orthodenticle

EVOLUTIONARY HOMOLOGS

Table of contents

Orthodenticle homologs and the pituitary gland

Otx1, a homeobox-containing gene of the Otx gene family, is postnatally transcribed and translated in the pituitary gland. Cell culture experiments indicate that Otx1 may activate transcription of the growth hormone (GH), follicle-stimulating hormone (betaFSH), luteinizing hormone (betaLH) and (alpha)-glycoprotein subunit ([alpha]GSU) genes. Analysis of Otx1 null mice indicates that at the prepubescent stage they exhibit transient dwarfism and hypogonadism due to low levels of pituitary GH, FSH and LH hormones which, in turn, dramatically affect downstream molecular and organ targets. Nevertheless, Otx1(-/-) mice gradually recover from most of these abnormalities, showing normal levels of pituitary hormones with restored growth and gonadal function at 4 months of age. Expression patterns of related hypothalamic and pituitary cell type restricted genes, growth hormone releasing hormone (GRH), gonadotropin releasing hormone (GnRH) and their pituitary receptors (GRHR and GnRHR) suggest that, in Otx1(-/-) mice, hypothalamic and pituitary cells of the somatotropic and gonadotropic lineages appear unaltered. What is affected is the ability to synthesize GH, FSH and LH, rather than the number of cells producing these hormones. These data indicate that Otx1 is a new pituitary transcription factor involved at the prepubescent stage in the control of GH, FSH and LH hormone levels and suggest that a complex regulatory mechanism might exist to control the physiological need for pituitary hormones at specific postnatal stages (Acampora, 1998a).

The pituitary gland contains six distinct hormone-producing cell types that arise sequentially during organogenesis. The first cells to differentiate are those that express the pro-opiomelanocortin (POMC) gene in the anterior pituitary lobe. The other lineages, which appear later, include cells that are dependent on the POU factor Pit-1 and another POMC-expressing lineage in the intermediate pituitary lobe. Using AtT-20 cells as a model for early expression of POMC in the anterior pituitary, a regulatory element has been defined that confers cell specificity of transcription and a cognate transcription factor has been cloned. This factor, Ptx1 (pituitary homeo box 1; see Drosophila Ptx1), contains a homeo box related to those of the anterior-specific genes bicoid and orthodenticle in Drosophila, and Otx-1 and Otx-2 in mammals. Ptx1 activates transcription upon binding a sequence related to the Drosophila Bicoid target sites. Ptx1 is the only nuclear factor of this DNA-binding specificity that is detected in AtT-20 cells, and it is expressed at high levels in a subset of adult anterior pituitary cells that express POMC. However, Ptx1 is expressed in most cells of Rathke's pouch at an early time during pituitary development and before final differentiation of hormone-producing cells. Thus, Ptx1 may have a role in differentiation of pituitary cells; its early expression pattern suggests that it may have a role in pituitary formation. In the adult pituitary gland, Ptx1 appears to be recruited for cell-specific transcription of the POMC gene (Lamonerie, 1996).

A novel OTX-related homeodomain transcription factor has been identified on the basis of its ability to interact with the transactivation domain of the pituitary-specific POU domain protein, Pit-1. This factor, referred to as P-OTX (pituitary OTX-related factor), is expressed in primordial Rathke's pouch, oral epithelium, first branchial arch, duodenum, and hindlimb. In the developing anterior pituitary, it is expressed in all regions from which cells with distinct phenotypes will emerge in the mature gland. P-OTX is able to independently activate and to synergize with Pit-1 on pituitary-specific target gene promoters. Therefore, P-OTX may subserve functions in generating both precursor and specific cell phenotypes in the anterior pituitary gland and in several other organs (Szeto, 1996).

Orthodenticle homologs and ear development

The Otx1 and Otx2 genes are two murine orthologues of the Orthodenticle (Otd) gene in Drosophila. In the developing mouse embryo, both Otx genes are expressed in the rostral head region and in certain sense organs such as the inner ear. Previous studies have shown that mice lacking Otx1 display abnormal patterning of the brain, whereas embryos lacking Otx2 develop without heads. The inner ears of mice lacking both Otx1 and Otx2 genes were studied at different developmental stages. In wild-type inner ears, Otx1, but not Otx2, is expressed in the lateral canal and ampulla, as well as part of the utricle. Ventral to the mid-level of the presumptive utricle, Otx1 and Otx2 are co-expressed, in regions such as the saccule and cochlea. Paint-filled membranous labyrinths of Otx1-/- mutants show an absence of the lateral semicircular canal, lateral ampulla, utriculosaccular duct and cochleosaccular duct, and a poorly defined hook (the proximal part) of the cochlea. Defects in the shape of the saccule and cochlea are variable in Otx1-/- mice and are much more severe in an Otx1-/-;Otx2(+/)- background. Histological and in situ hybridization experiments of both Otx1-/- and Otx1-/-;Otx2(+/)- mutants reveal that the lateral crista is absent. In addition, the maculae of the utricle and saccule are partially fused. In mutant mice in which both copies of the Otx1 gene are replaced with a human Otx2 cDNA (hOtx2(1)/ hOtx2(1)), most of the defects associated with Otx1-/- mutants are rescued. However, within the inner ear, hOtx2 expression fails to rescue the lateral canal and ampulla phenotypes, and only variable rescues are observed in regions where both Otx1 and Otx2 are normally expressed. These results suggest that both Otx genes play important and differing roles in the morphogenesis of the mouse inner ear and the development of its sensory organs (Morsli, 1999).

Otx2 is a target of N-myc and acts as a suppressor of sensory development in the mammalian cochlea

Transcriptional regulatory networks are essential during the formation and differentiation of organs. The transcription factor N-myc is required for proper morphogenesis of the cochlea and to control correct patterning of the organ of Corti. The Otx2 gene, a mammalian orthologue of the Drosophila orthodenticle homeobox gene, is a crucial target of N-myc during inner ear development. Otx2 expression is lost in N-myc mouse mutants, and N-myc misexpression in the chick inner ear leads to ectopic expression of Otx2. Furthermore, Otx2 enhancer activity is increased by N-myc misexpression, indicating that N-myc may directly regulate Otx2. Inactivation of Otx2 in the mouse inner ear leads to ectopic expression of prosensory markers in non-sensory regions of the cochlear duct. Upon further differentiation, these domains give rise to an ectopic organ of Corti, together with the re-specification of non-sensory areas into sensory epithelia, and the loss of Reissner's membrane. Therefore the Otx2-positive domain of the cochlear duct shows a striking competence to develop into a mirror-image copy of the organ of Corti. Taken together, the work shows that Otx2 acts downstream N-myc and is essential for patterning and the spatial restriction of the sensory domain of the mammalian cochlea (Vendrell, 2015).

Ectodermal patterning by mutual repression between Gbx2 and Otx2 in sensory placodes

In the vertebrate head, central and peripheral components of the sensory nervous system have different embryonic origins, the neural plate and sensory placodes. This raises the question of how they develop in register to form functional sense organs and sensory circuits. This study shows that mutual repression between the homeobox transcription factors Gbx2 and Otx2 patterns the placode territory by influencing regional identity and by segregating inner ear and trigeminal progenitors. Activation of Otx2 targets is necessary for anterior olfactory, lens and trigeminal character, while Gbx2 function is required for the formation of the posterior otic placode. Thus, like in the neural plate antagonistic interaction between Otx2 and Gbx2 establishes positional information thus providing a general mechanism for rostro-caudal patterning of the ectoderm. These findings support the idea that the Otx/Gbx boundary has an ancient evolutionary origin to which different modules were recruited to specify cells of different fates (Steventon, 2012).

To form a functional nervous system its peripheral and central components must develop in register. In the head, the olfactory bulb, the retina and the targets and proximal parts of the sensory ganglia are derived from the central nervous system, while the olfactory epithelium, the lens, inner ear and distal cranial ganglia arise in the non-neural ectoderm from specialized structures, the sensory placodes. How is anterior-posterior patterning between both territories integrated? During development sensory placode precursors originate in the pre-placodal region, where cells of different fates are initially intermingled. Over time, they acquire distinct rostro-caudal identity leading to the alignment with their central counterparts suggesting that a global patterning mechanism imparts positional information to the entire ectoderm. This study shows that the transcription factors Otx2 and Gbx2 are important components of such a mechanism. In the pre-placodal region (PPR), they segregate otic and trigeminal progenitors, while they establish a compartment boundary at the midbrain-hindbrain boundary (MHB) in the neural plate and prevent mixing of cells with different fates. In both regions, Otx2 and Gbx2 seem to play a dual role: they repress each other to endow cells with unique identities and to suppress the alternative fate (trigeminal vs otic; midbrain vs rhombomere1), while simultaneously mediating sorting. Initially, both genes partially overlap and mutual repression at the transcriptional level is likely to form a gene expression the boundary. Subsequently, cell sorting ensures compartmentalization to restrict cells of the same fate to a contiguous domain. Accordingly, in the brain, Otx2 deficient cells segregate from wild type neighbors as do cells expressing exogenous Otx2 in rhombomere. Likewise, the results show that Otx2 and Gbx2 expressing cells sort out in the non-neural ectoderm. The degree of cell mixing in the placode territory is still under debate with more cell mixing observed in chick than in Xenopus. As fate maps may introduce some error due to variability between different embryos, ultimately live imaging over long time periods will be required to resolve this question. Nevertheless, together with previous studies on neural and neural crest cells the current findings establish cross-regulatory interactions between Otx2 and Gbx2 as key components for global ectodermal patterning. Both factors establish anterior-posterior identity across the embryonic ectoderm and mediate cell sorting to segregate cells of different fates (Steventon, 2012).

These observations also suggest that signals that establish anterior-posterior identity not only pattern the neural plate, but the entire ectoderm with transcription factors like Otx2 and Gbx2 as a read-out. Among these Fgfs, Wnts, Retinoic Acid, Nodals and BMPs provide posteriorizing factors, while their antagonists protect anterior identity. Indeed, elevated Wnt activity in zebrafish leads to an expansion of posterior neural and placodal fates. Wnt signaling also promotes derivatives of the posterior neural plate border, neural crest cells, and Gbx2, a direct Wnt target, mediates its activity. In addition to such global patterning mechanisms local signaling and downstream transcriptional networks subsequently fine tune allocation of different cell fates (Steventon, 2012).

In the PPR, the Otx2/Gbx2 boundary roughly separates prospective otic and trigeminal fates suggesting that olfactory, lens and trigeminal precursors receive different transcriptional inputs from otic progenitors. The transcriptional regulation of the PPR marker Six1 supports this idea. Although Six1 is expressed in a contiguous domain containing all sensory progenitors, different enhancers control its expression along the anterior-posterior axis. Cells from the anterior Six1 domain contribute to the olfactory, lens and trigeminal placodes, but not to the otic. These findings suggest that one of the first subdivisions of the placode territory occurs between trigeminal and otic precursors clearly grouping trigeminal precursors together with other anterior placodes unlike an earlier suggestion to group profundal and trigeminal placodes in Xenopus with posterior progenitors. Shortly thereafter, the PPR begins to express other transcription factors in nested domains to subdivide this territory further (Steventon, 2012).

Otx2 and Gbx2 are already expressed at gastrula stagesand act early during placode specification. Gbx2 is required for the onset of otic-specific genes, where it appears to act as transcriptional activator: the constitutive repressor Gbx2-EnR mimics MO-mediated knock-down. This is in contrast to its earlier role as repressor during boundary formation suggesting that the availability of cofactors determines the final outcome as observed for other homeobox factors. After initial specification, otic development is Gbx2 independent, although it is later involved in ear morphogenesis. The lack of an early ear phenotype in Gbx2 mutant mice is likely due to functional redundancy with Gbx1. In contrast, Otx2 is necessary for both formation and maintenance of lens and olfactory identity consistent with its continued expression in both placodes. In the trigeminal placode, Otx2 is downregulated shortly after its specification probably due to repression by Pax3, which also inhibits Pax6 in this territory. Like Gbx2, Otx2 switches from a transcriptional repressor at early stages to an activator later. In summary, like in the neural plate, in the PPR Otx2 and Gbx2 are among the earliest factors that subdivide a contiguous territory along the anterior-posterior axis (Steventon, 2012).

Although Otx2 and Gbx2 are required for early placode specification, neither factor alone is sufficient to endow cells with new regional character or to induce ectopic placodes. This appears to differ considerably from their activity in the neural tube, where ectopic expression of either factor respecifies anterior-posterior identity . However, here Otx2 and Gbx2 mainly function to position the MHB, an organizer region that itself patterns the brain. Thus, changes in regional identity are likely to be a consequence of MHB induction. Whether a similar organizing center forms at the Otx2/Gbx2 boundary in the PPR remains to be established, however, so far the results argue against this notion. The finding that neither Gbx2 nor Otx2 is sufficient to induce ectopic placodes suggests that additional factors cooperate to control the expression of placode-specific downstream targets. This is indeed the case in the lens, where Otx2 directly binds to the lens-specific FoxE3 enhancer and together with intracellular effectors of Notch signaling activates its transcription (Steventon, 2012).

The development of cranial sensory placodes and neural crest is considered to be a key step in the evolution of the vertebrate head. Like in vertebrates Gbx and Otx form a boundary within the Amphioxus ectoderm raising the question whether, at an early stage of their evolution, neural crest and placodes co-opted an already existing gene expression boundary to position themselves along the anterior-posterior axis. Despite Gbx2/Otx2 apposition in Amphioxus, MHB specific genes such as En, Wnt1, FGF8/17/18 and Pax2/5/8 are not restricted to this boundary, indicating that MHB organizer genes were recruited to the Otx/Gbx border in early vertebrates. A Gbx/Otx boundary appears to have been present in the early bilaterian ancestor as Unpg/Gbx and Otd/Otx also negatively regulate one another to form a boundary that positions En and Pax2/5/8 in Drosophila. In addition, Gbx2 and Otx2 form a boundary in the annelid Platynereis dumerilii that corresponds to a band of En expression. Together these findings raise the possibility that Otx2 and Gbx2 form an ancient boundary of gene expression responsible for anterior-posterior patterning of both the neural plate and neural plate border. However, this boundary has been utilized differently in each territory: to position an organizing region at the MHB, and to specify placodal fates in the PPR (Steventon, 2012).

Otx2 and endoderm development

Genetic and embryological experiments have demonstrated an essential role for the visceral endoderm in the formation of the forebrain; however, the precise molecular and cellular mechanisms of this requirement are poorly understood. Lineage tracing in combination with molecular marker studies has been used to follow morphogenetic movements and cell fates before and during gastrulation in embryos mutant for the homeobox gene Otx2. These results show (1) that Otx2 is not required for proliferation of the visceral endoderm, but is essential for anteriorly directed morphogenetic movement; (2) molecules that are normally expressed in the anterior visceral endoderm, such as Lefty1 and Mdkk1, are not expressed in Otx2 mutants. These secreted proteins have been reported to antagonise, respectively, the activities of Nodal and Wnt signals, which have a role in regulating primitive streak formation. The visceral endoderm defects of the Otx2 mutants are associated with abnormal expression of primitive streak markers in the epiblast, suggesting that anterior epiblast cells acquire primitive streak characteristics. Taken together, these data support a model whereby Otx2 functions in the anterior visceral endoderm to influence the ability of the adjacent epiblast cells to differentiate into anterior neurectoderm, indirectly, by preventing them from coming under the influence of posterior signals that regulate primitive streak formation (Perea-Gomez, 2001).

The development of inner ear innervation in mice was investigated in Otx1 null mutants, which lack a horizontal canal. Comparable to control animals, horizontal crista-like fibers are found to cross over the utricle in Otx1 null mice. In mutants these fibers extend toward an area near the endolymphatic duct, not to a horizontal crista. Most Otx1 null mutants had a small patch of sensory hair cells at this position. Measurement of the area of the utricular macula suggested it to be enlarged in Otx1 null mutants. Parts of the horizontal canal crista appear to remain incorporated in the utricular sensory epithelium in Otx1 null mutants. Other parts of the horizontal crista appear to be variably segregated to form the isolated patch of hair cells identifiable by the unique fiber trajectory as representing the horizontal canal crista. Comparison with lamprey ear innervation reveals similarities in the pattern of innervation with the dorsal macula, a sensory patch of unknown function. Scanning electron microscope data confirm that all foramina are less constricted in Otx1 null mutants. It is proposed that Otx1 is not directly involved in sensory hair cell formation of the horizontal canal but affects the segregation of the horizontal canal crista from the utricle. It also affects constriction of the two main foramina in the ear, but not their initial formation. Otx1 is thus causally related to horizontal canal morphogenesis as well as morphogenesis of these foramina (Fritzsch, 2001).

Inner ear sensory organs and VIIIth cranial ganglion neurons of the auditory/vestibular pathway derive from an ectodermal placode that invaginates to form an otocyst. In the mouse otocyst epithelium, Tbx1 suppresses neurogenin 1-mediated neural fate determination and is required for induction (Otx1) or proper patterning of gene expression (Bmp4) related to sensory organ morphogenesis. Tbx1 loss-of-function causes dysregulation of neural competence in otocyst regions linked to the formation of either mechanosensory or structural sensory organ epithelia. Subsequently, VIIIth ganglion rudiment form is duplicated posteriorly, while the inner ear is hypoplastic and shows neither a vestibular apparatus nor a coiled cochlear duct. It is proposed that Tbx1 acts in the manner of a selector gene to control neural and sensory organ fate specification in the otocyst (Raft, 2004).

Genetic and embryological experiments demonstrated that the visceral endoderm (VE) is essential for positioning the primitive streak at one pole of the embryo and head morphogenesis through antagonism of the Wnt and Nodal signaling pathways. The transcription factor Otx2 is required for VE anteriorization and specification of rostral neuroectoderm at least in part by controlling the expression of Dkk1 and Lefty1. This study investigated the relevance of the Otx2 transcriptional control in these processes. Otx2 protein is encoded by different mRNAs variants, which, on the basis of their transcription start site, may be distinguished in distal and proximal. Distal isoforms are prevalently expressed in the epiblast and neuroectoderm, while proximal isoforms prevalently in the VE. Selective inactivation of Otx2 variants reveals that distal isoforms are not required for gastrulation, but essential for maintenance of forebrain and midbrain identities; conversely, proximal isoforms control VE anteriorization and, indirectly, primitive streak positioning through the activation of Dkk1 and Lefty1. Moreover, in these mutants the expression of proximal isoforms is not affected by the lack of distal mRNAs and vice versa. Taken together these findings indicate that proximal and distal isoforms, whose expression is independently regulated in the VE and epiblast-derived neuroectoderm, functionally cooperate to provide these tissues with the sufficient level of Otx2 necessary to promote a normal development. Furthermore, in the VE the expression of Otx2 isoforms is tightly controlled at single cell level, and it is hypothesized that this molecular diversity may potentially confer specific functional properties to different subsets of VE cells (Acampora, 2009).

Otx2 and facial development

Mice heterozygous for the Otx2 mutation display a craniofacial malformation, known as otocephaly or agnathia-holoprosencephaly complex. The severity of the phenotype is dependent on the genetic background of a C57BL/6 (B6) strain; most of the offspring of Otx2 knock-out chimeras, which are equivalent to the F₁ of CBA and B6 strains, backcrossed with B6 females display reduction or loss of mandible, whereas those backcrossed with CBA females do not show noticeable phenotype at birth. The availability of phenotypically disparate strains renders identification of Otx2 modifier loci possible. In this study, a backcross of chimera with B6 was generated and genome-wide scans were conducted with polymorphic markers for non-Mendelian distribution of alleles in Otx2 heterozygous mutant mice displaying abnormalities in the lower jaw. One significant locus, Otmf18, was identified between D18Mit68 and D18Mit120 on chromosome 18, linked to the mandibular phenotype (LOD score 3.33). A similar replication experiment using a second backcross (N3) mouse demonstrated the presence of another significant locus, Otmf2 between D2Mit164 and D2Mit282 on chromosome 2, linked to the mandibular phenotype (LOD score 3.93). These two modifiers account for the distribution of the craniofacial malformations by the genetic effect between B6 and CBA strains. Moreover, the Otmf2 region contains a candidate gene for several diseases in mice and humans. These genetic studies involving an otocephalic mouse model appear to provide new insights into mechanistic pathways of craniofacial development. Furthermore, these experiments offer a powerful approach with respect to identification and characterization of candidate genes that may contribute to human agnathia-holoprosencephaly complex diseases (Hide, 2002).

Orthodenticle homologs interact with LIM proteins

Two highly homologous proteins specifically interact with the LIM domains of P-Lim/Lhx3 and several other LIM homeodomain factors (See Drosophila Apterous). Transcripts encoding these factors can be detected as early as mouse E8.5, with maximal expression observed in regions of the embryo in which the LIM homeodomain factors P-Lim/Lhx3, Isl-1, and LH-2 are selectively expressed. These proteins can potentiate transactivation by P-Lim/Lhx-3 and are required for a synergistic activation of the glycoprotein hormone alpha-subunit promoter by P-Lim/Lhx3 and a pituitary Otx class homeodomain transcription factor (P-OTX/Ptx1), with which they also specifically associate. The two new genes are referred to as CLIM-1 and CLIM-2 (cofactor of LIM phomeodomain proteins). The CLIM proteins are required for a transcriptional synergy between P-Lim/Lhx3 and P-OTX/Ptx1. The fact that CLIM-encoded mRNAs show a widely overlapping expression pattern with Otx1 and Otx2 in the developing mouse brain suggests that the CLIM protein family may play critical roles in the functional relationships of LIM homeoproteins and additional Otx factors (Bach, 1997).

Transcriptional regulation of Orthodenticle homologs

Vertebrate Hox and Otx genes encode homeodomain-containing transcription factors thought to transduce positional information along the body axis in the segmental portion of the trunk and in the rostral brain, respectively. Moreover, Hox and Otx2 genes show a complementary spatial regulation during embryogenesis. A 1821-base pair (bp) upstream DNA fragment of the Otx2 gene is positively regulated by co-transfection with expression vectors for the human HOXB1, HOXB2, and HOXB3 proteins in an embryonal carcinoma cell line (NT2/D1) and a shorter fragment of only 534 bp is able to drive this regulation. The HOXB1, HOXB2, and HOXB3 DNA-binding region on the 534-bp Otx2 genomic fragment has been demonstrated using nuclear extracts from Hox-transfected COS cells and 12.5 days postcoitum mouse embryos or HOXB3 homeodomain-containing bacterial extracts. HOXB1, HOXB3, and nuclear extracts from 12.5 day mouse embryos bind to a sequence containing two palindromic TAATTA sites, which bear four copies of the ATTA core sequence, a common feature of most HOM-C/HOX binding sites. HOXB2 protects an adjacent site containing a direct repeat of an ACTT sequence, quite divergent from the ATTA consensus. The region bound by the three homeoproteins is strikingly conserved through evolution and necessary (at least for HOXB1 and HOXB3) to mediate the up-regulation of the Otx2 transcription. Taken together, the data support the hypothesis that anteriorly expressed Hox genes might play a role in the refinement of the Otx2 early expression boundaries in vivo (Guazzi, 1998).

The mouse Otx2 gene is essential throughout head and brain development, from anterior-posterior polarity determination and neuroectoderm induction to post-natal sensory organ maturation. These numerous activities must rely on a very finely tuned regulation of expression. In order to understand the molecular control of the Otx2 gene, attempts were made to isolate its promoter. Three remote transcription start sites were found, two defining two new upstream exons and one mapping within the previously reported first exon. The three transcripts differ in their 5' non-coding region but encode the same protein. The transcription start nucleotides of each mRNA species have been mapped by RNase protection assays and by an RNA circularization technique. They are all used and linked to functional promoters. In addition to leader versatility, alternative splicing was detected within the coding sequence that gives rise to a new protein endowed with an 8 amino-acid insertion upstream of the homeodomain. Combined analysis of the relative abundance of Otx2 mRNA isoforms in representative tissues and in situ hybridization studies revealed distinct spatial and temporal (although partially overlapping) expression patterns of the mRNA isoforms. These findings provide new clues to a better understanding of the relationships between Otx2 gene architecture and its complex regulatory requirements (Courtois, 2003).

Cis-regulatory sequences have been identified acting on Otx2 expression in epiblast (EP) and anterior neuroectoderm (AN) at about 90 kb 5' upstream. The activity of the EP enhancer is found in the inner cell mass at E3.5 and the entire epiblast at E5.5. The AN enhancer activity is detected initially at E7.0 and ceases by E8.5; it is found later in the dorsomedial aspect of the telencephalon at E10.5. The EP enhancer includes multiple required domains over 2.3 kb, and the AN enhancer is an essential component of the EP enhancer. Mutants lacking the AN enhancer have demonstrated that these cis-sequences indeed regulate Otx2 expression in EP and AN. At the same time, this analysis indicates that another EP and AN enhancer must exist outside of the -170 kb to +120 kb range. In Otx2^DeltaAN/- mutants, in which one Otx2 allele lacks the AN enhancer and the other allele is null, anteroposterior axis forms normally and anterior neuroectoderm is normally induced. Subsequently, however, forebrain and midbrain are lost, indicating that Otx2 expression under the AN enhancer functions to maintain anterior neuroectoderm once induced. Furthermore, Otx2 under the AN enhancer cooperates with Emx2 in diencephalon development. The AN enhancer region is conserved among mouse, human and Xenopus; moreover, the counterpart region in Xenopus exhibits an enhancer activity in mouse anterior neuroectoderm (Kurokawa, 2004b).

Otx2 expression in the forebrain and midbrain was found to be regulated by two distinct enhancers (FM and FM2) located at 75 kb 5' upstream and 115 kb 3' downstream. The activities of these two enhancers were absent in anterior neuroectoderm earlier than E8.0; however, at E9.5 their regions of activity spanned the entire mesencephalon and diencephalon with their caudal limits at the boundary with the metencephalon or isthmus. In telencephalon, activities were found only in the dorsomedial aspect. Potential binding sites of OTX and TCF were essential to FM activity, and TCF sites were also essential to FM2 activity. The FM2 enhancer appears to be unique to rodent; however, the FM enhancer region is deeply conserved in gnathostomes. Studies of mutants lacking FM or FM2 enhancer demonstrate that these enhancers indeed regulate Otx2 expression in forebrain and midbrain. Development of mesencephalic and diencephalic regions is differentially regulated in a dose-dependent manner by the cooperation between Otx1 and Otx2 under FM and FM2 enhancers: the more caudal the structure the higher the OTX dose requirement. At E10.5 Otx1^-/-Otx2^{DeltaFM/DeltaFM} mutants (in which Otx2 expression under the FM2 enhancer remains) exhibit almost complete loss of the entire diencephalon and mesencephalon; the telencephalon does, however, develop (Kurokawa, 2004b).

Analysis of Otx homologs in lamprey suggests that the ancestral vertebrate originated with a single copy of the Otx gene and that divergence into the Otx2 and Otx1 lineages occurs in the gnathostome lineage. In the tetrapod lineage, Otx2 is thought to have retained its ancestral roles. By contrast, Otx1 appears to have been co-opted for more complex functions associated with vertebrate brains. However, the specialization of Otx2 and Otx1 functions appears quite divergent in the lineage leading to the extant teleosts. In light of the essential roles of Otx in head development, the question of how the organizations of Otx enhancers in each vertebrate are related is one of keen interest. The mouse Otx1 locus has no regions homologous to the mouse Otx2 enhancers identified (Kurokawa, 2004b).

In the genomic region surrounding Otx2-coding region, 22 domains are conserved in mouse, human and Xenopus. The organization of these domains is also conserved among these animals, and probably throughout all tetrapods. Furthermore, nine and five of these domains are conserved in FrOtx2a and zebrafish Otx2, respectively, in the same array. To consider these Otx2-cis elements in evolutional terms, Otx2 genome information is desired for coelacanth/lungfish, shark and lamprey. With the exception of the FM2 enhancer, all Otx2 enhancers identified correspond to one of these 22 domains; enhancer activities were confirmed in six domains in mouse. Many other domains may also exhibit enhancer activities at later stages (Kurokawa, 2004b).

Visceral endoderm is unique to mammals; the neural crest cells responsible for generation of cephalic mesenchyme arose with vertebrate evolution. Moreover, mesendoderm is common to chordates. In Xenopus Otx2 and zebrafish Otx2, the regions homologous to the mouse non-ectodermal enhancers (domain alpha) also exist near the coding region; it remains for future studies to determine whether the domains retain these non-ectodermal enhancer activities. By contrast, non-ectodermal enhancers occur at about 3' 15 kb downstream in FrOtx2a, and overall sequences are not conserved (Kurokawa, 2004b).

EP and AN enhancer regions are not conserved in either zebrafish or pufferfish. The lack of conservation of the AN enhancer region in these fish genomes was particularly unexpected. Otx2 alone in tetrapod and all Otx genes in zebrafish are expressed in the anterior neuroectoderm. By contrast, the FM enhancer region (domain ß) is deeply conserved in gnathostome Otx2. Obviously, discussion of the phylogenetic significance of the non-conservation of the EP/AN enhancer region and the conservation of the FM region requires the identification of the second mouse EP/AN enhancer. Nevertheless, the differences in Otx2 expression in anterior neuroectoderm in teleost and tetrapod is notable. In tetrapods, Otx2 is initially expressed in the entire neuroectoderm (under AN enhancer) and subsequently the expression is lost in the anterior region, which corresponds to the dorsal telencephalon (under FM enhancer). In zebrafish, Otx2 is not expressed in the most anterior part of the neuroectoderm, even at the earliest phase. It is tempting to speculate that in fish the FM enhancer might function also as the AN enhancer. A comprehensive analysis of Otx enhancers in fish is awaited (Kurokawa, 2004b).

Analysis of mouse Otx2 transcripts has revealed the existence of three different promoters and suggests that the corresponding mRNAs could exhibit specific expression patterns. Studied were the precise dynamics of their expression throughout mouse development. Their spatial distribution was determined by isoform-specific in situ hybridization and their relative abundance by real-time reverse transcriptase-polymerase chain reaction. Although the three promoters may be used in the same areas, transcription preferentially occurs from the proximal promoter at onset of gene activity in early embryogenesis, and switches to the more distal one in most of the sites of expression in the adult brain. During gestation, their relative utilization becomes inverted. The third promoter, which shows no activity in embryonic stem cells and is moderately expressed during embryogenesis, is mostly used in specific areas derived from the rostral part of the neural tube (Fossat, 2005).

Otx2 plays essential roles in each site at each step of head development. Previously identifed enhancers include the AN1 enhancer at 91kb 5' upstream that regulates Otx2 expressions in anterior neuroectoderm (AN) at neural plate stage before E8.5, and the FM1 enhancer at 75kb 5' upstream and the FM2 enhancer at 122kb 3' downstream regulates expression in forebrain/midbrain (FM) at brain vesicle stage after E8.5. The present study identified a second AN enhancer (AN2) at 88kb 5' upstream; the AN2 enhancer also recapitulates the endogenous Otx2 expression in choroid plexus, cortical hem and choroidal roof. However, the enhancer mutants indicated the presence of another AN enhancer. The study also identified a third FM enhancer (FM3) at 153kb 5' upstream. Thus, the Otx2 expressions in anterior neuroectoderm and forebrain/midbrain are regulated by more than six enhancers located far from the coding region. The enhancers identified are differentially conserved among vertebrates; none of the AN enhancers has activities in caudal forebrain and midbrain at brain vesicle stage after E8.5, nor do any of the FM enhancers in anterior neuroectoderm at neural plate stage before E8.5 (Kurokawa, 2014).

Otx transcriptional targets

Transcription of the human interphotoreceptor retinoid binding protein (IRBP) gene is strictly tissue specific, being restricted to retinal photoreceptors and pinealocytes. A sequence named A element, in the IRBP promoter is essential for IRBP gene transcription in vivo. The human homeodomain protein OTX2 is present in nuclear extracts of IRBP expressing cells and specifically interacts with the IRBP A promoter element in vitro. OTX2, as well as CRX, a homeodomain protein very similar to OTX2, activates the human IRBP promoter in co-transfection experiments (Bovola, 1999).

A short (30 base pair) element has been characterized from the Xenopus Wnt-5a promoter which is nearly identical to one located in the human Wnt-5a promoter, and has the same position relative to the transcription start site. When placed in front of a LacZ gene, this element can reproduce the same expression pattern observed for Wnt-5a at the late gastrula stage. Gastrula stage Wnt-5a expression is repressed by otx2, something that is reflected by the mutually exclusive expression patterns of these two genes. The isolated promoter sequence contains an OTX- consensus binding site. This binding site's activity in embryos is repressed by ectopically expressed otx2 (Morgan, 1999a).

Otx2, a vertebrate homolog of the Drosophila orthodenticle gene, coordinates two processes in early embryonic development. Not only does it specify cell fate in the anterior regions of the embryo, it also prevents the cells that express it from participating in the convergence extension movements that shape the rest of the body axis. In Xenopus, this latter function is mediated by XclpH3, transcription of which is directly stimulated by Xotx2. XclpH3 is a Xenopus homolog of the mammalian calponin gene, the product of which binds both actin and myosin and prevents the generation of contractile force by actin filaments (Morgan, 1999b).

Interestingly, the transcription of Spec2a, a sea urchin gene that shares a high degree of sequence identity with the vertebrate calponin genes, is also activated directly by Spotx, a sea urchin homolog of the orthodenticle gene. Sea urchin larvae do not form heads and hence the function of Spotx is assumed not to be comparable with that of otd or the vertebrate otx genes. However, Spotx activates the expression of Spec2a in the aboral ectoderm (Mao, 1996). It is cells of this ectodermal region that undergo the first, and highly characteristic, morphogenetic change in sea urchin development. A proportion of them change shape and move into the center of the embryo to form the primary mesenchyme. This change of cell shape is coincident with the downregulation of Spotx and Spec2a. Overexpression of Spotx prevents the formation of primary mesenchyme; the cells that should change shape fail to do so and remain within the ectoderm. Thus the functions of Spotx and Spec2a may be to prevent inappropriate morphogenetic changes in the aboral ectoderm, in a manner comparable with that by which Xotx2 excludes anterior neuroectoderm from convergence (Morgan, 1999b).

Lamonerie, T. and Tam, P. P. (2014). Head formation: OTX2 regulates Dkk1 and Lhx1 activity in the anterior mesendoderm

The Otx2 gene encodes a paired-type homeobox transcription factor that is essential for the induction and the patterning of the anterior structures in the mouse embryo. Otx2 knockout embryos fail to form a head. Whereas previous studies have shown that Otx2 is required in the anterior visceral endoderm and the anterior neuroectoderm for head formation, its role in the anterior mesendoderm (AME) has not been assessed specifically. This study shows that tissue-specific ablation of Otx2 in the AME phenocopies the truncation of the embryonic head of the Otx2 null mutant. Expression of Dkk1 and Lhx1, two genes that are also essential for head formation, is disrupted in the AME of the conditional Otx2-deficient embryos. Consistent with the fact that Dkk1 is a direct target of OTX2, it was shown that OTX2 can interact with the H1 regulatory region of Dkk1 to activate its expression. Cross-species comparative analysis, RT-qPCR, ChIP-qPCR and luciferase assays have revealed two conserved regions in the Lhx1 locus to which OTX2 can bind to activate Lhx1 expression. Abnormal development of the embryonic head in Otx2;Lhx1 and Otx2;Dkk1 compound mutant embryos highlights the functional intersection of Otx2, Dkk1 and Lhx1 in the AME for head formation (Ip, 2014).

Table of contents