Tissue-specific RNA interference in postimplantation mouse embryos with endoribonuclease-prepared short interfering RNA

Abstract

RNA interference (RNAi) using double-stranded RNA has been used for the systematic analysis of gene function in invertebrate organisms. Here we have explored the use of short interfering RNA (siRNA) to knock down gene expression during the development of mammalian postimplantation embryos. The developing CNS system of embryonic day 10 mouse embryos was used as a model tissue. siRNA prepared by endoribonuclease digestion (esiRNA) was injected into the lumen of the neural tube at specific regions and delivered into neuroepithelial cells by directed electroporation. Injected and electroporated embryos were grown for 1 day in whole-embryo culture and the effects of RNAi were examined. esiRNA directed against beta-galactosidase (beta-gal), coelectroporated into neuroepithelial cells together with reporter plasmids expressing GFP and beta-gal, abolished expression of beta-gal but not GFP, showing the specificity of the esiRNA-mediated RNAi. To demonstrate RNAi of endogenous gene expression, we used heterozygous embryos of a knock-in mouse line expressing GFP from the Tis21 locus, a gene turned on in neuroepithelial cells that switch from proliferation to neurogenesis. GFP-directed esiRNA electroporated into neuroepithelial cells of such embryos blocked the GFP expression normally occurring on the onset of neurogenesis. Taken together, our data indicate that esiRNA delivered in a tissue-specific manner by topical injection followed by directed electroporation can efficiently silence endogenous gene expression in mammalian postimplantation embryos.

Fig 1.

Approach used for tissue-specific RNAi in postimplantation mouse embryos. (a) Cartoon illustrating the region-specific injection (Cap, capillary) of siRNA into the lumen of the neural tube of an E10 mouse embryo, with the electroporation electrodes (El) in the lateral, cathode-right/anode-left orientation (Upper), and the uptake of siRNA into the left side of the neuroepithelium on electroporation (Lower). (b–e) Directed uptake of nucleic acids into neuroepithelial cells on region-specific injection into the lumen of the anterior neural tube of E10 mouse embryos followed by electroporation, exemplified by the use of pEGFP-N2 and expression of GFP on subsequent whole-embryo culture for 24 h. (b and c) Anterior region of whole unfixed embryos showing GFP expression in the left ventral telencephalon on injection into the telencephalic neural tube and electroporation by using the lateral, cathode-right/anode-left orientation (b), and in the dorsal telencephalon (c, arrowhead) and dorsal mesencephalon (c, arrow) on injection into the mesencephalic neural tube and electroporation by using a cathode-caudal/anode-rostral orientation. (Bar = 500 μm.) (d) Coronal-horizontal cryosection through the mesencephalon of the embryo in c showing GFP expression in the neuroepithelium (combined phase-contrast and fluorescence microscopy). The lumen of the neural tube is indicated by *. (Bar = 200 μm.) (e) Higher magnification fluorescence micrograph of a horizontal cryosection through the hindbrain of another embryo on injection into the rhombencephalic neural tube and electroporation by using a cathode-dorsal/anode-ventral orientation, showing GFP expression in individual neuroepithelial cells and neurons derived therefrom. The luminal surface of the neural tube is indicated by the dashed line. (Bar = 10 μm.)

Fig 2.

Specificity of esiRNA-mediated RNAi in the neuroepithelium of postimplantation mouse embryos developing in whole-embryo culture. E10 mouse embryos were injected into the lumen of the telencephalic neural tube with the GFP-expressing plasmid pEGFP-N2 plus the β-gal-expressing plasmid pSVpaXΔ either without (a–c and g, Control) or with (d–f and g, siRNA) β-gal-directed esiRNAs, followed by directed electroporation (lateral, cathode-right/anode-left orientation) and whole-embryo culture for 24 h. (a–f) Horizontal cryosections through the left telencenphalon were analyzed by double fluorescence for expression of GFP (green; a and d) and β-gal immunoreactivity (red; b and e). Neuroepithelial cells expressing both GFP and β-gal (arrowheads) appear yellow in the merge (c and f). Note the lack of β-gal expression in neuroepithelial cells in the presence of β-gal-directed esiRNAs. Upper and lower dashed lines indicate the luminal (apical) surface and basal border of the neuroepithelium, respectively. Asterisks in b and e indicate the basal lamina and underlying mesenchymal cells, which cross-react with the secondary antibody used to detect β-gal immunoreactivity. (Bar in f = 20 μm.) (g) Quantitation of the percentage of GFP-expressing neuroepithelial cells that also express β-gal without (Control) or with (siRNA) application of β-gal-directed esiRNAs. Data are the mean of three embryos analyzed as in a–f. (Bars indicate SD.)

Fig 3.

esiRNA-mediated RNAi of a gene endogenously expressed during the development of postimplantation mouse embryos. Heterozygous E10 Tis21^{+/tm2(Gfp)Wbh} mouse embryos were injected with GFP-directed esiRNAs into the lumen of the telencephalic (a and b) or diencephalic (c, d, f, and g) neural tube, followed by directed electroporation (lateral, cathode-right/anode-left orientation) and whole-embryo culture for 24 h. (a–d) Low-power dark-field (a and c) and fluorescence (b and d) micrographs of horizontal vibratome sections through the telencenphalon (a and b) and diencephalon (c and d). The left half of the brain is on the right side of the panels. Note the silencing of GFP expression in defined regions of the neuroepithelium (dashed lines) in the left, anode-facing half of the embryo. Asterisks in a and c indicate the lumen of the neural tube. (Bar in c = 200 μm.) (f and g) Higher-magnification fluorescence micrographs of horizontal cryosections through the right (f, cathode-facing) and left (g, anode-facing) half of the diencephalon at boundary to the telencephalon, prepared from the vibratome section shown in d. Dashed lines indicate the luminal (apical) surface of the neuroepithelium. (Bar in g = 20 μm.) (e and h) Quantitation of GFP fluorescence (e) and GFP-expressing cells (h) in the right (cathode-facing) and left (anode-facing) half of the neuroepithelium. Data in e are the mean of three distinct regions of the neuroepithelium, each from an independent embryo, two of which are the regions shown in b and d. Data in h are the mean of two cryosections prepared from the vibratome section shown in d; the region counted was the entire diencephalic neuroepithelium to the boundary to the telencephalon, including the fields shown in f and g. For each embryo/determination, values obtained for the left (anode-facing) half of the neuroepithelium (siRNA) were expressed as percentage of that for the right (cathode-facing) half (Control); the latter was arbitrarily set to 100. [Bars indicate SD (e) or the variation of the duplicates from the mean (h).]