© 2003 Nature Publishing Group http://www.nature.com/naturebiotechnology
TECHNICAL REPORT
Transgenic RNA interference in ES cell–derived embryos recapitulates a genetic null phenotype
A
Tilo Kunath1,2†, Gerald Gish1†, Heiko Lickert1, Nina Jones1, Tony Pawson1,2*, and Janet Rossant1,2*
B
Published online 7 April 2003; doi:10.1038/nbt813
Gene targeting via homologous recombination in murine embryonic stem (ES) cells has been the method of choice for deciphering mammalian gene function in vivo1. Despite improvements in this technology, it still remains a laborious method. Recent advances in RNA interference (RNAi) technology have provided a rapid loss-of-function method for assessing gene function in a number of organisms2,3. Studies in mammalian cell lines have shown that introduction of small interfering RNA (siRNA) molecules mediates effective RNA silencing4,5. Plasmid-based systems using RNA polymerase III (RNA pol III) promoters to drive short hairpin RNA (shRNA) molecules were established to stably produce siRNA6–8. Here we report the generation of knockdown ES cell lines with transgenic shRNA. Because of the dominant nature of the knockdown, embryonic phenotypes could be directly assessed in embryos completely derived from ES cells by the tetraploid aggregation method9. Such embryos, in which endogenous p120-Ras GTPase-activating protein (RasGAP), encoded by Rasa1 (also known as RasGAP), was silenced, had the same phenotype as did the previously reported Rasa1 null mutation10.
We constructed a drug-selectable shRNA expression plasmid with a chemically synthesized H1 RNA pol III promoter11. This promoter efficiently drives expression of short RNA molecules in mammalian cells, as previously shown with the pSUPER construct6. We introduced downstream of the H1 promoter a Rasa1-specific sequence consisting of sense and antisense 23 base pair (bp) regions homologous to part of the mouse and human Rasa1 coding region, separated by a 7 bp spacer. This was followed by a stretch of five thymidines to terminate RNA pol III transcription (Fig. 1A). We tested the RasGAP shRNA construct by transient transfection into a human bladder carcinoma cell line. Efficient downregulation of RasGAP protein was observed; levels of other proteins, such as Syk and Diaphanous, were not affected (Fig. 1B). These results show that the shRNA construct specifically silences RasGAP and does not elicit a global inhibition of protein synthesis. We generated stable ES cell lines by electroporation of the linearized RasGAP shRNA transgene into R1 ES cells9. We established 12 drug-resistant ES cell lines, referred to as RasGAP shRNA ES cells, and determined their RasGAP expression by western blotting. Although RasGAP protein expression was only slightly reduced in one line (clone 1), it was significantly downregulated in the 11 other 1Samuel
Lunenfeld Research Institute, Mount Sinai Hospital, Toronto, ON, Canada M5G 1X5. 2Department of Molecular and Medical Genetics, University of Toronto, Toronto, ON, Canada M5S 1A8. *Corresponding authors (
[email protected],
[email protected]). †These authors contributed equally to this study. www.nature.com/naturebiotechnology
•
MAY 2003
C
Figure 1. RasGAP shRNA transgene and its ability to silence RasGAP expression. (A) Schematic of RasGAP shRNA transgene showing the sense and antisense regions that target the Rasa1 gene. Also shown is the predicted transcript in a hairpin structure with a 7 bp loop. (B) Western blot analysis of whole cell lysates from mock-treated or RasGAP shRNA–transfected human bladder carcinoma cell line 5637. RasGAP was drastically downregulated in the RasGAP shRNA transfected cells, whereas Syk and Diaphanous (Dia) were unaffected. (C) Western blot analysis of whole cell lysates from 12 RasGAP shRNA ES cell lines (1–12) and wild-type ES cells (ES) for RasGAP and tubulin. An enhanced exposure revealed clone 3 had weak expression of RasGAP. Longer exposures revealed very weak expression in clones 5, 6, and 8 (data not shown).
cell lines (Fig. 1C). An enhanced exposure showed that some cell lines (clones 3, 5, 6, and 8) had a very low level of expression that was barely detectable above the background, whereas the others had no detectable expression (Fig. 1C). Levels of tubulin expression were unaffected, and the morphology and growth characteristics of all twelve cell lines appeared normal, indicating that the expression of siRNA molecules in ES cells is not detrimental. These observations also indicated that RasGAP function is not essential for ES cell maintenance, consistent with the previous finding that Rasa1 homozygous null ES cells can be derived and do not exhibit a phenotype different from wild type in culture10. The developmental capacity of four ES cell lines (clones 1, 2, 3, and 8) with differing levels of RasGAP suppression was tested in completely ES cell–derived embryos generated by the tetraploid aggregation method9. We formed aggregates of RasGAP shRNA ES cells and B5/EGFP tetraploid embryos, derived from a transgenic line that ubiquitously expresses enhanced green fluorescent protein •
VOLUME 21
•
nature biotechnology
559
TECHNICAL REPORT
A
C
Table 1. Categorization of RasGAP shRNA embryos. RasGAP shRNA ES cell clone
© 2003 Nature Publishing Group http://www.nature.com/naturebiotechnology
D
Qualitative No. ES cell– Phenotypic severity RasGAP derived embryosb in comparison to a expression Rasa1 –/–
no. 2
–
26
19 more 7 equal
no. 8
+/–
22
1 more 21 equal
no. 3
+
14
1 more 8 equal 5 less
no. 1
+++
10
10 wild type
B aAs
determined by western blotting of ES cell clones (Fig. 1C). generated by aggregation of ES cells to B5/EGFP tetraploid embryos. Any embryos with even minor contributions of EGFP host cells were excluded.
E
bEmbryos
F
Figure 2. RasGAP shRNA embryos derived from aggregations to wild-type EGFP tetraploid embryos. (A) Schematic of the tetraploid aggregation method illustrating the tissues that the tetraploid embryo and the ES cells give rise to. (B) Embryo and yolk sac from RasGAP shRNA ES cell clone 1 (left) and clone 3 (right). The embryo and yolk sac from clone 1 resembled wild type, with a normal heart (asterisks), tail (arrow), and head (dots). These structures were defective in the embryos derived from the RasGAP-silenced ES cell clones. (C) Embryo and yolk sac (ys) from RasGAP shRNA ES cell clone 8. (D) Two embryos from RasGAP shRNA ES cell clone 2 exhibiting a typical Rasa1 null phenotype. (E) Two embryos from ES cell clone 2 with more severe phenotypes. Bar, 200 µm in (B–E). (F) Western blot analysis for RasGAP and tubulin of whole cell lysates from wild-type E9.5 embryos and six RasGAP shRNA embryos (two independent embryos from each clone) that appeared equal in severity to the genetic Rasa1 null phenotype.
(EGFP) but is wild type in all other respects12. The tetraploid host embryos contribute to the trophoblast tissue of the placenta and the extraembryonic endoderm component of the yolk sac, but owing to their increased ploidy, they are rigorously excluded from the embryo proper. If any host cells should contribute to the embryonic portion of the conceptus, they would be readily apparent from the presence of the EGFP marker. In direct contrast to the tetraploid cells, ES cells cannot make trophoblast tissue in vivo and very rarely contribute to extraembryonic endoderm. As a result, the aggregated ES cells will make the entire embryo proper, as well as the mesodermal compartment of the yolk sac (Fig. 2A). RasGAP shRNA ES cell–derived embryos were dissected and analyzed at embryonic day 9.5 (E9.5). The Rasa1 mutants exhibit a number of characteristic morphological defects at this stage and they die 1 day later at E10.5 of vascular defects10. A total of 72 embryos without contributions from the host EGFP tetraploid cells were analyzed. The embryos were grouped qualitatively into four categories: more severe, less severe, or equal in severity to the Rasa1 genetic null embryos, and wild type (Table 1). The single ES cell line that exhibited a slight decrease in RasGAP expression 560
nature biotechnology
•
VOLUME 21
•
(clone 1) produced embryos that were all wild type in appearance (Fig. 2B, Table 1). By contrast, three cell lines with significantly reduced RasGAP levels produced embryos with phenotypes strikingly similar to the null phenotype (Fig. 2B–E). A number of morphological defects that are characteristic of the Rasa1 genetic null embryos were observed in the RasGAP shRNA embryos. Distention of the pericardial region and enlarged hearts were immediately apparent. This defect was accompanied by posterior truncations, abnormal development of the head region (Fig. 2B–E), and a ruffled appearance of the yolk sac (Fig. 2B, C). The yolk sac defect was of particular interest because the endoderm portion was wild type (from the EGFP-positive tetraploid cells) and the mesoderm was from the RasGAP-silenced ES cells. The persistence of the ruffled yolk sac phenotype indicated that the defect in Rasa1 mutants reflects a cell-autonomous role for RasGAP in the mesoderm, but not the endoderm, of the yolk sac. Individual embryos were screened for RasGAP expression by western blot analysis to confirm that the gene silencing initiated in ES cells continued into later development. We compared equivalent amounts of protein lysate from transgenic embryos derived from three silenced ES cell clones to wild-type embryos. RasGAP protein was significantly, but not completely, reduced in these transgenic embryos (Fig. 2F). The severity of the phenotypes observed in RasGAP shRNA ES cell-derived embryos was variable and correlated roughly with the levels of RasGAP expression in ES cell clones (Table 1). A few embryos with less severe phenotypes were observed from clone 3, the ES cell line with the highest RasGAP expression among the 11 silenced lines. The intermediate cell line (clone 8) produced mostly typical-looking Rasa1 null embryos. The ES cell line in which no RasGAP expression could be detected (clone 2) produced a large percentage of embryos with phenotypes that appeared similar in substance to, but more severe, than the genetic null (Fig. 2E, Table 1). In some embryos, the posterior truncation and head defects were more pronounced and the pericardia were more grossly distended. The explanation for this observation is not entirely clear. There could be phenotypic variation resulting from the differences in genetic background, because the R1 ES cells are from an F1 hybrid of two 129 substrains9 and the Rasa1 null mutation was analyzed on different backgrounds10. These results suggest that different levels of shRNA-mediated gene silencing may be useful in generating the phenotypic equivalent of a hypomorphic allelic series for dissecting gene function. This can also be accomplished by targeting different regions of the same transcript with shRNA, as shown for the Trp53 gene13. MAY 2003
•
www.nature.com/naturebiotechnology
© 2003 Nature Publishing Group http://www.nature.com/naturebiotechnology
TECHNICAL REPORT
With the completion of the mouse genome sequence, a wealth of sequence information is available and awaiting functional characterization. There is an urgent need to develop rapid, high-throughput approaches to dissect mammalian gene function in vivo. The merging of three established technologies—plasmid-based shRNA, transgenic ES cells, and tetraploid-hosted embryo production—addresses this need. In a rapid and efficient manner, we reproduced the null phenotype of the Rasa1 mutation, identified a cell-autonomous role for RasGAP in the yolk sac mesoderm, and produced a range of phenotypic effects at different levels of RasGAP suppression. Identification of homologous genomic regions, construction of complex targeting vectors, and germline transmission of genetic mutations were not required. Indeed, only accurate transcript sequence information is necessary to construct a specific shRNAbased transgene to mediate gene silencing. There are limitations to the application of this shRNA-based method. Genes essential for ES cell growth or maintenance cannot be suppressed without affecting the viability of the cultures, although an inducible-shRNA system should circumvent this problem. In addition, genes required in the trophoblast or the yolk sac endoderm cannot be analyzed by this method because ES cells do not make these tissues in vivo. In such cases, injection or lentiviral infection of an shRNA transgene into zygotes could solve this problem, and indeed, DNA injection of an shRNA transgene against EGFP suppressed gene expression in mouse embryos, in adults, and into the next generation14. Germline transmission of ES cells harboring an shRNA transgene has also been accomplished15. In addition, lentiviral infection was used to suppress gene expression in embryos in two recent reports16,17. The feasibility of rapidly assaying mammalian gene function in vivo without gene targeting is now evident.
50 mM Tris HCl, pH 7.4, 150 mM NaCl, 2 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 10 µg/ml aprotinin, 10 µg/ml leupeptin, and 1 mM sodium vanadate). Equivalent amounts of lysate, as determined by the Bradford assay, were resolved using SDS-PAGE and transferred to a polyvinylidene difluoride membrane (Westran, Scheicher & Schuell, Keene, NH). Polyclonal antiRasGAP antibody (171; Santa Cruz, Santa Cruz, CA), monoclonal anti-Syk antibody (4D10; Santa Cruz) and goat polyclonal anti-Dia-1 antibody (L-15; Santa Cruz) were used at 1 µg/ml in 5% skim milk in TBS-T (20mM Tris HCL, 150 mM NaCl, 0.05% (v/v) Tween-20). Monoclonal anti-tubulin antibody (DMIA, Sigma) was used at 1:5,000 and secondary antibodies (Bio-Rad, Mississauga, Canada) were used at a dilution of 1:10,000. Tetraploid aggregations. Four RasGAP shRNA transgenic ES cell clones (1, 2, 3, and 8) were aggregated with B5/EGFP tetraploid embryos to generate completely ES cell-derived embryos as described9. Embryos were dissected at E9.5 and assessed for EGFP contribution under fluorescence optics. Embryos with any EGFP-positive cells were excluded from the analysis, although most exhibited the Rasa1 null phenotype. Experimental animals were treated according to guidelines approved by the Canadian Council for Animal Care. Rasa1 genetic null embryos. Rasa1 mutant embryos were obtained from intercrosses between heterozygous Rasa1+/– mice10. Rasa1–/– embryos were visibly identifiable at E9.5 by morphological features described in the text and confirmed by Rasa1-specific PCR of yolk sac DNA using PCRGap1, PCRGap2, and PCRNeoβ2 oligonucleotide primers as previously described10. Acknowledgments We thank Sue MacMaster and Lois Byers for tetraploid aggregations and uterine transfers and Andras Nagy for critical discussions. This research was supported by the Canadian Institutes of Health Research and a Terry Fox Programme Project grant from the National Cancer Institute of Canada. Competing interests statement The authors declare that they have no competing financial interests.
Experimental protocol Construction of RasGAP shRNA transgene. DNA encoding the human H1 RNA pol III promoter (position –100 to –7)11 and a RasGAP shRNA sequence were prepared by oligonucleotide synthesis (DNA/RNA Synthesizer Model 394; Applied Biosystems, Foster City, CA). The H1 promoter was cloned into the BglII and XbaI sites of pcDNA3.1(+) (Invitrogen, Burlington, Canada), replacing the cytomegalovirus (CMV) promoter. A unique Asp718 restriction site was introduced at positions –6 to –1 of the RNA pol III promoter upstream of the shRNA sequence. The target siRNA sequence, corresponding to bases 2460–2482 of the Homo sapiens RASA1 gene (NCBI accession no.: NM_002890), was introduced into the Asp718 and XbaI sites using the senseand antisense-strand oligonucleotides 5′ GTACCAAGATGAAGCCACTACCCTATTTCAAGAGAAAATAGGGTAGTGGCTTCATCTTTTTTTGGAAAT 3′ and 5′ CTAGATTTCCAAAAAAAGATGAAGCCACTACCCTATTTTCTCTT GAAATAGGGTAGTGGCTTCATCTTG 3′, respectively. This sequence also corresponds to the murine Rasa1 3′ gene from bases 1640–1662 (NCBI accession no.: NM_145452). The transcriptional unit is predicted to form a stem-loop structure with a 7 bp loop. This construct maintains the neomycin-resistance gene, driven by the SV40 early promoter, found in pcDNA3.1(+). The authenticity of the resulting vector was confirmed by DNA sequencing. Establishment of transgenic ES cell lines. ScaI-linearized RasGAP shRNA transgene (25 µg) was electroporated into R1 ES cells (5 × 106 cells) in 0.8 ml electroporation medium (Specialty Media, Phillipsburg, NJ). The electroporation parameters were 0.25 kV and 500 µFD, followed by an incubation on ice (20 min). Cells were plated in high-glucose DMEM Gibco-BRL, Burlington, ON, Canada) with 20% fetal bovine serum (HyClone, Logan, UT), 1 mM sodium pyruvate, 2 mM L-glutamine, 0.1 mM nonessential amino acids (all from Gibco-BRL), 100 µM β-mercaptoethanol (Sigma, Oakville, Canada), 1000 U/ml leukemia inhibitory factor (LIF), and 90 µg/ml G418 (Gibco-BRL) on 0.1% (w/v) gelatin. G418-resistant colonies were picked to a 96-well plate after 8 d of selection and twelve clonal lines were expanded and further analyzed. Western blotting. Whole cell lysates were prepared from human bladder carcinoma cell line 5637, ES cells, and E9.5 embryos (lysis buffer: 1% NP-40, www.nature.com/naturebiotechnology
•
MAY 2003
Received 22 January 2003; accepted 3 March 2003
1. Joyner, A.L. Gene Targeting: A Practical Approach (Oxford University Press, Oxford, 2000). 2. Fire, A. et al. Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans. Nature 391, 806–811 (1998). 3. Hannon, G.J. RNA interference. Nature 418, 244–251 (2002). 4. Caplen, N.J., Parrish, S., Imani, F., Fire, A. & Morgan, R.A. Specific inhibition of gene expression by small double-stranded RNAs in invertebrate and vertebrate systems. Proc. Natl. Acad. Sci. USA 98, 9742–9747 (2001). 5. Elbashir, S.M. et al. Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature 411, 494–498 (2001). 6. Brummelkamp, T.R., Bernards, R. & Agami, R. A system for stable expression of short interfering RNAs in mammalian cells. Science 296, 550–553 (2002). 7. Lee, N.S. et al. Expression of small interfering RNAs targeted against HIV-1 rev transcripts in human cells. Nat. Biotechnol. 20, 500–505 (2002). 8. Paul, C.P., Good, P.D., Winer, I. & Engelke, D.R. Effective expression of small interfering RNA in human cells. Nat. Biotechnol. 20, 505–508 (2002). 9. Nagy, A., Rossant, J., Nagy, R., Abramow-Newerly, W. & Roder, J.C. Derivation of completely cell culture-derived mice from early-passage embryonic stem cells. Proc. Natl. Acad. Sci. USA 90, 8424–8428 (1993). 10. Henkemeyer, M. et al. Vascular system defects and neuronal apoptosis in mice lacking ras GTPase-activating protein. Nature 377, 695–701 (1995). 11. Myslinski, E., Ame, J.C., Krol, A. & Carbon, P. An unusually compact external promoter for RNA polymerase III transcription of the human H1RNA gene. Nucleic Acids Res. 29, 2502–2509 (2001). 12. Hadjantonakis, A.-K., Gertsenstein, M., Ikawa, M., Okabe, M. & Nagy, A. Generating green fluorescent mice by germline transmission of green fluorescent ES cells. Mech. Dev. 76, 79–90 (1998). 13. Hemann, M.T. et al. An epi-allelic series of p53 hypomorphs created by stable RNAi produces distinct tumor phenotypes in vivo. Nat. Genet. 33, 396–400 (2003). 14. Hasuwa, H., Kaseda, K., Einarsdottir, T. & Okabe, M. Small interfering RNA and gene silencing in transgenic mice and rats. FEBS Lett. 532, 227–230 (2002). 15. Carmell, M.A., Zhang, L., Conklin, D.S., Hannon, G.J. & Rosenquist, T.A. Germline transmission of RNAi in mice. Nat. Struct. Biol. 10, 91–92 (2003). 16. Rubinson, D.A. et al. A lentivirus-based system to functionally silence genes in primary mammalian cells, stem cells and transgenic mice by RNA interference. Nat. Genet. 33, 401–406 (2003). 17. Tiscornia, G., Singer, O., Ikawa, M. & Verma, I.M. A general method for gene knockdown in mice by using lentiviral vectors expressing small interfering RNA. Proc. Natl. Acad. Sci. USA 100, 1844–1848 (2003). •
VOLUME 21
•
nature biotechnology
561
