Date Published: January 26, 2012
Publisher: Hindawi Publishing Corporation
Author(s): A. E. Evans, C. M. Kelly, S. V. Precious, A. E. Rosser.
The central nervous system is composed of the brain and the spinal cord. The brain is a complex organ that processes and coordinates activities of the body in bilaterian, higher-order animals. The development of the brain mirrors its complex function as it requires intricate genetic signalling at specific times, and deviations from this can lead to brain malformations such as anencephaly. Research into how the CNS is specified and patterned has been studied extensively in chick, fish, frog, and mice, but findings from the latter will be emphasised here as higher-order mammals show most similarity to the human brain. Specifically, we will focus on the embryonic development of an important forebrain structure, the striatum (also known as the dorsal striatum or neostriatum). Over the past decade, research on striatal development in mice has led to an influx of new information about the genes involved, but the precise orchestration between the genes, signalling molecules, and transcription factors remains unanswered. We aim to summarise what is known to date about the tightly controlled network of interacting genes that control striatal development. This paper will discuss early telencephalon patterning and dorsal ventral patterning with specific reference to the genes involved in striatal development.
The striatum plays a vital role in the coordination of movement (primary motor control), emotions, and cognition [1–3]. In humans, the striatum is divided into two nuclei, the caudate and the putamen, by the internal capsule, whereas in mice it is one structure. This is shown in Figure 1. The complexity and importance of the striatum is best highlighted when it is impaired. There are a number of diseases that may produce striatal damage, including acquired conditions such as a stroke and genetically inherited conditions such as Huntington’s disease (HD). HD is a condition that is characterised by neuronal dysfunction and neuronal loss that principally affects the medium spiny neurons (MSNs) of the striatum. MSNs are the major projection neuron and constitute the vast majority of neurons in this structure. HD results in progressive deterioration of movement and cognition and, in many cases, additional behavioural deficits over a period of 15–30 years, and eventually renders an individual unable to care for themselves. To date there is little in the way of symptomatic treatment and no disease-modifying agents available. A better understanding of striatal development is likely to accelerate our understanding of the pathogenic processes underlying conditions such as HD and is central to the development of protocols to engineer stem cells to be suitable as donor tissue for cell replacement therapy [3–5].
Development of the nervous system starts with neural induction, followed by neurulation that gives rise to the neural tube, and finally, patterning of this tube along the anterior-posterior (AP) axis. Following AP patterning, the neural tube folds and is subdivided into the prosencephalon (forebrain), the most anterior (rostral) part of the neural tube, which consists of the telencephalon and diencephalon, the mesencephalon (midbrain), and the rhombencephalon (hindbrain) . These major subdivisions are shown in Figure 2. Regional patterning of the putative brain regions is then controlled by a series of interacting gene networks, of which the ones controlling telencephalic development are the most complex.
The embryonic telencephalon, which is located at the most rostral end of the neural tube, is divided into the dorsal telencephalon (also called pallium), which gives rise to the neocortex, and the ventral telencephalon (also called the subpallium), which forms the striatum and is the origin of cells that populate the olfactory bulb, globus pallidus (GP), and some cells that also populate the cortex . This paper will concentrate on the development of the ventral telencephalon. Although the adult striatum is different between all mammalian species, the initial subdivisions observed in the telencephalon are comparable [8, 9].
The telencephalon is the most complex region of the mammalian brain and shows vast heterogeneity in terms of its various neuronal populations, structures, and function. Telencephalon development requires a variety of signals secreted from surrounding signalling centres to ensure the correct positional identity of the neurons, which will populate the adult forebrain. Several gene families are involved in coordinating the initial events for telencephalon patterning: principally fibroblast growth factors (FGFs), bone morphogenic proteins (Bmps), Wnts (originated from the drosophila gene wingless), retinoic acid (RA), and sonic hedgehog (Shh). These genes are responsible for activating downstream factors that enable signalling cascades to be initiated that allow cells to gain a positional and molecular identity . It is likely that only a proportion of the factors required for neuronal identity have so far been identified, and the precise way in which such factors interact to specify the timing and terminal differentiation of particular neuronal subpopulations is not yet defined. However, there have been clear advances in knowledge in this area over the last few decades and we summarise here what is known about some of the key factors so far identified as being involved in striatal development.
FGFs are ligands that activate several pathways, for example, Ras Map kinase (MAPK) pathway upon binding FGF receptors (FGFR) 1, 2, or 3  and at least 5 FGF proteins have been indentified in CNS development . Fgf8 is expressed rostrally from the anterior neural ridge (ANR) in mammals and has roles in proliferation and cell survival. In addition, it has been shown that Fgf8 regulates the expression of forkhead box protein G1 (FOXG1) (previously bf1), a rostral forebrain marker. It was shown that FGF8 could renew Foxg1 expression in mouse explants that had the ANR removed and secondly that inhibitors of Fgf8 reduced Foxg1 expression in neural plate explants [20, 21]. In addition, reduction in Fgf8 leads to rostral truncations and midline defects in the developing forebrain . In Fgf8 null mice (Fgf8−/−), the telencephalon was smaller than in wild-type (WT) littermates and exhibited patterning abnormalities [22–24]. Specifically, loss of function studies showed that the MGE and LGE are absent, and there was loss of genes found in the ventral regions, for example, Nkx2.1 and Dlx2 and an expansion of the dorsal marker Pax6 . These results suggest a role for Fgf8 in ventralisation of the telencephalon. However, it seems unlikely that Fgf8 is the sole factor involved in the induction of rostral forebrain development due to the continued presence of the telencephalon in Fgf8−/−or FGFR null mutants . One explanation for the telencephalon still remaining in these mutants is that compensation is achieved by other Fgfs expressed at the same time. However, others in the field are of the opinion that overlapping Fgf expression profiles do not exist and that each Fgf has exclusive roles in telencephalon development [25, 26]. Therefore, the reason the telencephalon is not lost completely in the Fgf8 mutant is that Fgf8 alone is not essential for telencephelon generation [26, 27]. However, the fact that beads soaked in FGF8 added to anterior neural explants lacking an ANR promote expression of Foxg1 suggests FGFs are necessary for telencephalon induction .
SHH is a member of the hedgehog (Hh) family of secreted proteins and acts as a morphogen that is first secreted from the notochord, which underlies the posterior structures of the brain, following which expression is from the overlying neural plate [6, 27, 28]. By E9.5 Shh is expressed in neural epithelium of the ventral telencephelon , and by E12 it is expressed in the mantle zone and is no longer detectable in the neuroepithelium . Shh operates through a concentration gradient that spans the DV axis at different time points to confer different neuronal identities on the developing precursors, and expression is first seen in the ventral telencenphelon from E11.5 [31, 32]. SHH expression directs neural progenitors to a ventral fate and importantly is both necessary and sufficient to induce specific ventral forebrain markers [32–34]. Shh is expressed in the ventral telencephalon and is thought to maintain Fgf8 expression [32, 35] as well as induce expression of forebrain markers. Specifically, SHH activates several TFs including Nkx2.1 [36, 37], Gsx2 (formerly Gsh2) [38–40], and Pax6 .
RA is the biologically active form of vitamin A and has been implicated in survival, specification, proliferation, and differentiation during forebrain development [58–60]. Two oxidation events occur to ensure RA is successfully derived to function as a ligand for either RA receptors (RARs) (RARα, RARβ and RARγ) or retinoid X receptors (RXRα, RXRβ, and RXRγ) that belong to the steroid/thyroid receptor superfamily . Initially retinol dehydrogenases oxidate retinol to retinaldehyde and then the rate-limiting enzymes, retinaldehyde dehydrogenases (Raldh), are required to oxidate retinaldehyde to RA .
Wnts belong to the wingless protein family and are a class of ligands that are crucial in embryogenesis and have been implicated in CNS development. Wnts can signal through three different pathways: the canonical pathway, the planar cell polarity pathway, and the calcium pathway and it is the canonical pathway that is important in telencephalon development. In the canonical pathway, β-catenin is indirectly activated by a WNT ligand binding to the cell surface receptor, Frizzled. Upon binding, frizzled activates its intracellular component dishevelled (Dsh) that dephosphorylates β-catenin preventing its degradation by the axin-glycogen synthase kinas 3β (GSK3β) complex (in the absence of WNT signalling, β-catenin is phosphorylated by the GSK3β complex and degraded). β-catenin then translocates to the nucleus where it can activate the transcription of Wnt target genes such as T-cell factors (TCF), which in turn regulate genes such as c-myc. This pathway is shown in Figure 5.
Although BMP inhibition is required for neuronal development, graded concentrations are necessary in the neural plate to establish medial-to-lateral patterning . BMPs belong to the TGFβ family of secreted proteins. It is thought that BMPs are also needed to dorsalize the telencephalon and restrict ventral telencephalic development. Forebrain patterning was repressed in forebrain explant cultures when BMPs were added as shown by inhibition of Foxg1, Nkx2.1, and Dlx2 . Similarly beads soaked in BMP4 or BMP5 that were implanted into the neural tube of a chick forebrain induced dorsal markers, for example, Wnt4 and repressed ventral markers . Additionally, when the telencephalic roof plate (a source of BMPs) was ablated, there was a reduction in cortical size and a decrease of one of the most dorsal cortical markers, Lhx2 . BMPs are inhibited by several factors including chordin and noggin. In mice that lacked both copies of the chordin gene (chordin−/−) and one copy of the noggin gene (noggin+/−), a dorsal, rather than ventral telencephalon was evident. However, this effect may not be direct because of an increase in BMP and may be in part due to the decreased levels of Shh and Fgf8 expression in the forebrain caused by increased BMP levels .
Foxg1 is a member of the winged helix family of TFs and is the earliest recognised marker of the telencephalon . This TF was first identified in the rat brain, where it was shown that its expression was restricted to the telencephelon . By E8.5 Foxg1 is expressed in the neural tube, specifically in the anterior plate cells that are fated to contribute to the telencephalon [20, 79], and functions to establish and subdivide the telencephalon.
As discussed already, the developing telencephalon is divided into the dorsal region, and the ventral region and these areas can be defined on a morphological and genetic basis. In the dorsal telencephalon Pax6, Neurogenin (Ngn) 1/2 and Emx1/2 are expressed; in the ventral telencephalon, Gsx2, Asc1, Dlx1/2, and Nkx2.1 are expressed, as shown in Figure 6 . The pallial genes will only be discussed briefly in the context of them as markers, not to fully elucidate their role in cortical development. Emx1/2 expression profiles are restricted to the most dorsal region of the cortex with no expression seen in the ventral cortical region. Ngn 1 and 2 are basic helix loop helix (bHLH) TFs which are expressed throughout the cortex together with Pax6. In the absence of Ngn expression, Ascl1 is ectopically expressed in the dorsal telencephalon thus priming these cells to adopt a ventral fate and becoming GABAergic rather than glutamatergic neurons. Therefore, the role of Ngn1/2 is to maintain the DV boundary in the developing telencephalon and to inhibit ventral gene expression such as Ascl1 [24, 85].
This paper has aimed to summarise and organise research that is being carried out to understand the genetic mechanisms controlling striatal development. However, it is clear that there are still many pieces of the jigsaw to be found and fitted into the gaps of this puzzle. The more that is known about the development of the striatum, and, importantly, the development and differentiation of striatal MSN neurons, the more precise the protocols can be to direct the fate of renewable cell sources, such as embryonic stem cells, to a functional MSN phenotype for use in cell replacement therapy for HD. An additional aspiration is to identify specific genes to detect MSN precursors rather than relying on markers of terminally differentiated MSNs, such as DARPP-32. Earlier markers of putative MSNs could be used to facilitate the generation and refinement of neuronal differentiation protocols as well as tracking neuronal differentiation in grafts.