**3. Results**

#### *3.1. The Fate of OB Cells After Targeting Cell Progenitors at Distinct Ventricular Sites*

Using UbC-StarTrack plasmids (Figure 2A), we performed different IUEs at E12 that targeted different ventricular areas (dorsal, ventral, medial) and the most rostral portion of LV, the OV (Figure 2B). Animals were injected with Tx at P5 to remove the episomal copies of the constructs and analyzed at adult stages (from P30 onwards). Rostral IUE, restricted to the rostral OV, labeled glial cells, mitral cells, and some interneurons in the OB (Figure 2C). Interestingly, these glial cells were radially disposed in the different layers of the OB close to the electroporation area. Mitral cells in the mitral cell layer (MCL) were identified through their morphology and the presence of reelin (data not shown). These results indicated that glial and mitral cells originated from progenitor cells located in the most rostral part of the LV at E12. By contrast, when the dorsal, medial, and ventral walls of the LV were targeted, the labeled cells in the OB were periglomerular and granular interneurons, not glial or mitral cells (Figure 2D–I).

After targeting E12 progenitor cells within the dorsal LV, different neural cells were labeled in the adults, spread throughout the corpus callosum and cortex (Figure 2D), although only interneurons were labeled in the OB (Figure 2E). Likewise, ventral electroporation at E12 labeled neurons in the striatum, piriform cortex, and corpus callosum and interneurons in the dorsal cortex (Figure 2F) and the OB (Figure 2G). Dorsal and ventral electroporation mostly labeled interneurons in the GcL, with a few periglomerular cells also labeled. By contrast, medial electroporation labeled cells in the septal area of the telencephalon (Figure 2H), although most cells were located in the GL of the OB (Figure 2I). Finally, IUE of the third ventricle did not label glia or neurons in the OB (data not shown).

In summary, after targeting different ventricular areas at E12, the adult labeled cell-progeny displayed different morphologies at different locations in both OB and forebrain. Thus, the origin of the progenitor cells in specific areas determines their cell fate in the adult telencephalon.

**Figure 2.** (**A**) Diagram of the UbC-StarTrack vectors, 12 different plasmids encoding six different fluorescent proteins at two different locations, cytoplasmic and nuclear according to the H2B sequence. All vectors were driven by the Ubiquitin C promoter. (**B**) Summary of the IUE procedure, where E12 embryos were injected with UbC-StarTrack mixture and electroporated. After birth Tamoxifen (Tx) was injected at around P5, and the adult tissue was analyzed (>P30). Four different orientations of the electrodes were used for electroporation: olfactory ventricle (OV-IUE), dorsal (D-IUE), ventral (V-IUE), and medial (M-IUE). The red line illustrates the electroporation area. UbC-StarTrack OV-IUE labeled both neurons and glia in the olfactory bulb (OB, **C**). Targeted cells in each lateral ventricular (LV) zone gave rise to different labeled neural cells in the dorsal cortex (**D**), piriform cortex (**F**), and septum (**H**). By contrast, dorsal-, ventral-, and medial- IUE did not produce any labeled glia in the OB; only interneurons were targeted (**E**,**G**,**H**). Dorsal and ventral-IUE targeted progenitors that gave rise to labeled cells in the GcL and eventually, the GL. However, M-IUE produced more labeled cells in the GL. The white squares represent the electroporation area in the telencephalon and OB (**C–I**). IUE, in utero electroporation; OB, olfactory bulb; LV, lateral ventricle; Cx, cerebral cortex; Pir, piriform cortex; St, striatum; Sp, septum.

#### *3.2. The Fate of Olfactory Bulb Cells After Targeting Specific Progenitors with StarTrack*

We analyzed the fate of progenitor cells using a novel UbC-StarTrack strategy based on the combination of UbC-StarTrack plasmids with different PiggyBac transposases driven by specific promoters. This strategy drives the integration of the plasmids exclusively into the progenitors that express the specific promoters chosen at the time of electroporation (Figure 3A). As such, we specifically targeted NSCs using the CMV, NG2, Gsh-2, and GFAP promoters. First, the UbC-StarTrack and CMV-transposase (CMV-hyPBase: Figure 3B) incorporated copies of the plasmids ubiquitously, labelling all the progenitor cells and their progeny. Subsequently, the PiggyBac transposase encoding the NG2 promoter (NG2-hyPBase: Figure 3C) was used to target only those progenitor cells with an active NG2 promoter, integrating copies of the plasmids and labeling their progeny. In another approach, the PiggyBac transposase was driven by the subpallial promoter Gsh-2 promoter (Gsh2-hyPBase: Figure 3D) to only label the progenitors located in the ganglionic eminences at early developmental stages and consequently, their adult cell progeny. Finally, the GFAP promoter was incorporated into a transposase (GFAP-hyPBase) and co-electroporated with UbC-StarTrack to label GFAP-progenitor cells (Figure 3E). All these IUEs were directed at the dorso-lateral ventricle walls, except for the Gsh2-hyPBase, which was ventrally orientated. As a result of these manipulations, all the labeled cells in the OB corresponded to interneurons situated in the GL and GcL, with no glial cells or projection neurons. This comparative analysis of the di fferent StarTrack vectors involved 12 animals (*n* = 3 for each transposase driven by a di fferent promoter) and the study of 26,685 labeled interneurons in the OB, of which 12,236 were generated by progenitors expressing NG2; 8308 were from CMV progenitors; 5035 were from progenitors electroporated with the Gsh-2 transposase; and only 1,106 cells were from progenitor cells expressing GFAP. However, no significant di fferences were evident for each construct in terms of the average of labeled interneurons in the OB (Figure 3F).

**Figure 3.** Diagram of the UbC-StarTrack strategy based on transposase promoter expression ( **A**). The concept that is focused in the transposase only integrates copies of the UbC-StarTrack vectors into progenitor cells with the corresponding promoter active, labeling all their progeny (**a**). Progenitor cells with the inactive promoter do not integrate copies into the NSCs (**b**). For these experiments, the CMV, NG2, Gsh-2, and GFAP promoters were chosen to target specific NSCs. The first strategy with CMV-hyPBase labeled OB interneurons in the di fferent layers (**B**). The NG2 progeny labeled cells in the GcL and GL ( **C**), resembling the Gsh-2 progeny ( **D**). GFAP progenitors gave rise to granular cells and periglomerular cells (**E**). GFAP progenitors produced fewer labeled cells in the OB than the other vectors (**F**). All data were normalized; the box plot represents the percentage of labeled cells after targeting each set of progenitors with a specific transposase (whiskers represent 5th/95th percentile, horizontal line displays the median of the data; *n* = 3 for each transposase). Data showed no statistically significant di fference between groups (ns). CMV-progenitors are shown in soft pink; NG2-progenitors in blue; Gsh-2 in yellow; GFAP-progenitors in red.

Therefore, these results indicate that the pool of progenitor cells committed to give rise to OB interneurons was quite heterogeneous. Accordingly, NG2 and CMV progenitors at E12 produced a larger proportion of adult OB cells compared to those produced from progenitors expressing GFAP.

#### *3.3. Diversity of Olfactory Bulb Interneurons in Relation to Progenitor Cell Identity*

Considering the molecular profile of specific NCSs, we studied the di fferences between the interneurons generated by the di fferent pools of progenitor cells. The UbC-StarTrack plasmids and the CreERT2 vector were injected along with one of the specific transposases (CMV, NG2, GFAP or Gsh-2) (Figure 4A–E). The distribution of the labeled cells in the adult OB was analyzed and correlated with their progenitor cell profile. All the labeled cells were interneurons, periglomerular, and granular cells, even though some immature cells were found close to the subependymal zone (data not shown). Of the cells labeled by the transposase driven by the CMV promoter, 14% were located in the GL, while 86% were located in the GcL (Figure 4B). When NG2 and Gsh2 drove transposase expression, a similar proportion of cells was found in the GcL (88% NG2, 89% Gsh-2) and GL (12% NG2, 11% Gsh-2: Figure 4C,D). However, after targeting the GFAP progenitors, the cell-derived progeny was preferentially sited within the GcL (93%) rather than in the GL (7%: Figure 4E). Besides, these GFAP-progenitors are committed preferably to external areas of GL compared with those that express other promoters. In summary, progenitor cells were committed to preferentially generate granule cells more than periglomerular cells (Figure 4F). Otherwise, there were no significant di fferences between the distinct types of progenitor cells committed to generate periglomerular and granular cells (Figure 4G).

**Figure 4.** IUE at E12 with the UbC-StarTrack constructs ( **A**) and CAG-Cre-recombinase, along with the transposase (**B–E**). All animals were injected with Tx at P5 to remove the episomal copies of the

UbC-constructs, and the brain was analysed from P30 onwards. Of the cells produced by CMV-progenitors, 86% were in the GcL and 14% in the GL (**B**). The NG2-progenitors produced 88% GcL cells and 12% GL cells ( **C**), similar to the Gsh-2 progenitors ( **D**), while GFAP progenitors gave rise to only 7% of GL cells (**E**). Nevertheless, more labeled cells were located in the GcL (green) than in the GL (soft pink). The box plot represents the percentage of labeled cells after targeting each set of progenitors with a specific transposase, and the line displays the mean of the data (box and whisker 5th/95th percentile plot). A confidence interval of 95% (*p* < 0.05) was used to determine statistically significant values (\*\*\**p* < 0.001). (**F**). Pallial and subpallial electroporations into the LV produced more cells in the GcL than GL (three animals were analyzed per experiment: CMV; NG2; Gsh-2; GFAP (*n* = 12). Data are shown as average data points ( **G**).

Accordingly, these results sugges<sup>t</sup> that E12 progenitors in dorsal and ventral LV produce more granule cells than periglomerular cells, and this cell fate is independent of the molecular profile of NSC.

In brief, these results summarize the importance of the genes targeted, the location, and the identity of progenitor cells when studying the heterogeneity of specific populations, in this case, adult OB cells (Figure 5). The data obtained open the window for further transcriptomic and clonal studies of these populations in order to define the heterogeneous lineages present in the adult brain.

**Figure 5.** Summary of the importance of age, location, and cell identity to reveal the heterogeneity of NSCs after targeting with the UbC-StarTrack mixture. E12 progenitor cells lining the lateral ventricles can give rise to different neural cell types in the OB depending on their location in the neurogenic/gliogenic niches. The identity of the progenitor is crucial to define the potential fate of the progenitor cells.
