Conventional loss-of-function studies of genes using knockout animals have often been costly and time-consuming. Electroporation-based CRISPR-mediated somatic mutagenesis is a powerful tool to understand gene functions in vivo. Here, we report a method to analyze knockout phenotypes in proliferating cells of the cerebellum.
Brain malformation is often caused by genetic mutations. Deciphering the mutations in patient-derived tissues has identified potential causative factors of the diseases. To validate the contribution of a dysfunction of the mutated genes to disease development, the generation of animal models carrying the mutations is one obvious approach. While germline genetically engineered mouse models (GEMMs) are popular biological tools and exhibit reproducible results, it is restricted by time and costs. Meanwhile, non-germline GEMMs often enable exploring gene function in a more feasible manner. Since some brain diseases (e.g., brain tumors) appear to result from somatic but not germline mutations, non-germline chimeric mouse models, in which normal and abnormal cells coexist, could be helpful for disease-relevant analysis. In this study, we report a method for the induction of CRISPR-mediated somatic mutations in the cerebellum. Specifically, we utilized conditional knock-in mice, in which Cas9 and GFP are chronically activated by the CAG (CMV enhancer/chicken ß-actin) promoter after Cre-mediated recombination of the genome. The self-designed single-guide RNAs (sgRNAs) and the Cre recombinase sequence, both encoded in a single plasmid construct, were delivered into cerebellar stem/progenitor cells at an embryonic stage using in utero electroporation. Consequently, transfected cells and their daughter cells were labeled with green fluorescent protein (GFP), thus facilitating further phenotypic analyses. Hence, this method is not only showing electroporation-based gene delivery into embryonic cerebellar cells but also proposing a novel quantitative approach to assess CRISPR-mediated loss-of-function phenotypes.
Brain diseases are one of the most dreadful mortal diseases. They often result from genetic mutations and subsequent dysregulation. To understand molecular mechanisms of brain diseases, ever-lasting efforts to decipher the genomes of human patients have discovered a number of potential causative genes. So far, germline genetically engineered animal models have been utilized for in vivo gain-of-function (GOF) and loss-of-function (LOF) analyses of such candidate genes. Due to the accelerated development of functional validation studies, a more feasible and flexible in vivo gene assay system for studying gene function is desirable.
The application of an in vivo electroporation-based gene transfer system to the developing mouse brain is suitable for this purpose. In fact, several studies using in utero electroporation have shown their potential to conduct functional analyses in the developing brain1,2,3. Actually, several regions of the mouse brain, such as the cerebral cortex4, retina5, diencephalon6, hindbrain7, cerebellum8, and spinal cord9 have been targeted by somatic gene delivery approaches, so far.
Indeed, transient gene expression by in vivo electroporation on embryonic mouse brains has long been used for GOF analysis. Recent transposon-based genomic integration technologies further enabled long-term and/or conditional expression of genes of interest10,11, which is advantageous to dissect gene function in a spatial and temporal manner during development. In contrast to GOF analysis, LOF analysis has been more challenging. While transient transfection of siRNAs and shRNA-carrying plasmids was performed, long-term effects of LOF of genes are not guaranteed due to eventual degradation of exogenously introduced nucleic acids, such as plasmids and dsRNAs. However, the CRISPR/Cas technology provides a break-through in LOF analyses. Genes encoding fluorescent proteins (e.g., GFP) or bioluminescent proteins (e.g., firefly luciferase) have been co-transfected with CRISPR-Cas9 and sgRNAs to label the cells exposed to CRISPR-Cas9-mediated somatic mutations. Nevertheless, this approach might have limitations in functional studies on proliferating cells, since exogenous marker genes are diluted and degraded after long-term proliferation. While the transfected cells and their daughter cells undergo CRISPR-induced mutations in their genomes, their footprints might get lost over time. Thus, genetic labeling approaches would be suitable to overcome this issue.
We recently developed a CRISPR-based LOF method in cerebellar granule cells that undergo long-term proliferation during their differentiation12. To genetically label the transfected cells, we constructed a plasmid carrying a sgRNA together with Cre and introduced the plasmid into the cerebella of Rosa26-CAG-LSL-Cas9-P2A-EGFP mice13 using in utero electroporation. Unlike regular plasmid vectors encoding EGFP, this approach successfully labeled transfected granule neuron precursors (GNPs) and their daughter cells. This method provides great support in understanding in vivo function of genes of interest in proliferating cells in normal brain development and a tumor-prone background.
All animal experiments were conducted according to animal welfare regulations and have been approved by the responsible authorities (Regierungspräsidium Karlsruhe, approval numbers: G90/13, G176/13, G32/14, G48/14, and G133/14).
1. Generate pU6-sgRNA-Cbh-Cre Plasmids
2. Test the Efficiency of the sgRNAs Using the EGxxFP Plasmid System
NOTE: The efficiency of the sgRNAs is normally tested by Surveyor or T7E1 (T7 endonuclease I) assays. In this protocol, an easy and efficient alternative approach is used. The key of this approach is to use the pCAG-EGxxFP plasmid which contains overlapping EGFP fragments separated by a DNA sequence containing the sgRNA targeting site17. Upon the expression of pCAG-EGxxFP together with the sgRNA and Cas9 in the transfected cells, the Cas9-mediated double strand break (DSB) in the target sequence is repaired by endogenous homology-dependent mechanisms, which reconstitutes the EGFP expression cassette.
3. Perform In Utero Electroporation
4. Prepare Cryosections from the Electroporated Cerebella
5. Immunostaining of the Cryosections
6. Imaging and Analysis
For in vivo functional analysis, it is critical to identify the cells into which exogenous gene(s) have been introduced. While the expression of a marker, such as GFP in non-proliferating cells can be followed-up for a long period of time, the signal gets sequentially lost in proliferating cells. An illustration of this effect is demonstrated in Figure 1. To circumvent losing the footprint of transfected cells in LOF analyses, we developed a novel approach by combining electroporation-based gene delivery with the CRISPR/Cas9 technology.
Representative results are shown in Figure 2. Functionality of the plasmid constructs is tested by transient transfection of pU6-sgTop2b-Cbh-Cre and pCAG-EG-Top2b-FP17, carrying the sgTop2b target sequence, into HEK293T cells stably expressing SpCas9 (Figure 2A). The sgRNA-guided Cas9 endonuclease activity induces DSB-mediated homology-directed repair of the EGFP expression cassette. Hence, the function of the sgRNA was directly analyzed by detection of GFP expression. According to Mashiko et al., a transfection efficiency of more than 30% determines an sgRNA as effective17.
GNPs originate from the rhombic lip (RL), a region bordering the roof of the fourth ventricle at embryonic stages E12.5 until E16.5. These cells have been known to undergo massive proliferation after birth in the outer external granule layer (oEGL) and shift to the inner EGL (iEGL) after exiting cell cycle20. Thus, GNPs are an appropriate example to test the advantages of our genetic labeling approach.
By following this protocol, in utero electroporation of GNPs allows tracing of edited cells with GFP. We introduced the pU6-sgRNA-Cbh-Cre plasmid expressing sgControl into the cerebellar primordium of Rosa26-CAG-LSL-Cas9-P2A-EGFP mice at embryonic stage E13.5. Mice were sacrificed at postnatal day P7, and sagittal sections of cerebellar tissue sections were stained by immunohistochemistry. Outer and inner EGL were defined by Ki67 expression (oEGL) and p27 expression (iEGL), respectively. Despite undergoing proliferation and maturation, mature cerebellar granule neurons (CGNs) still retained strong GFP expression (Figure 2B).
Emphasizing the utility of this protocol, we furthermore used the sgRNA targeting DNA Top2b as a representative example. Experimental procedures were performed as described above (see Figure 2B). Most of the GFP-positive cells transfected with sgRNAs for Top2b exhibited a loss of Top2b expression in the internal granule layer (IGL), while the introduction of control sgRNAs did not result in clear reduction of its expression (Figure 2C). A quantification of this result is shown in Figure 2D. These data present a valuable tool for tracing edited cells for a long period of time during development or tumor formation.
Figure 1: Schematic representation of GFP expression in proliferating and non-proliferating cells. (A) When an exogenous gene encoding GFP (e.g., pCAG-EGFP) is transfected in non-proliferating cells, the cells may express GFP for a long time (upper lane). Meanwhile, GFP expression could disappear in proliferating cells due to dilution of exogenous GFP (middle lane). In contrast, the cells that carry the LoxP-Stop-LoxP-GFP (LSL-GFP) transgene can keep the GFP expression during proliferation after transfection with Cre recombinase (lower lane). (B) The principle of SpCas9-mediated gene silencing and GFP expression after Cre and sgRNA transfection in a Rosa26-CAG-LSL-Cas9-P2A-EGFP mouse strain. Please click here to view a larger version of this figure.
Figure 2: Representative images of results from this protocol. (A) HEK293T cells stably expressing SpCas9 were transfected with pU6-sgTop2b-Cbh-Cre and pCAG-EG-Top2b-FP plasmids. GFP expression in live cells was monitored using a fluorescent cell imager (see Table of Materials) 48 h after transfection. Left and right panels show an effective and non-effective sgRNA sequence based on GFP expression, respectively. Scale bars: 100 µm. (B) Immunostaining of GFP, Ki67, and p27 on the P7 cerebella from a Rosa26-CAG-LSL-Cas9-P2A-EGFP mouse subjected to electroporation at E13.5 with pU6-sgRNA-Cbh-Cre plasmid constructs expressing control sgRNA. The section is counterstained with DAPI (blue). Scale bars: 50 µm; 20x magnification. Note GFP-expressing cells in the oEGL marked by Ki67. (C) Immunostaining of GFP (green) and Top2b (magenta) on the P7 cerebella from a Rosa26-CAG-LSL-Cas9-P2A-EGFP mouse that was electroporated at E13.5 with pU6-sgRNA-Cbh-Cre plasmid constructs expressing sgRNA against Top2b (sgTop2b) and a control sequence (sgControl). Arrows indicate Top2b expression in GFP+ cells in the same field. Scale bars: 20 µm; 20x magnification. (D) Quantification of Top2b in electroporated (GFP-positive) cells in sgControl and sgTop2b experiments. Two and three pups were investigated for sgControl and sgTop2b, respectively, and 25 cells from each brain were analyzed. Error bars represent ± standard error of the mean (SEM) and p-values were calculated by unpaired t-test, *p = 0.0002. Please click here to view a larger version of this figure.
Name | Sequence | Application | |
sgTop2b-1 | CTTCGTCCTGATACATACAT | sgRNA target sequence | |
sgTop2b-2 | AGCTGTCCAAAAATTAAAGC | sgRNA target sequence | |
sgControl | GCGACCAATACGCGAACGTC | sgRNA target sequence | |
EGxxFP-Top2b-F | gctgcccgacaaccactgagTACCTTGATATCTTAGAGAGCTG | Cloning into pCAG-EGxxFP | |
EGxxFP-Top2b-R | gggtcagcttgccgatatcgCTCGCGCATTGTCTTAGC | Cloning into pCAG-EGxxFP | |
hU6-F | GAGGGCCTATTTCCCATGATT | Sequencing for sgRNA |
Table 1: Sequences of oligos being used in this study.
Using exo utero electroporation, we have previously reported siRNA-based in vivo functional analyses of Atoh1 at an early stage of cerebellar granule cell differentiation8. Due to siRNA dilution/degradation and exposure of embryos outside the uterine wall, phenotypic analysis of the electroporated granule cells was limited to embryonic stages. However, the current method enabled analysis of the phenotype of postnatal animals.
Our previous study demonstrated that Cas9-mediated knockout of tumor suppressor Ptch1 via in utero electroporation successfully induced medulloblastoma2. In comparison, the current approach exhibits two major advantages: 1) Cas9 does not have to be delivered with the plasmid, allowing for multiple gene targeting by carrying multiple sgRNA expression cassettes in a single plasmid instead; and 2) the target cells and their progenies are permanently labeled with GFP, enabling visualization of the live cells and analysis of the behavior of transforming tumor cells in vivo.
The most critical steps of this protocol are the proper injection of plasmid DNAs into the correct location and the delivery of electric pulses into the appropriate places in the brain. Cerebellar neurons are born from the cerebellar neuroepithelium sequentially in a birthdate-dependent manner. Not all types of cells in the cerebellar primordium can be targeted via electroporation, because only a specific time window at E12.5-14.5 is technically feasible. Deep cerebellar nuclei neurons and Purkinje cells have been known to arise at E10.5-11.5, when the extra-embryonic membranes are not transparent, and the embryo is not clearly visible for DNA injection. Later, unipolar brush cells are derived from E17.5 upper RL (uRL), when the fourth ventricle is too narrow to inject a sufficient amount of DNA in the vicinity of the uRL. Thus, our method is only applicable at E12.5-14.5, which mainly targets mid- and later-born progenitors, such as GNPs and inhibitory interneurons. Another drawback of the method is that a full knockout phenotype may not be feasible when compared to Cre/LoxP-mediated germline GEMMs, since only thousands of cerebellar cells can be targeted by electroporation in each embryo.
In an earlier study, approximately 80% of GFP-positive cerebellar granule cells lost expression of the targeted gene12. While the identity of the granule cells was confirmed by molecular markers and their distribution, this identification approach might not be always applicable, as genes of interest could be involved in the marker expression and neuronal migration. Conditional activation of Cre using a cell-specific promoter would solve the problem in this case. Therefore, the current ubiquitous Cbh promoter in the pU6-sgRNA-Cbh-Cre plasmid can be replaced with a cell-specific promoter.
The authors have nothing to disclose.
We appreciate Laura Sieber, Anna Neuerburg, Yassin Harim, and Petra Schroeter for technical assistance. We also thank Drs. K. Reifenberg, K. Dell and P. Prückl for helpful assistance for animal experiments at DKFZ; the Imaging Core Facilities of the DKFZ and the Carl Zeiss Imaging Center in the DKFZ for confocal microscopy imaging. This work was supported by the Deutsche Forschungsgemeinschaft, KA 4472/1-1 (to D.K.).
Alexa 488 Goat anti-Chicken | ThermoFisher | A11039 | 1:400 dilution |
Alexa 568 Donkey anti-Mouse | Life Technologies | A-10037 | 1:400 dilution |
Alexa 594 Donkey anti-Rabbit | ThermoFisher | A21207 | 1:400 dilution |
Alexa 647 Donkey anti-Rabbit | Life Technologies | A31573 | 1:400 dilution |
Alkaline Phosphatase (FastAP) | ThermoFisher | EF0654 | |
Autoclave band | Kisker Biotech | 150262 | |
BamHI (HF) | NEB | R3136S | |
BbsI (FastDigest) | ThermoFisher | FD1014 | |
Cellulose Filter Paper (Whatman) | Sigma-Aldrich | WHA10347525 | |
Cloth | Tork | 530378 | |
Confocal laser scanning microscope | Zeiss | LSM800 | |
D-Luciferin | biovision | 7903-1 | |
DAPI | Sigma-Aldrich | D9542 | 1:1000 dilution |
Disposable plastic molds (Tissue-Tek Cyromold) | VWR | 4566 | |
DMEM Glutamax | ThermoFisher | 31966047 | |
Donkey serum | Sigma-Aldrich | D9663 | |
EcoRI (HF) | NEB | R3101S | |
Electro Square Porator | BTX | ECM830 | |
Endofree Maxi Kit | Qiagen | 12362 | |
Ethanol | Merck | 107017 | |
Eye ointment (Bepanthen) | Bayer | 81552983 | |
Fast Green | Merck | 104022 | |
FBS | ThermoFisher | 10270-016 | |
Filter (0.22 µm) | Merck | F8148 | |
Fluorescent cell imager (ZOE) | Biorad | 1450031 | |
Forceps straight | Fine Science Tools | 91150-20 | |
Gauze (X100 ES-pads 8f 10 x 10 cm) | Fisher Scientific | 15387311 | |
GFP antibody | Abcam | ab13970 | 1:1000 dilution |
Gibson Assembly Master Mix | NEB | E2611S | |
Glass Capillary with Filament | Narishige | GD1-2 | |
Heating Pad | ThermoLux | 463265 / -67 | |
Image Processing software (ImageJ and Fiji) | NIH | – | |
Insulin syringe (B. Braun OMNICAN U-100) | Carl Roth | AKP0.1 | |
Isoflurane | Zoetis | TU061219 | |
IVIS Lumina LT Series III Caliper | Perkin Elmer | CLS136331 | |
Kalt Suture Needles | Fine Science Tools | 12050-02 | |
KAPA HIFI HOTSTART READY mix | Kapa Biosystems | KK2601 | |
Ki67 antibody | Abcam | ab15580 | 1:500 dilution |
Light Pointer | Photonic | PL3000 | |
Liquid blocker pen | Kisker Biotech | MKP-1 | |
Metamizol | WDT | – | |
Microgrinder | Narishige | EG-45 | |
Microinjector | Narishige | IM300 | |
Micropipette Puller | Sutter Instrument Co. | P-97 | |
Microscope software ZEN | Zeiss | – | |
Non-sterile Silk Suture Thread (0.12 mm) | Fine Science Tools | 18020-50 | |
O.C.T. Compound (Tissue-Tek) | VWR | 4583 | |
p27 antibody | BD bioscience | 610241 | 1:200 dilution |
Paraformaldehyde | Roth | 335.3 | |
PBS (1x) | Life Technologies | 14190169 | |
pCAG-EGxxFP | Addgene | 50716 | |
Polyethylenimine | Sigma-Aldrich | 408727 | |
pX330 plasmid | Addgene | 42230 | |
QIAprep Spin Miniprep Kit | Qiagen | 27104 | |
QIAquick Gel Extraction Kit | Qiagen | 28704 | |
Quick Ligation Kit | NEB | M2200S | |
Ring Forceps | Fine Science Tools | 11103-09 | |
Slides (SuperFrost) | ThermoFisher | 10417002 | |
Software for biostatistics (Prism 7) | GraphPad Software, Inc | – | |
Spitacid | EcoLab | 3003840 | |
Stereomicroscope | Nikon | C-PS | |
Sucrose | Sigma-Aldrich | S5016 | |
Surgical scissors | Fine Science Tools | 91460-11 | |
Surgical scissors with blunt tip | Fine Science Tools | 14072-10 | |
Suture (Supramid schwarz DS 16, 1.5 (4/0)) | SMI | 220340 | |
T4 DNA Ligation Buffer | NEB | B0202S | |
T4 PNK | NEB | M0201S | |
Tissue scissors Blunt (11.5 cm) | Fine Science Tools | 14072-10 | |
TOP2B antibody | Santa Cruz | sc13059 | 1:200 dilution |
Trypsin (2.5 %) | ThermoFisher | 15090046 | |
Tweezers w/5mm Ø disk electrodes Platinum | Xceltis GmbH | CUY650P5 | |
Vaporizer | Drägerwerk AG | GS186 |