The Role of Foxg1 in the Development of Neural Stem Cells of the Olfactory Epithelium

The olfactory epithelium (OE) of the mouse is an excellent model system for studying principles of neural stem cell biology because of its well‐defined neuronal lineage and its ability to regenerate throughout life. To approach the molecular mechanisms of stem cell regulation in the OE, we have focused on Foxg1, also known as brain factor 1, which is a member of the Forkhead transcription factor family. Foxg1−/− mice show major defects in the OE at birth, suggesting that Foxg1 plays an important role in OE development. We find that Foxg1 is expressed in cells within the basal compartment of the OE, the location where OE stem and progenitor cells are known to reside. Since FoxG1 is known to regulate proliferation of neuronal progenitor cells during telencephalon development, we performed bromodeoxyuridine pulse–chase labeling of Sox2‐expressing neural stem cells during primary OE neurogenesis. We found the percentage of Sox2‐expressing cells that retained bromodeoxyuridine was twice as high in Foxg1−/− OE cells as in the wild type, suggesting that these cells are delayed and/or halted in their development in the absence of Foxg1. Our findings suggest that the proliferation and/or subsequent differentiation of Sox2‐expressing neural stem cells in the OE is regulated by Foxg1.

source of treatment for injured or diseased nervous system tissue.
In order to understand the basic principles that govern the generation and regeneration of neurons in mammals, we have studied the molecular regulation of neurogenesis in a wellcharacterized neurogenic epithelium, the olfactory epithelium (OE) of the mouse. 1, 3 We use OE as a model system, both because of its capacity for continual neurogenesis 6 and because several of its properties simplify the study of neurogenesis. [1][2][3] The OE consists of one major type of neuron, the olfactory receptor neuron (ORN), as well as the stem and progenitor cells that give rise to it. Together, these cell types comprise the OE neuronal lineage, the stages  of which can be clearly defined by molecular markers expressed by cells at different stages of neuronal differentiation. In addition, the OE contains an intrinsic glial cell population, the sustentacular cells, which recent evidence suggests may be derived from the same early stem cells as are ORNs. 1,6-10 Figure 1 shows a schematic of OE neurogenesis. This process begins with OE neural stem cells, which retain a self-renewing ability. These stem cells give rise to the first transit-amplifying progenitors (TA progenitors), which express the Mammalian Achaete Scute Homolog 1 gene (Mash1; also known as Asl1), a proneural gene that encodes a basic helix-loop-helix (bHLH) transcription factor. [11][12][13][14][15][16] Mash1-expressing progenitors in turn give rise to the second TA progenitors, immediate neuronal precursors (INPs), which express a different bHLH proneural gene, Neurogenin 1 (Ngn1). [15][16][17] INPs divide and give rise to daugh-ter cells that undergo terminal differentiation into ORNs, which express the neuronal cell adhesion molecule Ncam 18,19 and subsequently mature to express the olfactory marker protein Omp. 18,20 Each stage of neurogenesis is defined both by the expression of particular marker genes and by a generally consistent histological arrangement of the cells within the OE, which becomes apparent after about day 14 of gestation (E14) in the mouse. [21][22][23] Studies suggest that stem cell and TA progenitors are components of the so-called "globose" basal cell population and reside in the basal compartment of the OE atop the "horizontal" basal cells, which are adjacent to the basal lamina. 1,5,18,24 Once TA progenitors commit to the ORN lineage and undergo postmitotic differentiation, the immature ORNs start migrating from the basal compartment toward a more apical position in the OE and finally become the bipolar mature ORNs, extending dendrites to the nasal cavity and axons through the OE to the olfactory bulb (Fig. 1).
The OE also provides a useful model for studying principles of stem cell maintenance and regeneration of the nervous system. [1][2][3][4][5] The OE retains not only the ability to generate neurons throughout life (it has been estimated that the ORN turnover rate is 3-12 months in rodents 25,26 ) but also a distinct regenerative ability following injury. Studies have shown that, after surgical or chemical ablation the OE undergoes massive cell division and restores almost complete OE formation within 2 weeks. 6,11,27 These characteristics strongly suggest that the OE maintains its neural stem cells not only during embryonic development but also throughout adult life. Thus, unlike other regions of the nervous system, the OE provides a system in which it is possible to investigate neural stem cell behavior during regeneration in the adult nervous system as well as during development.
Despite these advantages, the detailed molecular regulation of the expansion and differentiation of OE neural stem cells is not fully understood. To date, Sox2 is the best molecular marker of the neuronal stem cell of the OE. 3,28 Sox2 is an SRY transcription factor of the SoxB1 family, and it is expressed in multipotent stem cells throughout the neural primordia. [29][30][31] It has been shown that cells expressing Sox2 are capable of both self-renewal and differentiation along different developmental pathways, suggesting that Sox2 expression identifies a stem cell pool. 32 Sox2 is expressed in the OE as well, and importantly it is detected in the basal layer where stem cells are known to reside. 1,3

Role of Foxg1 in OE Stem Cell Development
To investigate the molecular regulation of stem cells in the OE, we have focused on Foxg1 (also known brain factor 1), a member of the Forkhead or Fox proteins, which comprise a large family of winged-helix transcription factors that regulate diverse developmental processes in mammals. 33 Foxg1 has been shown to be important in regulating the development of numerous anterior neural structures, including the cerebral cortex, ventral telencephalon, retina, and OE. [34][35][36][37][38][39] In the OE, Foxg1 expression is detected early during development, when the olfactory placode is forming. 39,40 In Foxg1 null (Foxg1 −/− ) embryos, OE formation is initiated; however, the OE is greatly reduced in size or absent at birth. [39][40][41] Such findings suggest that Foxg1 is expressed by early progenitors and/or stem cells of the ORN lineage and that it plays a role in controlling their expansion. To test this hypothesis, we performed in situ hybridization (ISH) using a Foxg1 cRNA probe at day E14.5 in wild-type mouse OE (Fig. 2). Foxg1 expression in the OE is restricted to the basal layer, where early progenitors such as Ngn1-expressing INPs reside. 2,15,16,42 Several lines of study have demonstrated a decrease in proliferation of neuronal progenitor cells in Foxg1 −/− mice, and this is thought to contribute to the severe hypoplasia observed in the Foxg1 −/− telencephalon. 37,43 Thus, we hypothesized that the early failure of OE development in Foxg1 −/− embryos is due to a reduction of stem cell proliferation and/or successive stages of differentiation, resulting in the loss of OE neuronal cells by birth.
To test this hypothesis, pulse-chase bromodeoxyuridine (BrdU) incorporation experiments were performed. We reasoned that if rapidly dividing neural stem cells were blocked from dividing or differentiating in the OE of Foxg1 −/− mice, an increase in the number of BrdU-retaining cells would be observed in Foxg1 −/− OE but not in the OE of wild-type littermates. BrdU was injected into pregnant dams at E10, when the olfactory pit has formed and cells at all stages of the lineage can be observed in the invaginating neuroepithelium (primary OE neurogenesis). 28 Embryos were collected 52 h later and processed for Sox2 ISH and anti-BrdU immunofluorescence. 44 We chose this paradigm because (1) OE neurogenesis initiates at E9-10, 17,28 (2) OE neuronal progenitor cell cycle length is estimated to be ≤17 h (with approximately 8 h of S phase) in embryonic OE, 19 and (3) the OE of Foxg1 −/− mice is already noticeably hypoplastic and much thinner than normal by E12.5. 39, 42 We quantified how many cells were double positive for Sox2 ISH and BrdU immunofluorescence in the OE, as shown in Figure 3. When these cells were counted, it was found that the percentage of Sox2-expressing cells that retained BrdU in the pulse-chase paradigm was twice as high in Foxg1 −/− OE as in wild-type OE (Fig. 3C). These results indicate that proliferation and/or differentiation of Sox2-expressing neural stem cells is reduced in the absence of Foxg1. The fact that subsequent neuronal cell stages fail to develop in Foxg1 −/− mice 42 further adds to the idea that the absence of Foxg1 results in an early block in OE neurogenesis.

Animals
Foxg1 cre/cre mice, in which the Foxg1 coding sequence is replaced by Cre, 36 were obtained by intercrossing Foxg1 +/cre mice maintained on a Swiss Webster (Harlan, Indianapolis, IN) background. Because Foxg1 cre/cre mice have been used previously for studying Foxg1 null phenotypes and show a phenotype identical to another Foxg1 null allele, Foxg1 lacz/lacz , 37,45 Foxg1 cre/cre mice are designated Foxg1 −/− in the text of this paper. The middle of the day on which a vaginal plug was detected was designated E0.5. All protocols for animal use were approved by the Institutional Animal Care and Use Committee of the University of California, Irvine.

ISH
Dissected tissues were fixed in 4% paraformaldehyde-phosphate-buffered saline (PBS), cryoprotected in 30% sucrose-PBS, and embedded as described elsewhere. 15 Embedded tissue was sectioned on a cryostat at 12-20 μm, depending on the particular analysis. ISH using digoxigenin-labeled cRNA probes was performed according to published protocols. 2

Detection of BrdU Incorporation
For the 52-h BrdU pulse-chase experiment, pregnant mice were given two injections of BrdU (Sigma, St. Louis, MO; 50 μg/g of body weight; 1-h interval) at E10.0 and euthanized 52 h later. Dissected embryo heads were fixed in 4% paraformaldehyde-PBS and cryoprotected in 30% sucrose-PBS. Embedded heads were sectioned on a cryostat at 12 μm. Cryosections were processed for anti-BrdU immunoreactivity according to published methods. 15 For the double staining, Sox2 ISH was performed prior to BrdU immunostaining. BrdU + cells were labeled with monoclonal rat anti-BrdU (clone BU1/75-ICR1; 1:1000 dilution of ascites fluid; Harlan) and detected with Texas Red-conjugated goat anti-rat immunoglobulin G (1:100; Jackson ImmunoResearch, West Grove, PA). For intensity quantification, the pixel fluorescent intensity sum of individual cells was measured using Axio Vision software (Carl Zeiss Inc., Thornwood, NY). The cells exhibiting more than 50% of the highest intensity of the field were scored.

Conclusions and Future Directions
Our laboratory's research focuses on understanding how interacting developmental signaling pathways govern OE neurogenesis. We have shown that multiple factors, including fibroblast growth factor (FGF) and transforming growth factor-β (TGF-β) superfamily members are involved in regulating OE neurogenesis. 16,19,28,47,48 FGFs are important proneurogenic factors for many cells of neuroectodermal origin. Among the FGFs, we have shown Fgf8 to be expressed and involved in the maintenance of the OE neural stem cell population. 28 Fgf8 appears to exert its major neurogenic effect early in development, during the initial invagination of the olfactory pit and the establishment of the neuronal lineage during primary OE neurogenesis. In contrast, our studies indicate that TGF-βs are negative regulators of neurogenesis in the OE and other sensory epithelia and that TGF-βs mediate feedback inhibition of stem and progenitor cell proliferation and/or fate choice in these tissues (submit-ted for publication). 16,44 One member of the TGF-β superfamily, GDF11, is made by ORNs and INPs within the OE proper. GDF11 has the ability to inhibit OE neurogenesis by reversibly arresting cell division by INPs, an effect that is accompanied by increased expression of the cyclin-dependent kinase inhibitor p27 Kip1 . Moreover, Gdf11 null mice show an increase in OE neurogenesis in vivo, with increased numbers of proliferating progenitors (INPs) as well as increased OE thickness and increased numbers of ORNs. 16,44 Interestingly, FoxG1 has been shown to be involved in modulating both FGF and TGF-β signaling pathways. FoxG1 can be phosphorylated by FGF signaling, which promotes the nuclear exportation of FoxG1 and consequently promotes neuronal differentiation. 49 FoxG1 has also been shown to interact with Smadcontaining transcriptional complexes, which mediate canonical TGF-β signaling, and this can regulate the expression of downstream target genes, such as the CKI p21 Cip1 . [50][51][52] From these studies, we hypothesize that stem cell-and progenitor cell-specific intrinsic factors control and modulate extrinsic signals to direct these cells in their choice of fates within their developmental niche. Currently, we are investigating the possibility that FoxG1 modulates these signaling pathways in OE development.