Dichaete/Sox box protein 70D

EVOLUTIONARY HOMOLOGS (part 1/2)

Phylogeny of Sox factors

Members of the SOX family of transcription factors are found throughout the animal kingdom, are characterized by the presence of a DNA-binding HMG domain, and are involved in a diverse range of developmental processes. Support is found for subdivision of the family into groups A-H, and for the assignment of two new groups, I and J. For vertebrate genes, it appears that relatedness as suggested by HMG domain sequence is congruent with relatedness as indicated by overall structure of the full-length protein and intron-exon structure of the genes. Most of the SOX groups identified in vertebrates are represented by a single SOX sequence in each invertebrate species studied. An HMG domain signature motif has been identifed which may be considered representative of the SOX family. Based on this data, a robust phylogeny of SOX genes is presented that reflects their evolutionary history in metazoans (Bowles, 2000).

By convention, SOX proteins are more than 50% identical to SRY in the HMG domain. This definition is now inaccurate, with the identification of new SOX genes that do not conform to this rule. For instance, the group H sequence hu-SOX30 is 48% and ce-SOXJ is only 46% identical to hu-SRY in the HMG domain. By comparison, mo-LEF1 is 24% identical to hu-SRY. Based on these figures, it seems that classification based on a strict 50% identity to SRY may not be a suitable indicator of SOX family membership. Reference to the SRY sequence was a historical, arbitrary, and, in retrospect, poor choice for such SOX family comparisons, since SRY has arisen only in the mammalian lineage and is clearly very divergent. It may be more appropriate to compare identity to another SOX sequence or to the SOX consensus sequence. However, even if this is done, a 50% cutoff value may be too stringent (e.g., human SOX30, 46% identity to the SOX consensus). The results provide an alternative criterion to define SOX genes using the conservation of key motifs within the HMG domain. The HMG domain sequence RPMNAF (positions 5-10) appears to be conserved for all SOX sequences, including those of groups H, I, and J, but not for the most closely related outgroups fu-MATA1, mo-LEF1, and mo-TCF1. However, this sequence is also present in a recently defined SOX-like gene in Drosophila, capicua (cic), which has apparent orthologs in C. elegans and humans, suggesting that this 6-amino-acid motif is insufficient to strictly define SOX genes. The extended version, common to all non-SRY SOX members (RPMNAFMVW), appears to be the most reliable signature of the SOX family (Bowles, 2000).

Invertebrate representatives have been discovered for most of the SOX groups thus far identified. C. elegans proteins ce-SOXB1, ce-SOXB2, ce-SOXC, and ce-SOXD (known as COG-2) are associated with subgroups B1 and B2 and groups C and D, respectively. No C. elegans genes encoding proteins with homology to groups E, F, G, H, or I have been detected. An additional C. elegans SOX protein which cannot be assigned to any of the existing groups has been provisionally allocated to a new group J. A putative C. elegans ortholog of LEF/TCF (ce-LEF/TCF) has also been identified. Drosophila SOX proteins dr-SOXB1 (CG18024 or SoxNeuro), dr-SOXB2.1 (dichaete), dr-SOXB2.2, dr-SOXB2.3, dr-SOXC (Sox box protein 14 or CG17263), dr-SOXD, dr-SOXE (Sox100B or CG12098), and dr-SOXF (Sox box protein 15 or CG8404) are associated with groups B1, B2, C, D, E, and F as indicated by their names. No Drosophila sequences have been found for groups G, H, I, or J. Similarly, sea urchin sequences se-SOXB1, se-SOXB2, and se-SOXD1 are associated with subgroups B1 and B2 and group D. It is entirely likely that additional sea urchin SOX genes remain to be identified -- this genome is not yet completely sequenced. For each invertebrate species examined, only one representative sequence has been identified for each group -- the exception to this is group B. This suggests that for each of the currently recognizable SOX groups, a single ancestral form existed before the origin of the vertebrate lineage. In contrast to this general trend, four group B SOX sequences have been identified in Drosophila. These include a single group B1 representative (SoxB1) along with three group B2 genes (SoxB2.1, SoxB2.2, and SoxB2.3). The three SoxB2 genes are physically linked and it is possible that lineage-specific duplication and diversification have occurred in this case. In support of this possibility, the HMG boxes of SoxB2.2 and SoxB2.3 are approximately 50 and 70 kb downstream of the HMG box of SoxB2.1 (dichaete) on chromosome 3. This relatively recent divergence is confirmed by maximum likelihood analysis: the four group B sequences clustered in 100% of analyses (Bowles, 2000).

Based on phylogenetic considerations, it is not possible to define the invertebrate orthologs of specific mammalian genes. It has been suggested that dichaete (here called SOXB2.1) is the Drosophila equivalent of mammalian SOX2. Although mouse SOX2 can functionally substitute for dichaete, the analysis suggests that the Drosophila protein might more reasonably be considered to represent an ancestral form of the entire SOXB or SOXB2 group. dichaete is similar to vertebrate SOX2 sequences only in the HMG domain and a short C-terminal region that does not appear to be essential for the gene's function, suggesting that rescue of the dichaete mutant might be possible also with other vertebrate group B proteins (Bowles, 2000).

Invertebrate Sox factors

In C. elegans, the anchor cell (AC) plays multiple, essential roles during vulval cell development. Among other roles, the AC signals six of twelve granddaughters of the ventral uterine cells, causing them to adopt a fate, pi, different from the default fate of their six sisters and cousins. The six pi precursors (three per side) divide to form twelve pi cells. Eight of the pi cells (four per side) fuse to form the large multinucleate uterine seam cell that lies over the top of the vulva orifice and underlies the developing uterus. The remaining four pi cells (two per side) adopt the uv1 fate and also make connections with vulval cells. In screens for mutants defective in vulval morphogenesis, multiple mutants were isolated in which the uterus and the vulva fail to make a proper connection. Five alleles are described that define the gene cog-2 (connection of gonad defective-2). To form a functional connection between the vulva and the uterus, the AC must fuse with the multinucleate uterine seam cell, derived from uterine cells that adopt a pi lineage. In cog-2 mutants, the anchor cell does not fuse to the uterine seam cell and remains instead at the apex of the vulva, blocking the connection between the vulval and uterine lumens, resulting in an egg-laying defective phenotype. According to lineage analysis and expression assays for two pi-cell-specific markers, induction of the pi fate occurs normally in cog-2 mutants. cog-2 is shown to encode a Sox family transcription factor that is expressed in the pi lineage. Thus, it appears that COG-2 is a transcription factor that regulates a late-stage aspect of uterine seam cell differentiation that specifically affects anchor cell-uterine seam cell fusion (Hanna-Rose, 1999).

Amphioxus (Cephalochordata within the phylum Chordata) , as the closest living invertebrate relative of the vertebrates, can provide insights into the evolutionary origin of the vertebrate body plan. Therefore, to investigate the evolution of genetic mechanisms for establishing and patterning the neuroectoderm, two amphioxus transcription factors, AmphiSox1/2/3 and AmphiNeurogenin were cloned and characterized. These genes are the earliest known markers for presumptive neuroectoderm in amphioxus. Genes in the Sox1/2/3 group are Sry-related HMG box transcription factors most closely related to Drosophila Dichaete. Until the discovery of target of Pox-n (tap), it was thought that Drosophila had no close relative of neurogenin and NeuroD, the nearest relative being atonal, most closely related to vertebrate MATH-1. However, Drosophila tap has sequence affinities with neurogenins, not with NeuroD, suggesting that neurogenin and NeuroD arose by gene duplication before the deuterostome-protostome split. Thus, both amphioxus and Drosophila may have homologs of NeuroD that are yet to be discovered (Holland, 2000).

By the early neurula stage, AmphiNeurogenin expression becomes restricted to two bilateral columns of segmentally arranged neural plate cells, which probably include precursors of motor neurons. This is the earliest indication of segmentation in the amphioxus nerve cord. Later, expression extends to dorsal cells in the nerve cord, which may include precursors of sensory neurons. By the midneurula, AmphiSox1/2/3 expression becomes limited to the dorsal part of the forming neural tube. These patterns resemble those of their vertebrate and Drosophila homologs. Taken together with the evolutionarily conserved expression of the dorsoventral patterning genes of chordates and Drosophila, BMP2/4 and chordin, respectively in nonneural and neural ectoderm, these results are consistent with the evolution of the chordate dorsal nerve cord and the insect ventral nerve cord from a longitudinal nerve cord in a common bilaterian ancestor. However, AmphiSox1/2/3 differs from its vertebrate homologs in not being expressed outside the CNS, suggesting that additional roles for this gene have evolved in connection with gene duplication in the vertebrate lineage. In contrast, expression in the midgut of AmphiNeurogenin, together with the gene encoding the insulin-like peptide, suggest that amphioxus may have homologs of vertebrate pancreatic islet cells, which express neurogenin3. In addition, AmphiNeurogenin, like its vertebrate and Drosophila homologs, is expressed in apparent precursors of epidermal chemosensory and possibly mechanosensory cells, suggesting a common origin for protostome and deuterostome epidermal sensory cells in the ancestral bilaterian (Holland 2000).

AmphiSox1/2/3 is a highly specific marker for presumptive neuroectoderm. Its expression is first detectable by in situ hybridization in the early gastrula (cap-shaped stage) in the dorsal epiblast, including the presumptive neuroectoderm. Expression remains uniformly strong in the presumptive neuroectoderm throughout gastrulation. Whether the expression domain includes any cells in adjacent nonneural ectoderm at these early stages is not clear because of the lack of anatomical markers. However, at the onset of neurulation when the neural plate flattens and becomes distinct from nonneural ectoderm, it is evident that expression is limited to neural ectoderm and is excluded from nonneural ectoderm. Since the edges of the neural plate curl dorsally, expression initially remains panneural. Later, before the edges of the neural plate have fused dorsally, expression becomes down-regulated along the midline and in the anterior part of the CNS. At the midneurula stage, expression is limited to the extreme posterior-dorsal portion of CNS and then ceases entirely. However, in much later larvae at the one- and two-gill-slit stages (2-5 days), prolonged staining reveals extremely weak expression in a few cells scattered in the CNS, rostral to the pigment spot at the level of somite 5. Expression is not detected in tissues other than neuroectoderm (Holland, 2000).

The chordate central nervous system has been hypothesized to originate from either a dorsal centralized, or a ventral centralized, or a noncentralized nervous system of a deuterostome ancestor. In an effort to resolve these issues, the hemichordate Saccoglossus kowalevskii was examined and the expression of orthologs of genes that are involved in patterning the chordate central nervous system was examined. All 22 orthologs studied are expressed in the ectoderm in an anteroposterior arrangement nearly identical to that found in chordates. Domain topography is conserved between hemichordates and chordates despite the fact that hemichordates have a diffuse nerve net, whereas chordates have a centralized system. It is proposed that the deuterostome ancestor may have had a diffuse nervous system, which was later centralized during the evolution of the chordate lineage (Lowe, 2003).

The adult S. kowalevskii has tripartite, tricoelomic organization. At the anterior is the muscular proboscis or prosome, used for burrowing and collecting food particles. It contains the heart, kidney, a section of the dorsal nerve cord, and the protocoel. The middle region, which is the collar or mesosome, contains the mouth, a section of dorsal nerve cord formed by neurulation, the paired mesocoels, and the base of the stomochord, which projects forward into the prosome. The posterior region or metasome contains the gill slits, the remainder of the dorsal nerve cord, the entire ventral nerve cord, paired metacoels, gonads, a long through-gut, and terminal anus. At juvenile stages, a ventral post-anal extension (called a tail or sucker) is present (Lowe, 2003).

Gastrulation entails uniform and simultaneous inpocketing of the vegetal half of the hollow blastula. As the blastopore closes, a gumdrop-shaped gastrula is formed. As the embryo lengthens, two circumferential grooves indent and divide the length into prosome, mesosome, and metasome regions. Mesodermal coeloms outpouch from the gut anteriorly and laterally. The first gill slit pair appears externally by day 5, and the animal bends from the dorsal side. The hatched juvenile elongates and adds further pairs of gill slits successively. The animal is nearly bilaterally symmetric, except that the prosome excretory pore (the proboscis pore) from the kidney is reliably on the left, defining a left-right asymmetry (Lowe, 2003).

The hemichordate adult nervous system is not centralized but is a diffuse intraepidermal, basiepithelial nerve net. Nerve cells are interspersed with epidermal cells and account for 50% or more of the cells in the proboscis and collar ectoderm and a lower percentage in the metasome. Axons form a meshwork at the basal side of the epidermis. The two nerve cords are through-conduction tracts of bundled axons and are not enriched for neurogenesis. This general organizational feature of the nervous system has been largely underemphasized in recent literature that focuses on possible homologies between chordate and hemichordate nerve cords (Lowe, 2003).

Although neurons are dispersed throughout the epidermis in the adult, it has not been demonstrated that neurogenesis in the embryo is uniform. To determine the site of neurogenesis, the domains of expression were localized for three orthologs of pan-neural genes of chordates and Drosophila -- nrp/ musashi, sox1/2/3/ soxneuro, and hu/elav. The first two are markers of proliferating neuron precursors, whereas the third is a marker of differentiating neurons. All are expressed in the neural plate of various chordates, but not in the epidermis. nrp/musashi and sox1/2/3 /soxneuro are expressed in the entire ectoderm of the early S. kowalevskii embryo (except for the ciliated band, which all probes except emx fail to stain). In later stages, the expression remains strong in the prosome and declines in the metasome, correlating with Bullock's observation of decreasing neuron density posteriorly. In sections, weak expression of nrp/musashi can be detected in the posterior endoderm, possibly correlated with a sparse endodermal nerve net. Hu/elav exhibits similar diffuse staining throughout the ectoderm in early stages. Additionally, Hu/elav staining remains strong along the posterior dorsal midline at later stages, in a punctate pattern perhaps reflecting a concentration of early-differentiating nerves at this site. In sagittal sections of embryos, hu/elav expression appears localized toward the basal side of the ectoderm (basiepithelial); it is absent from the mesoderm. Thus, S. kowalevskii shows pervasive neurogenesis with no large, contiguous nonneurogenic subregion, as occurs in chordates (Lowe, 2003).

Fish and Xenopus Sox factors

Waardenburg-Shah syndrome combines the reduced enteric nervous system characteristic of Hirschsprung's disease with reduced pigment cell number, although the cell biological basis of the disease is unclear. A zebrafish Waardenburg-Shah syndrome model has been analyzed. The colourless gene encodes a sox10 homolog, sox10 lesions have been identified in mutant alleles and the mutant phenotype has been rescued by ectopic sox10 expression. Using iontophoretic labelling of neural crest cells, it has been demonstrated that colourless mutant neural crest cells form ectomesenchymal fates. By contrast, neural crest cells which in wild types form non-ectomesenchymal fates generally fail to migrate and do not overtly differentiate. These cells die by apoptosis between 35 and 45 hours post fertilization. Evidence is provided that melanophore defects in colourless mutants can be largely explained by disruption of nacre/mitf expression. It is proposed that all defects of affected crest derivatives are consistent with a primary role for colourless/sox10 in specification of non-ectomesenchymal crest derivatives. This suggests a novel mechanism for the etiology of Waardenburg-Shah syndrome in which affected neural crest derivatives fail to be generated from the neural crest (Dutton, 2001).

From early stages of development, Sox2-class transcription factors (Sox1, Sox2 and Sox3) are expressed in neural tissues and sensory epithelia. Sox2 function is required for neural differentiation of early Xenopus ectoderm. Microinjection of dominant-negative forms of Sox2 (dnSox2) mRNA inhibits neural differentiation of animal caps caused by attenuation of BMP signals. Expression of dnSox2 in developing embryos suppresses expression of N-CAM and regional neural markers. Temporal requirements of Sox2-mediated signaling were analyzed by using an inducible dnSox2 construct fused to the ligand-binding domain of the glucocorticoid receptor. Attenuation of Sox2 function both from the late blastula stage and from the late gastrula stage onwards causes an inhibition of neural differentiation in animal caps and in whole embryos. Additionally, dnSox2-injected cells that fail to differentiate into neural tissues are not able to adopt epidermal cell fate. These data suggest that Sox2-class genes are essential for early neuroectoderm cells to consolidate their neural identity during secondary steps of neural differentiation (Kishi, 2000).

Sox2 signaling is required during secondary stages of neural differentiation, starting from late gastrula. However, it remains unclear whether Sox2 function is necessary for the initial step of neural induction, which occurs around stage 10. Overexpression of Sox2 mRNA per se has little effects on animal cap ectoderm. When combined with FGF, Sox2 can modify the responsiveness of animal cap cells to FGF neuralizing signals. Sox2 alone is not sufficient to direct cells to neural fate but rather plays a role in changing the competence of the ectoderm. When 100 pg of Sox2 mRNA is injected into all the animal cells of 8-cell embryos, overexpression of Sox2 had very weak effects, if any, on the expression of neural markers at neurula stages. Higher doses of Sox2 mRNA causes non-specific effects such as exogastrulation. Therefore, as of yet, there is no particular evidence for instructive roles of Sox2 in early neural development. Sox2 and/or its close relatives are necessary for the ectoderm to develop into neural tissues during secondary steps of neural differentiation. However, it remains to be elucidated exactly which members of the Sox2-subfamily are responsible for this role. Being so closely related in structure and in expression pattern, it is likely that Sox2-subfamily genes have largely redundant functions. Additionally, it also remains to be clarified precisely which signaling steps in the downstream cascade of neural induction require the presence of Sox2-class factors. The transgenic frog technique may prove useful in the study of detailed gene interaction between early regulators of neural differentiation and analysis of the promoters of early neural marker genes. This new technique should provide information about the mode of action of Sox2 on neural-specific transcription (Kishi, 2000).

Cranial placodes, which give rise to sensory organs in the vertebrate head, are important embryonic structures whose development has not been well studied because of their transient nature and paucity of molecular markers. Markers of pre-placodal ectoderm (PPE) (six1, eya1) have been used to determine that gradients of both neural inducers and anteroposterior signals are necessary to induce and appropriately position the PPE. Overexpression of six1 expands the PPE at the expense of neural crest and epidermis, whereas knock-down of Six1 results in reduction of the PPE domain and expansion of the neural plate, neural crest and epidermis. Using expression of activator and repressor constructs of six1 or co-expression of wild-type six1 with activating or repressing co-factors (eya1 and groucho, respectively), it has been demonstrated that Six1 inhibits neural crest and epidermal genes via transcriptional repression and enhances PPE genes via transcriptional activation. Ectopic expression of neural plate, neural crest and epidermal genes in the PPE demonstrates that these factors mutually influence each other to establish the appropriate boundaries between these ectodermal domains (Brugmann, 2004).

These studies predict complex in vivo interactions between six1 and other ectodermal genes. sox2 and sox3 are both induced in presumptive neural ectoderm by neural inductive signaling, and promote stabilization of a neural fate. Later, they are both expressed in neural stem cells. Endogenous six1 expression in the LNE precedes sox2/3 placodal expression, indicating that sox2/3 are not upstream of six1. six1 overexpression in the lateral neurogenic ectoderm (LNE) has no significant effect on sox2/3 neural plate expression, indicating that its effects in this border domain are cell autonomous and not due to intermediate signaling. However, when six1 is reduced in the lateral ectoderm, sox2/3 expression expands laterally. This could be a secondary result of the expansion of foxD3, which in turn expands sox2/3 and/or a mutual antagonism between six1 and sox2/3. The latter possibility is supported by the observations that six1 expression in the neural plate dramatically represses sox2/3 and that expression of sox2 in the LNE represses six1. At later stages sox2/3 are expressed in placodal domains that presumably overlap with six1 expression. How these genes interact at this later phase of placode development remains to be determined (Brugmann, 2004).

Formation of the organizer is one of the most central patterning events in vertebrate development. Organizer-derived signals are responsible for establishing the CNS and patterning the dorsal ventral axis. The mechanisms promoting organizer formation are known to involve cooperation between Nodal and Wnt signalling. However, the organizer forms in a very restricted region, suggesting the presence of mechanisms that repress its formation. This study shows in zebrafish that the transcription factor Sox3 represses multiple steps in the signalling events that lead to organizer formation. Although beta-catenin, Bozozok and Squint are known to play major roles in establishing the dorsal organizer in vertebrate embryos, overexpression of any of these is insufficient to induce robust expression of markers of the organizer in ectopic positions in the animal pole, where Sox3 is strongly expressed. A dominant-negative nuclear localisation mutant of Sox3 can cause ectopic expression of organizer genes via a mechanism that activates all of these earlier factors, resulting in later axis duplication including major bifurcations of the CNS. It was also found that the related SoxB1 factor, Sox19b, can act redundantly with Sox3 in these effects. It therefore seems that the broad expression of these SoxB1 genes throughout the early epiblast and their subsequent restriction to the ectoderm is a primary regulator of when and where the organizer forms (Shih, 2010).

Progenitors in the developing central nervous system acquire neural potential and proliferate to expand the pool of precursors competent to undergo neuronal differentiation. Both the formation and maintenance of neural-competent precursors are regulated by SoxB1 transcription factors, and evidence that their expression is regionally regulated suggests that specific signals regulate neural potential in subdomains of the developing nervous system. The frizzled (Fz) transmembrane receptor Xfz5 selectively governs neural potential in the developing Xenopus retina by regulating the expression of Sox2. Blocking either Xfz5 or canonical Wnt signaling within the developing retina inhibits Sox2 expression, reduces cell proliferation, inhibits the onset of proneural gene expression, and biases individual progenitors toward a nonneural fate, without altering the expression of multiple progenitor markers. Blocking Sox2 function mimics these effects. Rescue experiments indicate that Sox2 is downstream of Xfz5. Thus, Fz signaling can regulate the neural potential of progenitors in the developing nervous system (Van Raay, 2005).

Progenitor cells in the central nervous system must leave the cell cycle to become neurons and glia, but the signals that coordinate this transition remain largely unknown. Wnt signaling, acting through Sox2, promotes neural competence in the Xenopus retina by activating proneural gene expression. This study reports that Wnt and Sox2 inhibit neural differentiation through Notch activation. Independently of Sox2, Wnt stimulates retinal progenitor proliferation and this, when combined with the block on differentiation, maintains retinal progenitor fates. Feedback inhibition by Sox2 on Wnt signaling and by the proneural transcription factors on Sox2 mean that each element of the core pathway activates the next element and inhibits the previous one, providing a directional network that ensures retinal cells make the transition from progenitors to neurons and glia (Agathocleous, 2009).

Wnt/β-catenin signaling acting through Sox2 activates proneural gene expression in the frog retina. This study shows that Wnt and Sox2 inhibit proneural action through Notch, thereby blocking neuronal differentiation. In addition, Wnt signaling stimulates proliferation independently of Sox2, maintaining the progenitor fate, while Sox2 pushes retinal progenitors to Müller glial fates. Concurrent activation of Sox2 and the cell cycle can recapitulate the effects of Wnt in maintaining the retinal precursor cell (RPC) fate. Finally, inhibition of Wnt signaling by Sox2, and of Sox2 by the proneural transcription factors, facilitates a transition from proliferation to differentiation, thereby ensuring that progenitors progress forwards to a differentiated state (Agathocleous, 2009).

These results tie together disparate strands in the function of Wnt/β-catenin and Sox2 signaling as investigated in various vertebrate models. Sox2 both sets up neural potential and inhibits terminal neuronal differentiation. The present study shows that Sox2 plays a central role in suppressing retinal neurogenesis downstream of Wnt/β-catenin signaling, but it enhances Müller glial differentiation and does not maintain progenitors. Similarly, Sox2 overexpression increases Müller cells in mouse retinal explants and promotes the in vitro differentiation of neocortical progenitors into astroglial cells. Notch activation by Sox2 may be involved in this gliogenic effect, as activated Notch signaling promotes gliogenesis. Therefore, either the absence of proneural gene expression or the suppression of proneural activity allows retinal progenitors to adopt the glial fate (Agathocleous, 2009).

The Wnt pathway is activated in the peripheral retina near the ciliary marginal zone in other species besides Xenopus. Yet, Wnt activation in the chick causes cells to be blocked in a proneural-negative progenitor state and in the mouse they assume non-neuronal peripheral fates. In chick and mouse, Wnt signaling does not appear to regulate Sox gene expression; however, the suppression of neurogenesis via activation of Wnt/β-catenin is common to the frog, chick and mammalian retina (Agathocleous, 2009).

There is strong evidence for connections between Wnt/β-catenin, SoxB1 and proneural genes in the regulation of neural differentiation in other tissues. In the zebrafish hypothalamus, canonical Wnt signaling, acting via Sox3, is necessary for the expression of proneural and neurogenic genes. LRP mutant mice exhibit dramatic hypoplasia of the developing neocortex owing to a reduction in neurogenesis as well as in proliferation. Similarly, in the adult hippocampus, Wnt activation promotes both neurogenesis and stem cell proliferation in a dissociable manner, which fits with the explanation that Wnt/β-catenin signaling sets up neuronal potential but then suppresses differentiation and maintains progenitor cells (Agathocleous, 2009).

The results suggest two parallel aspects of the progenitor cell fate: the suppression of neuronal differentiation and the maintenance of proliferative ability, controlled by two branches of Wnt signaling, one of which is Sox2 dependent. This model fits with findings in the spinal cord that Wnt activates proliferation, whereas Sox2 does not. The parallel control of differentiation and proliferation might be a more general feature of Wnt signaling; for example, in the developing limb, Wnt/β-catenin signaling and Sox9 interact to couple proliferation and chondrocyte differentiation (Agathocleous, 2009).

If Sox2 is not mediating the proliferative effects of Wnt/β-catenin signaling, other effectors must be involved. Although exogenous Cyclin E1 was able to cooperate with Sox2 in progenitor maintenance, little or no change was detected in Cyclin E1 retinal expression after Wnt signaling perturbations, nor in the expression of other cell cycle activators including Cyclin D1, Cyclin A2, n-Myc and c-Myc, suggesting that these genes might not be transcriptional targets in the frog retina. Perhaps other genes might function as Wnt-dependent effectors of proliferation here, or perhaps proliferation is regulated through post-transcriptional mechanisms or by changing the mode of progenitor division (Agathocleous, 2009).

Müller cells are transcriptionally very similar to neuroepithelial progenitor cells. They can divide after injury or provision of growth factors, at which point they may return to a neuroepithelial-like state, perhaps through a Wnt-dependent mechanism. These and results therefore suggest that a crucial distinction between RPCs and Müller cells is a Wnt-mediated capacity to proliferate (Agathocleous, 2009).

For the progression from a progenitor to a neuronal fate, both Wnt/β-catenin signaling and Sox2 must be switched off, relieving the inhibition of proneural activity and stopping proliferation. The inhibition of Wnt by Sox2 is likely to take place during retinogenesis, as Sox2 injections do not result in early defects in the specification of retinal progenitor identity. This therefore suggests a negative-feedback mechanism of Sox2 on Wnt signaling. Interestingly, mutations in human SOX2 associated with anophthalmia have been mapped to the C-terminal domain, which normally interacts with β-catenin, resulting in an inability of Sox2 to inhibit canonical Wnt signaling in vitro (Agathocleous, 2009).

For neuronal differentiation to proceed, Sox2 must also be switched off to relieve the inhibition of proneural activity. In the Xenopus retina, it was found that the proneural bHLH transcription factor Xath5 induced a dramatic reduction of the Sox2 protein. In the cortex, a serine protease cleaves Sox2 specifically in neuronal but not glial precursors, thus relieving the block on neurogenesis. It will be interesting to see whether in the retina, proneural genes feed back on Sox2 through this mechanism or through transcriptional repression (Agathocleous, 2009).

Wnt, Sox2 and the proneural genes appear to form a modular circuit in which each step activates the subsequent step and is in turn inactivated by it, driving cells towards differentiation, while limiting the ability of an external proliferation signal, such as Wnt, to continue signaling indefinitely. The relative levels of Wnt, Sox2 and proneural genes determine where a cell lies along the pathway from proliferation to differentiation and whether it assumes a progenitor, glial or neuronal fate. Understanding fully the function of each interaction in the cascade must await a more quantitative analysis of the relationship between the participating factors. This mechanism of transition from one cell state to another by the integration of directional interactions and feedback loops resembles that reported in diverse systems; for example, during sporulation of Bacillus subtilis, where a circuit with five basic nodes displays successive hierarchical gene activations, coupled with negative-feedback loops that switch off the previous state. Further investigations will reveal whether general aspects of the mechanism that is described here are at work in other neural and non-neural tissues, and how this directional pathway integrates with other factors that help to coordinate neuronal proliferation and differentiation (Agathocleous, 2009).

Chicken Sox factors

Group B Sox genes, Sox1, -2 and -3 are known to activate crystallin genes and to be involved in differentiation of lens and neural tissues. Screening of chicken genomic sequences for more Group B Sox genes has identified two additional genes, Sox14 and Sox21. Proteins encoded by Sox14 and Sox21 genes are similar to each other but distinct from those coded by Sox1-3 (subgroup B1) except for the HMG domain and Group B homology immediately C-proximal to the HMG domain. C-terminal domains of SOX21 and SOX14 proteins function as strong and weak repression domains, respectively, when linked to the GAL4 DNA binding domain. These SOX proteins strongly (SOX21) or moderately (SOX14) inhibit activation of delta1-crystallin DC5 enhancer by SOX1 or SOX2, establishing that Sox14 and Sox21 constitute a repressing subgroup (B2) of Group B Sox genes. This provides the first evidence for the occurrence of repressor SOX proteins. Activating (B1) and repressing (B2) subgroups of Group B Sox genes display interesting overlaps of expression domains in developing tissues (e.g. optic tectum, spinal cord, inner ear, alimentary tract, branchial arches). Within each subgroup, most expression domains of Sox1 and -3 are included in those of Sox2 (e.g. CNS, PNS, inner ear), while co-expression of Sox14 and Sox21 occurs in highly restricted sites of the CNS, with the likely temporal order of Sox21 preceding Sox14 (e.g. interneurons of the spinal cord). These expression patterns suggest that target genes of Group B SOX proteins are finely regulated by the counterbalance of activating and repressing SOX proteins (Uchikawa, 1999).

The epibranchial placodes are ectodermal thickenings that generate sensory neurons of the distal ganglia of the branchial nerves. Although significant advances in understanding of neurogenesis from the placodes have recently been made, the events prior to the onset of neurogenesis remain unclear. Chick Sox3 (cSox3) shows a highly dynamic pattern of expression before becoming confined to the final placodes: one pre-otic (geniculate) and three post-otic (one petrosal and two nodose) placodes. A fate-mapping study using lipophilic dyes has revealed that all post-otic placodes arise within a single broad cSox3-positive domain, where cSox3 expression and epithelial thickness are retained only in much smaller final neurogenic placodes. The data presented here suggest that post-otic placodes are remnants of a common primordium defined as a discrete domain of cSox3 expression (Ishii, 2001).

Sox2 expression marks neural and sensory primordia at various stages of development. A 50 kb genomic region of chicken Sox2 was isolated and scanned for enhancer activity utilizing embryo electroporation, resulting in identification of a battery of enhancers. Although Sox2 expression in the early embryonic CNS appears uniform, it is actually pieced together by five separate enhancers with distinct spatio-temporal specificities, including the one activated by the neural induction signals emanating from Hensen's node. Enhancers for Sox2 expression in the lens and nasal/otic placodes and in the neural crest were also determined. These functionally identified Sox2 enhancers exactly correspond to the extragenic sequence blocks conspicuously conserved between chicken and mammals; these blocks are not discernible by sequence comparison among mammals (Uchikawa, 2003).

The nucleotide sequences of the Sox2-flanking regions between chicken and mouse, and between chicken and human were compared using a stringent criterion to assess similarity (>60% identity in a stretch longer than 100 bp). The analysis revealed 25 blocks of sequence highly conserved between chicken and mammals and distributed in the region of analysis. Most remarkably, all ten identified neural and placodal enhancers (except for one) matched these blocks. The sequence alignment of the neural enhancers N-1 to N-5 derived from the three animal species confirms the high degree of conservation, including the putative transcription factor binding sites (Uchikawa, 2003).

Mammalian Sox factors

The mammalian genome contains a family of genes that are related to SRY, the mammalian sex determining gene. The homology is restricted to the region of SRY that encodes a DNA binding motif of the HMG-box class. These genes have been named SOX genes (SRY-related HMG-box genes). SOX3, a member of the human SOX gene family, maps to the X chromosome in the region Xq26-27. A mentally retarded male patient with hemophilia B is deleted for both the Factor IX gene and SOX3. This suggests that SOX3 is not essential for testis formation. The phenotype of the patient and the expression of SOX3 gene in neuronal tissues raises the possibility that this gene is a candidate gene for Borjeson-Forssman-Lehmann, an X-linked mental retardation syndrome (Stevanovic, 1993).

SRY and SOX9, members of the family of high-mobility group (HMG) domain transcription factors, are both essential for testis formation during human embryonic development. The HMG domain is a DNA-binding and DNA-bending motif comprising about 80 amino acid residues. It has been shown that SRY and SOX9 are nuclear proteins. Using normal or mutant SRY-beta-galactosidase and SOX9-beta-galactosidase fusion proteins in transfection studies involving COS-7 cells, two nuclear localization signals (NLSs) have been located within the HMG domains of both proteins that can independently direct the fusion proteins into the nucleus. Only mutational inactivation of both NLS motifs result in complete exclusion of the fusion proteins from the nucleus. The NLS sequences are located at the N and C termini of the HMG domain and are a bipartite NLS motif and a basic cluster NLS motif, respectively. Both NLS motifs are conserved in the HMG domains of other transcription factors. Implications of these results include (1) the apparent dual function of certain basic amino acid residues in the HMG domain of SRY in both DNA binding and in nuclear localization and (2) the possible control of SOX9 in early gonadal differentiation at the level of nuclear translocation (Südbeck, 1997).

cSox21 is a novel member of the Sox gene family of transcription factors. This gene is a member of the subgroup B, which includes Sox1, Sox2 and Sox3. Although all of these genes are expressed predominantly in the nervous system, only cSox21 expression is positionally restricted within the CNS. Longitudinal stripes are seen in the spinal cord; a more complex pattern is seen in the brain. The timing and position of cSox21 expression stripes provide further insight into dorsoventral patterning in the CNS. The expression of cSox21 as well as other genes (such as Delta, Serrate and Pax genes) may play a part in defining the developmental fate of cells along the dorsoventral axis (Rex, 1997).

Sox factors, sex determination and gonad development

Fibroblast growth factor 4 (FGF-4) is a signaling molecule whose expression is essential for postimplantation mouse development and, at later embryonic stages, for limb patterning and growth. The FGF-4 gene is expressed in the blastocyst inner cell mass, and later in distinct embryonic tissues. In tissue culture FGF-4 expression is restricted to undifferentiated embryonic stem (ES) cells and embryonal carcinoma (EC) cell lines. EC cell-specific transcriptional activation of the FGF-4 gene depends on a synergistic interaction between octamer-binding proteins Sox2, a member of the Sry-related Sox factors family. Sox2 can form a ternary complex with either the ubiquitous Oct-1 or the embryonic-specific Oct-3 protein on FGF-4 enhancer DNA sequences. However, only the Sox2/Oct-3 complex is able to promote transcriptional activation. These findings identify FGF-4 as the first known embryonic target gene for Oct-3 as well as for any of the Sox factors, and offers insights into the mechanisms of selective gene activation by Sox and octamer-binding proteins during embryogenesis (Yuan, 1995).

The Sry gene regulates sex determination in rodents and humans. Sry of mouse is expressed by germ cells in the adult testis and by somatic cells in the genital ridge. Transcripts in the former exist as circular RNA molecules of 1.23 kb, which are unlikely to be efficiently translated. SRY mRNA begins in the unique region 5' of the protein coding region and extends several kilobases into the 3' arm of the large inverted repeat that bounds the Sry genomic locus. This transcript, which is very different from that of the human SRY gene, reveals several features which may be involved in translational control. There appears to be two promoters for the Sry gene: a proximal one gives functional transcripts in the genital ridge, and a distal promoter is used in germ cells in the adult testis. Sry transcripts are first detectable just after 10.5 days post coitum, they reach a peak at 11.5 days and then decline sharply so that none are detected 24 hours later. This was compared with anti-Mullerian hormone gene expression, an early marker of Sertoli cells and the first known downstream gene of Sry. Amh expression begins 20 hours after the onset of Sry expression at a time when Sry transcripts are at their peak. While this result does not prove a direct interaction between the two genes, it defines the critical period during which Sry must act to initiate Sertoli cell differentiation (Hacker, 1995).

Murine Sox-3, located on the X chromosome, is most closely related to Sry. The main site of expression of Sox-3 is in the developing CNS, suggesting a role for Sox-3 in neural development. Sox-3, as well as Sox-1 and Sox-2 are expressed in the urogenital ridge, and their protein products are able to bind the same DNA sequence motif as Sry in vitro, but with different affinities. At 11.5 days of development, when Sry is thought to act, the indifferent gonad consists of primordial germ cells that are not required for testis determination, and two bipotential somatic cell lineages that give rise to supporting and steroidigenic cell types. Sry acts within the supporting cell precursors that produce Sertoli cells in the testis and granulosa cells in the ovary. Sox-3 is expressed in the somatic cells of genital ridges at a level equivalent to or greater than Sry. These findings raise two questions: what is the relationship between Sox-3 and Sry, and could Sox-3 also function in sex determination. One simple hypothesis would be that the action of SOX-3 protein on its gene target(s) is a critical step in the normal genetic pathway leading to differentiation of an ovary, and that in a male, SRY protein competes for the same target site(s) (Collignon, 1996).

Fgfs direct embryogenesis of several organs, including the lung, limb, and anterior pituitary. Male-to-female sex reversal occurs in mice lacking Fibroblast growth factor 9 (Fgf9), demonstrating a novel role for FGF signaling in testicular embryogenesis. Fgf9-/- mice also exhibit lung hypoplasia and die at birth. Reproductive system phenotypes range from testicular hypoplasia to complete sex reversal, with most Fgf9-/- XY reproductive systems appearing grossly female at birth. Fgf9 appears to act downstream of Sry to stimulate mesenchymal proliferation, mesonephric cell migration, and Sertoli cell differentiation in the embryonic testis. While Sry is found only in some mammals, Fgfs are highly conserved. Thus, Fgfs may function in sex determination and reproductive system development in many species (Colvin, 2001).

Male and female mouse gonads at embryonic day 11.0 (E11.0) are morphologically identical in different gonads medial to each mesonephros. By E13.5, the testis is twice the size of the ovary and exhibits morphologically complex testicular cords. Three male-specific events are known to direct early testiculogenesis: cell proliferation, cell migration, and testicular cord formation. An increase in proliferation at the coelomic lining of the gonad (the coelomic epithelium) occurs between E11.3 and E12.1. This proliferation gives rise to Sertoli cells (a supporting cell lineage) early on and to interstitial cells throughout this period. Cells contributing to the interstitium, including vascular endothelial cells and peritubular myoid cells, migrate into the testis from the mesonephros and are required for testicular cord formation. Testicular cord development begins at about E12.0 with clustering of Sertoli and germ cells, followed by rearrangement so that Sertoli cells surround the germ cells. Testicular cords isolate male germ cells from interstitial cells, and prevent male germ cells from entering meiosis. Ovarian germ cells, which are not enclosed by supporting cells, progress by E13.5 to the first meiotic division (Colvin, 2001 and references therein).

The testis regulates further male reproductive development. Until E13.5, both sexes have Mullerian and Wolffian ducts in each mesonephros. Sertoli cells produce Mullerian inhibiting substance (MIS). MIS causes regression of the Mullerian ducts, which, in the absence of MIS, form the oviducts, uterus, and upper vagina. Interstitial Leydig cells produce testosterone, which induces formation of Wolffian duct derivatives, including the epididymis, vas deferens, and seminal vesicles. In females, the absence of testicular MIS and testosterone results in development of Mullerian structures and regression of the Wolffian ducts. Targeted deletion of Mis or its receptor results in development of Mullerian structures in XY mice (Colvin, 2001 and references therein).

Testicular expression of Sry, a transcription factor gene on the Y chromosome, is essential for increased proliferation in, and mesonephric cell migration into, the mouse testis. Sry is expressed in mouse testis between E10.5 and E12.5 and is necessary and sufficient to induce male development. Deletion of Sry generates XY ovaries and mice with a female phenotype, and addition of an Sry transgene generates XX males. A potential downstream target of Sry is Sox9, an autosomal transcription factor expressed in Sertoli cells. Mutations in SRY and SOX9 have been identified in human XY females with gonadal dysgenesis (Colvin, 2001 and references therein).

Fgf9 appears to act downstream of Sry, but the signaling relationship between Sox9 and Fgf9 is unclear. Sry is essential for each mode of mesenchymal expansion in the early testis: proliferation and mesonephric cell migration. Thus, reduced mesenchyme in Fgf9-/- XY gonads suggests that Sry and Fgf9 act along the same developmental pathway. Testicular Fgf9 expression begins shortly after the onset of Sry expression at E10.5, consistent with Fgf9 acting downstream of Sry. Some Fgf9-/- XY gonads exhibit aberrant Sox9 expression, but Fgf9 is not required to induce Sox9 expression in the testis or to maintain Sox9 expression through E18.5. Analysis of Fgf9 expression in Sox9-deficient gonads would determine if Sox9 is required to induce testicular Fgf9 expression. Unfortunately, Sox9 heterozygous mice die at birth precluding the generation of homozygous mutant embryos, and embryos derived by introducing Sox9 homozygous mutant ES cells into tetraploid blastocysts, die by E11.5. Correlation between testicular cord formation and Sox9 expression in Fgf9-/- XY gonads suggests that Fgf9 may regulate Sox9 expression indirectly by facilitating testicular development (Colvin, 2001).

Fgf9 affects early steps in testiculogenesis, including Sertoli cell development, gonadal cell proliferation, and mesonephric cell migration. Pre-Sertoli cells originate from multipotential cells in the coelomic epithelium and proliferate at the coelomic epithelium between E11.3-E11.5. Impaired Fgf9-/- Sertoli cell development suggests that Fgf9 could directly induce Sertoli cell specification, proliferation, and/or maintenance of differentiation. Loss of signaling from Sertoli cells could then secondarily impair mesenchymal proliferation and mesonephric cell migration. Full Sertoli cell differentiation probably requires testicular cord formation, and maintenance of Sertoli differentiation may require contact with peritubular myoid cells and the basal lamina. Thus, Fgf9 could also facilitate Sertoli cell differentiation by promoting mesenchymal expansion and testicular cord formation (Colvin, 2001).

Proliferation at the coelomic epithelium gives rise to Sertoli and interstitial cells during an initial burst of proliferation (E11.3-E11.5), and to interstitial cells after this time. Proliferation below the coelomic epithelium in E12.5 Fgf9-/- XY gonads is reduced relative to controls, indicating that Fgf9 is essential for normal proliferation at this stage. Decreased numbers of Sertoli and interstitial cells are observed in Fgf9-/- gonads by E12.5. This, and the onset of testicular Fgf9 expression between E10.5-E11.5, suggests that Fgf9 may mediate the initial stage of proliferation as well (Colvin, 2001).

Mesonephric cell migration into the testis at E11.3-E16.5 contributes to interstitial cell populations, including vascular endothelial, myoepithelial, and peritubular myoid cells. Exogenous FGF9 induces mesonephric cell migration into E11.5 XX gonads, suggesting that FGF9 in the early testis could act as a chemotactic factor for mesonephric cells. When mesonephric migration into XX gonads is artificially induced, XX gonads exhibit testicular cord formation and increased Sox9 expression. Conversely, blocking mesonephric cell migration in culture impairs testicular cord formation, indicating that impaired mesonephric cell migration could contribute to Fgf9-/- sex reversal. Analysis of mesonephric cell migration into Fgf9-/- XY gonads will test this hypothesis. Mesonephric cells that migrate into the testis are proliferating, suggesting that one molecular signal could induce both migration and proliferation. In the embryonic lung, FGF10 stimulates both migration and proliferation of epithelial cells (Colvin, 2001).

Sox is a large family of genes related to the sex-determining region Y gene (designated as the SRY gene), In mammals, Sry expression in the bipotential, undifferentiated gonad directs the support cell precursors to differentiate as Sertoli cells, thus initiating the testis differentiation pathway. In the absence of Sry, or if Sry is expressed at insufficient levels, the support cell precursors differentiate as granulosa cells, thus initiating the ovarian pathway. The molecular mechanisms upstream and downstream of Sry are not well understood. The transcription factor GATA4 and its co-factor FOG2 are required for gonadal differentiation. Mouse fetuses homozygous for a null allele of Fog2 or homozygous for a targeted mutation in Gata4 (Gata4^ki) that abrogates the interaction of GATA4 with FOG co-factors exhibit abnormalities in gonadogenesis. Sry transcript levels are significantly reduced in XY Fog2^–/– gonads at E11.5, which is the time when Sry expression normally reaches its peak. In addition, three genes crucial for normal Sertoli cell function (Sox9, Mis and Dhh) and three Leydig cell steroid biosynthetic enzymes (p450scc, 3ßHSD and p450c17) are not expressed in XY Fog2^–/– and Gata^ki/ki gonads, whereas Wnt4, a gene required for normal ovarian development, is expressed ectopically. By contrast, Wt1 and Sf1, which are expressed prior to Sry and necessary for gonad development in both sexes, are expressed normally in both types of mutant XY gonads. These results indicate that GATA4 and FOG2 and their physical interaction are required for normal gonadal development (Tevosian, 2002).

Sox factors: protein interactions and interaction with DNA

The HMG box domain of the testis determining factor, SRY, includes a basic amphiphilic sequence common to calmodulin (CaM) binding proteins. SRY exhibits calcium-dependent binding to CaM. Binding occurs via the HMG box, and an SRY peptide of residues 57-80 binds CaM like the intact domain. SRY/CaM complex formation is specifically inhibited by the SRY DNA binding site sequence AACAAT. The binding exhibits a 1:1 stoichiometry and is accompanied by a conformational change in SRY. The A domain of HMG1 also binds CaM and it is proposed that CaM binding is a property of the wider HMG box family, including SOX and TCF/LEF proteins. These results suggest that CaM may regulate the DNA binding activity of HMG box transcription factors (Harley, 1996).

A PDZ domain protein (see Drosophila Discs large), termed SIP-1, interacts with human SRY. Interacting domains map to the C-terminal seven amino acids of SRY and to the PDZ domains of SIP-1. SIP-1, possessing two PDZ domains, could connect SRY to other transcription factors providing for SRY, trans-regulation function, as SRY itself has no trans-regulation domain (Poulat, 1997).

SRY, is a DNA binding protein that causes a large distortion of its DNA target sites. The DNA binding domains (HMG-boxes) of mutant SRY proteins have been analyzed from five patients with complete gonadal dysgenesis. The mutant proteins fall into three categories: two bind and bend DNA almost normally, two bind inefficiently but bend DNA normally and one binds DNA with almost normal affinity but produces a different angle. The mutations with moderate effect on complex formation can be transmitted to male progeny, the ones with severe effects on either binding or bending are de novo. The angle induced by SRY depends on the exact DNA sequence and thus adds another level of discrimination in target site recognition. These data suggest that the exact spatial arrangement of the nucleoprotein complex organized by SRY is essential for sex determination (Pontiggia, 1994).

The HMG domain of the SRY protein represents a DNA binding motif that displays rather unusually weak evolutionary conservation of amino acids between human and mouse sequences. The human (h) SRY gene is unable to induce a male phenotype in genetically female transgenic mice. The DNA binding and bending properties of the HMG domains of murine (m) SRY and hSRY differ from each other. In comparison, mSRY shows more-extensive major-groove contacts with DNA and a higher specificity of sequence recognition than hSRY. Moreover, the extent of protein-induced DNA bending differs from the HMG domains of hSRY and mSRY. These differences in DNA binding by hSRY and mSRY may, in part, account for the functional differences observed with these gene products (Giese, 1995).

SOX proteins bind similar DNA motifs through their high-mobility-group (HMG) domains, but their action is highly specific with respect to target genes and cell type. The mechanism of target selection was examined by comparing SOX1/2/3, which activates delta-crystallin minimal enhancer DC5, with SOX9, which activates Col2a1 minimal enhancer COL2C2. These enhancers depend on both the SOX binding site and the binding site of a putative partner factor. The DC5 site is equally bound and bent by the HMG domains of SOX1/2 and SOX9. The activation domains of these SOX proteins mapped at the distal portions of the C-terminal domains are not cell specific and are independent of the partner factor. Chimeric proteins produced between SOX1 and SOX9 show that to activate the DC5 enhancer, the C-terminal domain must be that of SOX1, although the HMG domains are replaceable. The SOX2-VP16 fusion protein, in which the activation domain of SOX2 was replaced by that of VP16, activates the DC5 enhancer still in a partner factor-dependent manner. The results argue that the proximal portion of the C-terminal domain of SOX1/2 specifically interacts with the partner factor, and this interaction determines the specificity of the SOX1/2 action. Essentially the same results were obtained in the converse experiments in which COL2C2 activation by SOX9 was analyzed, except that specificity of SOX9-partner factor interaction also involves the SOX9 HMG domain. The highly selective SOX-partner factor interactions presumably stabilize the DNA binding of the SOX proteins and provide the mechanism for regulatory target selection (Kamachi, 1999).

The Pax6 gene plays crucial roles in eye development and encodes a transcription factor containing both a paired domain and a homeodomain. During embryogenesis, Pax6 is expressed in restricted tissues under the direction of distinct cis-regulatory regions. The head surface ectoderm-specific enhancer of mouse Pax6 directs reporter expression in the derivatives of the ectoderm in the eye, such as lens and cornea, but the molecular mechanism of its control remains largely unknown. A Pax6 protein-responsive element termed LE9 (52 bp in length) has been identified within the head surface ectoderm-specific enhancer. LE9, a sequence well conserved across vertebrates, acted as a highly effective enhancer in reporter analyses. Pax6 protein forms in vitro a complex with the distal half of LE9 in a manner dependent on the paired domain. The proximal half of the LE9 sequence contains three plausible sites of HMG domain recognition, and HMG domain-containing transcription factors Sox2 and Sox3 activate LE9 synergistically with Pax6. A scanning mutagenesis experiment indicates that the central site is most important among the three presumptive HMG domain recognition sites. Furthermore, Pax6 and Sox2 proteins form a complex when they are expressed together. Based on these findings, a model is proposed in which Pax6 protein directly and positively regulates its own gene expression, and Sox2 and Sox3 proteins interact with Pax6 protein, resulting in modification of the transcriptional activation by Pax6 protein (Aota, 2003).

Members of the POU and SOX transcription factor families exemplify the partnerships established between various transcriptional regulators during early embryonic development. Although functional cooperativity between key regulator proteins is pivotal for milestone decisions in mammalian development, little is known about the underlying molecular mechanisms. In this study, focus was placed on two transcription factors, Oct4 and Sox2, since their combination on DNA is considered to direct the establishment of the first three lineages in the mammalian embryo. Using experimental high-resolution structure determination, followed by model building and experimental validation, it was found that Oct4 and Sox2 were able to dimerize onto DNA in distinct conformational arrangements. The DNA enhancer region of their target genes is responsible for the correct spatial alignment of glue-like interaction domains on their surface. Interestingly, these surfaces frequently have redundant functions and are instrumental in recruiting various interacting protein partners (Reményi, 2003).

The interaction of Oct1 and Oct4 with Sox2 was investigated on two different DNA enhancers to test whether a previously discovered regulation mechanism of DNA-mediated swapping of the arrangement of homodimers may also be applicable for unrelated transcription factor assemblies. The crystal structure of the ternary Oct1/Sox2/FGF4 enhancer element complex was solved and then homology modeling tools were used to construct an Oct4/Sox2/FGF4 as well as an Oct4/Sox2/UTF1 structural model. These models reveal that the FGF4 and the Undifferentiated Transcription Factor 1 (UTF1) enhancers mediate the assembly of distinct POU/HMG complexes, leading to different quaternary arrangements by swapping protein-protein interaction surfaces of Sox2. Moreover, it has been demonstrated that Sox2 uses one of its two protein interacting surfaces to assemble a ternary complex with another unrelated transcription factor on a late-embryonic-stage-specific enhancer (Pax6/Sox2 on the DC5 element). These findings outline a simple mechanism for promiscuous yet highly specific assembly of transcription factors, in which the sequence of DNA enhancers governs a combinatorial use of redundant protein-protein interaction surfaces (Reményi, 2003).

Transcriptional targets of Sox factors

Fibroblast growth factor 4 (FGF-4) has been shown to be a signaling molecule whose expression is essential for postimplantation mouse development and, at later embryonic stages, for limb patterning and growth. The FGF-4 gene is expressed in the blastocyst inner cell mass and later in distinct embryonic tissues but is transcriptionally silent in the adult. In tissue culture, FGF-4 expression is restricted to undifferentiated embryonic stem cells and embryonal carcinoma (EC) cell lines. EC cell-specific transcriptional activation of the FGF-4 gene depends on a synergistic interaction between octamer-binding proteins and an EC-specific factor, Fx, that binds adjacent sites on the FGF-4 enhancer. This latter activity is carried out by Sox2, a member of the Sry-related Sox factors family. Sox2 can form a ternary complex with either the ubiquitous Oct-1 or the embryonic-specific Oct-3 protein on FGF-4 enhancer DNA sequences. However, only the Sox2/Oct-3 complex is able to promote transcriptional activation. These findings identify FGF-4 as the first known embryonic target gene for Oct-3 and for any of the Sox factors, and offer insights into the mechanisms of selective gene activation by Sox and octamer-binding proteins during embryogenesis (Yuan, 1995).

Octamer binding and Sox factors are thought to play important roles in development by potentiating the transcriptional activation of specific gene subsets. The proteins within these factor families are related by the presence of highly conserved DNA binding domains, the octamer binding protein POU domain or the Sox factors HMG domain. Fibroblast growth factor 4 (FGF-4) gene expression in embryonal carcinoma cells requires a synergistic interaction between Oct-3 and Sox2 on the FGF-4 enhancer. Sox2 and Oct-3 bind to adjacent sites within this enhancer to form a ternary protein-DNA complex (Oct-3*) whose assembly correlates with enhancer activity. Increasing the distance between the octamer and Sox binding sites by base pair insertion results in a loss of enhancer function. Significantly, those enhancer "spacing mutants" which fail to activate transcription are also compromised in their ability to form the Oct* complexes even though they can still bind both Sox2 and the octamer binding proteins, suggesting that a direct interaction between Sox2 and Oct-3 is necessary for enhancer function. Consistent with this hypothesis, Oct-3 and Sox2 can participate in a direct protein-protein interaction in vitro in the absence of DNA, and both this interaction and assembly of the ternary Oct* complexes require only the octamer protein POU and Sox2 HMG domains. Assembly of the ternary complex by these two protein domains occurs in a cooperative manner on FGF-4 enhancer DNA, and the loss of this cooperative interaction contributes to the defect in Oct-3* formation observed for the enhancer spacing mutants. These observations indicate that Oct-3* assembly results from protein-protein interactions between the domains of Sox2 and Oct-3 that mediate their binding to DNA, but it also requires a specific arrangement of the binding sites within the FGF-4 enhancer DNA. Thus, these results define one parameter that is fundamental to synergistic activation by Sox2 and Oct-3 and further emphasize the critical role of enhancer DNA sequences in the proper assembly of functional activation complexes (Ambrosetti, 1997).

Embryonic development requires a complex program of events that are directed by a number of signaling molecules whose expression must be rigorously regulated. Expression of Fgf4, which plays an important role in postimplantation development and growth and patterning of the limb, is regulated in embryonal carcinoma (EC) cells by the synergistic interaction of Sox2 and Oct-3 with the Fgf4 EC cell-specific enhancer. To verify whether this mechanism is also operating in vivo, and to identify new elements controlling Fgf4 gene expression in distinct developmental stages, the expression of LacZ reporter plasmids containing different fragments of the Fgf4 gene have been analyzed in transgenic mouse embryos. Utilizing these transgenic constructs Fgf4 gene expression could be recapitulated, for the most part, during embryonic development. Most of the cis-acting regulatory elements determining Fgf4 embryonic expression are located in conserved regions within the 3' UTR of the gene. The EC cell-specific enhancer is required to drive gene expression in the ICM of the blastocyst, and its activity requires the Sox and Oct-proteins binding sites. Specific and distinct enhancer elements could be identified that govern postimplantation expression in the somitic myotomes and the limb bud AER. The myotome-specific elements contain binding sites for bHLH myogenic regulatory factors, which appear to be essential for myotome expression. Evidence is also presented that the very restricted pattern of expression of Fgf4 transcripts in the AER results from the combined action of positive and negative regulatory elements located 3' to the Fgf4 coding sequences. Thus the Fgf4 gene relies on multiple and distinct regulatory elements to achieve stage- and tissue-specific embryonic expression (Fraidenraich, 1998).

Delta 1-crystallin gene activation occurs early in lens cell differentiation. An essential element of the delta 1-crystallin enhancer is bound by chicken SOX-2 protein (cSOX-2), which is structurally related to the sex-determining factor SRY. Sox-2 is expressed at high levels in the early developing lens in both chicken and mouse embryos. Overexpression of delta cSOX-2 increased delta 1-crystallin enhancer activity to a plateau in lens cells, but not in fibroblasts, consistent with the previously drawn conclusion that SOX-2 activates transcription only in concert with another factor present in the lens. This result supports the model that SOX proteins act as architectural components in the activating complex formed on an enhancer, as indicated for lymphoid enhancer binding factor 1 (LEF-1), another HMG domain protein. SOX protein binding is essential for lens-specific promoter activity of the mouse gamma F-crystallin gene. This work is the first to show delta- and gamma-crystallin genes as examples of direct regulatory targets of SOX proteins and provides evidence that diversified crystallin genes are regulated, at least in part, by a common mechanism (Kamachi, 1995).

Gamma-crystallins are major structural components of the lens fiber cells in amphibians and mammals. Many dominant inherited cataracts in humans and mice have been shown to map within the gamma-crystallin gene cluster. Several transcription factors, including PAX6 and SOX proteins, have been suggested as candidates for crystallin gene regulation. The targeted deletion of Sox1 in mice causes microphthalmia and cataract. Mutant lens fiber cells fail to elongate, probably as a result of an almost complete absence of gamma-crystallins. It appears that the direct interaction of the SOX1 protein with a promoter element conserved in all gamma-crystallin genes is responsible for their expression (Nishiguchi, 1998).

The POU transcription factor Oct-4, which has no known Drosophila homolog, is expressed specifically in the germ line, pluripotent cells of the pregastrulation embryo and stem cell lines derived from the early embryo. Osteopontin (OPN) is a protein secreted by cells of the preimplantation embryo and contains a GRGDS motif that can bind to specific integrin subtypes and modulate cell adhesion/migration. Oct-4 and OPN are coexpressed in the preimplantation mouse embryo and during differentiation of embryonal cell lines. Immunoprecipitation of the first intron of OPN (i-opn) from covalently fixed chromatin of embryonal stem cells by Oct-4-specific antibodies indicates that Oct-4 binds to this fragment in vivo. The i-opn fragment functions as an enhancer in cell lines that resemble cells of the preimplantation embryo. It contains a novel palindromic Oct factor recognition element (PORE) that is composed of an inverted pair of homeodomain-binding sites separated by exactly 5 bp (ATTTG +5 CAAAT). POU proteins can homo- and hetero-dimerize on the PORE in a configuration that has not been described previously. Strong transcriptional activation of the OPN element requires an intact PORE. In contrast, the canonical octamer overlapping with the downstream half of the PORE is not essential. Sox-2 is a transcription factor that contains an HMG box and is coexpressed with Oct-4 in the early mouse embryo. Sox-2 represses Oct-4 mediated activation of i-opn by way of a canonical Sox element that is located close to the PORE. Repression depends on a carboxy-terminal region of Sox-2 that is outside of the HMG box. Expression, DNA binding, and transactivation data are consistent with the hypothesis that OPN expression is regulated by Oct-4 and Sox-2 in preimplantation development (Botquin, 1998).

Early neural patterning along the anteroposterior (AP) axis appears to involve a number of signal transducing pathways, but the precise role of each of these pathways for AP patterning and how they are integrated with signals that govern neural induction step is not well understood. The nature of Fgf response element (FRE) has been investigated in a posterior neural gene, Xcad3 (Xenopus caudal homolog), which plays a crucial role of posterior neural development. Evidence suggests that FREs of Xcad3 are widely dispersed in its intronic sequence and that these multiple FREs comprise Ets-binding and Tcf/Lef-binding motifs that lie in juxtaposition. Functional and physical analyses indicate that signaling pathways of Fgf, Bmp and Wnt are integrated on these FREs to regulate the expression of Xcad3 in the posterior neural tube through positively acting Ets and Sox family transcription factors and negatively acting Tcf family transcription factor(s) (Haremaki, 2003).

Sox2 is de-repressed by Bmp antagonists in the neurogenic region of ectoderm during neural induction. Sox2, which shares a cognate DNA bindings motif with Tcf/Lef family members, is required as a co-activator for the Fgf response of Xcad3. Sox2 is likely to compete with XTcf3 for TLBMs in the composite FREs to cooperate with Ets proteins that bind to adjacent EBMs. Physical analysis supports this idea. Both Sox and Ets family transcription factors interact with specific partner factors to direct signals to target genes, but direct partnership between them has not been reported. Collectively, these results indicate that signaling pathways of Fgf, Bmp and Wnt are integrated on the FREs to regulate the expression of Xcad3 in the posterior neural tube through positively acting Ets and Sox proteins and negatively acting Tcf protein (Haremaki, 2003).

Sox factors and development

Continued: Dichaete Evolutionary homologs part 2/2

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.