Induction of Protein Deletion Through In Utero Electroporation to Define Deficits in Neuronal Migration in Transgenic Models

Different regions of the brain can be targeted for electroporation by varying the orientation of the paddle electrodes over the embryo’s head (Figure 2). For all parts of the forebrain, the plasmid is injected into the lateral ventricle (the hind brain can be targeted by injecting into the fourth ventricle). In general, place the electrodes so that the straight line between them passes through the region of interest, with DNA being drawn toward the positive electrode. To target the lateral cortex (Figure 2A), place the positive electrode on top of the green crescent, with a slight rostral angle. For the medial cortex (Figure 2B), switch the orientation so that the positive electrode is over the contralateral non-injected ventricle, this time with an even larger forward angle. To target the hippocampus (Figure 2C), place the positive electrode over the contralateral side, but instead angle the positive electrode caudally (electroporations must be done at an age when the neural precursors are still part of the cortical hem, prior to migration to the final location of the hippocampus). To target the ganglionic eminence (Figure 2D) place the paddles on the embryo in line with the ears, with the positive electrode slightly lower than the negative electrode. Representative images of these electroporated regions are shown in Figure 2.

It is important to validate Cre-mediated excision in electroporated brains. The ideal approach is to perform in situ hybridization or immunohistochemistry against the transcript or protein in question. Unfortunately, as is the case for pRb, reliable probes or antibodies are not always available against the deleted gene. In these cases, alternative approaches, such as staining for a downstream target regulated by the deleted gene (Figure 3A) or confirming excision of the targeted allele in electroporated sections by PCR (Figure 3B), can be used. For example, Figure 3A confirms Cre mediated excision by reproducing a well-documented effect of pRb/p107/p130 gene deletion in electroporated neurons. Following electroporation of pRb f/f; p107 -/-; p130 f/f neurons with the pCig2-Cre plasmid, these triple knock out GFP+ cells continue to express the cell cycle progression marker Ki67, which is a known effect following excision of pocket proteins²⁵. Control GFP+ neurons electroporated with the empty pCig2 plasmid do not express Ki67. The Cre plasmid should also be tested in vitro through co-transfection with a plasmid carrying a reporter gene behind a floxed stop codon, such as the NL-dsRed plasmid (Figure 3C) in a cell line, but this should only be used in combination with other in vivo techniques.

Due to the variation between electroporated brains, analysis of electroporated neurons requires calculating the percentage of total GFP+ neurons that portray a specific characteristic, such as expression of a protein marker or location within a given cortical layer. It is crucial to compare electroporated sections from similar anatomical locations within the brain relative to known landmarks. For example, images for Figure 4 were taken at the level of the anterior commissure. Furthermore, due to horizontal migration, images taken near the edge of the electroporated regions have a higher proportion of mature neurons that have reached the cortex, while images taken near the middle of the electroporated region have both mature and immature GFP+ cells. To insure consistency, it is important to quantify images selected from similar places within the electroporated patch.

Following Emx1-Cre mediated deletion of pRb and p107 in pyramidal neurons in the dorsal telencephalon, a population of neurons expressing the upper cortical layer marker FoxP1 were mislocalized to the intermediate zone (IZ) (Figure 4A). There are significantly more FoxP1+ cells in the IZ of double knock out (DKO) brains compared to control brains with a corresponding decrease in the number of FoxP1+ cells in the upper cortical layers (above layer IV) of DKO brains than found in the upper layers of control brains (Figure 4B). There was no corresponding defect in the radial glial scaffold (Figure S2 of Svoboda et al., 2012) ⁶. Here we compare expression of cortical layer markers Tbr1, FoxP1 and Cux1 in pRb f/+; p017 +/-; Emx1-Cre + (wild type (WT)) brains to expression of these same markers pRb f/f; p107 -/-; Emx1-Cre + (DKO) brains. Heterozygote embryos can be used as controls since pocket proteins exhibit autoregulation at the protein level and heterozygote animals are aphenotypic ²⁶.

It was unclear whether these FoxP1+ cells were found in the IZ because these cells were immature neurons inappropriately expressing FoxP1, or whether they were earlier born neurons that had failed to migrate. To distinguish between these possibilities, we electroporated GFP into the wild type (WT) and double knock out (DKO) embryos at E14.5 and tracked the progress of these neurons over 4 days. Although this technique does not conclusively identify the final cell division, as with BrdU birthdating, it does allow the tracking of newborn neurons since it marks the discrete set of neurons derived from cells lining the ventricle at the time of electroporation ¹⁷. As shown in Figure 4C, all of the GFP+ neurons reach the uppermost cortical layer in the WT brains, but only a subset of these neurons reach the top of the cortex in DKO brains, indicating that there is in fact a defect in the migration of upper layer neurons, rather than improper expression of neuronal markers (similar results were obtained through traditional BrdU birthdating ⁶).

Neuronal migration is a highly complex process that is regulated by a multitude of interconnected mechanisms ^1-3. As it is very difficult to identify the mechanism underlying a migration defect in an uncharacterized transgenic model, distinguishing cell autonomous versus non-cell autonomous defects can facilitate phenotype characterization. If following electroporation of Cre-GFP, GFP+ neurons fail to migrate in an otherwise wild type brain, this indicates that the mechanism disrupted by gene deletion is specific to the electroporated cell. Alternatively, if the electroporated neurons migrate normally in the wild type brain, this suggests that the cause of aberrant migration in the transgenic embryo is due to an environmental defect. Following electroporation of Cre-GFP in control pRb f/+; p107 +/- and double knock out pRb f/f; p107 -/- embryos, GFP+ neurons were found to migrate into the cortex in pRb f/f; p107 -/- brains, but many of these neurons failed reach the upper cortical layers as they did in control brains (Figure 4D) ⁶. There were significantly lower percentage of GFP+ neurons within the layer III/IV, marked by expression of FoxP1²⁷, in pRbf/f; p107 -/- electroporated brains than in control brains (Figure 4E). A Cre-GFP fusion protein was used to limit GFP fluorescence to the cell body and facilitate localization of the neuron to a specific cortical layer. Since the phenotype following electroporation of Cre is less severe than the phenotype following gene deletion throughout the cortex, this suggests that the original phenotype shown in Figure 4A is a combination of multiple mechanisms, both cell autonomous and not cell autonomous. Alternatively, it is possible that the time delay following electroporation, during which Cre is expressed, DNA recombination is performed, pRb is downregulated and the effects are manifested, is long enough to allow the electroporated neurons to reach the cortex before a migration defect becomes apparent.

It should be noted that the distinction between cell autonomous and non-cell autonomous mechanisms should be interpreted with caution in brains with high electroporation efficiencies, where 100-200 or more neurons are electroporated in a single 14 µm thick section.If a high enough proportion of neurons are electroporated, GFP+ neurons may influence each other and create a small microenvironment. To be truly cell autonomous, migration defects must be detected even when electroporation efficiency is very low (10-30 GFP+ neurons per section). If necessary, the number of transfected cells can be lowered by decreasing voltage, pulse number or DNA concentration. All of these factors must be taken into consideration when analyzing the results following electroporation of Cre in a transgenic model.

Figure 1: Performing In utero electroporation. (A) Sterile surgical set up. (B) Surgical tools from left: fine scissors for cutting needle tips; stapler; forceps; wound retractor; surgical scissors; needle holder. (C) Prepped pregnant mouse with shaved and disinfected abdomen. (D) Mouse covered with sterile drape prior to incision. (E) Uterus with E13.5 embryos removed from abdominal cavity. (F) Injecting pups in the ventricle. (G) Close up of injected pup; note the blue crescent on the embryo’s head indicating that the ventricle was successfully injected with the plasmid solution (arrow). (H) Paddles are placed to electroporate the lateral cortex. (I) Mouse post-surgery. (J) Preparing sharp needles is a crucial part of in utero electroporation. Examples of uncut needles (i, iv), good needles (ii, v) and poor needles (iii, vi) are shown at low (i-iii) and high (iv-vi) magnification. Please click here to view a larger version of this figure.

Figure 2: Targeting different brain regions. Diagrams A-D show paddle placement for targeting different regions of the forebrains. The dark blue circle represents the paddle in front of the embryo while the light blue circle represents the paddle behind the embryo. Positive and negative electrodes are indicated with the appropriate sign. (A) Electrodes angled to target the lateral cortex. (A’, A”) Example of electroporation of the lateral cortex, three days after surgery. (B) Electrodes angled to target the medial cortex. (B’, B”) Example of electroporation of the medial cortex. (C) Electrodes angled to target the hippocampus. (C’, C”) Example of electroporation of the hippocampus. (D) Electrodes angled to target the ganglionic eminence (GE). (D’, D”) Example of electroporation of the GE. Please click here to view a larger version of this figure.

Figure 3: Confirming Cre-mediated gene excision in electroporated neurons. (A) Cre mediated excision can be confirmed by co-staining for a downstream target of the deleted gene. pRb f/f; p107 -/-; p130 f/f embryos were electroporated with either the empty pCig2 plasmid or the pCig2-Cre plasmid and stained for GFP and Ki67, a marker of cell cycle entry. Double stained cells were seen in neurons electroporated with Cre (arrows in inset) but not in neurons electroporated with pCig2 alone (arrowheads). (B) Excision of pRb was also confirmed by performing PCR with genotyping primers on DNA extracted from the tissue sections removed from the slides. Sections were divided down the midline and the electroporated side of the section was pooled. The upper band represents the floxed allele and the lower band represents the wild type allele. The arrow indicates the excised allele in the electroporated side of the brain only. Because pRb f/+; p107 +/- heterozygotes were used as controls, an excised allele can be seen in the electroporated side of the brain in both the control and knockout brains. (C) Cre-mediated excision can also be confirmed in vitro using plasmids with floxed stop codons upstream of RFP, such as NL-dsRed. RFP is only expressed in HEK cells when co-transfected with Cre-GFP in Hek293 cells (bottom row), indicating that excision was successful. Please click here to view a larger version of this figure.

Figure 4: Using IUE to study migration in transgenic mouse models. (A) Expression of FoxP1, a marker for layer III/IV cortical neurons, was examined in pRb f/+; p107 +/-; Emx1-Cre + (WT), and pRb f/f; p107 -/-; Emx1-Cre + (DKO) brains at E17.5. Overlay includes Cux1 (blue – marker for layers II-IV) and Tbr1 (green – marker for layers V/VI). A population of cells expressing FoxP1 was observed in the intermediate zone of DKO embryos (arrow), but not WT embryos, with a concurrent decrease in the number of FoxP1+ neurons in layers II-IV of the cortex in DKO brains relative to WT brains (B). The white line indicates the delineation between the cortical plate and intermediate zone, which was determined using DAPI staining (not shown). (C) To trace the migration of a discrete population of late born neurons, WT and DKO embryos were electroporated with GFP at E14.5 and collected the brains at E18.5. Brains were co-stained with FoxP1 (red) and Tbr1 (blue) to show that GFP+ neurons are found almost exclusively in layer II in the WT brain but are delayed in the intermediate zone (IZ) and lower cortical layers in the DKO. (D) To determine if pRb and p107 regulate radial migration cell autonomously, pRb f/f; p107-/- and pRb f/+; p107-/+ embryos were electroporated with Cre-GFP at E13.5 and brains were collected at E17.5. The location of GFP+ neurons was quantified according to cortical layer. Layer II was defined as being above FoxP1 staining (red), layer III/IV was defined as within FoxP1 staining, above Tbr1 (blue) and layer V/VI was defined as within the Tbr1 positive layer (E). We show that a portion pRb and p107 deficient neurons fail to reach layer II but instead remain in layer III/IV. The figure has been modified from Svoboda et al., 2012. Please click here to view a larger version of this figure.

Table 1: Equipment and Materials.

Table 1: Electroporation voltage by embryo age.