Fibroblast growth factor signaling in ... - Wiley Online Library

Develop. Growth Differ. (2009) 51, 299–323

doi: 10.1111/j.1440-169X.2009.01104.x

Review Blackwell Publishing Asia

Fibroblast growth factor signaling in development of the cerebral cortex Tomoko Iwata1,* and Robert F. Hevner2 1

Division of Cancer Sciences & Molecular Pathology, University of Glasgow, Beatson Laboratories, Garscube Estate, Switchback Road, Glasgow G61 1BD, UK; and 2Department of Neurological Surgery, University of Washington, Seattle Children’s Hospital Research Institute, Seattle, Washington 98101, USA

Despite substantial and exciting recent progress in our understanding of developmental processes in the cerebral cortex, there is still much to be learned about the molecular and cellular mechanisms that account for formation of the cortical structures, and in turn, how the regulation of these mechanisms is linked to cortical functions and behaviors in animals and humans. Fibroblast growth factors (FGFs) are a classic family of growth factors that are important in neural development and whose structures and signaling have been well-studied molecularly and biochemically. Recent advances have revealed their diverse but specific functions in patterning and neurogenesis during cortical development, as evidenced by multiple experimental approaches using in vivo models. Importantly, changes in FGF signaling during development have been shown to influence structure and function of the cerebral cortex as well as animal behavior, and have been implicated in disorders of nervous system function and intellectual development in humans. For example, disturbance of FGF pathways during development has been implicated in the pathogenesis of autism spectrum disorders. Experimental models with altered cortical structure due to perturbations of FGF signaling present a unique opportunity whereby molecular and cellular mechanisms that underlie cortical function and animal behavior can be directly studied and linked to each other. Key words: fibroblast growth factor, forebrain, hippocampus, neurogenesis, patterning.

Introduction Our understanding of the development of the cerebral cortex has progressed exponentially in recent years (Rakic 2006; Bystron et al. 2008). The mammalian cerebral cortex consists of multiple cortical areas with specific functions, such as motor, somatosensory, auditory and visual areas. Positioning and size of each cortical area, as well as their cellular composition and organization, must be strictly controlled for the formation of fully functional wiring network of the cerebrum. Defects in any of the processes of cerebral cortex development cause cortical malformation disorders and neurological syndromes (Francis et al. 2006; Hevner 2007; Guerrini et al. 2008). Furthermore, even a subtle *Author to whom all correspondence should be addressed. Email: [email protected] Received 10 November 2008; revised 06 February 2009; accepted 06 February 2009 © 2009 The Authors Journal compilation © 2009 Japanese Society of Developmental Biologists

change in area patterning and size may drastically influence behavior (O’Leary and Sahara 2008), and could lead to human conditions such as autism spectrum disorders (Vaccarino et al. 2009). Fibroblast growth factors (FGFs) are a family of 22 polypeptides known to play various roles in the control of cell behavior such as cell proliferation, differentiation, and migration, as well as in organogenesis, tissue repair and cancer pathology (Eswarakumar et al. 2005; Mason 2007). FGF signals are mainly mediated by high-affinity receptor tyrosine kinases, FGF receptors (FGFRs). Multiple reviews have already described recent discoveries of FGF functions in neural development and the pathogenesis of developmental diseases (FordPerriss et al. 2001; Dono 2003; Reuss and von Bohlen und Halbach 2003; Hevner 2005; Mason 2007; Hebert and Fishell 2008; Kim et al. 2008). The aim of this review is to summarize what is known so far of FGF signaling specific to development of the cerebral cortex, and to dissect its pleiotropic roles in patterning and neurogenesis, forming the structural bases of cortical function and animal behavior.

300

T. Iwata and R. F. Hevner

Fig. 1. Overview of key events during early development of the cerebral cortex. Initially, patterning defines the positioning of the telencephalon, such as dorso-ventral, rostral (anterior)-caudal (posterior) and medial-lateral, through the actions of morphogens secreted from the signaling centers early in development. Four signaling centers identified so far are; anterior neural ridge (ANR)/commissural plate (CoP), which expresses Fgf3, Fgf8, Fgf15, Fgf17, and Fgf18, the cortical hem expressing Bone morphogenic proteins (BMP)/Wnt, the antihem or pallial-subpallial boundary (PSB) expressing Fgf7, Fgf15, transforming growth factor-α (TGF-α), and secreted frizzled-related protein 2 (Sfrp2). Sonic hedgehog (Shh) defines the forth signaling center in the ventral telencephalon. These morphogens/signaling molecules induce the transcription factors, including Pax6, Emx2, Sp8, and Coup-Tf1, in each specific graded pattern. The combination of these gradients is proposed to form prospective area boundaries, based on which area-specific cytoarchitecture and wiring are built. In the meantime, generation of cortical cells, neurogenesis, defines the size of the telencephalon. Surface area of the cortex expands during the early phase of neurogenesis, followed by a further period of neurogenesis producing diverse populations of glutamatergic projection neurons organized into layered laminar structures. GABAergic interneurons have their origin in the ventral telencephalon and tangentially migrate to settle in the dorsal cortex.

A brief overview of developmental processes in the cerebral cortex Patterning and neurogenesis are two key events that define the structural foundations of the developing cerebral cortex (Fig. 1). Early patterning events establish three main telencephalic domains; the dorsal, dorsomedial, and ventral telencephalon (Sur and Rubenstein 2005). The dorsal cortical primordium gives rise to excitatory neurons of the cerebral cortex, while the dorsomedial telencephalon becomes the hippocampus. The ventral telencephalon produces inhibitory interneurons that migrate to the dorsal cortex and the hippocampus. Several signaling molecules and morphogens are focally expressed and secreted from the signaling centers of the telencephalon early in development; namely, the anterior neural ridge (ANR) at the most rostral edge of the telencephalon, which later becomes the commissural plate (CoP), the cortical hem at the dorsomedial region, and the most recently proposed ‘anti-hem’, or pallial-subpallial boundary (PSB) in the

lateral region of the telencephalon (Sur and Rubenstein 2005; Mallamaci and Stoykova 2006; Rash and Grove 2006; O’Leary and Sahara 2008). The signaling molecules/morphogens known to be expressed in the signaling centers include FGF ligands, especially Fgf8, Fgf15, Fgf17 and Fgf18 in the CoP; bone morphogenic proteins (BMP)/Wnt in the cortical hem; and Fgf7, Fgf15, transforming growth factor-α (TGF-α), and secreted frizzled-related protein 2 (Sfrp2) in the anti-hem. Furthermore, the ventral signaling center is defined by expression of Sonic hedgehog (Shh). Subsequently, the signaling molecules/morphogens induce or regulate the expression of transcription factors, such as Pax6, Emx2, Coup-Tf1, and Sp8, in neuroepithelial cells of the cortical ventricular zone (VZ) according to unique gradient patterns. The combinations of these transcription factor gradients ultimately lead to formation of the prospective area boundaries, based on which the area-specific cytoarchitecture and cortical wiring is established forming the well-defined cortical areas.

© 2009 The Authors Journal compilation © 2009 Japanese Society of Developmental Biologists

Control of patterning and neurogenesis

Following early stages of patterning, the cortex grows drastically in size by neurogenesis. Radial glia and intermediate progenitor cells (IPCs) represent the two main subpopulations of proliferating cells in the developing cortex, which are located in the VZ and subventricular zone (SVZ) (Gotz and Huttner 2005; Dehay and Kennedy 2007; Kowalczyk et al. 2009). The balance of proliferation and cell cycle exit defines the founder population of radial glia, which produce IPCs, neurons and finally glia, leading to the final number of overall cortical cells. The initial proliferation of radial glia contributes to an expansion of surface area of the cortex. In the following phase, cells that have exited the cell cycle migrate outwards to form the cortical plate and mature as excitatory glutamatergic projection neurons within a well-known inside-out laminar structure (Gupta et al. 2002; Cooper 2008). Neurogenesis at this phase, mainly from IPCs but also from radial glia, evidently contributes to the thickness of the cortex (Pontious et al. 2008; Kowalczyk et al. 2009). Diversity of neuronal populations is associated with the timing of cell cycle exit of progenitor cells and laminar- and subtypespecific gene expression (Molyneaux et al. 2007). In contrast, the majority of inhibitory GABA (γ-aminobutyric acid)-ergic interneurons are generated from the ventral telencephalon (ganglionic eminences) and tangentially migrate to settle in the cortical layers during development (Corbin et al. 2001; Marin and Rubenstein 2001).

Biochemical properties of FGF signaling Fibroblast growth factors are a classic class of growth factors, well-studied at the molecular and biochemical levels in vitro. The ligand specificities and downstream signaling pathways are well documented in previous reviews (Ford-Perriss et al. 2001; Reuss and von Bohlen und Halbach 2003; Itoh and Ornitz 2004; Mason 2007; Kim et al. 2008). FGFs are a family of 22 polypeptides that varies in size from 150 to 300 amino acids in humans and can be subdivided into seven subfamilies based on sequence phylogeny (Itoh and Ornitz 2004) (Fig. 2). Extracellular signals from FGFs are transduced into cells by high-affinity FGFR subtypes, Fgfr1–4, encoded by four independent genes (Itoh and Ornitz 2004). FGFRs share a common domain structure consisting of three immunoglobulin-like (Ig) domains and an acidic box in the extracellular side, a single transmembrane domain, and a split-type tyrosine kinase domain in the intracellular side (Eswarakumar et al. 2005) (Fig. 2A). Alternatively-spliced isoforms are generated by mutually-exclusive inclusion of an exon encoding a part of the Ig-III domain, which results in a receptor isoform of either IIIb or IIIc (i.e., ‘FGFR3IIIc’ is an IIIc alternative-spliced isoform of FGFR3). Although

301

similar in domain compositions, FGFRs do vary in (a) ligand binding and (b) intracellular domain that mediates the downstream signals. Thorough studies carried out in vitro using BaF3, mouse pro B-cell lymphoma line, indeed showed different responsiveness of FGFRs to each ligand (Fig. 2B,C) (Ornitz et al. 1996; Zhang et al. 2006). Downstream signaling is initiated by the binding of FGFs to the extracellular ligand-binding domain of FGFRs (Eswarakumar et al. 2005). Receptor dimerization and subsequent autophosphorylation leads to intracellular signaling cascades (Fig. 3A). Various cell surface proteins are known to bind FGFs and FGFRs including heparan sulfate proteoglycans (HSPG) (Ford-Perriss et al. 2002; Allen and Rapraeger 2003), Klotho proteins (Kuro-o 2008), and cell adhesion molecules (CAMs)/ Cadherins (Cavallaro and Christofori 2004; SanchezHeras et al. 2006). The mitogen-activated protein kinase (MapK) pathway is one of the major signaling pathways that control cell proliferation and gene expression downstream of FGFs (Eswarakumar et al. 2005; Mason 2007). Additionally, activation of phosphatidylinositol-3 (PI3) kinase results in phosphorylation of Akt and promotion of cell survival and proliferation (Brader and Eccles 2004). PLCγ (Phospholipase Cγ) signaling is another major pathway classically described as downstream of FGFs, leading to calcium signaling and activation of protein kinase C (PKC). FGF signals also lead to activation of the Rac-cdc42-Rho pathway, modulating cytoskeletal proteins leading to regulation of cell adhesion and migration. Finally, increasing numbers of negative feedback pathways have recently emerged, which include the CbI pathway leading to ubiquitindependent turnover of FGFRs (Monsonego-Ornan et al. 2002), as well as feedback by Sprouty proteins (Mason et al. 2006) and dual-specificity phosphatases (Jeffrey et al. 2007), such as MapK phosphatase 3 (Mkp3)/dual-specificity MapK phosphatase 6 (Dusp6) (Ekerot et al. 2008). However, translation of FGF signaling and function from in vitro studies to in vivo models requires caution, because signaling is context-dependent. FGF signals in the forebrain are mainly mediated by only three out of the four high-affinity FGFRs, Fgfr1–3, in the mammalian forebrain (Mason 2007) (Fig. 4). Of note, Fgfr4 is known to be able to mediate the effect of Fgf8 in BaF3 cells (Fig. 2B). However, its lack of expression during the mammalian forebrain development makes this receptor subtype an unlikely candidate as a receptor for Fgf8 in vivo. Furthermore, in vitro signaling studies often use Fgfr1 as a model and therefore the signaling pathways are often illustrated based on the Fgfr1 data. But as there are differences in the intracellular domain sequences of each FGFR, likewise the cellular components that


302


Fig. 2. Biochemical properties of fibroblast growth factor (FGF) ligand subfamilies and FGF receptor (FGFR) subtypes. (A) Schematic representation of the FGFR domain structure. The high-affinity FGFRs share common domain structure consisting of an acidic box, three immunoglobulin-like domains (Ig-I, Ig-II, and Ig-III), single transmembrane domain, and tyrosine kinase domains (kinase-I and kinase-II). Alternative-splicing in the C-terminal half of Ig-III domain generates receptor isoforms, type IIIb and IIIc. (B) Summary of FGF ligand-receptor specificity. Potential FGF-FGFR combinations are determined by thorough assays of mitogenic responses in BaF3 cell line expressing various FGFR subtypes and alternatively-spliced isoforms upon addition of each FGF ligand. FGF/FGFRs whose roles have been indicated in cortical development are in blue. Higher response > lower response. Summarized based on (Mason 2007) and on the data in (Ornitz et al. 1995; Zhang et al. 2006). (C) Preference of FGF subfamilies by FGFR lllb/lllc alternatively-spliced isoforms. FGF is a family of 22 polypeptides which is subdivided into seven subfamilies based on the phylogeny (Itoh and Ornitz 2004). FGF19 is a human ortholog of mouse FGF15. FGF7 subfamily show preference to IIIb isoform, while FGF4, FGF8, FGF15/19 and FGF9 subfamilies to IIIc. FGF1 subfamily binds to both IIIb/IIIc. FGF11 subfamily (FGF11, FGF12, FGF13, and FGF14) signaling was shown to be independent of FGFRs. Adopted from (Chaffer et al. 2007).

are available for each signaling event may differ among tissues, cell lineages and developmental stages. Therefore, the downstream signaling pathways must be tested in each context in vivo. Indeed, experiments using FGFR constructs in which intracellular domains are swapped showed different signaling abilities of each receptor (Wang et al. 2001). Adding further complexity to FGF signaling in vivo, specific combinations of multiple FGFs and FGFRs are expected within a defined telencephalic and cortical region (Fig. 4). However, the responses of FGFRs, singly or in cooperation, mediating the signals of each ligand to downstream pathways, are largely unknown in vivo

at the molecular level. It is noteworthy that it is unlikely that FGFRs will form a heterodimer (e.g. Fgfr1 and Fgfr3) from a structure point of view (Professor Moosa Mohammadi, pers. comm., 2008). Hetero-dimerization is not structurally preferred in terms of kinase domain, and therefore unlikely to occur under physiologicallyrelevant conditions. From a structural perspective, homodimerization should be strongly favored over heterodimerization, because FGFR-FGFR contacts may be more specific and thus tighter in the context of homodimers than in heterodimers. Nevertheless, in vitro, one could force formation of heterodimers by overexpressing the receptors (Bellot et al. 1991).



303

Fig. 3. Signaling pathways downstream of fibroblast growth factors (FGFs). (A) In vitro studies demonstrated that various signaling pathways mediate FGF signals leading to take effects in downstream gene transcription, cell proliferation, survival and changes in cytoskeleton (cell migration and adhesion). Heparan sulfate proteoglycans (HSPGs), Cell adhesion molecule (CAM)/cadherins, and klotho proteins are examples of molecules known to regulate FGF signaling at the plasma membrane. Major downstream pathways illustrated here are: mitogen-activated protein kianse (MapK), Protein kinase B (PKB)/Akt, Phospholipase Cγ (PLCγ), and Rac-cdc42-Rho pathways. The Cbl pathway that is involved in FGFR turnover is also shown. DAG, diacylglycerol; FRS2, FGF receptor substrate 2; IP3, inositol tris phosphate; PI3 K, Phosphatidylinositol-3-kinase; PKC, protein kinase C; Sos, son of sevenless. Adopted from (Chaffer et al. 2007; Mason 2007; Kim et al. 2008). (B) In contrast, only limited information is available so far towards the underlying signaling mechanisms controlling telencephalic development regulated by FGFs in vivo. A study using primary telencephalic progenitors in culture indicates that Fgf2 and Fgf8 activate both MapK and PKB/Akt pathways as well as its downstream glycogen synthase kinase-3 (GSK3) and p70 S6 kinase/mTor pathways. However the effects by Fgf15 were only apparent in the MapK pathway. Study in primary progenitor cells from Fgfr3 gain-of-function mice indicated that progenitor proliferation downstream of activated Fgfr3 is MapK-dependent, but not PKB/Akt. Based on data from (Thomson et al. 2007; Borello et al. 2008).

FGF signaling is required for the specification of ventral telencephalon and the generation of ventral cell types After initial formation of the telencephalic primordium, patterning takes place to define the dorsal and ventral

subdivisions (Fig. 5) (Hebert and Fishell 2008). Shh, expressed focally in the ventral signaling center, promotes ventralization of the telencephalon by restricting the dorsalizing function of Gli3, a transcription factor, in the ventral region. In the next step, another transcription factor, Forkhead box G1 (Foxg1), together


304


Fig. 4. Overview of expression patterns of fibroblast growth factor (FGF)/FGF receptors (FGFRs) during early development of the cerebral cortex. The data area based on; Fgf2 (Raballo et al. 2000), Fgf3 (Theil et al. 2008), Fgf7 (Assimacopoulos et al. 2003), Fgf8 (Shimamura and Rubenstein 1997) (Cholfin and Rubenstein 2008), Fgf15 (Gimeno et al. 2003; Cholfin and Rubenstein 2008), Fgf17 (Fukuchi-Shimogori and Grove 2003; Cholfin and Rubenstein 2008), Fgf18 (Hasegawa et al. 2004; Cholfin and Rubenstein 2008), Fgfr1 (Vaccarino et al. 1999; Bansal et al. 2003), Fgfr2 (Bansal et al. 2003; Hebert et al. 2003), Fgfr3 (Bansal et al. 2003; Fukuchi-Shimogori and Grove 2003; Garel et al. 2003). Note that the data described here are rather representative and may not show all regions of expression at each developmental stage. A, anterior; D, dorsal; P, posterior; V, ventral.

with Gli3, plays a role in generating the ventral and dorsal identities, respectively (Hebert and Fishell 2008). Foxg1 is essential for the development of ventral cell types and promotes expression of Fgf8 in the rostral signaling center (Fig. 5). In Foxg1 knockout mice, expression of Fgf8 in the rostral telencephalon was dramatically reduced at embryonic day 10 (E10) (Martynoga et al. 2005). In turn, Fgf8 promotes ventral fate by inducing the ventral genes and repressing the dorsal genes (Kuschel et al. 2003). Expression of Foxg1 itself is also induced by Fgf8 (Shimamura and Rubenstein 1997) (Fig. 5), which is further supported by obvious reduction of Foxg1 expression in the rostral telencephalon observed in mice with reduced Fgf8 expression, Fgf8Null/neo and Fgf8TelKO (Foxg1-Cre;Fgf8flox/Null) at E9 (Storm et al. 2006) (Table 1). In addition, Fgf8 expression is regulated downstream of Shh and Gli3.

In Shh−/− mice, Fgf8 expression is present until E8.75, then lost after E9.0 (Aoto et al. 2002; Ohkubo et al. 2002). In contrast, Fgf8 expression is upregulated in Gli3−/− as well as Shh−/−; Gli3−/− mice, indicating that reduction of Fgf8 expression in the absence of Shh is mediated through de-suppression of Gli3 by Shh (Aoto et al. 2002) (Fig. 5). Fgf15 has also recently been shown to be induced by Shh (Rash and Grove 2007; Borello et al. 2008). The importance of FGF signaling in generation of ventral telencephalic cells is further evidenced by a study in which combinations of FGFRs were deleted in mouse telencephalon (Gutin et al. 2006). First, deletion of Fgfr2 and Fgfr3, with remaining one copy of Fgfr1 (Foxg1-Cre;Fgfr1+/flox;Fgfr2flox/flox;Fgfr3−/−) (Table 1) showed a largely normal telencephalon with no obvious dorsoventral specification defects at E12.5, indicating that



Fig. 5. Interactions among fibroblast growth factors (FGFs), Sonic hedgehog (Shh), Gli3 and Forkhead box G1 (Foxg1) (A) FGF8 promotes expression of Foxg1 In turn, Foxg1 and Shh promote FGF8 expression. The effect of Shh is mediated through suppression of Gli3. See text for details. (B) Expression patterns of FGF8 (purple), Shh (brown), Gli3 and Foxg1 (green) in E10.5 mouse rostral forebrain. FGF8 is expressed in the midline commissural plate (CoP), corresponding to the rostral limit of the roof plate. Other FGFs, including FGF15, FGF17 and FGF18 also show expression in or near this zone (Cholfin and Rubenstein 2008). FGF8 and other FGFs are also expressed in the optic stalks and eye primordia (cut off at dotted lines). Shh is expressed in the medial ganglionic eminence (MGE) portion of the ventral telencephalon (VT), as well as the ventromedial hypothalamus (HT) or floor plate. Foxg1 is expressed in a gradient from high ventral to low dorsal within the telencephalon (Yu et al. 2009). Gli3 is expressed throughout the telencephalon (not depicted). Cerebral cortex is derived from the dorsal telencephalon (DT), and FGFs have important roles in not only dorsal-ventral, but also rostralcaudal patterning of the cortex. Note that the expression patterns and developmental neuroanatomy are highly dynamic and this figure represents only a single time point at the onset of cortical neurogenesis.

Fgfr1 plays a major and mostly sufficient role in mediating FGF signaling in early development of the telencephalon. Second, in mice with deletion of both Fgfr1 and Fgfr3 (Foxg1-Cre;Fgfr1flox/flox;Fgfr3−/−), a dorsal marker, Pax6, and ventral markers, DIx2 and Nkx2.1, were expressed with a normal dorsoventral border, indicating that the specification of the ventral telencephalon was normal. However, the morphology of the ventral telencephalon was abnormal without the usual sulcus between lateral and medial ganglionic eminences (LGE and MGE). Decreased expression of Lhx6 and Lhx7, which are MGE markers and downstream targets of Nkx2.1, indicated a failure of proper differentiation in MGE cells at E12.5. In addition, a loss of NPY

305

(Neuropeptide Y) at E18.5 revealed a loss of dorsal MGE and caudal ganglionic eminence (CGE)-derived interneurons. In contrast, deletion of both Fgfr1 and Fgfr2 (Foxg1-Cre; Fgfr1flox/flox;Fgfr2flox/flox) led to a much more drastic reduction of ventral markers, including the loss of Nkx2.1 as early as E9.25, with dorsal makers extended to the ventral midline, indicating a failure of ventral specification (Gutin et al. 2006). The additional deletion of Gli3 in the absence of Fgfr1;Fgfr2 did not rescue its ventral phenotype, providing further evidence that FGF signaling is downstream of Gli3. Third, the study also indicated that double deletions of Fgfr1;Fgfr3 and Fgfr2:Fgfr3 lead to a similar phenotype as Fgfr1 and Fgfr2 single deletions, respectively. Therefore, the roles of Fgfr1–3 are at least partially overlapping, with Fgfr1 and Fgfr2 playing dominant roles over Fgfr3 within the phenotype analyzed (Gutin et al. 2006). Finally, in mice lacking all three FGFRs (Foxg1-Cre;Fgfr1flox/flox;Fgfr2flox/ flox ;Fgfr3−/−), ventral and dorsal telencephalic domains are both lost, indicating the absolute requirement of FGF signaling in telencephalic development (Hebert and Fishell 2008). In summary, the in vivo evidence shows that FGF signaling, occurring downstream of the Shh-Gli3 signaling axis and interacting with Foxg1, is essential for specification of the ventral telencephalon and appropriate generation of ventral cell types, as well as overall telencephlic development. Fgfr1 and Fgfr2 are likely the major FGFR subtypes that mediate the effects of Fgf8 and Fgf15 in the rostral signaling center in these early specification processes.

Multiple FGFs in the rostral signaling center regulate area formation in the dorsal cortex Fgf8 was found to be expressed in the ANR, a strip between neuroectoderm and ectoderm in the most anterior (rostral) edge of the neural plate at E8.5 (Crossley and Martin 1995). Subsequently, the neural plate folds and fuses at the midline, forming the commissural plate (CoP), which continues to express Fgf8 at the most rostral edge of telencephalic primordium. This region has been defined as the rostral signaling center and is important for patterning of the dorsal telencephalon and area formation in the cortex (Shimamura and Rubenstein 1997; Crossley et al. 2001; Ohkubo et al. 2002). The rostral signaling center indeed expresses several FGFs, including Fgf3, Fgf8, Fgf15, Fgf17 and Fgf18 (Fig. 4). Among them, Fgf8, Fgf15, and Fgf17 have been shown to regulate specification of the frontal cortex (Sur and Rubenstein 2005; Mallamaci and Stoykova 2006; Rash and Grove 2006; O’Leary et al. 2007; O’Leary and Sahara 2008). Three FGF members, Fgf8, Fgf17, and Fgf18 belong to the same phylogenic


Mouse models of altered fibroblast growth factor (FGF) signaling that show phenotype in the cerebral cortex and hippocampus

Mouse model

Region-specificity

Single knockout, or gain-of-function of single gene Fgf2−/−

Fgf3−/− Fgf8neo/neo

Fgf8Null/neo

> E8.5, forebrain

Fgf17−/− Fgf18−/− Foxg1-Cre; Fgfr1flox/flox

> E8.5, forebrain

Nestin-Cre; Fgfr1flox/flox

> E11.5 all CNS

Nestin-Cre; Fgfr1ko/flox

> E11.5 all CNS

hGFAP-Cre; Fgfr1flox/flox

> E13.5, radial glia

Syn1-Cre; Fgfr1flox/flox Nestin-Cre; Fgfr2flox/flox, hGFAP-Cre; Fgfr2flox/flox

> E12.5, neurons > E11.5 all CNS (Nestin-Cre), > E13.5, radial glia (hGFAP-Cre)

Fgfr3−/− EIIa-Cre; Fgfr3+/K644E

> one-cell zygote, ubiquitous

Phenotype in the cerebral cortex and hippocampus

Reference

Viable and fertile. Various defects in neurogenesis. No change in cortical volume. Mildly thinner cortex. Reduced number and density of neurons and glia. Reduction in number of glutamatergic projection neurons, but not of GABAergic interneurons. Effects are specific to anterior region of the cortex. Neurogenesis in the ventral cortex is not affected. Reduced lighting reflex upon PTB treatment. No effect in patterning. Further reduction in the telencephalic size in Fgf3−/−;Fgf8TelKO. Fgf8neo allele leads to 40% of normal Fgf8 expression. Survive until birth. ‘Mild’ and ‘severe’ phenotype. No olfactory bulbs in ‘severe’. Rostral shift of presumptive areas (caudalization). Normal cortical size. Abnormal cortico-cortical projection. Rostral shift of presumptive areas (caudalization). Smaller size (75% wt) of the telencephalon, particularly rostral. No septum and preoptic nuclei, optic chiasm, and olfactory bulb. Thicker midline neuroepithelium than Fgf8TelKO. Severer phenotype than Fgf8Null/neo. Smaller size (50% wt) of the telencephalon.

(Dono et al. 1998; Vaccarino et al. 1999; Raballo et al. 2000; Korada et al. 2002; Chen et al. 2008) (Theil et al. 2008) (Garel et al. 2003)

Reduced viability before wean. Changes in expression of cortical patterning genes. Reduced cortical thickness. Survive to adult. Minor reduction in cortical surface area (7%) without overt morphological changes. Behavioral abnormalities in social interaction. Die at birth. Lacks expression of Pea3-Ets transcription factors at E15.5. No migration defects at P0. Homozygous mutants die within 24 h of birth, while Mendelian ratio is kept until E18. Lacks olfactory bulb, due to failed morphogenesis. Commissures fail to cross midline. The septum is missing. The corpus callosum and hippocampal commissures fail to cross midline. The anterior commissure is normal. No change in hippocampal size in adult; however, cell proliferation is reduced in DG. Deficits in memory consolidation, but not spatial learning. Reduced cortical size at P7.5 (15.5%) and adult (12.4%). Reduced number of various classes of interneurons. The corpus callosum and hippocampal commissures fail to cross midline. The anterior commissure is normal. Reduced hippocampus size. Spontaneous hyperactivity. No gross cortical defects. Normal formation of commissures. Viable and fertile. Decrease in cortical volume, pyramidal cell number and density in the cortex, pronounced in medial prefrontal area. Reduced volume of subcortical white matter. Postnatal skeletal phenotype and inner ear defect. Reduced viability at postnatal stages. No report on cortical phenotype so far. Conditional knock-in of activating kinase domain mutation (gain-of-function). Reflects endogenous Fgfr3 expression pattern. Lethal at birth. Skeletal phenotype. However, embryonic brain phenotype is unlikely to be secondary to skeletal defects. Enlarged brain and spinal cord. Increase in cortical thickness.

(Storm et al. 2006)

(Storm et al. 2006) (Wright et al. 2004; Borello et al. 2008) (Cholfin and Rubenstein 2007b) (Hasegawa et al. 2004) (Hebert et al. 2003; Tole et al. 2006) (Smith et al. 2006) (Zhao et al. 2007) (Ohkubo et al. 2004) (Smith et al. 2006; Muller Smith et al. 2008) (Smith et al. 2006) (Smith et al. 2006; Vaccarino et al. 2009) (Colvin et al. 1996) (Deng et al. 1996) (Inglis-Broadgate et al. 2005) (Thomson et al. 2007)


Fgf8TelKO (Foxg1-Cre; Fgf8flox/Null) Fgf15−/−

306


Table 1.

> E8.5, forebrain

Foxg1-Cre; Fgfr1+/flox; Fgfr2flox/flox; Fgfr3−/− Foxg1-Cre; Fgfr1flox/flox; Fgfr2flox/flox; Fgfr3−/−

CNS, central nervous system; DG, dentate gyrus; FGFR, fibroblast growth factor receptor; GABA, γ-aminobutyric acid; PTB, sodium pentobarbital; VZ, ventricular zone.

(Hebert and Fishell 2008)

(Gutin et al. 2006)

(Shin et al. 2004)

Transgenic mice that express dominant negative Fgfr1 that suppress all FGFRs under Otx1 promoter-enhancer. Various defects in neurogenesis specifically in frontal and temporal cortical areas, including thinner cortex and reduced glutamatergic neuron number. Spontaneous hyperactivity. Largely normal morphology, dorsoventral specification and differentiation of ventral cell types at E12.5. Both of the ventral and dorsal telencephalon is missing. > E8.5, dorsal telencephalon, mesencephalon, eyes > E8.5, forebrain Triple knockout and equivalent Otx1-tFgfr1

(Gutin et al. 2006) (Gutin et al. 2006) (Smith et al. 2006) > E8.5, forebrain > E8.5, forebrain > E13.5, radial glia Double knockout Foxg1-Cre; Fgfr1flox/flox; Fgfr2flox/flox Foxg1-Cre; Fgfr1flox/flox; Fgfr3−/− hGFAP-Cre; Fgfr1flox/flox; Fgfr2flox/flox

The ventral telencephalon is not specified at E12.5. Loss of the septum. Single ganglionic eminences at E12.5. Viable and fertile. Reduction in cortical size. Thickness of the VZ is reduced.

Region-specificity Mouse model

Table 1.

Continued

Phenotype in the cerebral cortex and hippocampus

Reference


307

subfamily of Fgf8, while Fgf15 (mouse homolog of human FGF19) belongs to the FGF19 subfamily together with Fgf21 and Fgf23 (Fig. 2C) (Itoh and Ornitz 2004; Mason 2007). The functions of Fgf18 in the rostral signaling center, and the expression patterns of Fgf21 and Fgf23 during cortical development, are unknown so far. The formation of cortical areas can be generally followed in early postnatal mice by appearance of the somatosensory ‘barrel field’ which reflects the somatotopic pattern of whisker representations in the cortex, by immunohistochemistry for serotonin transporter (5-HTT) that detects thalamocortical axon (TCA) projections, as well as Cytochrome C oxidase histochemistry (CO) that visualizes the high cellular metabolism associated with synaptic activities (Wong-Riley 1979; Hamasaki et al. 2004; O’Leary et al. 2007). In addition, area-related shifts in the expression of layer-specific genes, such as ephrin-A5, sFrp2, RZR-beta, Tbr1, Id2, cad6, cad8, EphA7, Lmo3, Lmo4, can be analyzed at late embryonic stages even before the actual appearance of cytoarchitectonic boundaries (Miyashita-Lin et al. 1999; Fukuchi-Shimogori and Grove 2001; Garel et al. 2003; O’Leary et al. 2007). The function of Fgf8 in areal patterning was first demonstrated by ectopic expression in the mouse cortical primordium by in utero electroporation, followed by analysis of area location in early postnatal cortices (Fukuchi-Shimogori and Grove 2001; Fukuchi-Shimogori and Grove 2003). Introduction of Fgf8 in the rostral cortical primordium at E11.5 led to a massive enlargement of rostral cortical region at postnatal day 6 (P6), shifting the barrel field caudally without changing the overall cortical size, indicating that Fgf8 is important for specification of rostral cortical regions (Fukuchi-Shimogori and Grove 2001). Further evidence was provided by introduction of soluble Fgfr3c construct, which is known to sequester Fgf8 with high affinity, leading to the opposite effect, shifting the cortical area rostrally. Finally, electroporation of Fgf8 at a posterior site of the cortical primordium where Fgf8 is not normally expressed, led to the formation of an additional, ectopic barrel field, further supporting the role of Fgf8 as a rostralregion specifying factor. Moreover, Fgf8 was shown to suppress the graded expression of Emx2 (FukuchiShimogori and Grove 2003) (Fig. 6), a key transcription factor expressed in rostrolateral-low caudomedial-high gradient in the cortical VZ and important for caudal region specification (Bishop et al. 2000; Mallamaci et al. 2000; Hamasaki et al. 2004). Conversely, Fgf8 expression is repressed by Emx2, as overexpression of Emx2 reduced Fgf8 expression, while Fgf8 expression was increased in Emx2−/− cortical primordium at E10.5 (Fukuchi-Shimogori and Grove 2003). © 2009 The Authors Journal compilation © 2009 Japanese Society of Developmental Biologists

308


Fig. 6. Regulatory interactions between fibroblast growth factor (FGF) signaling and transcription factors important for formation of cortical area boundaries. Expression gradient of the transcription factors, Pax6, Emx2, Coup-Tf1, and Sp8, is regulated by rostral FGF signaling and antagonizing bone morhogenetic protein (BMP)/ Wnt signaling from the cortical hem. Fgf8 suppresses Emx2 and Coup-Tf1 expression, while it increases Sp8 expression. The effects of Fgf17 in regulation of these genes are smaller, which is in accordance with the more localized phenotype of Fgf17−/− in subdivisions of the frontal area. Fgf15 regulates Coup-Tf1 and Sp8 in a fashion opposite to Fgf8/17. Interactions between transcription factors are also reviewed in (O’Leary and Sahara 2008).

In genetic models with reduced Fgf8 expression, area formation was also investigated by analyses of transcription factor gradients and layer-specific genes that are expressed in presumptive cortical areas (Garel et al. 2003) (Fig. 6). In Fgf8neo/neo (FGF8 hypomorphic) mice, expression of area-specific layer genes was shifted rostrally. In addition, there was an increase of caudal gene expression, including Emx2, CoupTf1, Dbx1, Fgfr3, at E11.5-E14.5, while little change was observed in a rostral gene, Pax6. Moreover, the study of mouse models with further reduction of Fgf8 expression in Fgf8Null/neo and Fgf8TelKO (Foxg1-Cre;Fgf8flox/Null) showed rostral expansion of Emx2 at E9.5 and E11.5, and similar rostral expansion of Coup-Tf1 in Fgf8Null/neo at E11.5 (Storm et al. 2006). This suggests that Fgf8 specifies frontal cortical areas by regulating caudal gene expression beginning from early stages of cortical development. In addition, analysis of Cad6 and Id2 expressions in newborn Fgf8neo/neo pups showed that within the presumptive prefrontal, motor, and somatosensory areas, lateral regions were shifted medially, indicating that Fgf8 specifies medial regions of the rostral cortex (Garel et al. 2003). Fgf17 is expressed in CoP at E10, in a slightly broader area than Fgf8, and has been shown to specifically regulate the formation of area-related subdivisions within the rostral cortex (Fukuchi-Shimogori and Grove 2003; Cholfin and Rubenstein 2007b; Cholfin and Rubenstein 2008; Dominguez and Rakic 2008). Overall growth of the cortex was largely normal in Fgf17−/− mice

with relatively small reduction of cortical surface area (7%) in adult stage, without any obvious morphological defects. Nevertheless, Fgf17−/− cortices showed a rostral shift of cortical areas at P7 as well as adult stages, with corresponding changes in area-related laminar markers at P0. Compared with Fgf8, the effect of Fgf17 inactivation was relatively specific to the frontal cortex and its subdivisions (Cholfin and Rubenstein 2007b). Within frontal cortex, a 54% reduction in surface area was observed in the dorsal area marked by Lmo4 expression. Panels of area-specific layer gene expression examined within the frontal cortex showed a selective loss of dorsal frontal regions, with preserved medial and orbital (ventral) frontal cortex in Fgf17−/−. Why do Fgf8 and Fgf17 loss-of-function result in distinct outcomes with regard to area formation? Changes in gene expression were closely examined in comparison with the Fgf8neo/neo model (Cholfin and Rubenstein 2008; Dominguez and Rakic 2008). In Fgf8neo/neo mice with severe FGF hypomorphism, the cortex demonstrated caudalized expression of markers in the dorsal and medial frontal cortex, and reduction of orbital frontal cortex size at P0 (Cholfin and Rubenstein 2008). In accordance to the severity of the rostral shift of cortical areas, upregulation of caudal genes, Coup-Tf1 and Emx1, was more significant in Fgf8neo/neo than in Fgf17−/− (Fig. 6). In contrast, within the frontal cortex, there was a significant reduction of Sp8, Erm, and Er81 expression in Fgf17−/− null mice, while changes in these genes were small in Fgf8neo/neo. Furthermore, expression of genes downstream of FGF signaling, such as Spryl and Spry2, was more strongly reduced in Fgf8neo/neo in the core CoP region than in Fgf17−/− penumbra (‘shadow’ adjacent region) at E10 and E12. All together, the differential functions of Fgf8 and Fgf17 could be explained by their local expression domains (core vs. penumbra) and downstream signaling targets, possibly caused by the distinct responsiveness of FGFR subtypes (i.e. Fgfr1 and Fgfr2 in regions of Fgf8/Fgf17 expression at this stage) that are likely to mediate the effects of these ligands (Fig. 2B). Finally, Emx2−/−;Fgf17−/− double knockout rescues the phenotype and gene expressions of Fgf17−/−, as well as of Emx2−/−, indicating their antagonistic functions in regulating regionalization of the frontal cortex. Interestingly, Fgf15 has recently also been implicated in the rostral signaling center, but was shown to influence patterning in a fashion opposite to Fgf8 (Borello et al. 2008). In Fgf15−/− mice, CoupTf1 was downregulated at E9.5–E14.5 and Sp8 was upregulated at E12.5 (Fig. 6). However, changes in expression of Pax6 and Emx2 were subtle at E12.5 and E14.5. These findings indicate that FGFs do not signal in a simple, linear additive manner in the rostral signaling center, but



rather pattern the cortex through a complex interplay of effects that may differ both qualitatively and across different spatial domains. Fibroblast growth factors can also have complex interactions with upstream and downstream signaling molecules and pathways. For example, transcription factor Sp8 was recently shown to play both Fgf8dependent and -independent roles in cortical areal patterning (O’Leary et al. 2007; Sahara et al. 2007; Zembrzycki et al. 2007; O’Leary and Sahara 2008). Sp8 was shown to be a direct transcriptional activator of Fgf8 in the CoP, while Emx2 is likely to limit Fgf8 expression by suppressing Sp8 activity in the rest of the cortical primordium (Sahara et al. 2007) (Fig. 6). In addition, Sp8 expression was also reciprocally regulated by other FGFs, as its expression was reduced in Fgf17−/− (Cholfin and Rubenstein 2008) and increased in Fgf15−/− mice (Borello et al. 2008). The anti-hem, or pallium-subpallium boundary (PSB), has been proposed as a potential signaling center located at the lateral telencephalon (Fig. 1) (Mallamaci and Stoykova 2006; O’Leary et al. 2007). Fgf15 is additionally expressed in the anti-hem as well as in the caudal ganglionic eminence (CGE), the caudal region of ventral telencephalon (Gimeno et al. 2003; Cholfin and Rubenstein 2008). Expression of Fgf7 was reported in the anti-hem at E13.5 (Assimacopoulos et al. 2003). However, the functions of FGFs in these domains are still to be determined. Unique and redundant functions of Fgf3 and Fgf8 in the ANR have been demonstrated in zebrafish (Walshe and Mason 2003). In contrast, deletion of Fgf3 in mice (Fgf3−/−) showed no defects in the formation of signaling centers and regional specification, indicating that Fgf3 is dispensable for early telencephalic patterning (Theil et al. 2008). In summary, in vivo studies using both genetic models and gene manipulative methods show that ligands in the Fgf8 subgroup (Fgf8, Fgf17) play an important role in area formation. Despite the very similar expression patterns of Fgf8 and Fgf17, their functions in area formation are distinct. Function of Fgf15 is additionally indicated in area formation, however, that of Fgf18, Fgf21 and Fgf23, has not been reported so far. The role of the PSB as a signaling center, and that of FGFs expressed in the PSB, are much less defined. It remains possible that additional FGFs are expressed in the developing cortex, whose functions are still to be explored.

Cooperation of FGFRs is likely to mediate FGF signals in cortical area formation Fgfr1–3 are expressed in proliferating progenitors of the cortical VZ, although previous studies have not yet

309

distinguished expression by radial glia progenitors and/ or IPCs (Bansal et al. 2003; Hebert et al. 2003). Current knowledge of the specific roles of individual receptors is limited, mostly due to early lethality of the mouse models upon gene deletion in Fgfr1 (Yamaguchi et al. 1994; Deng et al. 1997) and Fgfr2 (Arman et al. 1998; Xu et al. 1998), and due to functional redundancy. Nonetheless, some of the functions of each FGFR have emerged. Forebrain-specific knockout of Fgfr1 (Foxg1-Cre; Fgfr1flox/flox) showed a specific loss of the most anterior telencephalic structure, the olfactory bulb (Hebert et al. 2003; Hebert and Fishell 2008). Except for downregulation of a rostral area marker cad8 at E16.5, specification of the rostral region occurred normally in Foxg1-Cre;Fgfr1flox/flox, as evidenced by normal expressions of ephrin A5 and Pou3f1 at E12.5, as well as ephrin A5 and Tbr1 at E16.5. Therefore, the lack of olfactory bulb in Foxg1-Cre;Fgfr1flox/flox is unlikely to be due to a failure of early rostrocaudal patterning or regionalization of the cortex. Instead, abnormal morphogenesis was associated with reduction of progenitor proliferation at the rostral end of the telencephalon normally associated with the initial bulb evagination at E12.5 (Hebert et al. 2003). Interestingly, a rostral shift of Emx2 and Pax6 gradients was observed nonetheless, although the effects were milder than those observed in genetic models of reduced Fgf8 (Garel et al. 2003; Storm et al. 2006). The possibility of other FGFRs compensating the lack of Fgfr1 is likely. The expression of Fgfr2 and Fgfr3 are largely unchanged in Foxg1-Cre; Fgfr1flox/flox at E12.5; however, lower responsiveness to Fgf2 treatment in inducing downstream gene, fos, was evident in telencephalic cells from Foxg1-Cre;Fgfr1flox/flox (Hebert et al. 2003). Also, it should be noted that the Foxg1-Cre allele itself causes a reduction of cortical progenitor proliferation as well as hippocampal abnormalities, which might interact with FGF receptor gene inactivation (Shen et al. 2006; Eagleson et al. 2007; Siegenthaler et al. 2008). A unique graded pattern of Fgfr3 expression in the cortical VZ, similar to those of transcription factors that plays a key role in cortical area formation, has been observed during E11.5–E13.5 (Muzio et al. 2002; FukuchiShimogori and Grove 2003; Garel et al. 2003; Thomson et al. 2007). Fgfr3 is expressed in progenitor cells of the neuroepithelium from E9.5 in mice (Peters et al. 1993). Fgfr3 is then expressed in the cortical VZ/SVZ throughout neurogenesis, but is not expressed in postmitotic neurons in the cortical plate. In vitro assays have shown that Fgfr3 responds highly to Fgf8, Fgf15, and Fgf17, the ligands indicated to play a role in area formation (Ornitz et al. 1996; Zhang et al. 2006) (Fig. 2B).


310


Fgfr3−/− mice show skeletal overgrowth and deafness owing to inner ear defects, indicating the role of Fgfr3 in skeletal growth and hearing (Colvin et al. 1996). However, the role of Fgfr3 in brain development has been uncertain. Indeed, it has long been believed that Fgfr3−/− has little phenotype in the brain, with an exception of findings on glial populations and brain regions other than the cerebral cortex. A temporal delay in differentiation of oligodendrocytes was reported in Fgfr3−/− at P2 (Oh et al. 2003), and increased expression of an astrocyte marker, glial fibrillary acidic protein (GFAP), was observed in the hindbrain, the cerebellum, and the spinal cord in young postnatal and adult Fgfr3−/− mice (Oh et al. 2003; Pringle et al. 2003). A twofold increase in TUNEL+ apoptotic cells was reported specifically in the cerebellar cortex of Fgfr3−/− at P9, but not in other brain regions (Oh et al. 2003). The number of tyrosine hydroxylase (TH)-positive dopaminergic (DA) neurons is reduced in heterozygous Fgfr3 knockout mice in the substantia nigra, indicating a role of Fgfr3 in midbrain development (Timmer et al. 2007). Nonetheless, the role of Fgfr3 in regulation of cerebral cortical area formation has never been conclusively reported (Grove and Fukuchi-Shimogori 2003). In our hands and others (Timmer et al. 2007), Fgfr3 knockout mice (Colvin et al. 1996) rarely survive beyond birth in C57/BI6 background (data not shown), hampering the investigation of the cortical areas. Conditional knockout strategy, similar to models of Fgfr1 and Fgfr2, would greatly facilitate the study and the understanding of Fgfr3 function. Interestingly, Fgfr3 expression itself is regulated downstream of Fgf8, Pax6 and Emx2. The rostral factors, Fgf8 and Pax6, downregulate Fgfr3 expression, while the caudal factor, Emx2, upregulates Fgfr3 (Muzio et al. 2002; Fukuchi-Shimogori and Grove 2003; Garel et al. 2003; Nomura et al. 2007). The relationship between Coup-Tf1 and Fgfr3 is of particular interest. Both Coup-Tf1 and Fgfr3 are expressed in similar, although not completely identical, rostromedial-low, caudolateralhigh expression gradients (Zhou et al. 2001). Coup-Tf1 was shown to be important for cortical area patterning and neurogenesis, particularly that of the parietal and occipital cortices (Zhou et al. 2001; Armentano et al. 2007; Faedo et al. 2008). Fgfr3 was shown to be upregulated in the D6-deriven Coup-Tf1 transgenic mice and downregulated in Coup-Tf1−/− at E11.5 and E13.5, suggesting that Coup-Tf1 normally promotes Fgfr3 expression in the embryonic cortex (Faedo et al. 2008). Therefore, it is an interesting possibility that Fgfr3 may play a role in patterning as a direct or indirect downstream target of Coup-Tf1, as well as other transcription factors involved in area formation.

Taken together, there has so far been no report on a single FGFR deletion leading to changes in area positions as drastic as in mice with altered expression of FGFs in the rostral signaling center. However, it is likely that a combination of, or all FGFRs are responsible in mediating the effects of rostral FGFs.

FGF signaling controls telencephalic progenitor populations by regulating diverse cell proliferation parameters The relation between neurogenesis and cell cycle length has been a subject of much interest, and is a potential site of FGF effects. It was hypothesized that the founder population of cortical neurons is generated by 11 cell cycles between E11 and E17 in mice, and that the length of the cell cycle, particularly that of the G1 phase, becomes longer as development progresses (Caviness et al. 1995; Takahashi et al. 1995; Dehay and Kennedy 2007). The ‘cell cycle length hypothesis’ has recently been proposed based on experimental work in vitro and in vivo (Lukaszewicz et al. 2002; Calegari and Huttner 2003; Gotz and Huttner 2005). According to this hypothesis, the lengthening of the G1 phase plays a causative role in the switch of progenitor division mode, from proliferative to neurogenic. Growth factors, or mitogens, such as FGFs, are a classic class of molecules that control the length of the G1 phase. Addition of Fgf2 in primary culture prepared from the developing cortex at E14–E16 showed a shortening of the G1 length and an increase in proliferative divisions, indicating that Fgf2 controls cell proliferation via its control of the G1 length (Lukaszewicz et al. 2002). Fibroblast growth factors have been shown to regulate the telencephalic size, but the underlying mechanism appears to be complex in vivo. In addition to patterning, Fgf8 expression levels also affected telencephalic size, particularly in rostral regions (Storm et al. 2003; Storm et al. 2006). The size of the telencephalon was largely normal in Fgf8neo/neo (Garel et al. 2003); however, Fgf8Null/neo (further reduction of Fgf8 than Fgf8neo/neo) and Fgf8TelKO (Foxg1;Fgf8flox/Null, total lack of Fgf8) showed significant reduction of the telencephalic size at E12.5 (75%wt and 50%wt, respectively) (Storm et al. 2006). Analysis using phosphohistone-H3 (pHistone-H3), a marker for mitosis, showed that the mitotic index was reduced to 50% in the rostroventral region in Fgf8Null/neo and Fgf8TelKO at E9.0. Furthermore, TUNEL assay showed a 10-fold increase of apoptosis in Fgf8Null/neo and Fgf8TelKO. In sum, the results indicate that Fgf8 controls the rostroventral telencephalic size by increasing cell proliferation and suppressing apoptosis. In contrast, cell survival at the rostral midline is maintained



by Fgf8 in a dose-specific manner. While total lack of Fgf8 (Fgf8TelKO) leads to an increase in apoptosis owing to the loss of downstream Foxg1 expression, reduced Fgf8 expression (Fgf8Null/neo) leads to an increase in cell survival owing to the positive balance of Foxg1dependent survival effects and anti-survival effect from negative feedback of Fgf8 signaling (Storm et al. 2003). Although deletion of Fgf3 alone does not lead to any apparent change in telencephalic size, deletion of Fgf3 in addition to Fgf8 (Fgf3−/−;Fgf8TelKO) leads to further reduction of the telencephalic size beyond Fgf8, indicating that Fgf3 may function in synergy with Fgf8 in this aspect (Theil et al. 2008). In contrast, in Fgf17−/− mice, changes in cell proliferation or apoptosis were not detectable at E9.5 and E10.5 (Cholfin and Rubenstein 2008). In Fgf15−/− mice, the cortical plate was thinner at E14.5 (Borello et al. 2008). The mitotic index was increased by 20–40% during E12.5–E14.5, with parallel increases in the VZ and SVZ (37% and 42%, respectively). In addition, total cell cycle length was decreased by 1 h at E12.5 (11 h in wt, 10 h in Fgf15−/−) and by 1.5 h at E14.5 (13.5 h in wt, 11 h in Fgf15−/−). Finally, cell cycle exit was reduced by 38% at E14.5. All together, this indicates that during the early phase of neurogenesis, Fgf15 promotes progenitor differentiation by lengthening the total cell cycle length and promoting cell cycle exit (Borello et al. 2008). In Foxg1-Cre;Fgfr1flox/flox;Fgfr2flox/flox, cell proliferation in the ventral telencephalon was reduced by 12% at E10.5 which may contribute to the loss of ventral regions observed in addition to defects in dorso-ventral specification (Gutin et al. 2006). A significant increase (57.8%) in apoptosis was also observed at the dorsal midline, which is likely to be directly induced by upregulation of BMP through de-suppression by Foxg1. In contrast, the reduction of ventromedial regions in Foxg1-Cre;Fgfr1flox/flox;Fgfr3−/− is likely owing to reduced cell cycle exit (44% of control) during E10.5–E11.5 (Gutin et al. 2006). The differential effects observed in deletion of different receptor combinations could be potentially explained by receptor-specific cellular responses (Marie et al. 2007). An in vitro study using rat E13.5-derived neural stem cell cultures at 7 days in vitro showed that the presence of both Fgfr1 and Fgfr3 drives self-renewal, while the presence of either single receptor drives neurogenic division (Marie et al. 2007). A potential role of Fgfr3 in cortical progenitor proliferation has been indicated by studies using a gainof-function model. Mice expressing constitutively active mutant Fgfr3 allele displayed an enlarged cerebral cortex with an increased cortical thickness and total cell number, mostly due to an increased progenitor

311

proliferation and a decreased apoptosis during the early stage of neurogenesis (Inglis-Broadgate et al. 2005; Thomson et al. 2007). The gain-of-function model carries a mutation substituting the amino acid Lysine 644 to Glutamic Acid (K644E) (Iwata et al. 2000). The mutant allele is knocked-in to the endogenous Fgfr3 locus, allowing preservation of normal Fgfr3 expression domains and levels. Biochemically, the K644E mutation highly activates Fgfr3 signaling in terms of kinase activity and autophosphorylation in vitro (Naski et al. 1996; Webster and Donoghue 1996; Iwata et al. 2001). BaF3 cells expressing Fgfr3 with this mutation show augmented mitogenic response both in the absence of ligand (−25% of full response) and in a concentrationdependent manner (an increasing response by a further addition of ligand) with an overall 2–10-fold increase compared with control (Naski et al. 1996). A graded increase in progenitor proliferation was observed in the cortical VZ along the rostrocaudal axis in the gain-of-function Fgfr3 mutant cortex relative to controls, with the highest effect in caudal cortex at E11–E13, consistent with the expression gradient of Fgfr3 (Thomson et al. 2007). Although a possible function of Fgfr3 in cortical neurogenesis is implicated by these studies, a drawback in gain-of-function approaches is that the possibility of triggering ectopic (non-physiological) pathways cannot be excluded. In order to address whether the above effects reflect the normal function of Fgfr3, and to further clarify Fgfr3 functions in cerebral cortex development, generation of a conditional lossof-function (knockout) Fgfr3 model, and its careful and detailed analysis, will be necessary. Further importance in the study of Fgfr3 is illustrated by its potential function relevant to human cortical development and mental health. The corresponding kinase-domain mutations in human FGFR3 are known to cause a cortical malformation that accompanies severe and lethal dwarfism, Thanatophoric Dysplasia (TD) (OMIM#187601). The cerebral neocortex in TD is markedly enlarged (megalencephaly) and excessively convoluted, especially in temporal and occipital lobes, where sulcation begins prematurely (Hevner 2005). In addition, neuropathological study demonstrated that the cortical abnormalities in TD are likely to be attributed to defects beginning in early corticogenesis. The hypothesis that drastic increase in the surface area of occipitotemporal cortex in TD reflects an early regional expansion of the cortical primordium (Hevner 2005) is supported by recent studies of mouse models (Inglis-Broadgate et al. 2005; Thomson et al. 2007; Thomson et al. 2009). In summary, various proliferative parameters, including cell cycle length, cell cycle exit, and apoptosis, is involved in FGF regulation of telencephalicsize. Fgf3,


312


Fgf8 and Fgf15 are so far shown to be involved in this process. Each of the cell proliferative parameters is independent, and is shown to be differentially regulated by distinct FGF ligands and FGFRs. Results from the studies of combinatory FGFR deletions further support the differential effects of ligand-specific signaling in the control of cell proliferative parameters during development of the telencephalon.

FGF signaling controls genesis of diverse subpopulations of cortical cells Fgf2 expression is detected in the neuroepithelium as early as E9 (Nurcombe et al. 1993). Fgf2 is expressed highly in the VZ/SVZ until as late as E17.5 in rat, and also in the meninges and choroid plexus; however, its expression is not detected in the cortical plate of embryonic or early postnatal rodent cortex (Dono et al. 1998; Vaccarino et al. 1999; Raballo et al. 2000). Furthermore, in the VZ/SVZ, Fgf2 expression is graded in a high-dorsal, low-ventral fashion and this may contribute to phenotypic differences along the dorsoventral axis (Raballo et al. 2000) (see below). Studies of Fgf2−/− mice indicated its important role in neurogenesis (Dono et al. 1998; Vaccarino et al. 1999; Raballo et al. 2000; Korada et al. 2002; Chen et al. 2008). Despite normal gross morphology, several abnormalities were indicated in the initial study of Fgf2−/− mice, including reduced density of Parvalbumin-positive neurons and reduced neuronal migration to layer II and III in motor and somatosensory areas of the adult cortex (Dono et al. 1998). In addition, a reduced cortical thickness at E15.5 to birth and very few neurons with large pyramidalshape was reported (Dono et al. 1998; Raballo et al. 2000). The subsequent studies revealed the role of Fgf2 more in detail, particularly in control of neuronal density and cortical progenitor proliferation (Vaccarino et al. 1999; Raballo et al. 2000; Korada et al. 2002; Chen et al. 2008). In adult Fgf2−/− mice, a change in the total volume of the cortex is small (15% reduction) or not significant; however, total cell number and cell density are significantly reduced (45% and 38% reduction, respectively) (Vaccarino et al. 1999; Chen et al. 2008). Both neuron number (assessed by NeuN) and glia number (assessed by S-100) are equally reduced (Table 2). This change is associated with as much as 41% reduction of the VZ volume and 60% reduction of total cell number in the early period of neurogenesis (E10–12.5) in Fgf2−/− embryos (Vaccarino et al. 1999; Raballo et al. 2000). Furthermore, the lack of Fgf2 leads to a reduced cell proliferation with a reduction of growth fraction (proportion of proliferating cells) by 6–17% at E10–12.5 and a similar cell cycle length (Raballo et al.

2000; Vaccarino et al. 1999). These effects of Fgf2 are specific to the dorsal cortex, and not observed in the ventral cortex (Raballo et al. 2000), which is likely to be owing to its dorsoventral expression gradient. Apoptosis was not influenced by the lack of Fgf2 (Raballo et al. 2000). Within neuronal populations, the number of glutamatergic projection neurons was reduced, but that of the interneurons was unchanged in adult Fgf2−/− mice (Korada et al. 2002) (Table 2). The reduction in the number of pyramidal neurons was observed in both upper and lower layers (60% and 48% reduction, respectively) and the effects were limited to anterior cortical regions (dorsolateral prefrontal and parietal, but not occipital cortex) (Korada et al. 2002). In addition to the rostral signaling center, Fgf18 is expressed in the cortical plate at E13.5–E16.5 (Hasegawa et al. 2004). Pea3-Ets transcription factors are downstream genes of FGF signaling and their suppression was shown to result in a failure of progenitor migration to the cortical plate (Hasegawa et al. 2004). Ectopic expression of Fgf18 in the cortical VZ at E13.5 induced Pea3-Ets transcription factors, while in Fgf18−/− mice, Pea3-Ets expression were greatly reduced. Therefore these results suggested the role of Fgf18 in lamination via control of Pea3-Ets during mid–late phase of cortical development. However, lamination was not significantly affected in Fgf18−/− mice at P0. Genetic compensation downstream of Pea3-Ets may have rescued the radial migration defect in Fgf18−/− mice. Fibroblast growth factors that play a role in neurogenesis are otherwise largely unexplored. However, the importance of FGF signaling in neurogenesis, not only in frontal but also in caudal areas, is clearly evidenced by the study in which signaling of all FGFs are suppressed by the dominant negative FGFR construct (Shin et al. 2004). A truncated form of Fgfr1, which lacks tyrosine kinase domain (tFgfr1) was shown to act as a dominant-negative suppressor of FGF signaling (Shin et al. 2004). Transgenic mice that expressed tFgfr1 under the Otx1 promoter-enhancer, which drives expression in the dorsal telencephalon and mesencephalon from E9.5 (Otx1-tFgfr1), revealed abnormalities in medial prefrontal and temporal cortical regions. Cortical thickness was reduced by 30–40%, particularly affecting lower cortical layers with reduced cortical cell number (38%). Projection neurons showed a reduced soma size and shorter dendritic processes. Furthermore, the number of glutamatergic pyramidal neurons, including Tbr1+ layer 6 neurons, was reduced in the frontal and temporal cortex; however, that of GABAergic interneurons remained largely unaffected (Shin et al. 2004) (Table 2). In addition, disorganized morphology of radial glia was observed in frontal and temporal


Table 2.

Cortical and hippocampal cell populations influenced by changes in fibroblast growth factor (FGF) signaling

Mouse model Cortex Fgf2−/−

Cell type

Marker used

Phenotype

Reference

P1-P3 Adult

Neuron Overall neuron Neuron Glutamatergic projection neurons

NeuN Nissl NeuN Glutamate, SMI-32, Latexin

(Raballo et al. 2000) (Vaccarino et al. 1999) (Vaccarino et al. 1999) (Korada et al. 2002)

Tbr1 (layer 6) Calbindin Parvalbumin

Overall glia Glia

Nissl S-100, GFAP

48% reduction 34–38% reduction

E18.5

Radial glia

RC2, GLAST, BLBP

Adult

Glutamatergic projection neurons GABAergic interneurons

Tbr1 (layer 6), SMI-32

Disorganized, fragmented, discontinuous in frontal and temporal areas 46% reduction (Tbr1)

GABA Calbindin Parvalbumin

hGFAP-Cre; Fgfr1flox/flox

hGFAP-Cre; Fgfr2flox/flox hGFAP-Cre; Fgfr1flox/flox Fgfr2flox/flox Hippocampus hGFAP-Cre; Fgfr1flox/flox

P48

Neuron Glutamatergic projection neurons GABAergic interneurons

NeuN Tbr1

Adult (11–12 months)

GABAergic interneurons

P7.5 E18.5

Astrocytes Radial glia

Parvalbumin Calretinin Parvalbumin, Calbindin, Somatostatin, Calretinin, Gad67 GFAP RC2, BLBP, GFAP

Adult hippocampus (VZ) Adult hippocampus (DG)

Neuron GABAergic interneurons Neuron GABAergic interneurons

NeuN Parvalbumin NeuN Parvalbumin

(Chen et al. 2008) (Korada et al. 2002) (Korada et al. 2002; Chen et al. 2008) (Vaccarino et al. 1999) (Vaccarino et al. 1999; Chen et al. 2008) (Shin et al. 2004)

(Shin et al. 2004)

No significant change Not affected in frontal area, but reduced in temporal cortex Slightly increased in both frontal and temporal areas No change. No significant change

(Shin et al. 2004) (Shin et al. 2004)

(Muller Smith et al. 2008) (Muller Smith et al. 2008)

28.3% reduction No change. 20–47% reduction

(Muller Smith et al. 2008) (Muller Smith et al. 2008) (Muller Smith et al. 2008)

40–60% reduction 40–60% reduction

(Smith et al. 2006) (Smith et al. 2006)

51.4% reduction 50.0% reduction 28.8% reduction 64.8% reduction

(Ohkubo et al. 2004) (Ohkubo et al. 2004) (Ohkubo et al. 2004) (Ohkubo et al. 2004)

(Shin et al. 2004)

DG, dentate gyrus; GABA, γ-aminobutyric acid; GFAP, glial fibrillary acidic protein; VZ, ventricular zone.

313


GABAergic interneurons

46% reduction 45% reduction 23% reduction 38% reduction. Prefrontal and parietal, but not in occipital. 19% reduction No change No significant change


Otx1-tFgfr1

Stage

314


areas in the Otx-tFgfr1 cortex at E18.5, associated with reductions of VZ thickness and proliferating cell numbers at E12.5. Reduction of neurogenesis was also reported upon single deletion of FGFRs, indicating that each FGFR could individually play a role. A conditional deletion of Fgfr1 using hGFAP-Cre that drives recombination in radial glia from E13.5 (hGFAP-Cre;Fgfr1flox/flox) leads to a 15.5% reduction of overall cortical size at P7.5, while only 12.4% in adult stage (Ohkubo et al. 2004). Whereas the numbers of overall neurons and Tbr1+ glutamatergic neurons were unaffected, that of Parvalbumin+ interneurons was specifically reduced at P48, and subsequently, numbers of various subclasses of interneurons, including Parvalbumin, Calbindin, Somatostatin, Calretinin and Gad67 were reduced by 20 –47% in 11–12-month-old mice (Muller Smith et al. 2008). Finally, recent reports have revealed a role of Fgfr2 in proliferation of radial glia and in radial glia-astrocyte transition at the end of neurogenesis in the dorsolateral cortex (Smith et al. 2006; Vaccarino et al. 2009). Deletion of Fgfr2 (Nestin-Cre;Fgfr2flox/flox and hGFAPCre;Fgfr2flox/flox) leads to a reduced cortical volume, as well as decreased pyramidal cell number and density in the cortex, most pronounced in medial prefrontal area. In addition, despite the normal number and density of radial glia at a late stage of neurogenesis (E18.5), the density of GFAP+ astrocytes at P7 was reduced in hGFAP-Cre;Fgfr2flox/flox, and this effect was most pronounced in the upper cortical layers at a farther distance from the SVZ. Additional evidence from hGFAP-Cre;Fgfr1flox/flox, in utero electroporation of small hairpin RNA (shRNA) targeting Fgfr1, and Fgf8 beads applied in E14.5 cortical slice culture elegantly showed that Fgf signaling promotes transition from radial glia to astrocytes at the end of neurogenesis by inducing the retraction of apical endfeet and somal translocation. In dorsolateral cortex, Fgfr2 is likely to play this role singly, as this change was observed similarly in hGFAPCre;Fgfr1flox/flox;Fgfr2flox/flox and in hGFAP-Cre;Fgfr2flox/flox (Smith et al. 2006). In summary, Fgf2 is indicated to mainly play a role in the genesis of both neuronal and glial populations. A role of Fgf18 in formation of appropriate cortical laminar structures is also indicated, however, requires further study. Moreover, studies of various FGFR constructs indicated the overall role of FGF signaling in production of GABAergic interneurons as well as glutamatergic projection neurons in frontal and temporal areas. However, the FGF ligands and FGFR subtype that plays a role in the caudal region of the dorsal cortex, as well as in genesis of interneuron populations is still unclear.

FGF signaling plays a role in axonal projections Both cortical and subcortical cues are known to be important for establishing connections between specific cortical areas and the distinct thalamic nuclei (Lopez-Bendito and Molnar 2003; Vanderhaeghen and Polleux 2004). In newborn Fgf8neo/neo cortex, despite the drastic effects of Fgf8 reduction on area positioning, the projections of thalamocortical axons (TCAs) between the dorsal thalamus and rostral cortex were normal (Garel et al. 2003). Moreover, in young postnatal Fgf17−/− mice, no major abnormalities were present in connectivity within frontal areas, or between frontal area to nucleus accumbens, internal capsule, and medial thalamus (Cholfin and Rubenstein 2007b). This is distinct from the cases of disrupted TCA connections upon deletion of transcription factors important for area specification, such as Emx2 and Coup-Tf1 (LopezBendito and Molnar 2003; Armentano et al. 2007), and it was speculated that the mechanisms controlling the initial topography of thalamocortical connection may rely on subcortical signals induced independently of Fgf8. In contrast, intra-neocortical connections are abnormal in Fgf8neo/neo cortex at E16.5, P0 and P8 (Huffman et al. 2004). Retrograde labeling of cortical axons at rostral and caudal areas has revealed that, while axon growth from rostral-to-caudal and caudal-to-rostral directions rarely cross these regional boundaries in wild-type mice, caudal neurons in Fgf8neo/neo mice aberrantly projected into the rostral cortex. Abnormal expression of area-specific laminar markers, Id2 and RZRβ, was associated with the abnormal projection pattern. A rostral shift of patterning as indicated by transcription factor expression, such as Emx2 and CoupTf1, was suggested as a potential cause for rostral misprojection by the caudal neurons (Huffman et al. 2004). Importantly, Fgf8 is able to specify rostro-caudal regional identities in a layer-specific fashion (Shimogori and Grove 2005). In this study, experiments were carried out to introduce Fgf8 at two different stages of embryonic development (E10.5 and E11.5) in order to direct the molecular shift of VZ progenitor cells, which later become subplate cells or the cortical plate, respectively. The results showed that the thalamic axon projection was directed towards the presumptive area in accordance to the induced molecular shift. All together, Fgf8 indeed regulates guidance cues within the cortex, mainly by patterning-related mechanisms, but the initial subcortical projection of TCA does not require Fgf8/ 17 signaling.



FGF signaling plays a role in the formation of the commissural tracts Fibroblast growth factor signaling plays an important role in the formation of the commissural tracts (Lindwall et al. 2007). In mice that lack cortical Fgfr1, defects were observed in midline crossing of the three forebrain commissural tracts, the corpus callosum (CC), the anterior commissure (AC), and the hippocampal commissure (HC) (Smith et al. 2006; Tole et al. 2006). The failure of commissural axon midline crossing in these models was attributed mainly to defective differentiation of midline glial cell types known as indusium griseum, subcallosal sling, glial wedge, and midline zipper glia, which are thought to act as guideposts for commissural axon guidance (Lindwall et al. 2007). Such guidepost defects could alter local environmental guidance cues in the region of the commissures, even if the commissural axons had no autonomous deficiencies. Notably, projections to other targets, for example to the olfactory cortex, TCA and cortico-cortical projections, seem unaffected by Fgfr1 deficiency (Hebert et al. 2003; Smith et al. 2006; Tole et al. 2006), indicating that commissural pathways are selectively dependent on Fgfr1 activity. Interestingly, the commissural and glial phenotypes differed somewhat between Fgfr1 deficient models, with more severe defects in conditional mutants driven by Foxg1-Cre (Tole et al. 2006) than by hGFAP-Cre, Nestin-Cre, or Syn1-Cre (Smith et al. 2006). In experiments using Foxg1-Cre, all three commissures were defective in Fgfr1 homozygous mutants (Foxg1-Cre; Fgfr1flox/flox), while only the CC and HC were defective in Fgfr1 heterozygous mutants (Foxg1-Cre;Fgfr1+/flox) (Tole et al. 2006). In experiments using hGFAP-Cre or Nestin-Cre, the CC and HC were variably affected while the AC was spared in Fgfr1 homozygous mutants

315

(hGFAP-Cre or Nestin-Cre;Fgfr1flox/flox), but all three commissures developed normally in Fgfr1 heterozygous mutants (Smith et al. 2006). Homozygous or heterozygous deletion of Fgfr1 from neurons using Syn1-Cre had no effect on the commissures. Also, the septum as well as all midline progenitor cell types were defective in Foxg1-Cre; Fgfr1 at E14.5–E16.5 (Tole et al. 2006). In contrast, in hGFAP-Cre;Fgfr1flox/flox, the loss was limited to cells in the indusium griseum (Smith et al. 2006). In retrospect, the results from Foxg1-Cre experiments (Tole et al. 2006) may have been complicated by significant effects of Foxg1 haploinsufficiency on neocortical and hippocampal development (Shen et al. 2006; Eagleson et al. 2007; Siegenthaler et al. 2008). Thus, unintended consequences of using the Foxg1-Cre driver may well explain the different phenotypes between studies (Tole et al. 2006; Smith et al. 2006). Taken together, the above studies indicate that FGF signaling is essential for the development of the HC and CC, due to the critical role of midline glial and perhaps other unidentified guideposts or intermediate targets. These non-autonomous effects on growing commissural axons are sensitive to Fgfr1 deficiency. Any role for FGFs in AC development, or by autonomous mechanisms in neurons, remains to be conclusively demonstrated by the generation and analysis of other in vivo models.

FGF signaling in hippocampus development The hippocampus is a part of the cerebral cortex that lies inside the temporal lobe along its medial boundary. The hippocampus plays an indispensable role in processing and consolidation of long-term memory and in navigation of space and location. As in the cortex, early patterning and neurogenesis are key processes in development of the hippocampus (Fig. 7). Patterning

Fig. 7. Overview of key events during development of the hippocampus. Signaling between the rostral (commissural plate, red) and dorsomedial (cortical hem, blue) signaling centers regulates hippocampus patterning. Next, dorsomedial ventricular zone (VZ) and dentate gyrus (DG, orange) are the two sites where neurogenesis occurs in the hippocampus. Progenitor cells in the hippocampus VZ migrate to become hippocampal pyramidal neurons to form CA1–CA3 (initial CA3 is indicated in green). At around E15.5, a subtraction of the progenitor cells migrates to form prospective DG and differentiates to become granule cells at early postnatal weeks. Interneurons in the hippocampus are originated form the ventral telencephalon and tangentially migrate to settle in the hippocampus (purple). © 2009 The Authors Journal compilation © 2009 Japanese Society of Developmental Biologists

316


of the hippocampus is regulated interactively by BMP and Wnt family molecules secreted from the cortical hem, identified as the dorsomedial signaling center of the telencephalon; and by a suppressive effect from the cortical determinant gene, Lhx2 (Hebert and Fishell 2008; Mangale et al. 2008). The dorsomedial VZ and the dentate gyrus (DG) are the two sites where neurogenesis occurs in the hippocampus. Progenitor cells in the hippocampal VZ migrate to become pyramidal neurons and form CA1–CA3 regions of the hippocampus. At around E15.5, a subpopulation of progenitor cells migrate to form the prospective DG and differentiate to become granule cells during late embryonic and early postnatal weeks. As in the cortex, interneurons in the hippocampus are originated form the ventral telencephalon and tangentially migrate to settle in the hippocampus (Pleasure et al. 2000). It has been shown that Fgf8 in the rostral signaling center is able to influence patterning and growth of the hippocampus by regulating Wnt signaling (Shimogori et al. 2004). Suppression of Wnt-dependent patterning at the cortical hem was demonstrated by rostral overexpression of Fgf8 by in utero electroporation (Shimogori et al. 2004). Increased rostral patterning by Fgf8 overexpression at E11.5 led to a hypoplastic hippocampus at E15.5, similar to Emx2−/− and Wnt−/− mutant phenotypes (Galceran et al. 2000; Lee et al. 2000; Shimogori et al. 2004). In contrast, a reduction of Fgf8 results in a loss of Bmp4 expression and rostrally expanded Wnt8b and Emx2 expression domains at E9–9.5 (Storm et al. 2006). Furthermore, sequestering Fgf8 signaling in Emx2−/− cortex at E9.5 partially rescued Wnt expression and hippocampal phenotype (Shimogori et al. 2004). As the gross morphology of the hippocampus is largely normal in any single and double FGFR knockout mice at least up to E16.5 (Gutin et al. 2006; Tole et al. 2006), regulation of hippocampus patterning may require signaling of all three FGFRs. In contrast, little information is available for FGF ligands that play a role in neurogenesis during hippocampus development. No significant difference was observed in the volume and neuron numbers in Fgf2−/− mice at P1 (Korada et al. 2002). However, a more definitive role of FGF signaling in hippocampal neurogenesis was identified by the study of Fgfr1 function (Ohkubo et al. 2004). Fgfr1 is expressed in 91% of differentiating neurons and 70% of GFAP+ glia. In hGFAP-Cre:Fgfr1flox/flox, the volume of hippocampal CA fields was reduced by 35% at P7.5 and 6 weeks, and that of the DG was reduced by 41% at P7.5 and 21% at 6 weeks. It was shown that Fgfr1 was responsible for genesis of 30–65% of total cells at both the hippocampal VZ and DG by P7.5, by regulating progenitor cell proliferation

at E16.5 and at P0. Progenitor cells born at E16.5 migrating to DG were also decreased by 53%. However, cell survival was not affected by the loss of Fgfr1 at these stages (Ohkubo et al. 2004; Tole et al. 2006). Interestingly, 50–65% reduction of Parvalbumin+ interneurons was observed in adult (Ohkubo et al. 2004) (Table 2). As hippocampal interneurons originate by migration from the ventral telencephalon (Pleasure et al. 2000), where Fgfr1 deletion had not been directed in this model, it was suggested that there may be an indirect mechanism that negatively influenced the interneuron populations within the hippocampus. Functional and behavioral studies of 6–8-week-old Nestin-Cre;Fgfr1ko/flox mice showed an impairment of long-term potentiation (LTP) at medial perforant path (MPP)-granule neurons and memory consolidation, without affecting spatial learning (Zhao et al. 2007). In these mice, the volume of the hippocampus was unchanged; however, proliferation in the DG was reduced by 50% in adults, implicating FGF signaling in adult neurogenesis that takes place in the subgranular zone of the DG (Zhao et al. 2008). Functions of FGFRs other than Fgfr1 in hippocampal development are largely unknown so far. In terms of expression, albeit showing different patterns and intensity, both Fgfr2 and Fgfr3 are expressed in all prospective CA1–3 regions and the DG throughout development (Bansal et al. 2003; Ohkubo et al. 2004). However, there is an evident lack of Fgfr3 expression in the cortical hem at E12.5 (Bansal et al. 2003; Hasegawa et al. 2004). Quantitative trait analysis has previously identified the Hipp loci that may modulate hippocampal weight and neuron number in the DG (Lu et al. 2001). Interestingly, the Fgfr3 gene is one of the candidate genes located within this locus (Pozniak and Pleasure 2006). Furthermore, hippocampal dysplasia is one of the salient characteristics of cortical malformation in human TD, which is caused by gain-of-function kinase-domain mutations in human FGFR3, indicating a potential function of Fgfr3 in hippocampal development in humans (Hevner 2005). Mechanisms underlying the hippocampus neuropathology of TD have not been studied to date. Taken together, the above studies indicate that FGF signaling plays a role in both hippocampal patterning by regulating Bmp/Wnt signals at the cortical hem and hippocampal neurogenesis during development. In addition, in vivo evidence is now provided for the role of FGF signaling in adult DG neurogenesis leading to a functional consequence. However, FGF/ FGFR subtypes that regulate hippocampus development are largely unknown so far and await future studies.



Functional and behavioral consequences of developmental defects caused by altered FGF signaling Several recent reports have demonstrated that the structural changes caused by altered FGF signaling during embryonic development could influence cortical functions and animal behavior. A deletion of Fgf2 causes reduction of glutamatergic excitatory neurons with relatively unaffected GABAergic inhibitory interneurons in rostral cortical regions (Korada et al. 2002). The duration in loss of the lighting reflex (sleeping time) upon treatment with GABA receptor agonist (sodium pentobarbital, PTB) was much longer in Fgf2−/− mice, suggesting that the imbalance of excitatory and inhibitory neuron number does have a behavioral consequence (Korada et al. 2002). Furthermore, defects in patterning of prefrontal cortex subdivisions in Fgf17−/− mice lead to behavioral abnormalities in a series of specific social interaction phenotypes, including isolation-induced ultrasonic vocalization, social recognition, and social interaction in novel environment (Cholfin and Rubenstein 2007a; Scearce-Levie et al. 2008). A loss of proper olfactory bulb morphology in the absence of Fgfr1 is linked to Kallmann syndrome in humans (Kim et al. 2008). One of the major characteristics in Kallmann syndrome is anosmia (diminished olfaction) due to dysgenesis of the olfactory bulb. So far, six genes have been associated with Kallmann syndrome. Among them, three are directly related to FGF signaling, FGF8, KAL2 (FGFR1), and KAL1. KAL1 is identified as anosmin-1, which is likely to regulate the function of FGFR1 through binding to heparan sulfate, which is known to facilitate binding of FGF ligand to FGFRs. Suppression of FGF signaling in mice with a dominant negative Fgfr1 construct (Otx1-tFgfr1) leads to spontaneous and persistent locomotor hyperactivity that develops 2 weeks after weaning (Shin et al. 2004). This is not owing to an alteration in the subcortical monoaminergic systems, such as dopaminergic, noradrenergic and cholinergic systems, since responses to a stimulant, amphetamine, and an α2A adrenergic receptor agonist, guanfacine, were similar to control groups. How altered neurogenesis phenotype observed in the frontal and temporal cortex in these mice leads to this behavioral consequence is currently unknown. Hyperactivity is also observed in mice in which Fgfr1 is singly deleted (hGFAP-Cre;Fgfr1flox/flox) (Muller Smith et al. 2008). Various pharmacological tests have demonstrated that monoaminergic and catecholaminergic signaling are not affected in these mice. Learning and memory is also not impaired, as results of cross motor coordination and learning (rota-rod), spatial learning

317

(fear conditioning and nose poke test) are similar to control groups. Interestingly, the more the reduction in number of Parvalbumin+ interneurons, the more hyperactivity the hGFAP-Cre;Fgfr1flox/flox mice displayed, suggesting a close association of this cytoarchitectural changes to behavioral consequences. Patients with the non-lethal form of TD with Fgfr3 gain-of-function mutation present severe limitations in motor and intellectual development (OMIM#187600) (Bellus et al. 1999, 2000). However, the changes in neural structure, connections and physiology that lead to this functional and behavioral pathology are unknown to date. A recent study of mice with altered expression of Emx2 showed that even a mild change in area size and position could lead to defects in modality-specific behavioral tasks (Leingartner et al. 2007). As a result of being genetically downstream of Emx2 (Muzio et al. 2002; Fukuchi-Shimogori and Grove 2003; Garel et al. 2003), Fgfr3 may play a role in the formation of cortical structures that are important in specific cognitive abilities and behavior. Finally, individuals affected with autism spectrum disorders (ASD) present diverse behavioral symptoms that affect their well-being from early childhood, which can include impairments in social interaction and communication, stereotypic behavioral patterns, intellectual development, seizures, and anxiety (Amaral et al. 2008). Based on the known neuroanatomical characteristics of ASD and the known FGF function in control of cortical size and neurogenesis, it was suggested that a defect in FGF signaling could play a role in neuropathological mechanism that underlie ASD (Vaccarino et al. 2009).

Understanding of FGF signaling mechanisms during cortical development is still in infancy It is far from clear what kind of intracellular signaling pathways are activated downstream of FGFs during development of the cerebral cortex. Nonetheless, a few attempts have been made to address this in vivo, which gives an interesting insight for molecular mechanisms underlying phenotypic changes observed in models with altered FGF signaling (Fig. 3B). Erk activation was observed in regions of known FGF signaling in the frontal forebrain early in embryogenesis (< E10.5) (Corson et al. 2003). Experiments using cortical progenitors in culture indeed showed that Fgf2, Fgf8 and Fgf15 resulted in transient increase of phosphorylated Erk (pErk) by Western blots (Fig. 3B). Glycogen synthase kinase-3 (GSK3) is an important pathway regulated by both MapK and Akt pathways and an inhibitor of Wnt/β-catenin intracellular signal transduction (Patel et al. 2004; Ding et al. 2005).


318


Phosphorylation of Akt (pAkt) as well as that of GSK and p70 S6 kinase/mTor (Patel et al. 2004; Ding et al. 2005) was only observed upon addition of Fgf2/Fgf8, but not Fgf15. Therefore, the differential effects between Fgf2/ Fgf8 and Fgf15 in promoting and inhibiting cell proliferation, respectively, could be due to the additional activation of the Akt-GSK-mTor signaling axis by Fgf2/ Fgf8, but not by Fgf15 (Fig. 3B). Furthermore, a use of pathway-specific reporter mice showed that the retinoic acid signaling pathway was reduced in Fgf15−/− mice consistent with the reduction in neuronal differentiation at E12.5–E14.5 (Borello et al. 2008). In contrast, Wnt/ β-catenin pathway was upregulated in Fgf15−/−, consistent with increased progenitor proliferation in these mice. The level of pErk was also analyzed on tissue sections in Fgf17−/− mice at E10.5; however, no obvious changes were detected (Cholfin and Rubenstein 2008). Studies in mice expressing the gain-of-function Fgfr3 allele demonstrated that downstream activation of MapK, rather than that of Pl3k/Akt signaling, was largely responsible for increased cell proliferation observed in the mutant caudal cortex at E11.5 (Thomson et al. 2007) (Fig. 3B). A basal level of progenitor proliferation did persist even when MAPK signaling was inhibited, which strongly suggests the participation of other pathways such as PLCγ (Eswarakumar et al. 2005; Mason 2007). Various genes have been identified downstream of FGF signaling during cortical development. In Fgf8neo/neo and Fgf17−/− mice, a reduced expression of downstream genes Spry1 and Spry2 is observed in the rostromedial telencephalon at E10.5, and of Pea3-Ets transcription factors, Pea3, Erm, and Er81, at E12.5 (Cholfin and Rubenstein 2008) (see section ‘Multiple FGFs in the rostral signaling center regulate area formation in the dorsal cortex’ above). In contrast, in Fgf15−/− mice, little or no change was observed in Spry2 and Erm expression at E9.5, while Mest was increased at E12.5 and subtle reduction of Erm and Spry2 at E12.5 and 14.5 (Borello et al. 2008). A loss of Pea3-Ets expression in Fgf18−/− in the cortical plate at E13.5–E16.5 (Hasegawa et al. 2004) has been also described earlier in the review (see ‘FGF signaling controls genesis of diverse subpopulations of cortical cells’). Fibroblast growth factors in the rostral signaling center, particularly Fgf8 and Fgf15, but not Fgf17, appear to regulate expressions of other FGFs. In Fgf8neo/neo, Fgf15 expression was reduced in the midbrain /hindbrain boundary, but was induced ectopically in the prosencephalic midline at E9.5 (Borello et al. 2008) and remained so at E10.5 and E12.5 (Cholfin and Rubenstein 2008). Expression of Fgf17 and Fgf18 was clearly reduced in Fgf8neo/neo at E10.5 and E12.5 (Cholfin and

Rubenstein 2008), indicating that these genes are downstream of Fgf8. In contrast, in Fgf15−/− mice, expression of Fgf8 appears to be somewhat increased at E9.5 (Borello et al. 2008). In Fgf17−/− mice, expression of Fgf8, Fgf15, and Fgf18 remained unchanged at E10.5 and E12.5 (Cholfin and Rubenstein 2007b; Cholfin and Rubenstein 2008). Finally, alterations in gene expression downstream of FGF signaling were systematically assessed by microarray approach in Foxg1-Cre;Fgfr1flox/flox cortex at E12.5 (Sansom et al. 2005). The study confirmed the involvement of FGF signaling in regulation of cortical patterning genes (Sansom et al. 2005). A microarray study was also carried out identifying downstream genes of Fgf2 using cortical neuron culture from E14.5 (Pellicano et al. 2006). Taken together, the signaling studies in vivo (and in primary cultures) clearly show that the signaling pathways established in the in vitro work are not always activated in vivo (compare Fig. 3A,B). The downstream signaling pathway is FGF- and context-specific (location and developmental stages). Relative differences in responsiveness of FGFR to each ligand (Fig. 2) may at least partly explain differential effects in downstream signaling pathway and gene expression. Downstream signaling in the presence of multiple and/or single FGFRs needs to be investigated in the in vivo context (see ‘Biochemical properties of FGF signaling’). Technological advances to enable visualization of activated signaling pathway in vivo more efficiently and easily would enhance this area of study.

Conclusions and perspectives It has been indicated that developmental changes in area position and size may drastically influence animal and human behavior (Cholfin and Rubenstein 2008; O’Leary and Sahara 2008). However, there are currently few models in which behavioral alterations can be associated with cellular and molecular mechanisms in early development, and therefore a foundation for development of medical intervention for neurodevelopmental pathology is critically missing. Although models with altered FGF signaling are unlikely to directly and generically represent known psychiatric disorders such as ASD (Abrahams and Geschwind 2008), they may serve as models that allow studies of cellular and molecular mechanisms underlying altered animal behaviors similar to those in human disorders. To this end, a causative role of FGFR3 activation in the cortical malformation in TD may present a unique opportunity whereby its biological significance in cortex and hippocampus development can be directly compared in mouse models and humans.



Understanding of the roles of morphogens and transcription factors in cortical area formation and neurogenesis has advanced drastically in the last two decades (Dehay and Kennedy 2007; O’Leary and Sahara 2008). However, currently little is known regarding their downstream signaling pathways and mechanism mediating the effects. Furthermore, most of the studies have been aimed towards structural development of frontal cortical areas in relation to function in a higher order regulation of cognition and behavior. The caudalmost cerebral cortex contains the temporal lobe, which is important for auditory/visual perception, language function, and memory. Damage to the temporal lobe leads to a disturbance of these essential functions. Temporal lobe seizures can also cause personality changes including those involving aggressive behaviors. Despite the strong indication that FGF signaling controls the caudal cortical structures, no specific FGFs have been found in regulating formation of the caudal cortical areas. Similarly, the role and mechanism of FGF signaling regulating interneuron subpopulations generated in the ventral cortex is far from clear. Investigation of such mechanisms towards better understanding of cortical development is an important contribution to an improvement of human health and well-being.

Acknowledgments We thank Drs Jean Hebert, Makoto Kuro-o, Moosa Mohammadi, Flora Vaccarino, Thomas Theil, David Ornitz, and Soo-Hyun Kim, for providing precious information. This work was supported by Neurosciences Foundation, Glasgow, UK, and the start-up fund from University of Glasgow (T.I.); and by National Institutes of Health grant R01 NS050248 (R.F.H.).

References Abrahams, B. S. & Geschwind, D. H. 2008. Advances in autism genetics: on the threshold of a new neurobiology. Nat. Rev. Genet. 9, 341–355. Allen, B. L. & Rapraeger, A. C. 2003. Spatial and temporal expression of heparan sulfate in mouse development regulates FGF and FGF receptor assembly. J. Cell Biol. 163, 637–648. Amaral, D. G., Schumann, C. M. & Nordahl, C. W. 2008. Neuroanatomy of autism. Trends Neurosci. 31, 137–145. Aoto, K., Nishimura, T., Eto, K. & Motoyama, J. 2002. Mouse GLI3 regulates Fgf8 expression and apoptosis in the developing neural tube, face, and limb bud. Dev Biol. 251, 320 –332. Arman, E., Haffner-Krausz, R., Chen, Y., Heath, J. K. & Lonai, P. 1998. Targeted disruption of fibroblast growth factor (FGF) receptor 2 suggests a role for FGF signaling in pregastrulation mammalian development. Proc. Natl Acad. Sci. USA 95, 5082 –5087. Armentano, M., Chou, S. J., Tomassy, G. S., Leingartner, A., O’Leary, D. D. & Studer, M. 2007. COUP-TFI regulates the

319

balance of cortical patterning between frontal/motor and sensory areas. Nat. Neurosci. 10, 1277–1286. Assimacopoulos, S., Grove, E. A. & Ragsdale, C. W. 2003. Identification of a Pax6-dependent epidermal growth factor family signaling source at the lateral edge of the embryonic cerebral cortex. J. Neurosci. 23, 6399 –6403. Bansal, R., Lakhina, V., Remedios, R. & Tole, S. 2003. Expression of FGF receptors 1, 2, 3 in the embryonic and postnatal mouse brain compared with Pdgfralpha, Olig2 and Plp/dm20: implications for oligodendrocyte development. Dev. Neurosci. 25, 83 –95. Bellot, F., Crumley, G., Kaplow, J. M., Schlessinger, J., Jaye, M. & Dionne, C. A. 1991. Ligand-induced transphosphorylation between different FGF receptors. EMBO J. 10, 2849 – 2854. Bellus, G. A., Bamshad, M. J., Przylepa, K. A., Dorst, J., Lee, R. R., Hurko, O., Jabs, E. W., Curry, C. J., Wilcox, W. R., Lachman, R. S., Rimoin, D. L. & Francomano, C. A. 1999. Severe achondroplasia with developmental delay and acanthosis nigricans (SADDAN): phenotypic analysis of a new skeletal dysplasia caused by a Lys650Met mutation in fibroblast growth factor receptor 3. Am. J. Med. Genet 85, 53 –65. Bellus, G. A., Spector, E. B., Speiser, P. W., Weaver, C. A., Garber, A. T., Bryke, C. R., Israel, J., Rosengren, S. S., Webster, M. K., Donoghue, D. J. & Francomano, C. A. 2000. Distinct missense mutations of the FGFR3 Lys650 codon modulate receptor kinase activation and the severity of the skeletal dysplasia phenotype. Am. J. Hum. Genet. 67, 1411–1421. Bishop, K. M., Goudreau, G. & O’Leary, D. D. 2000. Regulation of area identity in the mammalian neocortex by Emx2 and Pax6. Science 288, 344 –349. Borello, U., Cobos, I., Long, J. E., Murre, C. & Rubenstein, J. L. 2008. FGF15 promotes neurogenesis and opposes FGF8 function during neocortical development. Neural Dev 3, 17. Brader, S. & Eccles, S. A. 2004. Phosphoinositide 3-kinase signalling pathways in tumor progression, invasion and angiogenesis. Tumori 90, 2– 8. Bystron, I., Blakemore, C. & Rakic, P. 2008. Development of the human cerebral cortex: Boulder Committee revisited. Nat. Rev. Neurosci. 9, 110 –122. Calegari, F. & Huttner, W. B. 2003. An inhibition of cyclindependent kinases that lengthens, but does not arrest, neuroepithelial cell cycle induces premature neurogenesis. J. Cell Sci. 116, 4947– 4955. Cavallaro, U. & Christofori, G. 2004. Cell adhesion and signalling by cadherins and Ig-CAMs in cancer. Nat. Rev. Cancer 4, 118 –132. Caviness, V. S. Jr, Takahashi, T. & Nowakowski, R. S. 1995. Numbers, time and neocortical neuronogenesis: a general developmental and evolutionary model. Trends Neurosci. 18, 379 –383. Chaffer, C. L., Dopheide, B., Savagner, P., Thompson, E. W. & Williams, E. D. 2007. Aberrant fibroblast growth factor receptor signaling in bladder and other cancers. Differentiation 75, 831– 842. Chen, K., Ohkubo, Y., Shin, D., Doetschman, T., Sanford, L. P., Li, H. & Vaccarino, F. M. 2008. Decrease in excitatory neurons, astrocytes and proliferating progenitors in the cerebral cortex of mice lacking exon 3 from the Fgf2 gene. BMC Neurosci. 9, 94. Cholfin, J. A. & Rubenstein, J. L. 2007a. Genetic regulation of prefrontal cortex development and function. Novartis Found. Symp. 288, 165–173; Discussion 173–167, 276 –181.


320


Cholfin, J. A. & Rubenstein, J. L. 2007b. Patterning of frontal cortex subdivisions by Fgf 17. Proc. Natl Acad. Sci. USA 104, 7652–7657. Cholfin, J. A. & Rubenstein, J. L. 2008. Frontal cortex subdivision patterning is coordinately regulated by Fgf8, Fgf17, and Emx2. J. Comp. Neurol. 509, 144 –155. Colvin, J. S., Bohne, B. A., Harding, G. W., McEwen, D. G. & Ornitz, D. M. 1996. Skeletal overgrowth and deafness in mice lacking fibroblast growth factor receptor 3. Nat. Genet. 12, 390 –397. Cooper, J. A. 2008. A mechanism for inside-out lamination in the neocortex. Trends Neurosci. 31, 113–119. Corbin, J. G., Nery, S. & Fishell, G. 2001. Telencephalic cells take a tangent: non-radial migration in the mammalian forebrain. Nat. Neurosci. 4(Suppl. ), 1177–1182. Corson, L. B., Yamanaka, Y., Lai, K. M. & Rossant, J. 2003. Spatial and temporal patterns of ERK signaling during mouse embryogenesis. Development 130, 4527– 4537. Crossley, P. H. & Martin, G. R. 1995. The mouse Fgf8 gene encodes a family of polypeptides and is expressed in regions that direct outgrowth and patterning in the developing embryo. Development 121, 439– 451. Crossley, P. H., Martinez, S., Ohkubo, Y. & Rubenstein, J. L. 2001. Coordinate expression of Fgf8, Otx2, Bmp4, and Shh in the rostral prosencephalon during development of the telencephalic and optic vesicles. Neuroscience 108, 183–206. Dehay, C. & Kennedy, H. 2007. Cell-cycle control and cortical development. Nat. Rev. Neurosci. 8, 438 – 450. Deng, C., Wynshaw-Boris, A., Zhou, F., Kuo, A. & Leder, P. 1996. Fibroblast growth factor receptor 3 is a negative regulator of bone growth. Cell 84, 911–921. Deng, C., Bedford, M., Li, C., Xu, X., Yang, X., Dunmore, J. & Leder, P. 1997. Fibroblast growth factor receptor-1 (FGFR-1) is essential for normal neural tube and limb development. Dev Biol. 185, 42–54. Ding, Q., Xia, W., Liu, J. C., Yang, J. Y., Lee, D. F., Xia, J., Bartholomeusz, G., Li, Y., Pan, Y., Li, Z., Bargou, R. C., Qin, J., Lai, C. C., Tsai, F. J., Tsai, C. H. & Hung, M. C. 2005. Erk associates with and primes GSK-3beta for its inactivation resulting in upregulation of beta-catenin. Mol. Cell 19, 159 – 170. Dominguez, M. H. & Rakic, P. 2008. Neuroanatomy of the FGF system. J. Comp. Neurol. 509, 141–143. Dono, R. 2003. Fibroblast growth factors as regulators of central nervous system development and function. Am. J. Physiol. Regul. Integr. Comp. Physiol. 284, R867–R881. Dono, R., Texido, G., Dussel, R., Ehmke, H. & Zeller, R. 1998. Impaired cerebral cortex development and blood pressure regulation in FGF-2-deficient mice. EMBO J. 17, 4213–4225. Eagleson, K. L., Schlueter McFadyen-Ketchum, L. J., Ahrens, E. T., Mills, P. H., Does, M. D., Nickols, J. & Levitt, P. 2007. Disruption of Foxg1 expression by knock-in of ere recombinase: effects on the development of the mouse telencephalon. Neuroscience 148, 385–399. Ekerot, M., Stavridis, M. P., Delavaine, L., Mitchell, M. P., Staples, C., Owens, D. M., Keenan, I. D., Dickinson, R. J., Storey, K. G. & Keyse, S. M. 2008. Negative-feedback regulation of FGF signalling by DUSP6/MKP-3 is driven by ERK1/2 and mediated by Ets factor binding to a conserved site within the DUSP6/MKP-3 gene promoter. Biochem. J. 412, 287– 298. Eswarakumar, V. P., Lax, I. & Schlessinger, J. 2005. Cellular signaling by fibroblast growth factor receptors. Cytokine Growth Factor Rev. 16, 139 –149.

Faedo, A., Tomassy, G. S., Ruan, Y., Teichmann, H., Krauss, S., Pleasure, S. J., Tsai, S. Y., Tsai, M. J., Studer, M. & Rubenstein, J. L. 2008. COUP-TFI coordinates cortical patterning, neurogenesis, and laminar fate and modulates MAPK/ERK, AKT, and beta-catenin signaling. Cereb Cortex 18, 2117–2131. Ford-Perriss, M., Abud, H. & Murphy, M. 2001. Fibroblast growth factors in the developing central nervous system. Clin. Exp. Pharmacol. Physiol. 28, 493 –503. Ford-Perriss, M., Guimond, S. E., Greferath, U., Kita, M., Grobe, K., Habuchi, H., Kimata, K., Esko, J. D., Murphy, M. & Turnbull, J. E. 2002. Variant heparan sulfates synthesized in developing mouse brain differentially regulate FGF signaling. Glycobiology 12, 721–727. Francis, F., Meyer, G., Fallet-Bianco, C., Moreno, S., Kappeler, C., Socorro, A. C., Tuy, F. P., Beldjord, C. & Chelly, J. 2006. Human disorders of cortical development: from past to present. Eur. J. Neurosci. 23, 877– 893. Fukuchi-Shimogori, T. & Grove, E. A. 2001. Neocortex patterning by the secreted signaling molecule FGF8. Science 294, 1071– 1074. Fukuchi-Shimogori, T. & Grove, E. A. 2003. Emx2 patterns the neocortex by regulating FGF positional signaling. Nat. Neurosci. 6, 825 –831. Galceran, J., Miyashita-Lin, E. M., Devaney, E., Rubenstein, J. L. & Grosschedl, R. 2000. Hippocampus development and generation of dentate gyrus granule cells is regulated by LEF1. Development 127, 469 –482. Garel, S., Huffman, K. J. & Rubenstein, J. L. 2003. Molecular regionalization of the neocortex is disrupted in Fgf8 hypomorphic mutants. Development 130, 1903 –1914. Gimeno, L., Brulet, P. & Martinez, S. 2003. Study of Fgf15 gene expression in developing mouse brain. Gene Expr Patterns 3, 473 – 481. Gotz, M. & Huttner, W. B. 2005. The cell biology of neurogenesis. Nat. Rev. Mol. Cell Biol. 6, 777–788. Grove, E. A. & Fukuchi-Shimogori, T. 2003. Generating the cerebral cortical area map. Annu. Rev. Neurosci. 26, 355–380. Guerrini, R., Dobyns, W. B. & Barkovich, A. J. 2008. Abnormal development of the human cerebral cortex: genetics, functional consequences and treatment options. Trends Neurosci. 31, 154 –162. Gupta, A., Tsai, L. H. & Wynshaw-Boris, A. 2002. Life is a journey: a genetic look at neocortical development. Nat. Rev. Genet. 3, 342–355. Gutin, G., Fernandes, M., Palazzolo, L., Paek, H., Yu, K., Ornitz, D. M., McConnell, S. K. & Hebert, J. M. 2006. FGF signalling generates ventral telencephalic cells independently of SHH. Development 133, 2937–2946. Hamasaki, T., Leingartner, A., Ringstedt, T. & O’Leary, D. D. 2004. EMX2 regulates sizes and positioning of the primary sensory and motor areas in neocortex by direct specification of cortical progenitors. Neuron 43, 359 –372. Hasegawa, H., Ashigaki, S., Takamatsu, M., Suzuki-Migishima, R., Ohbayashi, N., Itoh, N., Takada, S. & Tanabe, Y. 2004. Laminar patterning in the developing neocortex by temporally coordinated fibroblast growth factor signaling. J. Neurosci. 24, 8711– 8719. Hebert, J. M. & Fishell, G. 2008. The genetics of early telencephalon patterning: some assembly required. Nat. Rev. Neurosci. 9, 678 –685. Hebert, J. M., Lin, M., Partanen, J., Rossant, J. & McConnell, S. K. 2003. FGF signaling through FGFR1 is required for olfactory bulb morphogenesis. Development 130, 1101–1111.



Hevner, R. F. 2005. The cerebral cortex malformation in thanatophoric dysplasia: neuropathology and pathogenesis. Acta Neuropathol. (Berl) 110, 208 –221. Hevner, R. F. 2007. Layer-specific markers as probes for neuron type identity in human neocortex and malformations of cortical development. J. Neuropathol. Exp. Neurol. 66, 101– 109. Huffman, K. J., Garel, S. & Rubenstein, J. L. 2004. Fgf8 regulates the development of intra-neocortical projections. J. Neurosci. 24, 8917– 8923. Inglis-Broadgate, S. L., Thomson, R. E., Pellicano, F., Tartaglia, M. A., Pontikis, C. C., Cooper, J. D. & Iwata, T. 2005. FGFR3 regulates brain size by controlling progenitor cell proliferation and apoptosis during embryonic development. Dev Biol. 279, 73 – 85. Itoh, N. & Ornitz, D. M. 2004. Evolution of the Fgf and Fgfr gene families. Trends Genet. 20, 563–569. Iwata, T., Chen, L., Li, C., Ovchinnikov, D. A., Behringer, R. R., Francomano, C. A. & Deng, C. X. 2000. A neonatal lethal mutation in FGFR3 uncouples proliferation and differentiation of growth plate chondrocytes in embryos. Hum. Mol. Genet. 9, 1603 –1613. Iwata, T., Li, C. L., Deng, C. X. & Francomano, C. A. 2001. Highly activated Fgfr3 with the K644M mutation causes prolonged survival in severe dwarf mice. Hum. Mol. Genet. 10, 1255–1264. Jeffrey, K. L., Camps, M., Rommel, C. & Mackay, C. R. 2007. Targeting dual-specificity phosphatases: manipulating MAP kinase signaling and immune responses. Nat. Rev. Drug Discov. 6, 391– 403. Kim, S. H., Hu, Y., Cadman, S. & Bouloux, P. 2008. Diversity in fibroblast growth factor receptor 1 regulation: learning from the investigation of Kallmann syndrome. J. Neuroendocrinol. 20, 141–163. Korada, S., Zheng, W., Basilico, C., Schwartz, M. L. & Vaccarino, F. M. 2002. Fibroblast growth factor 2 is necessary for the growth of glutamate projection neurons in the anterior neocortex. J. Neurosci. 22, 863 –875. Kowalczyk, T., Pontious, A., Englund, C., Daza, R. A. M., Bedogni, F., Hodge, R., Attardo, A., Bell, C., Huttner, W. B. & Hevner, R. F. 2009. Intermediate neuronal progenitors (basal progenitors) produce pyramidal-projection neurons for all layers of cerebral cortex. Cereb Cortex Jan 2009 [Epub ahead of print]. Kuro-o, M. 2008. Endocrine FGFs and Klothos: emerging concepts. Trends Endocrinol. Metab. 19, 239 –245. Kuschel, S., Ruther, U. & Theil, T. 2003. A disrupted balance between Bmp/Wnt and Fgf signaling underlies the ventralization of the Gli3 mutant telencephalon. Dev. Biol. 260, 484 – 495. Lee, S. M., Tole, S., Grove, E. & McMahon, A. P. 2000. A local Wnt-3a signal is required for development of the mammalian hippocampus. Development 127, 457– 467. Leingartner, A., Thuret, S., Kroll, T. T., Chou, S. J., Leasure, J. L., Gage, F. H. & O’Leary, D. D. 2007. Cortical area size dictates performance at modality-specific behaviors. Proc. Natl Acad. Sci. USA 104, 4153 – 4158. Lindwall, C., Fothergill, T. & Richards, L. J. 2007. Commissure formation in the mammalian forebrain. Curropin Neurobiol. 17, 3 –14. Lopez-Bendito, G. & Molnar, Z. 2003. Thalamocortical development: how are we going to get there? Nat. Rev. Neurosci. 4, 276 – 289. Lu, L., Airey, D. C. & Williams, R. W. 2001. Complex trait analysis of the hippocampus: mapping and biometric analysis of two novel gene loci with specific effects on hippocampal structure in mice. J. Neurosci. 21, 3503 –3514.

321

Lukaszewicz, A., Savatier, P., Cortay, V., Kennedy, H. & Dehay, C. 2002. Contrasting effects of basic fibroblast growth factor and neurotrophin 3 on cell cycle kinetics of mouse cortical stem cells. J. Neurosci. 22, 6610 –6622. Mallamaci, A. & Stoykova, A. 2006. Gene networks controlling early cerebral cortex arealization. Eur. J. Neurosci. 23, 847– 856. Mallamaci, A., Muzio, L., Chan, C. H., Parnavelas, J. & Boncinelli, E. 2000. Area identity shifts in the early cerebral cortex of Emx2−/− mutant mice. Nat. Neurosci. 3, 679 –686. Mangale, V. S., Hirokawa, K. E., Satyaki, P. R., Gokulchandran, N., Chikbire, S., Subramanian, L., Shetty, A. S., Martynoga, B., Paul, J., Mai, M. V., Li, Y., Flanagan, L. A., Tole, S. & Monuki, E. S. 2008. Lhx2 selector activity specifies cortical identity and suppresses hippocampal organizer fate. Science 319, 304 – 309. Marie, D., Fiorio Pla, A., Chang, Y. H. & Barker, J. L. 2007. Selfrenewing and differentiating properties of cortical neural stem cells are selectively regulated by basic fibroblast growth factor (FGF) signaling via specific FGF receptors. J. Neurosci. 27, 1836 –1852. Marin, O. & Rubenstein, J. L. 2001. A long, remarkable journey: tangential migration in the telencephalon. Nat. Rev. Neurosci. 2, 780 –790. Martynoga, B., Morrison, H., Price, D. J. & Mason, J. O. 2005. Foxg1 is required for specification of ventral telencephalon and region-specific regulation of dorsal telencephalic precursor proliferation and apoptosis. Dev Biol. 283, 113 –127. Mason, I. 2007. Initiation to end point: the multiple roles of fibroblast growth factors in neural development. Nat. Rev. Neurosci. 8, 583 –596. Mason, J. M., Morrison, D. J., Basson, M. A. & Licht, J. D. 2006. Sprouty proteins: multifaceted negative-feedback regulators of receptor tyrosine kinase signaling. Trends Cell Biol. 16, 45 – 54. Miyashita-Lin, E. M., Hevner, R., Wassarman, K. M., Martinez, S. & Rubenstein, J. L. R. 1999. Early neocortical regionalization in the absence of thalamic innervation. Science 285, 906 – 909. Molyneaux, B. J., Arlotta, P., Menezes, J. R. & Macklis, J. D. 2007. Neuronal subtype specification in the cerebral cortex. Nat. Rev. Neurosci. 8, 427– 437. Monsonego-Ornan, E., Adar, R., Rom, E. & Yayon, A. 2002. FGF receptors ubiquitylation: dependence on tyrosine kinase activity and role in downregulation. FEBS Lett. 528, 83–89. Muller Smith, K., Fagel, D. M., Stevens, H. E., Rabenstein, R. L., Maragnoli, M. E., Ohkubo, Y., Picciotto, M. R., Schwartz, M. L. & Vaccarino, F. M. 2008. Deficiency in inhibitory cortical interneurons associates with hyperactivity in fibroblast growth factor receptor 1 mutant mice. Biol. Psychiatry 63, 953 –962. Muzio, L., DiBenedetto, B., Stoykova, A., Boncinelli, E., Gruss, P. & Mallamaci, A. 2002. Emx2 and Pax6 control regionalization of the pre-neuronogenic cortical primordium. Cereb Cortex 12, 129 –139. Naski, M. C., Wang, Q., Xu, J. & Ornitz, D. M. 1996. Graded activation of fibroblast growth factor receptor 3 by mutations causing achondroplasia and thanatophoric dysplasia. Nat. Genet. 13, 233 –237. Nomura, T., Haba, H. & Osumi, N. 2007. Role of a transcription factor Pax6 in the developing vertebrate olfactory system. Dev. Growth. Differ. 49, 683 –690. Nurcombe, V., Ford, M. D., Wildschut, J. A. & Bartlett, P. F. 1993. Developmental regulation of neural response to FGF-1 and FGF-2 by heparan sulfate proteoglycan. Science 260, 103 – 106.


322


Oh, L. Y., Denninger, A., Colvin, J. S., Vyas, A., Tole, S., Ornitz, D. M. & Bansal, R. 2003. Fibroblast growth factor receptor 3 signaling regulates the onset of oligodendrocyte terminal differentiation. J. Neurosci. 23, 883 –894. Ohkubo, Y., Chiang, C. & Rubenstein, J. L. 2002. Coordinate regulation and synergistic actions of BMP4, SHH and FGF8 in the rostral prosencephalon regulate morphogenesis of the telencephalic and optic vesicles. Neuroscience 111, 1–17. Ohkubo, Y., Uchida, A. O., Shin, D., Partanen, J. & Vaccarino, F. M. 2004. Fibroblast growth factor receptor 1 is required for the proliferation of hippocampal progenitor cells and for hippocampal growth in mouse. J. Neurosci. 24, 6057– 6069. O’Leary, D. D. & Sahara, S. 2008. Genetic regulation of arealization of the neocortex. Curr. Opin. Neurobiol. 18, 90 –100. O’Leary, D. D., Chou, S. J. & Sahara, S. 2007. Area patterning of the mammalian cortex. Neuron 56, 252–269. Ornitz, D. M., Herr, A. B., Nilsson, M., Westman, J., Svahn, C. M. & Waksman, G. 1995. FGF binding and FGF receptor activation by synthetic heparan-derived di- and trisaccharides. Science 268, 432–436. Ornitz, D. M., Xu, J., Colvin, J. S., McEwen, D. G., MacArthur, C. A., Coulier, F., Gao, G. & Goldfarb, M. 1996. Receptor specificity of the fibroblast growth factor family. J. Biol. Chem. 271, 15292–15297. Patel, S., Doble, B. & Woodgett, J. R. 2004. Glycogen synthase kinase-3 in insulin and Wnt signalling: a double-edged sword? Biochem. Soc. Trans. 32, 803–808. Pellicano, F., Inglis-Broadgate, S. L., Pante, G., Ansorge, W. & Iwata, T. 2006. Expression of coiled-coil protein 1, a novel gene downstream of FGF2, in the developing brain. Gene Expr Patterns 6, 285 –293. Peters, K., Ornitz, D., Werner, S. & Williams, L. 1993. Unique expression pattern of the FGF receptor 3 gene during mouse organogenesis. Dev Biol. 155, 423 – 430. Pleasure, S. J., Anderson, S., Hevner, R., Bagri, A., Marín, O., Lowenstein, D. H. & Rubenstein, J. L. 2000. Cell migration from the ganglionic eminences is required for the development of hippocampal GABAergic interneurons. Neuron 28, 727–740. Pontious, A., Kowalczyk, T., Englund, C. & Hevner, R. F. 2008. Role of intermediate progenitor cells in cerebral cortex development. Dev. Neurosci. 30, 24 –32. Pozniak, C. D. & Pleasure, S. J. 2006. Genetic control of hippocampal neurogenesis. Genome. Biol. 7, 207. Pringle, N. P., Yu, W. P., Howell, M., Colvin, J. S., Ornitz, D. M. & Richardson, W. D. 2003. Fgfr3 expression by astrocytes and their precursors: evidence that astrocytes and oligodendrocytes originate in distinct neuroepithelial domains. Development 130, 93 –102. Raballo, R., Rhee, J., Lyn-Cook, R., Leckman, J. F., Schwartz, M. L. & Vaccarino, F. M. 2000. Basic fibroblast growth factor (Fgf2) is necessary for cell proliferation and neurogenesis in the developing cerebral cortex. J. Neurosci. 20, 5012–5023. Rakic, P. 2006. A century of progress in corticoneurogenesis: from silver impregnation to genetic engineering. Cereb Cortex 16 (Suppl 1), i3–17. Rash, B. G. & Grove, E. A. 2006. Area and layer patterning in the developing cerebral cortex. Curr. Opin. Neurobiol. 16, 25 –34. Rash, B. G. & Grove, E. A. 2007. Patterning the dorsal telencephalon: a role for sonic hedgehog? J. Neurosci. 27, 11595– 11603. Reuss, B. & von Bohlen und Halbach, O. 2003. Fibroblast growth factors and their receptors in the central nervous system. Cell Tissue Res. 313, 139–157. Sahara, S., Kawakami, Y., Izpisua Belmonte, J. C. & O’Leary, D. D.

2007. Sp8 exhibits reciprocal induction with Fgf8 but has an opposing effect on anterior-posterior cortical area patterning. Neural Dev 2, 10. Sanchez-Heras, E., Howell, F. V., Williams, G. & Doherty, P. 2006. The fibroblast growth factor receptor acid box is essential for interactions with N-cadherin and all of the major isoforms of neural cell adhesion molecule. J. Biol. Chem. 281, 35208 – 35216. Sansom, S. N., Hebert, J. M., Thammongkol, U., Smith, J., Nisbet, G., Surani, M. A., McConnell, S. K. & Livesey, F. J. 2005. Genomic characterisation of a Fgf-regulated gradient-based neocortical protomap. Development 132, 3947–3961. Scearce-Levie, K., Roberson, E. D., Gerstein, H., Cholfin, J. A., Mandiyan, V. S., Shah, N. M., Rubenstein, J. L. & Mucke, L. 2008. Abnormal social behaviors in mice lacking Fgf17. Genes. Brain. Behav. 7, 344 –354. Shen, L., Nam, H. S., Song, P., Moore, H. & Anderson, S. A. 2006. FoxG1 haploinsufficiency results in impaired neurogenesis in the postnatal hippocampus and contextual memory deficits. Hippocampus 16, 875 – 890. Shimamura, K. & Rubenstein, J. L. 1997. Inductive interactions direct early regionalization of the mouse forebrain. Development 124, 2709 –2718. Shimogori, T. & Grove, E. A. 2005. Fibroblast growth factor 8 regulates neocortical guidance of area-specific thalamic innervation. J. Neurosci. 25, 6550 –6560. Shimogori, T., Banuchi, V., Ng, H. Y., Strauss, J. B. & Grove, E. A. 2004. Embryonic signaling centers expressing BMP, WNT and FGF proteins interact to pattern the cerebral cortex. Development 131, 5639 –5647. Shin, D. M., Korada, S., Raballo, R., Shashikant, C. S., Simeone, A., Taylor, J. R. & Vaccarino, F. 2004. Loss of glutamatergic pyramidal neurons in frontal and temporal cortex resulting from attenuation of FGFR1 signaling is associated with spontaneous hyperactivity in mice. J. Neurosci. 24, 2247– 2258. Siegenthaler, J. A., Tremper-Wells, B. A. & Miller, M. W. 2008. Foxg1 haploinsufficiency reduces the population of cortical intermediate progenitor cells: effect of increased p21 expression. Cereb. Cortex 18, 1865 –1875. Smith, K. M., Ohkubo, Y., Maragnoli, M. E., Rasin, M. R., Schwartz, M. L., Sestan, N. & Vaccarino, F. M. 2006. Midline radial glia translocation and corpus callosum formation require FGF signaling. Nat. Neurosci. 9, 787–797. Storm, E. E., Rubenstein, J. L. & Martin, G. R. 2003. Dosage of Fgf8 determines whether cell survival is positively or negatively regulated in the developing forebrain. Proc. Natl Acad. Sci. USA 100, 1757–1762. Storm, E. E., Garel, S., Borello, U., Hebert, J. M., Martinez, S., McConnell, S. K., Martin, G. R. & Rubenstein, J. L. 2006. Dose-dependent functions of Fgf8 in regulating telencephalic patterning centers. Development 133, 1831–1844. Sur, M. & Rubenstein, J. L. 2005. Patterning and plasticity of the cerebral cortex. Science 310, 805 – 810. Takahashi, T., Nowakowski, R. S. & Caviness, V. S. Jr. 1995. The cell cycle of the pseudostratified ventricular epithelium of the embryonic murine cerebral wall. J. Neurosci. 15, 6046 –6057. Theil, T., Dominguez-Frutos, E. & Schimmang, T. 2008. Differential requirements for Fgf3 and Fgf8 during mouse forebrain development. Dev. Dyn. 237, 3417–3423. Thomson, R. E., Pellicano, F. & Iwata, T. 2007. Fibroblast growth factor receptor 3 kinase domain mutation increases cortical progenitor proliferation via mitogen-activated protein kinase activation. J. Neurochem. 100, 1565 –1578.



Thomson, R. E., Kind, P. C., Graham, N. A., Etherson, M. L., Kennedy, J., Fernandes, A. C., Marques, C. S., Hevner, R. F. & Iwata, T. 2009. Fgf receptor 3 activation promotes selective growth and expansion of occipitotemporal cortex. Neural Dev 4, 4. Timmer, M., Cesnulevicius, K., Winkler, C., Kolb, J., LipokaticTakacs, E., Jungnickel, J. & Grothe, C. 2007. Fibroblast growth factor (FGF)-2 and FGF receptor 3 are required for the development of the substantia nigra, and FGF-2 plays a crucial role for the rescue of dopaminergic neurons after 6-hydroxydopamine lesion. J. Neurosci. 27, 459 – 471. Tole, S., Gutin, G., Bhatnagar, L., Remedios, R. & Hebert, J. M. 2006. Development of midline cell types and commissural axon tracts requires Fgfr1 in the cerebrum. Dev Biol. 289, 141–151. Vaccarino, F. M., Schwartz, M. L., Raballo, R., Nilsen, J., Rhee, J., Zhou, M., Doetschman, T., Coffin, J. D., Wyland, J. J. & Hung, Y. T. 1999. Changes in cerebral cortex size are governed by fibroblast growth factor during embryogenesis. Nat. Neurosci. 2, 246 –253. Vaccarino, F. M., Grigorenko, E. L., Smith, K. M. & Stevens, H. E. 2009. Regulation of cerebral cortical size and neuron number by fibroblast growth factors: implications for autism. J. Autism Dev. Disord. 39, 511–520. Vanderhaeghen, P. & Polleux, F. 2004. Developmental mechanisms patterning thalamocortical projections: intrinsic, extrinsic and in between. Trends Neurosci. 27, 384 –391. Walshe, J. & Mason, I. 2003. Unique and combinatorial functions of Fgf3 and Fgf8 during zebrafish forebrain development. Development 130, 4337– 4349. Wang, Q., Green, R. P., Zhao, G. & Ornitz, D. M. 2001. Differential regulation of endochondral bone growth and joint development by FGFR1 and FGFR3 tyrosine kinase domains. Development 128, 3867–3876. Webster, M. K. & Donoghue, D. J. 1996. Constitutive activation of fibroblast growth factor receptor 3 by the transmembrane domain point mutation found in achondroplasia. EMBO J. 15, 520–527.

323

Wong-Riley, M. 1979. Changes in the visual system of monocularly sutured or enucleated cats demonstrable with cytochrome oxidase histochemistry. Brain Res. 171, 11–28. Wright, T. J., Ladher, R., McWhirter, J., Murre, C., Schoenwolf, G. C. & Mansour, S. L. 2004. Mouse FGF15 is the ortholog of human and chick FGF19, but is not uniquely required for otic induction. Dev Biol. 269, 264 –275. Xu, X., Weinstein, M., Li, C., Naski, M., Cohen, R. I., Ornitz, D. M., Leder, P. & Deng, C. 1998. Fibroblast growth factor receptor 2 (FGFR2)-mediated reciprocal regulation loop between FGF8 and FGF10 is essential for limb induction. Development 125, 753 –765. Yamaguchi, T. P., Harpal, K., Henkemeyer, M. & Rossant, J. 1994. fgfr-1 is required for embryonic growth and mesodermal patterning during mouse gastrulation. Genes Dev. 8, 3032– 3044. Yu, T., Fotaki, V., Mason, J. O. & Price, D. J. 2009. Analysis of early ventral telencephalic defects in mice lacking functional Gli3 protein. J. Comp. Neurol. 512, 613–627. Zembrzycki, A., Griesel, G., Stoykova, A. & Mansouri, A. 2007. Genetic interplay between the transcription factors Sp8 and Emx2 in the patterning of the forebrain. Neural Dev 2, 8. Zhang, X., Ibrahimi, O. A., Olsen, S. K., Umemori, H., Mohammadi, M. & Ornitz, D. M. 2006. Receptor specificity of the fibroblast growth factor family. The complete mammalian FGF family. J. Biol. Chem. 281, 15694 –15700. Zhao, C., Deng, W. & Gage, F. H. 2008. Mechanisms and functional implications of adult neurogenesis. Cell 132, 645 – 660. Zhao, M., Li, D., Shimazu, K., Zhou, Y. X., Lu, B. & Deng, C. X. 2007. Fibroblast growth factor receptor-1 is required for longterm potentiation, memory consolidation, and neurogenesis. Biol. Psychiatry 62, 381–390. Zhou, C., Tsai, S. Y. & Tsai, M. J. 2001. COUP-TFI: an intrinsic factor for early regionalization of the neocortex. Genes Dev. 15, 2054 –2059.
