and S. Guo

Table S3 The efficacy and specificity of different oep-shRNAs based on morphological phenotypes. miR-oep-shRNA RNA (in pCS2) was in vitro transcribed and injected into wild-type embryos at 1-cell stage and phenotypes were analyzed at ~24 hpf. The shRNA-resistant oep mRNA (coding region only) was used for rescue. The miR-30e vector –injected embryos served as controls. **: P < 0.01; ***: P < 0.001 vs Ctrl., ##: P < 0.01 vs oep-shRNA4, Z-test.

Z. Dong, J. Peng, and S. Guo 11 SI

Table S4 Summary of stable transgenic lines carrying either UAS-TdTom-miR-shRNA1apkcλ or UAS-TdTom-miR-shRNA5apkcλ.

12 SI Z. Dong, J. Peng, and S. Guo

Table S5 Summary of morphological phenotypes of UAS-TdTom-miR-shRNA1apkcλ or UAS-TdTom-miR-shRNA5apkcλ transgenic lines when crossed with Ubi-GFF transgenic line.

Z. Dong, J. Peng, and S. Guo 13 SI

Table S6 Summary of morphological phenotypes of UAS-TdTom-miR-shRNA1apkcλ-A transgenic lines injected with KalTA4 RNA. ***: P< 0.001, vs Sibling + KalTA4 RNA.

14 SI Z. Dong, J. Peng, and S. Guo

Table S7 Summary of morphological phenotypes of UAS-TdTom-miR-shRNA1apkcλ-A transgenic lines injected with Pef1α-GFF plasmid. ***: P< 0.001, vs Sibling + Pef1α-GFF plasmid.

Z. Dong, J. Peng, and S. Guo 15 SI