Over the past decade, EP has become a powerful tool for genetic manipulation of neural stem and progenitor cells during embryonic and postnatal development [12, 23]. It has most often been utilized to target germinal zones surrounding the ventricles [13, 15, 16]. We have adapted and extended this technique to target the superficial layers of both the cortex and tectum at P2. In so doing, we provide a method for rapidly and specifically targeting these cell types. This technology will allow for elucidation of the genetic regulators of migration and differentiation of this population as well as their contribution to histogenesis during development and disease.
Our results demonstrate that postnatal EP can non-invasively target plasmids into the superficial progenitor domains rapidly and with a high degree of specificity. This methodology is given added importance, as we are not aware of any reporter mice or Cre recombinase lines capable of specifically targeting this unique pool of neural progenitors. Furthermore, while retroviruses have been used to specifically mark these cells , retroviral cloning and packaging is technically challenging and time-consuming. In addition, there are safety concerns associated with viral handling. EP simply requires purified plasmid DNA, which can be rapidly prepared. Furthermore, multiple genes or shRNAs can quickly be delivered, without the need for production of viral particles or complex mouse breeding strategies. However, it should be noted that the use of EP methodology and Cre-reporter or Cre-driver lines are not mutually exclusive.
We have provided a concise report of pial surface EP. The circumstantial evidence that we provide supports the progressive differentiation of immature neural progenitor populations into neurons and astroglia. Our results will need to be extended to rigorously determine the natural history of pial surface cell types and to exclude spurious or unexpected labeling mechanisms or transdifferentiation, which can be observed with such methods . Thus, we envision that the electroporation of plasmids with cell type-specific promoter elements driving Cre, or conversely, the delivery of Cre-dependent reporter plasmids into cell-lineage dependent Cre lines, will be invaluable for the definitive determination of cell potential and lineage commitment at the pial surface.
EP is most easily performed during the perinatal period, due to visibility of brain structures and the injection site through the skull. Furthermore, this technique most efficiently targets proliferating cells that are actively undergoing breakdown of the nuclear envelope, ostensibly allowing for episomal expression of the plasmid . Thus, during aging, EP efficiency would markedly decrease , as fewer and fewer cells are proliferating and targeting therefore becomes less accurate. Other potential caveats to EP include the episomal nature of most plasmids, whereby proliferation of cells may lead to plasmid dilution. Interestingly though, we did not observe much evidence of plasmid dilution within electroporated PS progenitors when compared to electroporated cells surrounding the lateral ventricles of the forebrain (data not shown). However, it did appear that astroglial EGFP expression declined between 14 days and 2 to 2.5 months post-EP, which is suggestive of further proliferation (data not shown). Nevertheless, these potential limitations could be overcome by using tools that allow for stable plasmid integration into the genome (for example, piggyBac as in , Tol2 , or Sleeping Beauty transposon systems).
Electroporation of the dorsal surface of the brain appears to target differentiating progenitor cells that initially do not express markers of mature neurons or astroglia. It is important to note that there are reports of in vivo postmitotic neuronal electroporation (reviewed in [23, 26–28]). However, by and large, these electroporation approaches are of the more invasive microelectroporation variety, where electrodes are inserted into the tissue vs. the external application used in this report. Notably, our BrdU labeling experiments did not label 100% of electroporated cells, and only weakly labeled some cells. However, the cell cycle lengthens throughout the course of embryogenesis  and throughout postnatal life , making BrdU labeling less efficient. In addition, there are likely temporal and regional differences in cell cycle length between the postnatal pial surface and cortical VZ, where previous EP/BrdU double-labeling experiments were carried out [17, 31]. Nevertheless, it is possible that we are targeting postmitotic populations, but these cells are likely to be immature, migrating, or otherwise in transition at the time of EP based on their progressive morphological differentiation. For example, EP of membrane-tagged EGFP demonstrated that the morphology of most cells was more undifferentiated, often in the form of multipolar glia or bipolar cells. At 14 days, cells displaying neuronal or astroglial morphologies did not strongly express the de facto markers of these respective cell types: NeuN (in both cortex and tectum) or Gfap (in the tectum only, as most non-activated cortical astrocytes do not normally express Gfap).
In our EP system, immunostaining and morphologic evidence of differentiated astroglia and neurons was present later on at 2 to 2.5 months. Even at this time, we did note that many cells with neuronal morphology did not label with NeuN. Notably, NeuN does not label all neuronal subtypes . In particular, it has been shown that some neurons in the piriform cortex are negative for this antigen . Based on available evidence then, we conclude that these cells represent a unique population of neurons that clearly display neuronal properties (thin dendrites stretching hundreds of μm, varicosities, and so on) and do not express non-neuronal lineage markers such as NG2, Olig2, or Aldolase C (data not shown). Interestingly, the longer time course of development of these unique pial progenitors seems to differ from typical VZ-derived neural cells. In summary, the perinatal EP technique detailed in this report should be amenable for rapid elucidation of the molecular mechanisms regulating this fascinating pial neural progenitor lineage.