atonal


EVOLUTIONARY HOMOLOGS (Part 2/2)

Mammalian bHLH family members: Distinct domains within the bHLH motif have different activities

Many members of the basic helix-loop-helix (bHLH) family of transcription factors play pivotal roles in the development of a variety of tissues and organisms. Activities have been identified for the neural bHLH proteins Mash1 and Math1 in inducing neuronal differentiation, and in inducing the formation of distinct dorsal interneuron subtypes in the chick neural tube. Although both factors induce neuronal differentiation, each factor has a distinct activity in the type of dorsal interneuron that forms: overexpression of Math1 increases dI1 interneurons; Mash1 increases dI3 interneurons. Math1 and Mash1 function as transcriptional activators for both of these functions. Furthermore, discrete domains within the bHLH motif have been defined that are required for these different activities in neural development. Helix 1 of the Mash1 HLH domain is necessary for Mash1 to be able to promote neuronal differentiation, and is sufficient to confer this activity to the non-neural bHLH factor MyoD. In contrast, helix 2 of Math1, and both helix 1 and 2 of Mash1, are the domains required for the neuronal specification activities of these factors. The requirement for distinct domains within the HLH motif of Mash1 and Math1 for driving neuronal differentiation and cell-type specification probably reflects the importance of unique protein-protein interactions involved in these functions (Nakada, 2004).

Mash1 and Math1 are class II bHLH factors that characteristically have tissue-specific expression, and bind E-box DNA (CANNTG) as heterodimers with class I bHLH factors such as E12, E47, HEB and E2-2. Crystal structures of a non-neural class II bHLH factor MyoD, or a class I bHLH factor E47, demonstrate that the basic region interacts with DNA and the HLH forms an amphipathic helix that is involved in protein-protein interactions in the formation of homo- or heterodimers. Studies with other bHLH factors have demonstrated the importance of the basic region for specific functions. Studies of the Xenopus neural bHLH factors Xash1 (Mash1 homolog) and Xngnr1 (Ngn homolog) have identified the HLH domain as the region encoding information for induction of specific downstream targets. The current findings are similar to these latter experiments, demonstrating that the HLH domain, and not the basic region, encodes the necessary information for the specific functions in neurogenesis of the neural bHLH factors Mash1 and Math1 in the chick neural tube (Nakada, 2004).

Since the HLH domain functions in protein-protein interactions, it is reasonable to propose that specificity of function may be conferred by interactions with specific cofactors that vary between the different bHLH proteins. It is known that in vitro, DNA binding activity is most efficient with heterodimers of the neural bHLH with an E-protein such as E12. There are no reports that the individual E-proteins (E2a, HEB, E2-2) can confer specificity of function on the neural bHLH/E-protein heterodimer, and knockout studies with the E-protein genes suggest they have a high level of functional redundancy. Thus, it may be that context-dependent co-factors that form higher order complexes with bHLH heterodimers are important for specific functions. Further support for the importance of cellular context for bHLH factor function is provided by the chick electroporation experiments. For example, electroporation of pMiWIII-Math1 results in overexpression of Math1 along the extent of the dorsoventral axis, however, the increase in the dI1 interneurons is biased to the dorsal regions. This is in contrast to the neuronal differentiation phenotype that is seen throughout the dorsoventral axis. Invoking context-dependent protein-protein interactions to explain the specificity of bHLH function along the dorsoventral axis of the spinal neural tube is also helpful in explaining how bHLH factors are required for different types of neurons in different regions of the nervous system. Consistent with the involvement of specific co-factor interactions is the presence of conserved surface amino acids within each neural bHLH sub-family that are distinct between the different sub-families (Nakada, 2004).

Besides the E-proteins, a number of other co-factors have been described that form complexes with Class II bHLH factors. In Drosophila, Chip, a LIM homeodomain binding protein (homolog of Ldb factors in vertebrates), has been shown to form a complex with Achaete (Mash1 homolog) and Daughterless (E12 homolog), possibly as an adapter protein bringing in another transcription factor, Pannier, to form a higher order transcriptional complex. In studies of Tal1, a class II bHLH factor important in hematopoiesis, a LIM only factor (LMO) and Ldb factors were seen to form a complex with Tal1 and E12. In the neural tube, it has been shown that higher order complex formation of homeodomain factors and bHLH factors control ventral cell fates. In this case, the homeodomain factor Lhx3 alone is not sufficient to generate motor neurons, but in combination with Ldb it specifies V2 interneurons. When co-expressed with islet1, a higher order complex is formed, resulting in different DNA binding characteristics, and the specification of motor neurons rather than V2 interneurons. Furthermore, two bHLH factors, Ngn2 and NeuroM, transcriptionally synergize with the homeodomain complex to specify the motor neurons. This type of combinatorial interactions of transcription factors is an attractive hypothesis for the context-dependent functions seen with the bHLH factors studied here for dorsal neural tube development. The identity and role of co-factors in forming higher order transcriptional complexes with Mash1 and Math1 is yet to be determined (Nakada, 2004).

Mammalian Atonal homologs

To determine the extent to which Drosophila atonal and its mouse homolog Math1 exhibit functional conservation, beta-galactosidase (lacZ) was inserted into the Math1 locus and its expression was analyzed. The consequences of loss of Math1 function were evaluated, and Math1 was expressed in atonal mutant flies. lacZ under the control of Math1 regulatory elements duplicates the known expression pattern of Math1 in the CNS (i.e., the neural tube, dorsal spinal cord, brainstem, and cerebellar external granule neurons) but also reveals new sites of expression: PNS mechanoreceptors (inner ear hair cells and Merkel cells) and articular chondrocytes. Expressing Math1 induces ectopic chordotonal organs (CHOs) in wild-type flies and partially rescues CHO loss in atonal mutant embryos. These data demonstrate that both the mouse and fly homologs encode lineage identity information and, more interestingly, that some of the cells dependent on this information serve similar mechanoreceptor functions (Ben-Arie, 2000).

Seven ato homologs have been cloned and analyzed in the mouse: Mouse Atonal Homologs (MATH) 1, 2, 3, 4A (also known as Ngn2), 4B (Ngn3), 4C (Ngn1), and 5. Most are expressed during neurogenesis in both the CNS and PNS. These homologs vary in the degree of their sequence conservation, and may be divided into three groups. The most distantly related group, the neurogenins, includes Ngn 1, 2 and 3. These gene products share, on average, 53% identity in the bHLH domain with Ato. They are expressed largely in mitotic CNS and sensory ganglia progenitor cells. Recent work suggests that these genes may play a role in neuroblast determination, and may therefore be true proneural genes. The second group includes MATH2 and MATH3, which share 57% identity in the bHLH domain with Ato. These proteins have been postulated to function in postmitotic neural cells. Math2 expression is confined to the CNS, while Math3 is expressed in both the CNS and the trigeminal and dorsal root ganglia. The third group includes MATH1 and MATH5, arguably the only true ato homologs by amino acid sequence criteria, sharing 67% and 71% identity with the bHLH domain of Ato, respectively. It is noteworthy that both genes encode a basic domain identical to that of Ato. Interestingly, the basic domain of Ato has been shown to be sufficient, in the context of another proneural protein (Scute), to substitute for the loss of ato function. Math1 was initially shown to be expressed in the precursors of the cerebellar EGL and in the dorsal spinal cord. Math5 is expressed in the dividing progenitors in the developing retina and in the vagal ganglion (Ben-Arie, 2000 and references therein).

With the exception of Math5 expression in the neural retina, these observations pose a paradox: none of the vertebrate homologs appears to be expressed in peripheral organs or tissues similar to those where ato is expressed. ato is expressed in the CNS. In addition to the inner proliferation center of the optic lobe, ato is expressed in a small anteriomedial patch of cells in each brain lobe. However, because it remains unclear precisely what role ato plays in Drosophila CNS development, it has been difficult to argue that ato and its vertebrate homologs display functional conservation. The experiments reported here reveal sites of previously uncharacterized Math1 expression. As expected, Math1/lacZ expression in the CNS corresponds to that of Math1, but Math1 is also expressed in the skin, the joints, and the inner ear, in striking parallel to ato expression in the fly. Moreover, the expression in the ear (sensory epithelium) and the skin (Merkel cells) is restricted to sensory structures whose function is to convert mechanical stimuli into neuronal electrochemical signals. It is important to point out that in Drosophila, ato appears to play two roles simultaneously. It is required not only to select the precursors of the CHOs (proneural role), but also to specify these precursors as CHO precursors (lineage identity role). The specificity of Math1 expression in the periphery makes it tempting to speculate that it, too, may endow specific cells with very specific lineage identities to distinguish them functionally from other sensory structures. The ability of Math1 to induce ectopic CHO formation and to restore CHOs to ato mutant embryos supports the notion that Math1, and particularly its basic domain, encodes lineage identity information not unlike that encoded by ato. This suggests that the mammalian cells expressing Math1, at least in the ear and the skin, are functionally similar and perhaps evolutionarily related to Drosophila cells that require ato. Furthermore, Math5 expression in the neural retina suggests that the functions of atonal in the fly are carried out by two genes in the mouse: the development of some mechanoreceptors is under the control of Math1 and retinal development is possibly under the control of Math5. It is interesting to note that in the fully sequenced nematode C. elegans, only one homolog of atonal, lin-32, has been identified. Mutants with the u282 allele are touch-insensitive, which strengthens the argument for evolutionary conservation of atonal function in mechanoreception (Ben-Arie, 2000 and references therein).

The pattern of Math1/lacZ expression in the pontine nuclei led to a careful evaluation of this region in null mutants. Although defects in the pons of Math1 null mice had not previous been detected, closer analysis revealed the lack of pontine nuclei at this site. These neurons derive from the rhombic lip as do the EGL neurons, which are also lacking in Math1 null mice. While it is possible to draw parallels between Math1 and ato expression in the skin and ear, it is not clear that such is the case for the joints. ato expression in the fly joints is required for the formation of leg CHOs. In contrast, Math1 is expressed in resting and articular chondrocytes that do not have any described neural function, and for which no parallels exist in the fly. It may be that Math1 expression in cartilage indicates a novel role for a mechanosensory gene, or it may simply reflect similarities in the molecular events underlying the development of the various Math1-expressing cell types. Alternatively, CHOs may also function as joint structural elements in the fly, or articular cartilage may have a mechanoreceptive or transducive capacity yet to be described. There is no evidence at this point to support one or another of these possibilities. It will be of interest to generate mice with a joint-specific Math1 mutation (Ben-Arie, 2000 and references therein).

The relationship between atonal (ato) in the fruit fly and its mouse homolog, Math1, has been analyzed. In flies, ato acts as a proneural gene that governs the development of chordotonal organs (CHOs), which serve as stretch receptors in the body wall and joints and as auditory organs in the antennae. In the fly CNS, ato is important not for specification but for axonal arborization. Math1, in contrast, is required for the specification of cells in both the CNS and the PNS. Furthermore, Math1 serves a role in the development of secretory lineage cells in the gut, a function that does not parallel any known role played by ato. ato and Math1 might be more functionally homologous than they appear. To examine this, Math1 was expressed in ato mutant flies and ato in Math1 null mice. Surprisingly, the two proteins are functionally interchangeable (Wang, 2002).

Ubiquitous expression of Math1 in flies under the control of a heat-shock promoter can induce the formation of chordotonal organs (CHOs) in the absence of ato. Can Math1 specify ectopic CHOs de novo or convert precursor cells that had already been selected by other proneural proteins such as Scute? The GAL4-UAS system was used to express Math1 in ato mutant embryos under the control of the ato embryonic enhancer. This approach permits control of Math1 expression in ectodermal cells that normally become sensory organ precursors of CHOs in stage 10/11 embryos. In wild-type flies, only three of the five sensory-organ precursors that give rise to lateral CHOs express ato; the other two CHOs are recruited from the surrounding ectodermal cells through activation of the epidermal growth factor (EGF) signaling pathway. Expression of one copy of Math1 restores three of the five lateral CHOs; two copies of Math1 result in the specification of five, and occasionally six, CHOs. The degree of Math1 rescue is the same as that achieved by expressing ato in these mutant flies under the same driver. Math1 thus fully compensates for the loss of ato in the fly; it allows both precursor cell selection and CHO specification. Math1, like ato, is able to activate EGF receptor signaling in the fly (Wang, 2002).

Math1 rescues the other known aspects, such as lack of photoreceptors, of the ato null phenotype as well. The markers senseless and alpha-chaoptin were examined in third-instar eye antennal disks of ato larva expressing MATH1. Math1 also induces ectopic photoreceptors in wild-type flies. In the fly CNS, ato is known to be important for axonal arborization of the dorsal-cluster (DC) neurons. Math1 overexpression induces the same axonal aborization in these two neuronal clusters as does ato overexpression. Importantly, no evidence was found that Math1 induces expression of the ato-related gene amos, which suggests that Math1 is directly regulating ato target genes (Wang, 2002).

To ascertain which functions of Math1 can be carried out by ato in the mouse, the Math1 coding region was replaced with ato. Targeted embryonic-stem (ES) cells were used to generate chimeras and heterozygous animals. Trans-heterozygote animals were generated carrying Math1ato and Math1lacZ (a null allele of Math1 in which the coding region is replaced by ß-galactosidase) by intercrossing the heterozygous animals (Wang, 2002).

Math1 null animals fail to initiate respiration and die very shortly after of birth, but Math1ato/lacZ animals carrying a single allele of ato survived to adulthood and appeared normal by observation and by histological analysis. Subtle developmental abnormalities were sought in tissues in which Math1 is known to play an important role in cell differentiation by comparing Math1ato/lacZ and Math1 null embryos (Wang, 2002).

At E18.5, the latest time-point at which they can be evaluated, Math1 null mice lack two rhombic lip derivatives: the external granular layer (EGL), which contains the granule neuron precursors, and the pontine nucleus. Both the EGL and pontine nuclei are normal in Math1ato/lacZ mice at this stage of development. The developing spinal cord was examined. Math1 governs the development of the D1 interneurons. The D1 interneuron precursors migrate ventrally from the dorsal spinal cord to settle in the deep dorsal horn, where they differentiate into neurons that give rise to some of the spinocerebellar tracts. In Math1 null mice the D1 neurons fail to migrate; instead, they accumulate in the dorsal aspect of the spinal cord. However, normal migration of these precursors can be seen in both Math1lacZ/+ and Math1ato/lacZ mice. Moreover, unlike the null animals, the D1 interneuron precursors in Math1ato/lacZ mice express several markers, such as Lh2A (Wang, 2002).

Math1 is essential for the formation of inner-ear hair cells. The inner-ear hair cells of the cochlea and balance organs appear normal by histological criteria in the Math1ato/lacZ animals; there are three rows of outer hair cells and a single row of inner hair cells in the cochlea. These hair cells express appropriate differentiation markers such as calretinin. To ascertain whether the Math1ato/lacZ mice might suffer functional deficits despite their apparently normal histology, hearing was tested in both Math1ato/lacZ and wild-type mice. In auditory brainstem response assays, the Math1ato/lacZ mice display a physiological response similar to that of control animals, indicating that their hearing is not impaired. The hair cells are thus both histologically normal and functional in Math1ato/lacZ mice (Wang, 2002).

To uncover any subtle deficits in coordination, the Math1ato/lacZ mice were tested on the rotating-rod apparatus. There were no significant differences between the performances of Math1ato/lacZ mice and their wild-type littermates. These data strongly suggest that the balance pathways in the Math1ato/lacZ mice are not only histologically normal but fully functional as well (Wang, 2002).

Outside of the hearing and proprioceptive pathways, ato rescues the null phenotype associated with the loss of Math1 in the developing gut. Math1 is essential for the differentiation and development of the secretory-cell lineage (enteroendocrine cells, goblet cells, and Paneth cells). Normal gut development in Math1ato/lacZ mice was evidenced by expression of markers such as chromogranin A that are normally expressed in enteroendocrine cells. Goblet cells were morphologically normal as well (Wang, 2002).

It is striking that ato and Math1 show such overlap in function despite their limited sequence similarity. There is only 68% identity in the 57 amino acid bHLH domain; the 12 amino acids that form the basic domain are identical between ato and Math1. There is no similarity, however, between the approximately 300 remaining amino acids of the two proteins. It appears that Math1's 'new' functions in the mouse (gut) seem to arise from expression in a new environment rather than the acquisition of a new domain. There is no other pair of orthologs beside ato and Math1 that are known to show full functional conservation between mouse and Drosophila. The data argue that Math1's new functions in the mouse were acquired through the evolution of cis-regulatory factors that result in its expression in new tissues. Interestingly, the two ato homologs in mice, Math1 and Math5, together recapitulate the expression pattern of ato in Drosophila. Further studies are necessary to determine whether these two paralogs are functionally equivalent as well (Wang, 2002).

math5 is a murine ortholog of atonal, a bHLH proneural gene essential for the formation of photoreceptors and chordotonal organs in Drosophila. The expression of math5 coincides with the onset of retinal ganglion cell (RGC) differentiation. Targeted deletion of math5 blocks the initial differentiation of 80% of RGCs and results in an increase in differentiated amacrine cells. Furthermore, the absence of math5 abolishes the retinal expression of brn-3b and the formation of virtually all brn-3b-expressing RGCs. These results imply that math5 is a proneural gene essential for RGC differentiation and that math5 acts upstream to activate brn-3b-dependent differentiation processes in RGCs (Wang, 2001).

To assess the potential loss of RGCs in math5-null retinas, analysis was performed of retinal sections prepared from postnatal mice at 3 wk of age, a time when all retinal neurons are fully developed. The overall laminar structure of the homozygous math5-deficient retinas resembles those of wild-type and heterozygous retinas, and no noticeable changes in the number of photoreceptor cells in the outer nuclear layer (ONL) are observed. However, the mutant retinas are ~15%-20% thinner. This difference is mostly caused by the loss of >40% of the cells in GCL and the absence of a defined nerve fiber layer (NFL), which consists entirely of bundles of RGC axons. In mice, RGCs and displaced amacrine cells each constitute about one-half of the cells in GCL. Therefore, the 40% or more loss of cells in GCL could account for a >80% loss of RGCs. Quantification of the number of axon bundles in retinas immunostained with SMI-32, a mouse monoclonal antibody that reacts with neurofilament H and labels predominantly the axons of large ganglion cells, showed an ~80% reduction of SMI-32-positive processes in math5-null retinas relative to wild-type controls. The remaining 20% of the SMI-32-positive processes appear to form axon bundles, project toward the optic disk region, and form a very thin optic nerve and optic chiasm. However, many of these axon bundles are abnormally curved, suggesting a defect in axon pathfinding during development (Wang, 2001).

The Class IV POU domain-containing brn-3b gene is required for the terminal differentiation of RGCs and null mutations in brn-3b result in axon growth defects and programmed cell death in ~70% of newly formed RGCs. The loss of RGCs observed in the math5-deficient mice resembles that of brn-3b-null mice. To test whether math5 and brn-3b are involved in the differentiation of the same population of RGCs, mice were generated with compound mutations in math5 (math5-GFP) and brn-3b (brn-3b-lacZ). Nuclear expression of lacZ in brn-3b-lacZ mice serves to effectively mark brn-3b-positive RGCs. Compared with normal retinas (brn-3b-lacZ/+, math5+/+), math5-null retinas (brn-3b-lacZ/+, math5-GFP/math5-GFP) show a loss of ~97% of brn-3b-lacZ-expressing cells, thus demonstrating that math5 is required for the differentiation of most brn-3b-expressing RGCs. The remaining RGCs in math5-deficient retina appear to be expressing brn-3b but are generated by a math5-independent pathway (Wang, 2001).

The absence of 80% of RGCs in mature math5-deficient retinas could be the result of an insufficient number of RGC progenitors or a failure of RGC differentiation or survival. Because a similar number of math5-lacZ-expressing cells are detected in heterozygous and homologous math5-mutant retinas between E11 and E15.5, it is unlikely that there are an insufficient number of progenitor cells. TUNEL analysis also reveals no overt difference in apoptosis in retinas at E12.5-E19 in math5-deficient and wild-type embryos. A reasonable hypothesis is that math5-deficient retinas are unable to form RGCs. To test whether math5-null retinal progenitor cells fail to differentiate into RGCs, the presence of newly differentiated RGCs was examined by detecting the expression of the early RGC markers, brn-3b and p75, the latter of which encodes a receptor for nerve growth factor (NGF) and neurotrophins. Section in situ hybridization with a brn-3b antisense probe and expression of the brn-3b-lacZ allele at E13.5 shows that brn-3b is highly expressed in wild-type retinas but is virtually absent in homozygous math5-null retinas. Similarly, immunostaining of E13.5 retinas with anti-p75 showed that the expression of p75 in math5-deficient retina is severely reduced. The absence of most RGCs in mature math5-deficient retinas, and the lack of brn-3b and p75 expression in math5-deficient embryonic retinas, argues strongly that in math5-null retinas, retinal progenitor cells are unable to differentiate into RGCs. Because brn-3b expression is essentially absent in math5-null retinas, the results also demonstrate that math5 is genetically upstream of brn-3b (Wang, 2001).

For mammalian cochlear hair cells, fate determination is normally completed by birth. Overexpression of Math1, a mouse homolog of the Drosophila gene atonal, in postnatal rat cochlear explant cultures results in extra hair cells. Surprisingly, the source of the ectopic hair cells is columnar epithelial cells located outside the sensory epithelium in the greater epithelial ridge, which normally give rise to inner sulcus cells. Moreover, Math1 expression also facilitates conversion of postnatal utricular supporting cells into hair cells. Thus Math1 was sufficient for the production of hair cells in the ear, and immature postnatal mammalian inner ears retain the competence to generate new hair cells (Zheng, 2000a).

To determine whether overexpression of Math1 in the inner ear would facilitate hair cell differentiation, cultured cochlear explants prepared from postnatal rats were transfected using a Math1-expressing plasmid. To identify the transfected cells, a plasmid co-expressing Math1 and the enhanced green fluorescent protein (EGFP) genes (pRK5-Math1-EGFP plasmid) were constructed using two separate CMV promoters, one driving the Math1 gene and the other driving the EGFP reporter gene. Overexpression of Math1 in the greater epithelial ridge (GER) cells in postnatal cochlear explant cultures leads to robust production of extra hair cells. In the presence of Math1, the GER cells not only express a hair cell-specific marker, but also acquire immature stereociliary bundles. The transfected GER cells change morphologically and acquire a hair cell fate. Additional experiments have demonstrated that overexpression of Math1 in supporting cells in the utricle, a vestibular organ, also facilitates their conversion into hair cells. These data provide 'gain-of-function' evidence that Math1 controls hair cell differentiation. These findings also suggest that enhancing Math1 expression levels by specific molecules or using various viral gene transfer technologies in the inner ear might be used to stimulate regeneration of inner ear hair cells, which could eventually benefit those suffering from hearing and balance disorders (Zheng, 2000a).

Hair cell fate determination in the inner ear has been shown to be controlled by specific genes. Recent loss-of-function and gain-of-function experiments have demonstrated that Math1, a mouse homolog of the Drosophila gene atonal, is essential for the production of hair cells. To identify genes that may interact with Math1 and inhibit hair cell differentiation, a focus was placed on Hes1, a mammalian hairy and enhancer of split homolog, which is a negative regulator of neurogenesis. Targeted deletion of Hes1 leads to formation of supernumerary hair cells in the cochlea and utricle of the inner ear. RT-PCR analysis shows that Hes1 is expressed in the inner ear during hair cell differentiation and its expression is maintained in adulthood. In situ hybridization with late embryonic inner ear tissue reveals that Hes1 is expressed in supporting cells, but not hair cells, of the vestibular sensory epithelium. In the cochlea, Hes1 is selectively expressed in the greater epithelial ridge and lesser epithelial ridge regions that are adjacent to inner and outer hair cells. Co-transfection experiments in postnatal rat explant cultures show that overexpression of Hes1 prevents hair cell differentiation induced by Math1. Therefore Hes1 can negatively regulate hair cell differentiation by antagonizing Math1. These results suggest that a balance between Math1 and negative regulators such as Hes1 is crucial for the production of an appropriate number of inner ear hair cells (Zheng, 2000b).

ath3, a vertebrate basic helix-loop-helix gene homologous to the Drosophila proneural gene atonal, can directly convert non-neural cells into neurons with anterior features. In the mouse, ath3 expression initially occurs widely in the developing nervous system and then gradually becomes restricted to the neural retina. The genomic organization and promoter activity of mouse ath3 (Math3) has now been characterized. The Math3 gene consists of two exons separated by an 8-kilobase intron; the whole protein-coding region is located in the second exon. Transcription starts at two sites, which are 75 nucleotides apart from each other, and there is no typical TATA box in the upstream region of either start site. Transient transfection analysis shows that the 5'-region of Math3 can direct efficient expression in neuroblastoma cells but not in glioma or fibroblast cells. Deletion studies reveal that the proximal 193-base pair region, which contains the downstream transcription initiation site but not the upstream site, is essential for the Math3 promoter activity and can direct efficient expression in neuroblastoma cells. In contrast, retrovirus-mediated promoter analysis demonstrates that a region further upstream is additionally necessary for retinal expression. These results indicate that the Math3 promoter contains two essential regulatory regions: the proximal 193-base pair region, which confers efficient neural-specific expression, and a region further upstream, required for retinal expression (Tsuda 1998).

MATH4A, a novel bHLH protein related to Atonal, is broadly expressed in neural precursor cells in the mouse embryonic CNS and PNS. Interaction assays demonstrate that MATH4A interacts effectively with both MASH1 and the ubiquitous bHLH protein E12 (Drosophila homolog: Daughterless). MATH4A-E12 heterodimers, but not MATH4A-MASH1, bind to a consensus E-box sequence. Math4A expression is restricted to undifferentiated neural precursors and is complementary to that of Mash1 in most regions of the nervous system. In particular, Math4A is transcribed at high levels in the cerebral cortex, dorsal thalamus, and epibranchial placodes, which present little or no Mash1 expression. However, expression of the two genes shows limited overlap in certain CNS regions (retina, preoptic area of the hypothalamus, midbrain, and hindbrain). Its structure and expression pattern suggest that MATH4A may regulate an early step in neural development, either as a partner of ubiquitous bHLH proteins or associated with other neural-specific bHLH proteins (Gradwohl, 1996).

MATH1 is a neural-specific basic helix-loop-helix transcription factor. Members of this family of transcription factors are involved in the development of specific subsets of neurons in the developing vertebrate nervous system. The cells expressing MATH1 have been examined with respect to their proliferative state and co-expression of cell-type-specific differentiation markers. The MATH1 protein is localized to the nucleus of cells in the dorsal neural tube and the external germinal layer (EGL) of the developing cerebellum. Using double-label immunofluorescence, it has been demonstrated that MATH1-expressing cells span both the proliferating and the differentiating zones within the dorsal neural tube, however, within the EGL of the cerebellum, these cells are restricted to the proliferating zone. The early differentiating MATH1-expressing cells in the dorsal neural tube co-express TAG-1 (a transiently expressed axonal surface glycoprotein that is found specifically on early motor and commissural neurons in the developing neural tube), DCC-1 (a homolog of Drosophila Frazzled) and LH2 (the ortholog of Drosophila Apterous); all three are markers of dorsal commissural interneurons. In addition, transgenic mice with lacZ under the transcriptional control of MATH1-flanking DNA sequences express beta-galactosidase specifically in the developing nervous system, in a manner that mimics subsets of the MATH1-expression pattern, including the dorsal spinal neural tube. Expression of the MATH1/lacZ transgene persists in differentiated dorsal commissural interneurons. Taken together, MATH1 expression is demonstrated in a differentiating population of neuronal precursors in the dorsal neural tube that appear to give rise specifically to dorsal commissural interneurons (Helms, 1998).

NeuroM, is a neural-specific helix-loop-helix transcription factor related to the Drosophila proneural gene atonal. The NeuroM protein most closely resembles the vertebrate NeuroD and Nex1/MATH2 factors, and is capable of transactivating an E-box promoter in vivo. NeuroM to NeuroD expression have been compared in the developing nervous system. In spinal cord and optic tectum, NeuroM expression precedes that of NeuroD. It is transient and restricted to cells lining the ventricular zone that have ceased proliferating but have not yet begun to migrate into the outer layers. In the retina, NeuroM is also transiently expressed in cells as they withdraw from the mitotic cycle, but persists in horizontal and bipolar neurons until full differentiation, assuming an expression pattern exactly complementary to NeuroD. In the peripheral nervous system, NeuroM expression closely follows cell proliferation, suggesting that it intervenes at a similar developmental juncture in all parts of the nervous system. It is proposed that availability of NeuroM defines a new stage in neurogenesis, at the transition between undifferentiated, premigratory and differentiating, migratory neural precursors (Roztocil, 1997).

While many neuronal differentiation genes have been identified, little is known about what determines when and where neurons will form and how this process is coordinated with the differentiation of neighboring tissues. In most vertebrates the onset of neuronal differentiation takes place in the spinal cord in a head to tail sequence. The changing signaling properties of the adjacent paraxial mesoderm control the progression of neurogenesis in the chick spinal cord. An inverse relationship is found between the expression of caudal neural genes in the prospective spinal cord, which is maintained by underlying presomitic mesoderm and FGF signaling, and neuronal differentiation, which is repressed by such signals and accelerated by somitic mesoderm. Key to this interaction is the ability of somitic mesoderm to repress Fgf8 transcription in the prospective spinal cord. These findings further indicate that attenuation of FGF signaling in the prospective spinal cord is a prerequisite for the onset of neuronal differentiation and may also help to resolve mesodermal and neural cell fates. However, inhibition of FGF signaling alone does not promote the formation of neurons, which requires still further somite signaling. A model is proposed in which signaling from somitic tissue promotes the differentiation of the spinal cord and serves to co-ordinate neural and mesodermal development. Two lines of evidence are presented that strongly suggest that signals from the differentiating somitic mesoderm regulate the onset of neuronal differentiation in the developing spinal cord: (1) somitic signals accelerate the appearance of neurons in caudal neural plate (CNP) explants, as indicated by the swift onset of Delta 1 in single cells, NeuroM expression and the appearance of cells with neurofilament-positive fine processes; and (2) these signals are also required in vivo for the normal onset of neuronal differentiation, as revealed by the strikingly few NeuroM-positive cells in neural tube forming in the absence of the differentiating somitic mesoderm. Removal of somites flanking the later neural tube also depletes the number of NeuroM-expressing cells, while Sox2 and Delta 1 expression remain unaltered, indicating that there is a continuing requirement for somite signals for neurogenesis progression. However, it is likely that somites become dispensable for the production of neurons at later stages, because their removal at stages 12-16 does not alter the number of motor neurons. This suggests that the influence of somite-derived factors is confined to the first born neurons (future reticular and spinal interneurons). Since explants of the neural tube readily form neurons in vitro, this requirement for somite signals in vivo suggests that these signals normally act to oppose other signals present in the neural tube that repress neuronal differentiation (del Corral, 2002).

Whereas vertebrate achaete-scute complex (as-c) and atonal (ato) homologs are required for neurogenesis, their neuronal determination activities in the central nervous system are not yet supported by loss-of-function studies, probably because of genetic redundancy. To address this problem, mice double mutant for the as-c homolog Mash1 and the ato homolog Math3 were generated. Whereas neurogenesis is only weakly affected in Mash1 or Math3 single mutants, in the double mutants, tectal neurons, two longitudinal columns of hindbrain neurons, and retinal bipolar cells are missing and, instead, those cells that normally differentiate into neurons adopt the glial fate. These results indicate that Mash1 and Math3 direct neuronal versus glial fate determination in the CNS and raise the possibility that downregulation of these bHLH genes is one of the mechanisms to initiate gliogenesis (Tomita, 2000).

To examine the possible fate switch in the double mutants, the retina was examined. The retina is an ideal system to analyze the cell fate because it has only six types of neuron and one type of glial cells, which can be clearly identified by position, cell morphology and specific markers. The mature retina consists of three cellular layers: the ganglion cell layer (GCL); the inner nuclear layer (INL), and the outer nuclear layer (ONL). The INL contains three types of interneuron (amacrine, bipolar and horizontal cells) and one type of glia (Müller glial cells). It has been shown that both Math3 and Mash1 are expressed by differentiating bipolar cells. Because most retinal cells including glial cells differentiate postnatally, the retinal explant culture, which mimics in vivo development well, was used. Retinal explants were prepared from E15.5 embryos and cultured for 2 weeks, during which period the majority of retinal cells finish differentiation. By this method, it was possible to monitor the later stage of cell differentiation well after the mutant hosts died. After 2 weeks of culture, the wild-type and mutant retina consisted of three cellular layers. Hematoxylin-eosin (HE) staining indicated that the cell number of each layer was normal in Math3(-/-), Mash1(-/-) and Math3(-/-)-Mash1(-/-) retina. However, whereas bipolar cells (PKC+, mGluR6+, L7+) are normally present in Math3(-/-), they are reduced in number in Mash1(-/-) retina and completely missing in the Math3(-/-)-Mash1(-/-) retina. Instead, Müller glial cells (vimentin+, glutamine synthetase+) are slightly increased in Mash1(-/-) and significantly increased in the double-mutant retina. The other cell types such as rods (rhodopsin+) and amacrine cells (HPC-1+) are not affected in the double-mutant retina. To determine whether any changes of birth date, which is important for cell type specification, are involved in the defects, the retinal explants were examined at days 5, 7 and 10 of culture. At day 5, no glial cells were detectable whereas at days 7 and 10 there were differentiating glial cells in both the wild type and the double mutants, indicating that the time course of gliogenesis is not affected in the double mutants. In addition, there were already more glial cells in the double mutants at day 7, suggesting that more cells initially adopt the glial fate. In contrast, bipolar cells were not detectable in the double mutants at any time points, excluding the possibility that once born, bipolar cells die in the double mutants. Furthermore, there was no significant difference in the Ki-67 staining and TUNEL assay between the wild type and double mutants. These results indicate that cell proliferation and apoptosis are not involved in the loss of bipolar cells and concomitant increase of Müller glial cells, supporting the idea that there is a fate switch from neurons to glial cells in the absence of Math3 and Mash1 (Tomita, 2000).

Thus, retinal explant assays demonstrate that Math3 and Mash1 are essential for bipolar cell development. It has been shown that the homeobox gene Chx10 is also essential for bipolar cell development: Chx10-null mice lack bipolar cells. Thus, at least three genes, Math3, Mash1 and Chx10, are involved in bipolar cell development although it is not clear at which stage Chx10 functions. Interestingly, in contrast to Math3(-/-)-Mash1(-/-) double-mutant retina, Müller glial cells are not increased and therefore a fate switch does not occur in Chx10-null mice, suggesting that the cells that normally differentiate into bipolar cells are likely to result in apoptosis in the absence of Chx10 rather than differentiate into glial cells. These results strongly suggest that the bHLH genes and Chx10 have distinct functions in bipolar cell development. Chx10 may not only confer the cell type-specific identities, as suggested by other homeobox genes, but may also regulate cell survival. In contrast, the main function of Math3 and Mash1 is determination of the neuronal fate but not cell survival. It remains to be analyzed whether the two bHLH genes can also directly confer some of the bipolar cell-specific identities. Further characterization of bHLH and homeobox genes will be necessary to decipher the roles of each factor in specification of particular neuronal subtypes (Tomita, 2000 and references therein).

Analysis of mutant mice has revealed that the bHLH genes Mash1 and Math3, and the homeobox gene Chx10 (Drosophila homologs CG15782 and CG4136) are essential for the generation of bipolar cells, the interneurons present in the inner nuclear layer of the retina. Thus, a combination of the bHLH and homeobox genes should be important for bipolar cell genesis, but the exact functions of each gene remain largely unknown. In Mash1-Math3 double-mutant retina, which exhibits a complete loss of bipolar cells, Chx10 expression does not disappear but remains in Müller glial cells, suggesting that Chx10 expression per se is compatible with gliogenesis. In agreement with this, misexpression of Chx10 alone with retrovirus in the retinal explant cultures induces generation of the inner nuclear layer cells, including Müller glia, but few of them are mature bipolar cells. Misexpression of Mash1 or Math3 alone does not promote bipolar cell genesis either, but inhibits Müller gliogenesis. In contrast, misexpression of Mash1 or Math3 together with Chx10 increases the population of mature bipolar cells and decreased that of Müller glia. Thus, the homeobox gene provides the inner nuclear layer-specific identity while the bHLH genes regulate the neuronal versus glial fate determination, and these two classes of genes together specify the bipolar cell fate. Moreover, Mash1 and Math3 promote the bipolar cell fate, but not the other inner nuclear layer-specific neuronal subtypes in the presence of Chx10, raising the possibility that the bHLH genes may be involved in neuronal subtype specification, in addition to simply making the neuronal versus glial fate choice (Hatakeyama, 2001).

The proprioceptive system provides continuous positional information on the limbs and body to the thalamus, cortex, pontine nucleus, and cerebellum. The basic helix-loop-helix transcription factor Math1 is essential for the development of certain components of the proprioceptive pathway, including inner-ear hair cells, cerebellar granule neurons, and the pontine nuclei. Math1 null embryos lack the D1 interneurons and these interneurons give rise to a subset of proprioceptor interneurons and the spinocerebellar and cuneocerebellar tracts. Three downstream genes of Math1 (Lh2A, Lh2B, and Barhl1) have been identified and Math1 is shown to govern the development of multiple components of the proprioceptive pathway (Bermingham, 2001).

Under normal patterning conditions, a gradient of BMP signaling is established along the dorsal spinal cord controlling differentiation of the neural crest, roof plate, and D1 and D2 interneurons. In this case, the roof plate expresses BMPs (e.g., Gdf7) and MSX1/2; the D1 interneuron precursors first express Math1, then Lh2A, Lh2B, Barhl1, and Brn3A; the D2 interneuron precursors express Ngn1, Isl1, and probably Brn3A. Expression of Math1 is required for the proper expression of Lh2A, Lh2B, and Barhl1, and specification of the D1 interneurons. The hypothesis that Math1 is involved in the fate determination of the D1 interneurons is favored. In the absence of Math1, however, neuronal differentiation in the dorsal spinal cord is altered. BMPs are expressed normally, inducing formation of the roof plate and D2 interneuron precursors, while the MSX1/2 domain expands laterally crossing the roof plate/D1 boundary. This is consistent with the hypothesis and indicates that some of these cells undergo a fate change. It is proposed that BMP concentration and negative regulation by Math1 define the extent of MSX1/2 expression. Because the BMP gradient is intact, the domain of the D2 neuron precursors does not come to occupy the Math1 domain (Bermingham, 2001).

During embryonic development of the inner ear, the sensory primordium that gives rise to the organ of Corti from within the cochlear epithelium is patterned into a stereotyped array of inner and outer sensory hair cells separated from each other by non-sensory supporting cells. Math1, a close homolog of the Drosophila proneural gene atonal, has been found to be both necessary and sufficient for the production of hair cells in the mouse inner ear. Math1 is not required to establish the postmitotic sensory primordium from which the cells of the organ of Corti arise, but instead is limited to a role in the selection and/or differentiation of sensory hair cells from within the established primordium. This is based on the observation that Math1 is only expressed after the appearance of a zone of non-proliferating cells that delineates the sensory primordium within the cochlear anlage. The expression of Math1 is limited to a subpopulation of cells within the sensory primordium that appear to differentiate exclusively into hair cells as the sensory epithelium matures and elongates through a process that probably involves radial intercalation of cells. Furthermore, mutation of Math1 does not affect the establishment of this postmitotic sensory primordium, even though the subsequent generation of hair cells is blocked in these mutants. Finally, in Math1 mutant embryos, a subpopulation of the cells within the sensory epithelium undergo apoptosis in a temporal gradient similar to the basal-to-apical gradient of hair cell differentiation that occurs in the cochlea of wild-type animals (Chen, 2002).

The Notch-regulated transcription factor mouse atonal homolog 1 (Math1) is required for the development of intestinal secretory cells, as demonstrated by the loss of goblet, endocrine and Paneth cell types in null mice. However, it was unknown whether Math1 is sufficient to induce the program of secretory cell differentiation. To examine the function of Math1 in the differentiation of intestinal epithelial cells, intestinal morphology and epithelial and mesenchymal cell fate were examined by histological staining and marker gene expression in transgenic mice expressing a villin-regulated Math1 transgene. Late prenatal transgenic founders exhibited a gross cellular transformation into a secretory epithelium. The expansion of secretory cells coupled with the almost complete loss of absorptive enterocytes suggested reprogramming of a bipotential progenitor cell. Moreover, Math1 expression inhibited epithelial cell proliferation, as demonstrated by a marked reduction in Ki67 positive cells and blunted villi. Unexpectedly, the transgenic mesenchyme was greatly expanded with increased proliferation. Several mesenchymal cell types were amplified, including smooth muscle and neurons, with maintenance of basic radial patterning. Since transgenic Math1 expression was restricted to the epithelium, these findings suggest that epithelial-mesenchymal signaling is altered by the cellular changes induced by Math1. Thus, Math1 is a key effector directing multipotential precursors to adopt secretory and not absorptive cell fate (VanDussen, 2010).

Independently specified Atoh1 domains define novel developmental compartments in rhombomere 1

The rhombic lip gives rise to neuronal populations that contribute to cerebellar, proprioceptive and interoceptive networks. Cell production depends on the expression of the basic helix-loop-helix (bHLH) transcription factor Atoh1. In rhombomere 1, Atoh1-positive cells give rise to both cerebellar neurons and extra-cerebellar nuclei in ventral hindbrain. The origin of this cellular diversity has previously been attributed to temporal signals rather than spatial patterning. This study shows that in both chick and mouse the cerebellar Atoh1 precursor pool is partitioned into initially cryptic spatial domains that reflect the activity of two different organisers: an isthmic Atoh1 domain, which gives rise to isthmic nuclei, and the rhombic lip, which generates deep cerebellar nuclei and granule cells. A combination of in vitro explant culture, genetic fate mapping and gene overexpression and knockdown was used to explore the role of isthmic signalling in patterning these domains. An FGF-dependent isthmic Atoh1 domain was shown to be the origin of distinct populations of Lhx9-positive neurons in the extra-cerebellar isthmic nuclei. In the cerebellum, ectopic FGF induces proliferation while blockade reduces the length of the cerebellar rhombic lip. FGF signalling is not required for the specification of cerebellar cell types from the rhombic lip and its upregulation inhibits their production. This suggests that although the isthmus regulates the size of the cerebellar anlage, the downregulation of isthmic FGF signals is required for induction of rhombic lip-derived cerebellar neurons (Green, 2014).

Regulation of expression of Atonal homologs

Development of the vertebrate nervous system requires the actions of transcription factors that establish regional domains of gene expression, which results in the generation of diverse neuronal cell types. MATH1, a transcription factor of the bHLH class, is expressed during development of the nervous system in multiple neuronal domains, including the dorsal neural tube, the external granule layer (EGL) of the cerebellum and the hair cells of the vestibular and auditory systems. MATH1 is essential for proper development of the granular layer of the cerebellum and the hair cells of the cochlear and vestibular systems, as shown in mice carrying a targeted disruption of Math1. Twenty-one kb of sequence flanking the Math1-coding region is sufficient for Math1 expression in transgenic mice. Two discrete sequences within the 21 kb region have been identified that are conserved between mouse and human, and are sufficient for driving a lacZ reporter gene in these domains of Math1 expression in transgenic mice. The two identified enhancers, while dissimilar in sequence, appear to have redundant activities in the different Math1 expression domains except the spinal neural tube. The regulatory mechanisms for each of the diverse Math1 expression domains must be tightly linked; separable regulatory elements for any given domain of Math1 expression were not found, suggesting that a common regulatory mechanism controls these apparently unrelated domains of expression. In addition, a role is demonstrated for autoregulation in controlling the activity of the Math1 enhancer, through an essential E-box consensus binding site (Helms, 2000).

Patterning of the dorsal neural tube involves Bmp signaling, which results in activation of multiple pathways leading to the formation of neural crest, roof plate and dorsal interneuron cell types. Constitutive activation of Bmp signaling at early stages (HH10-12) of chick neural tube development induces roof-plate cell fate, accompanied by an increase of programmed cell death and a repression of neuronal differentiation. These activities are mimicked by the overexpression of the homeodomain transcription factor Msx1, a factor known to be induced by Bmp signaling. By contrast, the closely related factor, Msx3, does not have these activities. At later stages of neural tube development (HH14-16), dorsal progenitor cells lose their competence to generate roof-plate cells in response to Bmp signaling and instead generate dorsal interneurons. This aspect of Bmp signaling is phenocopied by the overexpression of Msx3 but not Msx1. Taken together, these results suggest that these two different Msx family members can mediate distinct aspects of Bmp signaling during neural tube development (Liu, 2004).

Two lines of evidence suggest that the neural bHLH genes that are crucial for neuronal differentiation might be direct transcriptional targets of Msx1. (1) In pursuit of factors that regulate the expression of the neural bHLH genes by yeast one-hybrid screening, Msx1 was identified to be potential regulator for both Math1 and Mash1 and several consensus sites for Msx1 binding are present in the enhancer regions of Math1/Cath1 and Mash1/Cash1. (2) Both Msx1 and Msx3 can bind to these consensus sites in vitro. However, because in vivo Msx1 represses the bHLH factor expression and Msx3 induces Cath1 expression, additional in vivo co-factors or chromatin properties that modulate these activities must be invoked (Liu, 2004).

Shh signaling protects Atoh1 from degradation mediated by the E3 ubiquitin ligase Huwe1 in neural precursors

Signaling networks controlled by Sonic hedgehog (SHH) and the transcription factor Atoh1 regulate the proliferation and differentiation of cerebellar granule neuron progenitors (GNPs). Deregulations in those developmental processes lead to medulloblastoma formation, the most common malignant brain tumor in childhood. Although the protein Atoh1 is a key factor during both cerebellar development and medulloblastoma formation, up-to-date detailed mechanisms underlying its function and regulation have remained poorly understood. This study reports that SHH regulates Atoh1 stability by preventing its phosphodependent degradation by the E3 ubiquitin ligase Huwe1. These results reveal that SHH and Atoh1 contribute to a positive autoregulatory loop promoting neuronal precursor expansion. Consequently, Huwe1 loss in mouse SHH medulloblastoma illustrates the disruption of this developmental mechanism in cancer. Hence, the crosstalk between SHH signaling and Atoh1 during cerebellar development highlights a collaborative network that could be further targeted in medulloblastoma (Forget, 2014).

Compensational regulation of bHLH transcription factors in the postnatal development of BETA2/NeuroD1-null retina

The bHLH transcriptional factor BETA2/NeuroD1 is essential for the survival of photoreceptor cells in the retina. Although this gene is expressed throughout the retina, BETA2/NeuroD1 knockout mice show photoreceptor cell degeneration only in the outer nuclear layer of the retina; other retinal neurons are not affected. Previous studies on retina explants lacking three bHLH genes revealed that retinal neurons in the inner nuclear layer require multiple bHLH genes for their differentiation and survival. However, single- or double-gene mutations show no or a lesser degree of abnormalities during eye development, likely because of compensation or cooperative regulation among those genes. Because not all null mice survive until the retina is fully organized, no direct evidence of this concept has been reported. To understand the regulatory mechanisms between bHLH factors in retinal development, a detailed analysis of BETA2/NeuroD1 knockout mice was performed. BETA2/NeuroD1 was expressed in all 3 layers of the mouse retina, including all major types of neurons. In addition, a null mutation of BETA2/NeuroD1 resulted in up-regulation of other bHLH genes, Mash1, Neurogenin2, and Math3, in the inner nuclear layer. These data suggest that compensatory and cross regulatory mechanisms exist among the bHLH factors during retinal development (Cho, 2007).

During postnatal development, BETA2/NeuroD1 expression is observed mainly in photoreceptor cells in the outer nuclear layer (ONL). However, moderate levels of expression remain in the outer half and innermost layer of the inner nuclear layer (INL) of the retina as well as in a certain population of cells in the ganglion cell layer (GCL). Therefore, the possibility that BETA2/NeuroD1 has functions in cell type specification in the INL together with other bHLH genes could not be ruled out. Both gain-of-function and loss-of-function studies have demonstrated that BETA2/NeuroD1 participates in the neuron/glia cell fate decisions, similar to other bHLH genes, including Mash1, Math3, and Math5 in retinal explants. Thus, attempts were made to identify differences in the population of major cell types in BETA2/NeuroD1-null retina compared with wild-type littermate retina. However, no differences were found. This result may be due to compensational regulation by other bHLH genes, such as Mash1, Math3, Neurogenin2, and Math5 (Cho, 2007).

Mash1, Math3, and Neurogenin2 are known to be expressed in the developing retina and act as positive regulators. Together with homeodomain factors such as Pax6 and Crx, these factors play important roles in cell type specification during early development. For example, Mash1 and Math3 are expressed predominantly in bipolar cells, and double knockouts of these genes decreases the bipolar cell population while increasing the Müller glial cell population. Neurogenin2 is also transiently expressed in all major neuron types in the mouse retina, and its expression is required for photoreceptor cells, horizontal cells, and bipolar cells. In contrast, BETA2/NeuroD1 is transiently expressed in differentiating amacrine cells. Although BETA2/NeuroD1-null mutation shows delayed amacrine cell development at earlier stages, the number of amacrine cells eventually is the same as that found in wild-type retinas. However, in double-knockout mutations with BETA2/NeuroD1 and Math3, the number of amacrine cells is decreased and that of retinal ganglionic cells is increased. Interestingly, amacrine cells adopt the ganglion cells' fate in this BETA2/NeuroD1;Math3 double-knockout mutant. In addition, the triple bHLH knockouts Mash1;Neurogenin2;Math3 and Math3;Neurogenin2; BETA2/NeuroD1 have been shown to have fewer horizontal cells, but any combination of double mutations of Mash1 or Math3 or BETA2/NeuroD1 with Neurogenin2 display abnormalities in retina development. Furthermore, single-knockout mutations of the genes barely affect the neuronal cell population in the INL and show no retinal abnormalities. Taken together, these results suggest that the bHLH factors cross regulate each others' expression and can specify neuronal subtypes cooperatively during late retinogenesis, especially in progenitor cells in the INL, to generate various subtypes of retinal neurons. Although the precise mechanism for retina cell type specification remains to be determined, these results provide further support for cooperative and compensational regulatory specification during postnatal retinogenesis (Cho, 2007).

Conserved regulatory sequences in Atoh7 mediate non-conserved regulatory responses in retina ontogenesis

The characterisation of interspecies differences in gene regulation is crucial to understanding the molecular basis of phenotypic diversity and evolution. The atonal homologue Atoh7 participates in the ontogenesis of the vertebrate retina. This study reveals how evolutionarily conserved, non-coding DNA sequences mediate both the conserved and the species-specific transcriptional features of the Atoh7 gene. In the mouse and chick retina, species-related variations in the chromatin-binding profiles of bHLH transcription factors correlate with distinct features of the Atoh7 promoters and underlie variations in the transcriptional rates of the Atoh7 genes. The different expression kinetics of the Atoh7 genes generate differences in the expression patterns of a set of genes that are regulated by Atoh7 in a dose-dependent manner, including those involved in neurite outgrowth and growth cone migration. In summary, this study shows how highly conserved regulatory elements are put to use in mediating non-conserved functions and creating interspecies neuronal diversity (Skowronska-Krawczyk, 2009).

Given the crucial role of Atoh7 in the ontogenesis of the vertebrate retina, understanding how its gene is regulated should provide key insights into the transcriptional networks that specify and integrate the RGC lineage within the retina developmental programme. This study shows how evolutionarily conserved non-coding sequences mediate both the conserved and species-specific transcriptional features of the Atoh7 gene. The Ngn2 protein maintains the ability to initiate the retina-specific expression of Atoh7 across distant species, but diverges in its binding profile to evolutionarily conserved regulatory elements. This study reveals how such interspecies variations in transcription factor binding cause variations in the activity of the Atoh7 promoter and how these variations may underlie phenotypic differences between species (Skowronska-Krawczyk, 2009).

The onset of Ngn2 and Atoh7 expression in overlapping domains coincide in the chick and mouse retinas, consistent with the co-expression of Ngn2 and Atoh7 in individual chick and mouse progenitor cells. The downregulation of Atoh7 expression in Ngn2GFP/GFP mice and its upregulation in chick retina overexpressing Ngn2 reveal that the positive regulation of Atoh7 by Ngn2 is evolutionarily conserved. This regulation correlates with the variable in vivo occupancy by Ngn2 of the Atoh7 promoter as a function of developmental stage. Although Ngn2 is expressed in many regions of the nervous system anlage, it does not bind the promoter in tissues that do not express Atoh7. In the chick retina, the K4 di- and tri-methylation of histone H3, a modification known to reflect transcriptional competence, increases in exact register with the kinetics of Atoh7 promoter activity. Likewise, the binding of Ngn2 and its positive effect are associated with chromatin remodelling of the Atoh7 promoter in the early retina. Surprisingly, despite conservation of the proximal and distal elements, Ngn2 binds the distal sequences in the mouse and the proximal sequences in the chick. In yeast, sequence conservation does not readily predict transcription factor binding sites across related species. This study extends to vertebrates the notion that gene regulation resulting from the pattern of species-specific transcription factor binding to highly conserved regulatory elements may be a cause of divergence between species (Skowronska-Krawczyk, 2009).

Although the interplay of bHLH proteins at the proximal E-boxes E1, E2 and E4 determines the spatio-temporal specificity of Atoh7 expression in the chick retina, cooperation between E1, E2, E4 and the conserved distal E-box E9 is required for full-strength promoter activity. Consistent with this notion, mutation of E9 in chick does not alter cell specificity despite a much decreased promoter strength. The finding that in the mouse retina, E9 sets both the strength of the promoter and its specificity, highlights how the activity of conserved elements depends on the cellular context and may vary between species. In the chick, both Ngn2 and Atoh7 bind the proximal E-boxes and Atoh7 also binds E9. Although the possibility cannot be excluded that E9 could mediate competition between Ngn2 and Atoh7, the binding profiles of Atoh7 and Ngn2 proteins suggest that in the early retina, Ngn2 activates transcription of Atoh7 through the proximal promoter, whereas Atoh7 mediates a positive feedback through E9. This feedback by Atoh7 is moderated by the negative effect that the Hes1 protein exerts upon the proximal promoter, thus keeping the rate of Atoh7 transcription at a low level in proliferating progenitors. The downregulation of Hes1 in Atoh7-expressing cells that exit the cell cycle coincides with the rapid upregulation of Atoh7. This upregulation is mediated by Atoh7 and Ngn2, which drive the promoter at peak activity through the combinatorial use of E2, E4 and E9. In Hes1-/- mice, precocious peripheral expansion of the Atoh7 expression domain takes place in the retina (Skowronska-Krawczyk, 2009).

Consistent with the dominant effect of E9, Ngn2 exclusively binds the distal region in mouse. The striking difference in the functional properties of E9 between mouse and chicken does not result from sequence variations; it might reflect epigenetic differences, as chromatin modifications correlate with the binding of bHLH proteins. The fact that Ngn2 expression is low in the mouse retina, whereas it is strongly upregulated in the chick retina, suggests the interesting possibility that the proximal and distal promoter regions are selected by Ngn2 in a dose-dependent manner. Recruitment of the protein on both the distal and proximal promoter regions in mouse P19 cells that overexpress Ngn2 is consistent with this notion. The distal and proximal E-boxes have different sequence identities and their affinity for Ngn2 may thus be sequence dependent, as shown for Atoh7. The finding that the chick Atoh7 promoter in the mouse, or the mouse Atoh7 promoter in the chick, displays the activity features of the host suggests, however, that the cellular context can influence transcription in the developing nervous system (Skowronska-Krawczyk, 2009).

Consistent with the role of E9 ascertained in this report, a 0.6 kb sequence encompassing the proximal E-boxes has no promoter activity in transgenic mice, whereas a 2.3 kb sequence that includes E8 and E9 is sufficient to recapitulate in full the endogenous Atoh7 expression in the E11.5 mouse retina (Skowronska-Krawczyk, 2009).

The maintenance of Atoh7 expression in Math5-/- mice rules out a positive feedback in this species. Thus, the interplay of Ngn2 and Atoh7 at the proximal promoter seen in the chick might not occur in the mouse retina. Instead, the Ngn2 protein, acting through the distal promoter, mediates the low expression of Atoh7 seen in mouse during the period of development when RGCs are produced. The residual, but significant, expression of Atoh7 in Ngn2GFP/GFP mice suggests that other transcription factors might intervene in the regulation of this gene. Pax6 is necessary for Atoh7 expression in the mouse, but cannot by itself control the spatio-temporal and cell cycle phase expression of Atoh7 (Riesenberg, 2009). In Xenopus, Pax6 binding sites within a distal enhancer are required for its enhancer activity (Willardsen, 2009). However, Pax6 alone is not sufficient to induce ectopic expression of Xenopus Atoh7. The fact that the distal enhancer in Xenopus does not require E8 and E9 [respectively E3 and E4 in Willardsen (2009) extends to lower vertebrates the notion that conserved E-boxes assume different roles in different species (Skowronska-Krawczyk, 2009).

There is no obvious abnormality of the GCL in Ngn2GFP/GFP mice, despite the downregulation of Atoh7. However, the possibility cannot be excluded that the GCL might be populated with other cell types, such as amacrine cells, or that different ratios of the numerous RGC subclasses might be produced in response to different levels of Atoh7 (Skowronska-Krawczyk, 2009).

This study reveals that highly conserved non-coding sequences mediate non-conserved interplays of bHLH proteins at the Atoh7 promoter. The proximal E-boxes E1, E2 and E4 mediate activation by Ngn2 and, in addition, the positive feedback by Atoh7 acting upon E9 reinforces Atoh7 expression during the first phase of chick retina development. Expression of Atoh7 is at least 10-fold higher in chick than in mouse early retina, where mouse Ngn2 effects a weak activation through the distal promoter. RGCs are massively produced in the chick retina and the ratio of RGCs to photoreceptors is ~25-fold higher in the avian than in the mouse retina. It is suggested that the much-enhanced expression of Atoh7 in the chick promotes the recruitment to the RGC lineage of a larger set of retinal progenitors (Skowronska-Krawczyk, 2009).

Later, at the time when the majority of RGCs exit the cell cycle and differentiate, the interaction of the Atoh7 protein at E2, E4 and E9 mediates a strong positive feedback. This occurs in the chick but not in the mouse retina, raising the question of why such a transient Atoh7 upregulation is needed to produce RGCs in the chick. Part of the answer might reside in the coherent set of genes that are expressed in newborn RGCs and are regulated by Atoh7 in a dose-dependent manner. Proteins encoded by Stmn2, Snap25, Robo2 and Ptn are protagonists in signalling pathways that link external stimuli to processes such as growth cone protrusion, axonal pathfinding and initial formation of synaptic contacts. The stathmin proteins regulate microtubule dynamics and Stmn2 is highly expressed during neuronal development and is enriched in growth cones. The Snap25 proteins are components of the synaptic vesicle exocytotic machinery and participate in neurite outgrowth. Robo2 plays a role in the axonal pathfinding of RGCs. The regulation of the corresponding genes by Atoh7 suggests that the protein might directly control the development of dendritic arbors and axons in newborn RGCs (Skowronska-Krawczyk, 2009).

Ramon y Cajal noted early on that the avian retina is the most complicated with respect to the morphology of the RGCs. Screens of the dendritic patterns of RGCs have revealed that whereas the large majority of RGCs are monostratified in the mouse retina, bi- and tristratification patterns predominate in the chicken retina. A recent report brought into focus a novel mechanism whereby dendritic stratification of RGCs is achieved (Mumm, 2006). In that study, in vivo time-lapse imaging was used to show that zebrafish RGCs display early patterns of dendritic outgrowth that are predictive of their final lamination, rather than lamination resulting from the pruning of initially exuberant arbors, as generally accepted. It is proposed that mouse RGCs might develop simpler dendritic patterns as a result of the low expression of neurite outgrowth-associated proteins (Skowronska-Krawczyk, 2009).

Rewiring the retinal ganglion cell gene regulatory network: Neurod1 promotes retinal ganglion cell fate in the absence of Math5

Retinal progenitor cells (RPCs) express basic helix-loop-helix (bHLH) factors in a strikingly mosaic spatiotemporal pattern, which is thought to contribute to the establishment of individual retinal cell identity. This study asked whether a tightly regulated pattern is essential for the orderly differentiation of the early retinal cell types and whether different bHLH genes have distinct functions that are adapted for each RPC. To address these issues, one bHLH gene was replaced with another. Math5 is a bHLH gene that is essential for establishing retinal ganglion cell (RGC) fate. The retinas were examined of mice in which Math5 was replaced with Neurod1 or Math3, bHLH genes that are expressed in another RPC and are required to establish amacrine cell fate. In the absence of Math5, Math5Neurod1-KI was able to specify RGCs, activate RGC genes and restore the optic nerve, although not as effectively as Math5. By contrast, Math5Math3-KI was much less effective than Math5Neurod1-KI in replacing Math5. In addition, expression of Neurod1 and Math3 from the Math5Neurod1-KI/Math3-KI allele did not result in enhanced amacrine cell production. These results were unexpected because they indicated that bHLH genes, which are currently thought to have evolved highly specialized functions, are nonetheless able to adjust their functions by interpreting the local positional information that is programmed into the RPC lineages. It is concluded that, although Neurod1 and Math3 have evolved specialized functions for establishing amacrine cell fate, they are nevertheless capable of alternative functions when expressed in foreign environments (Mao, 2008).

Msx1 and Msx2 act as essential activators of Atoh1 expression in the murine spinal cord

Dorsal spinal neurogenesis is orchestrated by the combined action of signals secreted from the roof plate organizer and a downstream transcriptional cascade. Within this cascade, Msx1 and Msx2, two homeodomain transcription factors (TFs), are induced earlier than bHLH neuralizing TFs. Whereas bHLH TFs have been shown to specify neuronal cell fate, the function of Msx genes remains poorly defined. This study describes dramatic alterations of neuronal patterning in Msx1/Msx2 double-mutant mouse embryos. The most dorsal spinal progenitor pool fails to express the bHLH neuralizing TF Atoh1, which results in a lack of Lhx2-positive and Barhl2-positive dI1 interneurons. Neurog1 and Ascl1 expression territories are dorsalized, leading to ectopic dorsal differentiation of dI2 and dI3 interneurons. In proportion, the amount of Neurog1-expressing progenitors appears unaffected, whereas the number of Ascl1-positive cells is increased. These defects occur while BMP signaling is still active in the Msx1/Msx2 mutant embryos. Cell lineage analysis and co-immunolabeling demonstrate that Atoh1-positive cells derive from progenitors expressing both Msx1 and Msx2. In vitro, Msx1 and Msx2 proteins activate Atoh1 transcription by specifically interacting with several homeodomain binding sites in the Atoh1 3' enhancer. In vivo, Msx1 and Msx2 are required for Atoh1 3' enhancer activity and ChIP experiments confirm Msx1 binding to this regulatory sequence. These data support a novel function of Msx1 and Msx2 as transcriptional activators. This study provides new insights into the transcriptional control of spinal cord patterning by BMP signaling, with Msx1 and Msx2 acting upstream of Atoh1 (Duval, 2014).

Targets of mammalian Atonal homologs

Proneural basic helix-loop-helix (bHLH) proteins are key regulators of neurogenesis. However, downstream target genes of the bHLH proteins remain poorly defined. Mbh1 (Barhl2 - Mouse Genome Informatics) confers commissural neuron identity in the spinal cord. Enhancer analysis using transgenic mice reveals that Mbh1 expression requires an E-box 3' of the Mbh1 gene. Mbh1 expression is lost in Math1 knockout mice, whereas misexpression of Math1 induces ectopic expression of Mbh1. Moreover, Math1 binds the Mbh1 enhancer containing the E-box in vivo and activated gene expression. Generation of commissural neurons by Math1 is inhibited by a dominant negative form of Mbh1. These findings indicate that Mbh1 is necessary and sufficient for the specification of commissural neurons, as a direct downstream target of Math1 (Saba, 2005).

Gfi1 (Drosophila homolog: Senseless) is a transcriptional repressor implicated in lymphomagenesis, neutropenia, and hematopoietic development, as well as ear and lung development. This study demonstrates that Gfi1 functions downstream of Math1 in intestinal secretory lineage differentiation. Gfi1-/- mice lack Paneth cells, have fewer goblet cells, and supernumerary enteroendocrine cells. Gfi1-/- mice show gene expression changes consistent with this altered cell allocation. These data suggest that Gfi1 functions to select goblet/Paneth versus enteroendocrine progenitors. A model of intestinal cell fate choice is proposed in which beta-catenin and Cdx function upstream of Math1, and lineage-specific genes such as Ngn3 act downstream of Gfi1 (Shroyer, 2005).

Broadly permissive intestinal chromatin underlies lateral inhibition and cell plasticity

Cells differentiate when transcription factors bind accessible cis-regulatory elements to establish specific gene expression programs. In differentiating embryonic stem cells, chromatin at lineage-restricted genes becomes sequentially accessible, probably by means of 'pioneer' transcription factor activity, but tissues may use other strategies in vivo. Lateral inhibition is a pervasive process in which one cell forces a different identity on its neighbours, and it is unclear how chromatin in equipotent progenitors undergoing lateral inhibition quickly enables distinct, transiently reversible cell fates. This study reports the chromatin and transcriptional underpinnings of differentiation in mouse small intestine crypts, where notch signalling mediates lateral inhibition to assign progenitor cells into absorptive or secretory lineages. Transcript profiles in isolated leucine-rich repeat-containing receptor 5 (LGR5) positive intestinal stem cells and secretory and absorptive progenitors indicated that each cell population was distinct and the progenitors specified. Nevertheless, secretory and absorptive progenitors showed comparable levels of H3K4me2 and H3K27ac histone marks and DNase I hypersensitivity - signifying accessible, permissive chromatin - at most of the same cis-elements. Enhancers acting uniquely in progenitors were well demarcated in LGR5+ intestinal stem cells, revealing early priming of chromatin for divergent transcriptional programs, and retained active marks well after lineages were specified. On this chromatin background, ATOH1, a secretory-specific transcription factor, controls lateral inhibition through delta-like notch ligand genes and also drives the expression of numerous secretory lineage genes. Depletion of ATOH1 from specified secretory cells converted them into functional enterocytes, indicating prolonged responsiveness of marked enhancers to the presence or absence of a key transcription factor. Thus, lateral inhibition and intestinal crypt lineage plasticity involve interaction of a lineage-restricted transcription factor with broadly permissive chromatin established in multipotent stem cells (Kim, 2014).

Mammalian Atonal homologs and eye development

The vertebrate retina contains seven major neuronal and glial cell types in an interconnected network that collects, processes and sends visual signals through the optic nerve to the brain. Retinal neuron differentiation is thought to require both intrinsic and extrinsic factors, yet few intrinsic gene products have been identified that direct this process. Math5 (Atoh7) encodes a basic helix-loop-helix (bHLH) transcription factor that is specifically expressed by mouse retinal progenitors. Math5 is highly homologous to atonal, which is critically required for R8 neuron formation during Drosophila eye development. Like R8 cells in the fly eye, retinal ganglion cells (RGCs) are the first neurons in the vertebrate eye. Math5 mutant mice are fully viable, yet lack RGCs and optic nerves. Thus, two evolutionarily diverse eye types require atonal gene family function for the earliest stages of retinal neuron formation. At the same time, the abundance of cone photoreceptors is significantly increased in Math5-/- retinae, suggesting a binary change in cell fate from RGCs to cones. A small number of nascent RGCs are detected during embryogenesis, but these fail to develop further, suggesting that committed RGCs may also require Math5 function (Brown, 2001).

These results significantly advance the concept of functional conservation within the atonal gene family. In Drosophila, atonal controls development of retinal neurons and the chordotonal organs, which are internal mechanosensory structures that act as proprioceptors. In vertebrates, visual system and proprioceptive functions are separated, with Math5 controlling the former and Math1 (Atoh1) regulating the latter. Math1 is expressed in cochlear and vestibular hair cells, vibrissae, the dorsal spinal column, joint capsules, Merkel touch receptors, and the cerebellum. Each of these structures is associated with mechanosensory perception, locomotory coordination or the integration and processing of three-dimensional spatial information. Math5 and Math1 are the vertebrate bHLH genes most closely related in structure to atonal. This relationship, the absence of significant Math1 expression in the eye, and the findings presented here, strongly suggest that Math5 is the atonal ortholog, and thus the major proneural gene, for the mammalian eye. There are few other examples where two functions identified in the Metazoa are so cleanly partitioned by evolution in the vertebrate lineage. In other gene families, the expression of paralogous genes is largely overlapping so that phenotypes can arise only at the edges where incomplete redundancy is exposed. In contrast, Math1 and Math5 act in independent domains. Taken together they are functionally equivalent to atonal (Brown, 2001).

Math5 acts during retinal histogenesis, after primary pattern formation in the anterior neural tube, specification of the optic primordia, and the major period of the optic cup growth has occurred. Math5 expression is dependent upon Pax6, which has been positioned near the top of a hierarchy for metazoan eye development. Consequently, these results, like those for the hedgehog genes, show that functional homology in visual system development extends more deeply than Pax6. Finally, these results further establish an evolutionary parallel between vertebrate RGCs and fly R8 photoreceptors, the earliest-born neurons of bilaterian visual systems (Brown, 2001).

Distinct classes of neurons are generated from progenitor cells distributed in characteristic dorsoventral patterns in the developing spinal neural tube. Restricted neural progenitor populations are defined by the discrete, nonoverlapping expression of Ngn1, Math1, and Mash1. Crossinhibition between these bHLH factors is demonstrated and provides a mechanism for the generation of discrete bHLH expression domains. This precise control of bHLH factor expression is essential for proper neural development since as demonstrated in both loss- and gain-of-function experiments, expression of Math1 or Ngn1 in dorsal progenitor cells determines whether LH2A/B- or dorsal Lim1/2-expressing interneurons will develop. Together, the data suggest that although Math1 and Ngn1 appear to be redundant with respect to neurogenesis, they have distinct functions in specifying neuronal subtype in the dorsal neural tube (Gowan, 2001).

Five distinct classes of dorsal interneurons have been described by their expression of LIM homeodomain factors. These interneuron populations, D1A, D1B, D2, D3, and D4, express LH2A, LH2B, Islet1, Lim1/2, and Lmx1b, respectively. Defined here is a subclass of the D3 population, termed D3A, and characterized as the most dorsal Lim1/2-expressing cells that coexpress Brn3a (Gowan, 2001).

Expression in the dorsal neural tube of the bHLH transcription factors, Ngn1 and Math1, defines progenitor populations destined to be distinct types of neurons. Math1 marks progenitors that generate LH2A/B-expressing interneurons, and Math1 is required for these interneurons to form. Furthermore, ectopic expression of Math1 in chick can induce LH2 expression in the dorsal neural tube; however, it is not sufficient to induce LH2 expression in ventral regions. This suggests that other factors, specific to dorsal neural tube regions, work together with Math1 to form LH2 interneurons. Another phenotype in the Math1 mutant demonstrated here is the transition of the Math1 progenitors to Ngn1/Ngn2-expressing cells that give rise to D3A interneurons at the expense of LH2 interneurons. In contrast, examination of the lacZ-expressing cells in a lacZ knockin of Math1 reveals increased overlap with Msx1/2, and thus suggests that at least some Math1-expressing cells are becoming roof plate cells. These data taken together suggest that in the absence of Math1, expression of markers normally marking cells found at either border/roof plate (dorsally) and Ngn1/2 (ventrally) expand to fill the domain (Gowan, 2001).

A newly defined cell type (D3A), which expresses Lim1/2;Brn3a and accounts for the most dorsal Lim1/2 domain, arises from progenitor cells expressing Ngn1 or Ngn2, and either Ngn1 or Ngn2 is required for these cells to form. The correlation seen between the expansion of Ngn1- and Ngn2-expressing progenitors and the increase in the D3A interneuron population in the Math1 mutant support the idea that they are sufficient for Lim1/2 expression in the dorsal neural tube. It is surprising that the dorsal Islet1 population appears to be unaffected in the Ngn1 and Ngn2 double mutants since Islet1-expressing cells are completely absent in these mutants in the DRG, and are diminished in the ventral neural tube (J. E. J., unpublished). This suggests that Ngn1 or Ngn2, combined with other factors with differential expression in the dorsoventral axis of the neural tube, generate multiple cell types. This is consistent with conclusions from studies of Math1, and with the fact that loss of a specific bHLH results in the loss of different types of neurons in different regions of the nervous system (Gowan, 2001).

Another way to classify neurons, other than by expression of homeodomain markers, is by a description of their axonal projections. Using the Math1 and Ngn1 enhancers to drive reporter gene expression in transgenic mice, two nonoverlapping populations of commissural interneurons can be distinguished. Additional commissural interneurons have been seen in transgenic animals with a Ngn2 enhancer directing lacZ expression. Although the eventual function of the different neuronal populations from Ngn1 and Ngn2 progenitors is unknown, the dorsal commissural interneurons from Math1 progenitors have recently been defined as those carrying proprioceptive information in the spinocerebellar and cuneocerebellar pathways (Gowan, 2001).

Combining the results from loss- and gain-of-function studies of Ngn1, Ngn2, and Math1 with the timing of their expression, it is clear that the bHLH factors play a vital role in the transition from proliferating neural progenitor cells to differentiated neurons in multiple discrete neuronal lineages. The expansion of distinct bHLH factors in the different mutant backgrounds in the absence of cell death, taken with the wild-type appearance of the dorsal neural tube by morphology and general neural markers, suggests that Ngn1, Ngn2, and Math1 may compensate for each other's role in inducing neurons from stem cells in the dorsal neural tube. The redundant function in neurogenesis between different bHLH factors suggested here was recently suggested in studies of Mash1;Math3 or Mash1;Ngn2 double mutants. In these studies, neurogenesis was not significantly altered in the single mutants, but in the absence of both factors in different regions of the developing nervous system, there was excess gliogenesis at the expense of neurogenesis. Thus, the neural bHLH factors appear to share the function of inducing neurogenesis (Gowan, 2001).

A separate aspect of the function of bHLH factors is their role in specifying neuronal subtype. This was first investigated in Drosophila where atonal was shown to preferentially produce chordotonal neurons and achaete preferentially produced external sense organs when ectopically expressed. Furthermore, in vertebrates, forced expression of Ngn1 preferentially drives neural crest cells into the sensory neuron fate. These experiments clearly suggest a role for the bHLH factors in neuronal specification; however, the fact that each bHLH factor is present and required for multiple distinct lineages means that other factors and/or pathways modulate this activity. By demonstrating that Ngn1, Math1, and Mash1 are expressed in distinct progenitors, and are required for the formation of definable interneuron populations within a relatively discrete environment, it has been demonstrated that they function in specifying neuronal subtype in the dorsal spinal cord. Thus, although their roles in inducing neurogenesis appears to be a shared function, their roles in specifying neuronal cell type are distinct (Gowan, 2001).

Crossinhibitory regulation plays a role in boundary formation between the Class I and Class II homeodomain factors in the ventral neural tube. The first comparisons of Ngn1, Ngn2, and Mash1 mRNA expression domains demonstrate sharp boundaries in multiple regions of the developing nervous system, giving rise to the suggestion that they negatively regulate each other's expression. Mash1 expression in the dorsal telencephalon is increased in Ngn1;Ngn2 mutant embryos, and this ectopic Mash1 expression results in ventral telencephalon markers being expressed ectopically in the dorsal regions. In addition, in the absence of the roof plate, Math1 and Ngn1 expression are lost and Mash1 is found to extend to the dorsal edge of the neural tube. At the single cell level, at least Ngn1, Math1, and Mash1 define distinct progenitors in the spinal neural tube with only a single factor expressed in each progenitor cell. Indeed, even in the ventral neural tube where Ngn1 predominates, there are scattered cells that express only Mash1. This pattern of expression is at least in part due to an active inhibition between this set of bHLH factors. This is supported by two separate experimental paradigms. In the first paradigm, in mouse embryos lacking Math1, Ngn1- and Ngn2-expressing cells increase in number with no increase in apoptosis in this region of the neural tube. The converse is also true, in the absence of Ngn1 and Ngn2, Math1-expressing cells increase. In a second set of experiments using ectopic expression in chick neural tubes, Math1 is sufficient to inhibit the endogenous chick bHLH factors c-Ngn1 and Cash1, and Ngn1 is sufficient to inhibit the endogenous factors Cath1 and Cash1. The inhibition of Cath1 by Ngn1 appears to be cell autonomous because when examined at a single cell level, cells misexpressing Ngn1 do not coexpress Cath1. Notably, these crossinhibitory phenomena appear to be between different subclasses within the bHLH family. Mash1 is of the achaete/scute subclass; Math1 is of the atonal subclass, and Ngn1/Ngn2 are of the biparous/tap subclass (Gowan, 2001).

The crossinhibitory regulation within the early neural bHLH factor family shown here suggests a mechanism for how a progenitor ends up expressing only one factor, but does not address how the overall pattern of bHLH expression is initially set up. In the dorsal neural tube, BMP signaling likely plays a major role in setting up the bHLH pattern. Loss of the BMP GDF7 results in loss of a subset of Math1-expressing cells, while BMP7 induces Math1 expression in explanted intermediate regions of the neural tube. In a roof plate ablation paradigm, where BMP signaling is dramatically disrupted, both Math1 and Ngn1 are lost in the dorsal neural tube. Math1 is induced in chick neural tubes with a high level of expression of a constitutively active BMP receptor, BMPR-Ib. In contrast, at lower levels of ectopic expression of BMPR-Ib, c-Ngn1 is induced. Taken together, these data suggest that different levels of BMP signaling have different effects on bHLH expression and this pathway may set up the initial pattern of bHLH expression in the dorsal neural tube. Subsequently, other regulatory mechanisms such as the crossinhibition presented here, and differential autoregulation of the bHLH factors, modify and refine this initial pattern (Gowan, 2001).

The basic helix-loop-helix genes Math3 and NeuroD are expressed by differentiating amacrine cells, retinal interneurons. Previous studies have demonstrated that a normal number of amacrine cells is generated in mice lacking either Math3 or NeuroD. In Math3-NeuroD double-mutant retina, amacrine cells are completely missing, while ganglion and Müller glial cells are increased in number. In the double-mutant retina, the cells that would normally differentiate into amacrine cells do not die but adopt the ganglion and glial cell fates. Misexpression studies using the developing retinal explant cultures show that, although Math3 and NeuroD alone promote only rod genesis, they significantly increase the population of amacrine cells when either the homeobox genes Pax6 or Six3 is co-expressed. These results indicate that Math3 and NeuroD are essential, but not sufficient, for amacrine cell genesis, and that co-expression of the basic helix-loop-helix and homeobox genes is required for specification of the correct neuronal subtype (Inoue, 2002).

Thus, in Math3/NeuroD double-mutant retina, amacrine cells are completely missing without significant cell death. The cells that would normally differentiate into amacrine cells remain in the ganglion cell layer (GCL) and the inner nuclear layer (INL) of the double-mutant retina. The majority of the lacZ+ cells in the GCL and INL are small in size and show a ganglion cell phenotype, while others display a Müller glia phenotype. In accordance with this observation, ganglion and Müller glial cells are increased in number in the double-mutant retina. Normally, nearly 60% of the GCL cells are displaced amacrine cells and the others are ganglion cells. By contrast, in the double-mutant retina, all of the GCL cells are ganglion cells. Furthermore, there are many ectopic ganglion cells and extra Müller glial cells in the INL of the double-mutant retina. These results indicate that there is a fate switch from amacrine cells to ganglion and Müller glial cells in the absence of Math3 and NeuroD. Since amacrine cell genesis overlaps with ganglion and Müller glial cell genesis, it is likely that the double-mutant cells may adopt the alternatively available cell fates (Inoue, 2002).

The phenotype similar but opposite to the Math3/NeuroD double mutation is observed in the retina lacking the bHLH gene Math5 in mouse and its ortholog in zebrafish, which shows a fate switch from ganglion cells to amacrine cells. Thus, Math5 regulates ganglion versus amacrine cell fate, suggesting that this bHLH gene is involved in the neuronal subtype specification rather than the neuronal versus glial cell fate decision. By contrast, the retina lacking Mash1 and Math3 exhibit a fate switch from bipolar cells to Müller glial cells, indicating that Mash1 and Math3 regulate neuronal versus glial fate determination in the retina. The present data show that Math3-NeuroD double mutation leads to increase of both ganglion cells and Müller glia at the expense of amacrine cells, suggesting that Math3 and NeuroD regulate both neuronal subtype specification and neuronal versus glial fate determination. Although these bHLH genes seem to have distinct activities, it is speculated that the two types of fate switches, neurons to glia and neuronal subtype changes, may simply reflect the different competence of retinal precursors. Because ganglion cell genesis overlaps with amacrine cell genesis but not with Müller glial cell genesis, it is likely that the cells that would differentiate into ganglion cells have a potential to become amacrine cells but not others. Thus, in the absence of Math5, the cells that fail to differentiate into ganglion cells may predominantly become amacrine cells. Likewise, since bipolar and Müller glial cells are the last cell types to be generated, the cells that fail to differentiate into bipolar cells may have the only choice to become Müller glia in the absence of Mash1 and Math3. By contrast, since amacrine cell genesis overlaps with both ganglion and Müller glial cell genesis, the cells that fail to differentiate into amacrine cells may have a potential to become both ganglion and Müller glial cells and thereby adopt these two fates in the absence of Math3 and NeuroD. Thus, the two types of fate switches, neurons to glia and neuronal subtype changes, might mostly reflect the competence of retinal precursors, and it is likely that the cells that are blocked from differentiation to a particular cell type may simply adopt alternatively available cell fates (Inoue, 2002).

During vertebrate retinogenesis, seven classes of cells are specified from multipotent progenitors. To date, the mechanisms underlying multipotent cell fate determination by retinal progenitors remain poorly understood. The Foxn4 winged helix/forkhead transcription factor is shown to be expressed in a subset of mitotic progenitors during mouse retinogenesis. Targeted disruption of Foxn4 largely eliminates amacrine neurons and completely abolishes horizontal cells, while overexpression of Foxn4 strongly promotes an amacrine cell fate. These results indicate that Foxn4 is both necessary and sufficient for commitment to the amacrine cell fate and is nonredundantly required for the genesis of horizontal cells. Furthermore, evidence is provided that Foxn4 controls the formation of amacrine and horizontal cells by activating the expression of the retinogenic factors Math3, NeuroD1, and the Prospero-like transcription factor Prox1. These data suggest a model in which Foxn4 cooperates with other key retinogenic factors to mediate the multipotent differentiation of retinal progenitors (Li, 2004).

In the developing zebrafish retina, neurogenesis is initiated in cells adjacent to the optic stalk and progresses to the entire neural retina. It has been reported that hedgehog (Hh) signalling mediates the progression of the differentiation of retinal ganglion cells (RGCs) in zebrafish. However, the progression of neurogenesis seems to be only mildly delayed by genetic or chemical blockade of the Hh signalling pathway. cAMP-dependent protein kinase (PKA) effectively inhibits the progression of retinal neurogenesis in zebrafish. Almost all retinal cells continue to proliferate when PKA is activated, suggesting that PKA inhibits the cell-cycle exit of retinoblasts. A cyclin-dependent kinase (cdk) inhibitor p27 inhibits the PKA-induced proliferation, suggesting that PKA functions upstream of cyclins and cdk inhibitors. Activation of the Wnt signalling pathway induces the hyperproliferation of retinal cells in zebrafish. The blockade of Wnt signalling inhibits the PKA-induced proliferation, but the activation of Wnt signalling promotes proliferation even in the absence of PKA activity. These observations suggest that PKA inhibits exit from the Wnt-mediated cell cycle rather than stimulates Wnt-mediated cell-cycle progression. PKA is an inhibitor of Hh signalling, and Hh signalling molecule morphants show severe defects in cell-cycle exit of retinoblasts. Together, these data suggest that Hh acts as a short-range signal to induce the cell-cycle exit of retinoblasts. The pulse inhibition of Hh signalling revealed that Hh signalling regulates at least two distinct steps of RGC differentiation: the cell-cycle exit of retinoblasts and RGC maturation. This dual requirement of Hh signalling in RGC differentiation implies that the regulation of a neurogenic wave is more complex in the zebrafish retina than in the Drosophila eye (Masai, 2005).

Pulse treatment results suggest that Hh signalling between 24 and 29 hpf is required for the wave of atonal homolog ath5 expression. Furthermore, the introduction of shh and twhh morpholino-antisense oligonucleotides blocks neuronal production in the retina. These data suggest that Shh and Twhh regulate the progression of ath5 expression between 24 and 29 hpf. However, it is difficult to detect shh and twhh mRNA expression in the neural retina at this early stage, although it was reported that shh RNA is expressed in the retina at 28 hpf. One possibility is that Shh and Twhh expressed in the ventral forebrain may have a long-range action on progenitor cells of the optic cup. It was reported that Hh expressed in midline tissue is important for proliferation of the developing forebrain in chicks and mice, probably through its long-range actions. The most recent study on Hh signalling in the zebrafish retina also suggested that Hh signalling outside the optic cup regulates ath5 expression before 27 hpf. However, this is not considered as the most likely explanation, since the wave of ath5 expression normally occurs when the optic cups are dissected from the forebrain at 18 hpf, and cultured as an explant later, suggesting that a source of Hh signals is localised within the optic cup. Furthermore, when the dissected eye cup was divided into two (the nasal and temporal halves) only the nasal half expressed ath5, suggesting that short-range Hh signalling acts from the nasal to temporal regions across the neural retina. Transplantation of hh-MO-injected cells into wild-type host retinas demonstrated that ath5 expression is rarely observed in hh-MO-injected retinal columns, and that wild-type cells fail to express ath5 when they are located adjacent to the temporal side of Shh- and Twhh-deprived cells. These data suggest that a short-range action of Shh and Twhh expressed in the neural retina regulates the wave of ath5 expression and neuronal production. Low levels of Shh and Twhh expression may spread to the temporal retina up until 27 hpf and may be sufficient to regulate the wave of ath5 expression (Masai, 2005).

In the developing retina, the production of ganglion cells is dependent on the proneural proteins NGN2 and ATH5, whose activities define stages along the pathway converting progenitors into newborn neurons. Crossregulatory interactions between NGN2, ATH5 and HES1 maintain the uncommitted status of ATH5-expressing cells during progenitor patterning, and later on regulate the transition from competence to cell fate commitment. Prior to exiting the cell cycle, a subset of progenitors is selected from the pool of ATH5-expressing cells to go through a crucial step in the acquisition of a definitive retinal ganglion cell (RGC) fate. The selected cells are those in which the upregulation of NGN2, the downregulation of HES1 and the autostimulation of ATH5 are coordinated with the progression of progenitors through the last cell cycle. This coordinated pattern initiates the transcription of ganglion cell-specific traits and determines the size of the ganglion cell population (Matter-Sadzinski, 2005).

Spatial cell patterning and RGC commitment correlate with the two main phases of ATH5 expression. During the period of patterning, crossregulatory interactions between HES1, NGN2 and ATH5 keep ATH5 expression low, thereby maintaining the uncommitted status of ATH5-expressing cells and enabling the expansion and intermingling of pools of progenitors initially partitioned in distinct domains. Once progenitors are properly distributed throughout the retina, about one-third of ATH5-expressing cells become committed to acquire a definitive RGC fate immediately before exiting the cell cycle. This requires a tight coordination between downregulation of HES1, upregulation of NGN2, cell progression through the last S-phase and the upregulation of ATH5. Cells that upregulate ATH5 expression initiate transcription of early RGC-specific traits, then exit the cell cycle and express Neuro M and other post-mitotic RGC-specific genes. This study highlights how changes in the transcriptional patterns correlate with the progression of progenitors through the last cell cycle and with their commitment to the RGC fate, underlining the role of HES1 as a key prompt of the molecular events leading to RGC genesis (Matter-Sadzinski, 2005).

A specific feature of retinogenesis is that it proceeds from the centre to the periphery such that all seven retinal cell types are distributed at the proper ratio throughout the retina. At early stages of development, the retinal neuroepithelium is subdivided into two developmentally distinct territories. Low levels of HES1 transcripts outline a broad region of the posterior retina where ATH5, NGN2 and ASH1 are expressed, whereas a robust accumulation of HES1 transcripts throughout the anterior retina prevents the onset of proneural gene expression. HES1 functions similarly at the onset of neurogenesis in the olfactory placode, where it circumscribes a domain of Mash1 expression. It thus appears that HES1 is acting, much like hairy in Drosophila, as a prepattern gene. Neurogenesis starts within a rather broad central region defined by expression of ATH5, NGN2 and Neuro M. Cells expressing ATH5 at a high level and Neuro M-positive cells are evenly distributed throughout the neurogenic domain, indicating that the first newborn RGCs are produced with similar frequency throughout the central retina. In the posterior retina, cells that initiated expression of proneural genes are initially organized in two separate domains corresponding to two retinal lineages: cells that express NGN2/ATH5 constitute the progenitor pools from which early-born retinal neurons will emerge, whereas ASH1-expressing cells form a pool for late-born neurons. The opposite effects of NGN2 on ATH5 and ASH1 expression combined with the inhibitory activity of ASH1 on ATH5 transcription account for the distribution of ASH1 and ATH5/NGN2 cells in two distinct progenitor domains, the more peripheral expression of ASH1 perhaps reflecting its lower sensitivity towards HES1. The initial patterning of the posterior retina resembles the neuroepithelial partitioning detected in other areas of the developing CNS. However, whereas in other CNS regions the refining of borders is essential for the precise spatial generation of different classes of neurons along the dorsoventral axis, the blurring of borders and intermingling of initially distinct progenitor pools are necessary for a proper spatial distribution of neurons and glia throughout the retina. Although ATH5/NGN2 and ASH1 expressions are mutually exclusive, a small fraction of ATH5-expressing cells co-express ASH1, indicating that they are in a transient state prior to acquiring a definite progenitor status. Because the ATH5, NGN2 and ASH1 genes crossregulate and display different sensitivities towards HES1, it is supposed that various balances between these four factors may mediate alternate fate choices. Such dynamic regulatory interactions are, in part, responsible for the progressive loss of patterning in the posterior retina. The ATH5/NGN2 domain remains restricted to the posterior retina until E4 and expands to keep pace with growth of the whole retina at a rate similar to that reported for the differentiation of RGCs. Despite significant changes in the expression pattern of ATH5, similar proportions of retinal cells express this gene at stages 18 and 29-30, suggesting that ATH5-expressing cells propagate at a rate comparable with that of the other progenitors during the period of patterning (Matter-Sadzinski, 2005).

Even though the population of ATH5-expressing cells is established at E2.5, only a small fraction of these will differentiate into RGCs until E4. Retinogenesis is controlled by components of the Notch pathway, which may employ two strategies to keep the majority of cells in the central retina from differentiating during the patterning period. Cells that express proneural genes may promote the upregulation of HES1 in neighbouring cells, thereby preventing them from expressing proneural genes. The proximity in central retina of individual cells that highly express HES1 or ATH5 is indeed indicative of ongoing lateral inhibition. However, cells strongly expressing Notch effectors are rare in the posterior retina, whereas a high proportion of ATH5-expressing progenitors co-express HES1. Thus, it appears that the low level of HES1 in cells that have already initiated NGN2 and ATH5 expression suffices to prevent the upregulation of these genes. The proliferative state is thereby maintained in most ATH5-expressing cells, as required to ensure the proper ratio of RGC progenitors in the posterior retina and as expected of HES genes, which function to keep neuroepithelial cells undifferentiated, thereby regulating the size and cell architecture of brain structures and retina. In anterior retina, progenitor cell patterning becomes evident by E4 and the expansion of proneural gene expression proceeds, much as in zebrafish, in a wave-like fashion as HES1 expression recedes to the retinal margin. The ASH1 and NGN2 expression domains expand to the periphery at similar rates, whereas the progression of the ATH5 domain is slightly delayed. The full patterning of the retina accomplished around E6 coincides with the upregulation of proneural gene expression throughout the retina and with the peak of RGC production (Matter-Sadzinski, 2005).

To analyse how ATH5 is regulated along the course of RGC specification, a promoter region extending 775 bp upstream of the initiation codon was used. The cloned sequence accurately reproduces the activity and the mode of regulation of the endogenous promoter. It contains essential regulatory elements that are well conserved across distant vertebrate species, but it is unclear whether the different species use similar strategies to regulate ATH5 expression. Whereas a proximal cis-regulatory region of the Xenopus Xath5 gene suffices, much as in the chick retina, to drive retina specific reporter gene expression in a bHLH-dependent manner, the mouse ATH5 promoter appears to be regulated differently. It is tempting to speculate that the different modes regulating ATH5 across species may account for differences in the spatiotemporal progenitor patterning of the retinal neuroepithelium. Differences in the developments of the anterior and posterior retinas may have permitted the evolution of a specialized structure such as the macula (Matter-Sadzinski, 2005).

This study reveals that NGN2 acts at different regulatory levels during RGC specification. In early retina, NGN2 is a principal regulator of ATH5 expression and exerts this function through direct activation of ATH5 transcription and through crossregulatory interactions with HES1. In addition, NGN2 drives ATH5-expressing cells out of S phase. Whereas the capacity of NGN2 to promote cell cycle arrest is part of its panneuronal activities and is in evidence in other compartments of the developing CNS, its capacity to activate ATH5 expression is largely retina specific. The quasi-simultaneous onset of NGN2 and ATH5 expression in the central retina shortly after formation of the eye cup, the capacity of NGN2 to activate ATH5 transcription and to bind the ATH5 promoter at the early stages of development suggest that NGN2 may be directly involved in the activation of ATH5 expression. The finding that the expansion of the NGN2 domain towards the anterior edge of the retina precedes that of ATH5 argues in favour of this interpretation. In the retina of the Ngn2–/– mouse, the much increased expression of ASH1 and the downregulation of ATH5 when compared with the wild type, may result from an increase in the population of ASH1-expressing cells at the expense of the ATH5/NGN2 progenitors, thus underlining the importance of NGN2 in establishing and maintaining a pool of ATH5-expressing cells. Both the NGN2 and ATH5 genes fail to be activated in the retinal precursors of the Pax6–/– mouse and Pax6 has been proposed to regulate NGN2 directly in the mouse retina. There are multiple E-boxes but no consensus Pax6 binding site in the chicken ATH5 promoter, and therefore the idea is favored that Pax6 regulates ATH5 via NGN2. The expression of NGN2 in many regions of the nervous system anlage where ATH5 is not detected and the demonstration that recruitment of NGN2 on the ATH5 promoter is retina specific provide evidence that a retina-specific context accounts for the capacity of NGN2 to activate ATH5 expression. The ability of bHLH factors to regulate the development of distinct neurons has been proposed to depend upon the cellular contexts in which they function. In retina, this context may be determined, among other possibilities, by the balance between NGN2 and HES1: as show, HES1 inhibits the NGN2-mediated activation of ATH5 in a dose-dependent manner. Likewise, the upregulation of NGN2 correlates with the dowregulation of HES1. Moreover, single cell transcriptional analysis reveals that overexpressing NGN2 diminishes the pool of cells that co-express ATH5 and HES1, an indication that NGN2 may contribute to the downregulation of HES1 in early neural progenitors, thereby providing a cellular environment permissive for ATH5 autostimulation (Matter-Sadzinski, 2005).

The upregulation of both NGN2 and ATH5 occurs later in development, around E6, but by then ATH5 has become the main regulator of its own transcription. NGN2 occupies the ATH5 promoter similarly at E3 and at E6, suggesting that it still directly participates in the control of ATH5 transcription. However, its main contribution to ATH5 expression may occur through other, indirect regulatory pathways. As ATH5-expressing progenitors exit the cell cycle, NGN2 promotes the expression first of Neuro M and then of Neuro D, both stimulators of ATH5 promoter activity. These distinct functions of NGN2 in the ontogenesis of RGCs illustrate how, depending on specific combinations of transcription factors and of other cellular components, neurogenic proteins may contribute to neuronal identity (Matter-Sadzinski, 2005). <

During the development of the central nervous system, cell proliferation and differentiation are precisely regulated. In the vertebrate eye, progenitor cells located in the marginal-most region of the neural retina continue to proliferate for a much longer period compared to the ones in the central retina, thus showing stem-cell-like properties. Wnt2b is expressed in the anterior rim of the optic vesicles, and has been shown to control differentiation of the progenitor cells in the marginal retina. Stable overexpression of Wnt2b in retinal explants inhibits cellular differentiation and induces continuous growth of the tissue. Notably, Wnt2b maintained the undifferentiated progenitor cells in the explants even under the conditions where Notch signaling is blocked. Wnt2b downregulates the expression of multiple proneural bHLH genes as well as Notch. In addition, expression of Cath5 under the control of an exogenous promoter suppresses the negative effect of Wnt2b on neuronal differentiation. Importantly, Wnt2b inhibits neuronal differentiation independently of cell cycle progression. It is proposed that Wnt2b maintains the naive state of marginal progenitor cells by attenuating the expression of both proneural and neurogenic genes, thus preventing those cells from launching out into the differentiation cascade regulated by proneural genes and Notch (Kudo, 2005).

Mammalian atonal homologs and cerebellar development

The cerebellum is essential for fine motor control of movement and posture; its dysfunction disrupts balance and impairs control of speech, limb and eye movements. The developing cerebellum consists mainly of three types of neuronal cells: granule cells in the external germinal layer, Purkinje cells, and neurons of the deep nuclei. The molecular mechanisms that underlie the specific determination and the differentiation of each of these neuronal subtypes are unknown. Math1, the mouse homolog of the Drosophila gene atonal, encodes a basic helix-loop-helix transcription factor that is specifically expressed in the precursors of the external germinal layer and their derivatives. Mice lacking Math1 fail to form granule cells and are born with a cerebellum that is devoid of an external germinal layer. Math1 is the first gene that has been shown to be required in vivo for the genesis of granule cells, and hence the predominant neuronal population in the cerebellum (Beb-Arie, 1997).

Cerebellar granule cells (CGC) are the most abundant neurons in the mammalian brain, and an important tool for unraveling molecular mechanisms underlying neurogenesis. Math1 is a bHLH transcription activator that is essential for the genesis of CGC. To delineate the effects of Math1 on CGC differentiation, primary cultures of CGC progenitors were generated and studied from Math1/lacZ knockout mice. Rhombic lip precursors appeared properly positioned, expressed CGC-specific markers, and maintained Math1 promoter activity in vivo and in vitro, suggesting that Math1 is not essential for the initial stages of specification or survival of CGC. Moreover, the continuous activity of Math1 promoter in the absence of MATH1, indicates that MATH1 is not necessary for the activation of its own expression. After 6, but not 3, days in culture, Math1 promoter activity was downregulated in control cultures, but not in cells from Math1 null mice, thus implying that Math1 participates in a negative regulatory feedback loop that is dependent on increased levels of MATH1 generated through the positive autoregulatory feedback loop. In addition, Math1 null CGC did not differentiate properly in culture, and were unable to extend processes. All Notch signaling pathway receptors and ligands tested were expressed in the rhombic lip at embryonic date 14, with highest levels of Notch2 and Jag1. However, Math1-null rhombic lip cells presented conspicuous downregulation of Notch4 and Dll1. Moreover, of the two transcriptional repressors known to antagonize Math1, Hes5 (but not Hes1) was downregulated in Math1-null rhombic lip tissue and primary cultures, and was shown to bind MATH1, thus revealing a negative regulatory feedback loop. Taken together, these data demonstrate that CGC differentiation, but not specification, depends on Math1, which acts by regulating the level of multiple components of the Notch signaling pathway (Gazit, 2004).

The rhombic lip (RL) is an embryonic proliferative neuroepithelium that generates several groups of hindbrain neurons. However, the precise boundaries and derivatives of the RL have never been genetically identified. beta-galactosidase expressed from the Math1 locus in Math1-heterozygous and Math1-null mice was used to track RL-derived cells and to evaluate their developmental requirements for Math1. A Math1-dependent rostral rhombic-lip migratory stream (RLS) was uncovered that generates some neurons of the parabrachial, lateral lemniscal, and deep cerebellar nuclei, in addition to cerebellar granule neurons. A more caudal Math1-dependent cochlear extramural stream (CES) generates the ventral cochlear nucleus and cochlear granule neurons. Similarly, mossy-fiber precerebellar nuclei require Math1, whereas the inferior olive and locus coeruleus do not. It is proposed that Math1 expression delimits the extent of the rhombic lip and is required for the generation of the hindbrain superficial migratory streams, all of which contribute neurons to the proprioceptive/vestibular/auditory sensory network (Wang, 2005).

Over a century ago, His identified the region along the dorsal edge of the fourth ventricle of two-month-old human embryos as the 'Rautenlippe'(rhombic lip, RL) (His, 1891). Present in all vertebrates, the RL is the dorsal-most portion of the hindbrain proliferative neuroepithelium. It can be divided along the long axis of the hindbrain into rostral (rRL) and caudal (cRL) portions that assume dorsal and ventral positions, respectively, as the brainstem bends during development (Wang, 2005).

The rRL is classically thought to generate only cerebellar granule neurons. Granule neuron progenitors migrate superficially from the rRL starting at mouse embryonic day 13 (E13) to form the external granule layer (EGL) on the surface of the cerebellum. These cells proliferate to form the granule neurons, which subsequently descend into the cerebellum. Other cerebellar cell types such as inhibitory Purkinje cells (PCs) and efferent deep nuclear neurons (DNs) begin forming around E10 and are thought to derive from a more ventromedial portion of the ventricular neuroepithelium, classically called the 'ventricular zone'. Curiously, radioactive thymidine labeling suggests that DNs first aggregate in a superficial 'Nuclear Transitory Zone' (NTZ), similar in position to the later-forming EGL, before descending into the deep cerebellum (Wang, 2005).

In contrast, the cRL is thought to generate the entire cochlear nucleus (CN) . The CN is divided anatomically into dorsal (DC) and ventral (VC) subunits containing many different cell types, including cochlear granule neurons, inhibitory cartwheel cells, and deep efferent neurons. These cell types and the organization of the CN is similar to that of the cerebellum (Wang, 2005).

A more caudal portion of the cRL is thought to generate the five main brainstem 'precerebellar' nuclei that relay peripheral sensation and cortical input to the cerebellum. Four of these nuclei project mossy fibers to the cerebellar granule neurons and DNs and are thought to migrate in superficial 'extramural' streams, reminiscent of the EGL. In contrast, the inferior olive precerebellar nucleus (ION) projects climbing fibers to the PCs and is thought to form via an intramural migratory stream. Several recent studies suggest that mossy-fiber neurons may be more developmentally related to cerebellar granule neurons than to ION neurons (Wang, 2005)

Since His' initial analysis, many attempts have been made to further define the RL and to identify the cells that derive from it. Yet, the distinction between the RL and the adjacent ventricular neuroepithelium remains loosely defined by an anatomical bend that forms late in development. Hence, both a precise definition of the RL and a clear demonstration of its derivatives are lacking (Wang, 2005).

The basic helix-loop-helix transcription factor Mouse atonal homolog 1 (Math1, Atoh1) is expressed in the RL as early as E9.5. Math1 is required for the development of the cerebellar granule neurons and at least two of the mossy-fiber nuclei, but not for the ION. Interestingly, these known Math1-dependent RL derivatives do not begin to form until E13, when the RL becomes anatomically distinguishable, even though Math1 expression begins much earlier. Developmental studies show that molecular markers often identify regional domains more precisely than do anatomical landmarks because gene expression often precedes the anatomical changes. Thus, this earlier Math1 expression may identify the extent of the early RL and indicate a role for Math1 in previously unrecognized rhombic-lip derivatives (Wang, 2005).

To better define the rhombic lip and the neurons it generates, a LacZ reporter gene targeted to the Math1 locus was used. The transient LacZ mRNA expression within the RL tags the cells with β-galactosidase protein (β-gal) and allows their migration to be tracked for several days while the β-gal persists. This approach has worked well to trace other cell lineages in the central nervous system. This study demonstrates that Math1 identifies and is necessary for the development of both known and novel superficial migratory neurons of the hindbrain. These neurons include most cells thought to derive from the RL as well as novel RL derivatives. Based on these observations, a genetic definition of the rhombic lip is proposed as the Math1-expressing region of the hindbrain ventricular neuroepithelium and identify several novel rhombic-lip derivatives in the isthmus, pons, and cerebellum (Wong, 2005).

An in vivo-inducible genetic-fate-mapping strategy was used to permanently label cohorts of Math1-positive cells and their progeny that arise in the rhombic lip of the cerebellar primordium during embryogenesis. At stages prior to E12.5, with the exception of the deep cerebellar nuclei, Math1 cells migrate out of the cerebellar primordium into the rostral hindbrain to populate specific nuclei that include cholinergic neurons of the mesopontine tegmental system. Moreover, analysis of Math1-null embryos shows that this gene is required for the formation of some of these nuclei. Around E12.5, granule cell precursors begin to be labeled: first, those that give rise to granule cells that predominantly populate the anterior lobes of the adult cerebellum and later, those that populate progressively more caudally located lobes until labeling of all granule cell precursors is complete by E17. Thus, the cerebellar rhombic lip gives rise to multiple cell types within rhombomere 1 (Machold, 2005).

Inhibitory and excitatory subtypes of cochlear nucleus neurons are defined by distinct bHLH transcription factors, Ptf1a and Atoh1

The cochlear nucleus (CN), which consists of dorsal and ventral cochlear nuclei (DCN and VCN), plays pivotal roles in processing and relaying auditory information to the brain. Although it contains various types of neurons, the origins of the distinct subtypes and their developmental molecular machinery are still elusive. This study reveals that two basic helix-loop-helix transcription factors play crucial roles in specifying neuron subtypes in the CN. Pancreatic transcription factor 1a (Ptf1a) and atonal homolog 1 (Atoh1) were found to be expressed in discrete dorsolateral regions of the embryonic neuroepithelia of the middle hindbrain (rhombomeres 2-5). Genetic lineage tracing using mice that express Cre recombinase from the Ptf1a locus or under the control of the Atoh1 promoter revealed that inhibitory (GABAergic and glycinergic) or excitatory (glutamatergic) neurons of both DCN and VCN are derived from the Ptf1a- and Atoh1-expressing neuroepithelial regions, respectively. In the Ptf1a or Atoh1 null embryos, production of inhibitory or excitatory neurons, respectively, was severely inhibited in the CN. These findings suggest that inhibitory and excitatory subtypes of CN neurons are defined by Ptf1a and Atoh1, respectively and, furthermore, provide important insights into understanding the machinery of neuron subtype specification in the dorsal hindbrain (Fujiyama, 2009).

This study identified two neuroepithelial domains in the dorsal part of the hindbrain (~r2-5) that express Ptf1a and Atoh1, respectively. Using Cre-loxP-based genetic-fate-mapping studies, it was shown that GABAergic and glycinergic inhibitory neurons of both the DCN and the VCN are derived from the Ptf1a domain. Furthermore, two distinct origins were revealed for Gad67-GFP-positive inhibitory neurons in the middle hindbrain (~r2-5), one dorsal and the other ventral. The dorsally located origin corresponds to the Ptf1a neuroepithelial domain. These findings suggest that the Ptf1a neuroepithelial domain generates dorsally produced inhibitory neurons of the middle hindbrain (~r2-5), some of which give rise to inhibitory neurons of the CN (Fujiyama, 2009).

In the Ptf1a-null embryos, Gad67-GFP-labeled inhibitory neurons were hardly produced from the dorsal region, whereas ventrally generated ones were not affected. This confirms that the dorsal origin is precisely coincident with the Ptf1a neuroepithelial domain and suggests that Ptf1a is required for development of these inhibitory neurons (Fujiyama, 2009).

As to the CN, the dorsal region was severely disorganized in the Ptf1a null embryos, whereas the ventral part seemed to be relatively intact. However, even in the ventral part, Gad67-GFP-labeled inhibitory neurons were lost in the mutant embryos. By contrast, the glutamatergic population seemed to be maintained in the mutant CN. As inhibitory neurons were predominantly localized in the DCN, the DCN might be affected more severely than the VCN in the Ptf1a mutants. In the CN primordia at E14.5 and 16.5, increased apoptosis was present in the mutants, probably leading to a severe reduction of Ptf1a-lineage cells in the CN at later developmental stages. These facts indicate that Ptf1a is involved in differentiation and survival of inhibitory neurons in the CN (Fujiyama, 2009).

Consequently, Ptf1a participates in the development of inhibitory neurons in the cerebellum, the spinal cord and the CN. Moreover, ectopically expressed Ptf1a in the dorsal telencephalon neuroepithelium can confer GABAergic characteristics on the progeny neurons. However, the underlying molecular machinery to specify inhibitory fate has not been clarified. By contrast, this gene is also involved in the development of non-inhibitory neurons, such as climbing fiber neurons in the inferior olivary nucleus and some types of amacrine cells in the retina. Altogether, these facts suggest a very complex function of Ptf1a, through which cells can differentiate into inhibitory or other types of neurons, depending on the spatial information in the brain (Fujiyama, 2009).

Short-term lineage analyses in mice with a β-gal reporter inserted in the Atoh1 locus has reported that the Atoh1 neuroepithelial domain dominantly produced neurons of the VCN plus granule cells in the CN. However, the specific subtypes of cells in the Atoh1 lineage were not identified. Cre-loxP recombination-based lineage-trace analysis revealed that the Atoh1 neuroepithelial domain generates glutamatergic excitatory neurons in the CN (Fujiyama, 2009).

Furthermore,the phenotypes of Atoh1 null embryos were analyzed. The structure of the mutant VCN was severely disorganized, with loss of Mafb-positive cells, whereas the DCN was relatively intact. However, even in the relatively intact part of the CN (dorsal CN), granule cells, unipolar brush cells and VGluT2-expressing glutamatergic neurons were severely reduced in the Atoh1 mutants. These findings suggest that Atoh1 is involved in CN glutamatergic neuron development (Fujiyama, 2009).

Inhibitory and excitatory neurons in the CN are derived from distinct regions: the Ptf1a and Atoh1 neuroepithelial domains of the middle hindbrain, respectively. It is possible that this correlation may also be applied to cerebellar development. In the part of the cerebellum that corresponds to the dorsal region of rostral hindbrain (r1), inhibitory and excitatory neurons are produced from the Ptf1a- and Atoh1-expressing neuroepithelial regions, respectively, and their development is dependent on the corresponding bHLH proteins (Fujiyama, 2009).

By contrast, this rule is not applicable to the caudal hindbrain. The Ptf1a neuroepithelial domain in the caudal hindbrain (~r6-8) produces not only inhibitory but also excitatory neurons, such as inferior olivary neurons, the solo source of the climbing fibers. In addition, the dorsal regions of the middle (r2-5) and caudal (r6-8) hindbrain have been termed 'auditory lip' and 'pre-cerebellar lip', respectively, as these two subdomains generate distinct sets of neurons; the former generates CN neurons and the latter pre-cerebellar neurons, such as mossy and climbing fiber neurons. These facts suggest that middle (~r2-5) and caudal (~r6-8) hindbrain subdomains have distinct characteristics. Accordingly, in the middle hindbrain, no intervention of neurogenin 1-expressing neuroepithelial domain was seen between the Ptf1a and Atoh1 domains, which exists in the caudal hindbrain. Overall, throughout the hindbrain regions from r1 to r8, the bHLH type transcription factors, Ptf1a and Atoh1, seem to define neuroepithelial domains along the dorsoventral axis and participate in specifying distinct neuron subtypes according to the rostrocaudal spatial information (Fujiyama, 2009).

The DCN and the VCN have distinct anatomical structures. The DCN exhibits a laminar and cerebellum-like organization that includes a granule cell system. By contrast, the VCN does not have a laminar structure. Therefore, it has been proposed that developmental machinery for these two nuclei is quite different. On the one hand, in this study, it was revealed that neurons with the same electrophysical nature (e.g. inhibitory or excitatory) in both nuclei are derived from the same neuroepithelial domain, which is dorsoventrally defined by a bHLH transcription factor, Ptf1a or Atoh1. On the other hand, it has been shown that VCN and DCN neurons are produced from ~r2-4 and r4, 5 of the hindbrain, although some overlap was observed. It was observed that the Ptf1a domain is relatively narrower in ~r2-3 than ~r4-5. This may account for the smaller number of inhibitory neurons in the VCN that are derived from the Ptf1a domain. Thus, CN cell types seem to be determined according to rostrocaudal and dorsoventral spatial information of the neuroepithelium (Fujiyama, 2009).

The CN contains a variety of neurons, each of which has distinct morphological, histochemical and physiological characteristics. Although this study has clarified the molecular mechanisms to specify electrophysiological subtypes (inhibitory or excitatory) of CN neurons, the question still remains as to how distinct types of inhibitory neurons in the CN (e.g. cartwheel, Golgi, ML-stellate, etc.) are differentially generated from the Ptf1a neuroepithelial domain or how distinct types of excitatory ones (e.g. fusiform, granule, octopus cells) are differentially produced from the Atoh1 domain. Further studies are required to uncover the molecular machinery that specifies each cell type in the CN, which may lead to a better understanding of the common mechanisms that govern development of CNS neurons (Fujiyama, 2009).

Return Atonal Evolutionary homologs part 1/2


atonal: Biological Overview | Regulation | Targets of Activity and Protein Interactions | Developmental Biology | Effects of Mutation | References

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.