An historical perspective on the pioneering experiments of John Saunders.

John Saunders was a highly skilled embryologist who pioneered the study of limb development. His studies on chick embryos provided the fundamental framework for understanding how vertebrate limbs develop. This framework inspired generations of scientists and formed the bridge from experimental embryology to molecular mechanisms. Saunders investigated how feathers become organized into tracts in the skin of the chick wing and also identified regions of programmed cell death. He discovered that a region of thickened ectoderm that rims the chick wing bud - the apical ectodermal ridge - is required for outgrowth and the laying down of structures along the proximo-distal axis (long axis) of the wing, identified the zone of polarizing activity (ZPA; polarizing region) that controls development across the anteroposterior axis ("thumb to little finger "axis) and contributed to uncovering the importance of the ectoderm in development of structures along the dorso-ventral axis ( "back of hand to palm" axis). This review looks in depth at some of his original papers and traces how he made the crucial findings about how limbs develop, considering these findings both in the context of contemporary knowledge at the time and also in terms of their immediate impact on the field.

Dimeric combinations of MafB, cFos and cJun control the apoptosis-survival balance in limb morphogenesis.

Apoptosis is an important mechanism for sculpting morphology. However, the molecular cascades that control apoptosis in developing limb buds remain largely unclear. Here, we show that MafB was specifically expressed in apoptotic regions of chick limb buds, and MafB/cFos heterodimers repressed apoptosis, whereas MafB/cJun heterodimers promoted apoptosis for sculpting the shape of the limbs. Chromatin immunoprecipitation sequencing in chick limb buds identified potential target genes and regulatory elements controlled by Maf and Jun. Functional analyses revealed that expression of p63 and p73, key components known to arrest the cell cycle, was directly activated by MafB and cJun. Our data suggest that dimeric combinations of MafB, cFos and cJun in developing chick limb buds control the number of apoptotic cells, and that MafB/cJun heterodimers lead to apoptosis via activation of p63 and p73.

How the embryo makes a limb: determination, polarity and identity.

The vertebrate limb with its complex anatomy develops from a small bud of undifferentiated mesoderm cells encased in ectoderm. The bud has its own intrinsic polarity and can develop autonomously into a limb without reference to the rest of the embryo. In this review, recent advances are integrated with classical embryology, carried out mainly in chick embryos, to present an overview of how the embryo makes a limb bud. We will focus on how mesoderm cells in precise locations in the embryo become determined to form a limb and express the key transcription factors Tbx4 (leg/hindlimb) or Tbx5 (wing/forelimb). These Tbx transcription factors have equivalent functions in the control of bud formation by initiating a signalling cascade involving Wnts and fibroblast growth factors (FGFs) and by regulating recruitment of mesenchymal cells from the coelomic epithelium into the bud. The mesoderm that will form limb buds and the polarity of the buds is determined with respect to both antero-posterior and dorso-ventral axes of the body. The position in which a bud develops along the antero-posterior axis of the body will also determine its identity - wing/forelimb or leg/hindlimb. Hox gene activity, under the influence of retinoic acid signalling, is directly linked with the initiation of Tbx5 gene expression in the region along the antero-posterior axis of the body that will form wings/forelimbs and determines antero-posterior polarity of the buds. In contrast, Tbx4 expression in the regions that will form legs/hindlimbs is regulated by the homeoprotein Pitx1 and there is no evidence that Hox genes determine antero-posterior polarity of the buds. Bone morphogenetic protein (BMP) signalling determines the region along the dorso-ventral axis of the body in which both wings/forelimbs and legs/hindlimbs develop and dorso-ventral polarity of the buds. The polarity of the buds leads to the establishment of signalling regions - the dorsal and ventral ectoderm, producing Wnts and BMPs, respectively, the apical ectodermal ridge producing fibroblast growth factors and the polarizing region, Sonic hedgehog (Shh). These signals are the same in both wings/forelimbs and legs/hindlimbs and control growth and pattern formation by providing the mesoderm cells of the limb bud as it develops with positional information. The precise anatomy of the limb depends on the mesoderm cells in the developing bud interpreting positional information according to their identity - determined by Pitx1 in hindlimbs - and genotype. The competence to form a limb extends along the entire antero-posterior axis of the trunk - with Hox gene activity inhibiting the formation of forelimbs in the interlimb region - and also along the dorso-ventral axis.

Mathematical modelling of digit specification by a sonic hedgehog gradient.

BACKGROUND: The three chick wing digits represent a classical example of a pattern specified by a morphogen gradient. Here we have investigated whether a mathematical model of a Shh gradient can describe the specification of the identities of the three chick wing digits and if it can be applied to limbs with more digits. RESULTS: We have produced a mathematical model for specification of chick wing digit identities by a Shh gradient that can be extended to the four digits of the chick leg with Shh-producing cells forming a digit. This model cannot be extended to specify the five digits of the mouse limb. CONCLUSIONS: Our data suggest that the parameters of a classical-type morphogen gradient are sufficient to specify the identities of three different digits. However, to specify more digit identities, this core mechanism has to be coupled to alternative processes, one being that in the chick leg and mouse limb, Shh-producing cells give rise to digits; another that in the mouse limb, the cellular response to the Shh gradient adapts over time so that digit specification does not depend simply on Shh concentration.

Generation of mice with functional inactivation of talpid3, a gene first identified in chicken.

Specification of digit number and identity is central to digit pattern in vertebrate limbs. The classical talpid(3) chicken mutant has many unpatterned digits together with defects in other regions, depending on hedgehog (Hh) signalling, and exhibits embryonic lethality. The talpid(3) chicken has a mutation in KIAA0586, which encodes a centrosomal protein required for the formation of primary cilia, which are sites of vertebrate Hh signalling. The highly conserved exons 11 and 12 of KIAA0586 are essential to rescue cilia in talpid(3) chicken mutants. We constitutively deleted these two exons to make a talpid3(-/-) mouse. Mutant mouse embryos lack primary cilia and, like talpid(3) chicken embryos, have face and neural tube defects but also defects in left/right asymmetry. Conditional deletion in mouse limb mesenchyme results in polydactyly and in brachydactyly and a failure of subperisoteal bone formation, defects that are attributable to abnormal sonic hedgehog and Indian hedgehog signalling, respectively. Like talpid(3) chicken limbs, the mutant mouse limbs are syndactylous with uneven digit spacing as reflected in altered Raldh2 expression, which is normally associated with interdigital mesenchyme. Both mouse and chicken mutant limb buds are broad and short. talpid3(-/-) mouse cells migrate more slowly than wild-type mouse cells, a change in cell behaviour that possibly contributes to altered limb bud morphogenesis. This genetic mouse model will facilitate further conditional approaches, epistatic experiments and open up investigation into the function of the novel talpid3 gene using the many resources available for mice.

Interactions between Shh, Sostdc1 and Wnt signaling and a new feedback loop for spatial patterning of the teeth.

Each vertebrate species displays specific tooth patterns in each quadrant of the jaw: the mouse has one incisor and three molars, which develop at precise locations and at different times. The reason why multiple teeth form in the jaw of vertebrates and the way in which they develop separately from each other have been extensively studied, but the genetic mechanism governing the spatial patterning of teeth still remains to be elucidated. Sonic hedgehog (Shh) is one of the key signaling molecules involved in the spatial patterning of teeth and other ectodermal organs such as hair, vibrissae and feathers. Sostdc1, a secreted inhibitor of the Wnt and Bmp pathways, also regulates the spatial patterning of teeth and hair. Here, by utilizing maternal transfer of 5E1 (an anti-Shh antibody) to mouse embryos through the placenta, we show that Sostdc1 is downstream of Shh signaling and suggest a Wnt-Shh-Sostdc1 negative feedback loop as a pivotal mechanism controlling the spatial patterning of teeth. Furthermore, we propose a new reaction-diffusion model in which Wnt, Shh and Sostdc1 act as the activator, mediator and inhibitor, respectively, and confirm that such interactions can generate the tooth pattern of a wild-type mouse and can explain the various tooth patterns produced experimentally.

Integration of growth and specification in chick wing digit-patterning.

In the classical model of chick wing digit-patterning, the polarizing region--a group of cells at the posterior margin of the early bud--produces a morphogen gradient, now known to be based on Sonic hedgehog (Shh), that progressively specifies anteroposterior positional identities in the posterior digit-forming region. Here we add an integral growth component to this model by showing that Shh-dependent proliferation of prospective digit progenitor cells is essential for specifying the complete pattern of digits across the anteroposterior axis. Inhibiting Shh signalling in early wing buds reduced anteroposterior expansion, and posterior digits were lost because all prospective digit precursors formed anterior structures. Inhibiting proliferation also irreversibly reduced anteroposterior expansion, but instead anterior digits were lost because all prospective digit precursors formed posterior structures. When proliferation recovered in such wings, Shh transcription was maintained for longer than normal, suggesting that duration of Shh expression is controlled by a mechanism that measures proliferation. Rescue experiments confirmed that Shh-dependent proliferation controls digit number during a discrete time-window in which Shh-dependent specification normally occurs. Our findings that Shh signalling has dual functions that can be temporally uncoupled have implications for understanding congenital and evolutionary digit reductions.

eChickAtlas: an introduction to the database.

The precise control of gene expression is critical in embryonic development. Quantitative assays, such as microarrays and RNA sequencing, provide gene expression levels for a large number of genes, but do not contain spatial information. In contrast, in situ methods, such as in situ hybridization and immunohistochemistry, provide spatial resolution, but poor quantification and can only reveal the expression of one, or very few genes at a time. Furthermore, the usual methods of documenting the results, by photographing whole mounts or sections, makes it very difficult to assess the three-dimensional (3D) relationships between expressing and nonexpressing cells. Optical projection tomography (OPT) can capture the full 3D expression pattern in a whole embryo at a reasonable level of resolution and at moderately high throughput. A large database containing spatio-temporal patterns of expression for the mouse (e-Mouse Atlas Project, EMAP, www.emouseatlas.org) has been created, incorporating 3D information. Like the mouse, the chick is an important model in developmental biology and translational studies. To facilitate comparisons between these important model organisms, we have created a 3D anatomical atlas, accompanied by an anatomical ontology of the chick embryo and a database of gene expression patterns during chick development. This database is publicly available (www.echickatlas.org).

Retinoic acid signaling and the initiation of mammary gland development.

Retinoic acid receptors (RARs), which are involved in retinoic acid signal transduction, are essential for maintaining the differentiated state of epithelial tissues. Mammary glands are skin appendages whose development is initiated through continuous cell-cell interactions between the ectoderm and the adjacent mesenchyme. Considerable progress has been made in elucidating the molecular basis of these interactions in mammary gland formation in mouse embryos, including the network of initiating signals comprising Fgfs, Wnts and Bmps involved in gland positioning and the transcription factors, Tbx3 and Lef1, essential for mammary gland development. Here, we provide evidence that retinoic acid signaling may also be involved in mammary gland development. We documented the expression of gene-encoding enzymes that produce retinoic acid (Raldh2) and enzymes that degrade it (Cyp26a1, Cyp26b1). We also analyzed the expression of RAR-ß, a direct transcriptional target of retinoic acid signaling. Raldh2 and RAR-ß were expressed in E10-E10.5 mouse embryos in somites adjacent to the flank region where mammary buds 2, 3 and 4 develop. These expression patterns overlapped with that of Fgf10, which is known to be required for mammary gland formation. RAR-ß was also expressed in the mammary mesenchyme in E12 mouse embryos; RAR-ß protein was expressed in the mammary epithelium and developing fat pad. Retinoic acid levels in organ cultures of E10.5 mouse embryo flanks were manipulated by adding either retinoic acid or citral, a retinoic acid synthesis inhibitor. Reduced retinoic acid synthesis altered the expression of genes involved in retinoic acid homeostasis and also demonstrated that retinoic acid signaling is required for Tbx3 expression, whereas high levels of retinoic acid signaling inhibited Bmp4 expression and repressed Wnt signaling. The results of the experiments using RNAi against Tbx3 and Wnt10b suggested feedback interactions that regulate retinoic acid homeostasis in mammary gland-forming regions. We produced a molecular model for mammary gland initiation that incorporated retinoic acid signaling.

3D expression patterns of cell cycle genes in the developing chick wing and comparison with expression patterns of genes implicated in digit specification.

Sonic hedgehog (Shh) signalling controls integrated specification of digit pattern and growth in the chick wing but downstream gene networks remain to be unravelled. We analysed 3D expression patterns of genes encoding cell cycle regulators using Optical Projection Tomography. Hierarchical clustering of spatial matrices of gene expression revealed a dorsal layer of the wing bud, in which almost all genes were expressed, and that genes encoding positive cell cycle regulators had similar expression patterns while those of N-myc and CyclinD2 were distinct but closely related. We compared these patterns computationally with those of genes implicated in digit specification and Ptch1, 50 genes in total. Nineteen genes have similar posterior expression to Ptch1, including Hoxd13, Sall1, Hoxd11, and Bmp2, all likely Gli targets in mouse limb, and cell cycle genes, N-myc, CyclinD2. We suggest that these genes contribute to a network integrating digit specification and growth in response to Shh.

The chicken polydactyly (Po) locus causes allelic imbalance and ectopic expression of Shh during limb development.

Point mutations in the intronic ZRS region of Lmbr1, a limb specific cis-regulatory element of Sonic hedgehog (Shh), are associated with polydactyly in humans, cats, and mice. We and others have recently mapped the dominant preaxial polydactyly (Po) locus in Silkie chickens to a single nucleotide polymorphism (SNP) in the ZRS region. Using polymorphisms in the chicken Shh sequence, we confirm that the ZRS region directly regulates Shh expression in the developing limb causing ectopic Shh expression in the anterior leg, prolonged Shh expression in the posterior limb, and allelic imbalance between wt and Slk Shh alleles in heterozygote limbs. Using Silkie legs, we have explored the consequences of increased Shh expression in the posterior leg on the patterning of the toes, and the induction of preaxial polydactyly.

Comparison of Iroquois gene expression in limbs/fins of vertebrate embryos.

In Drosophila, Iroquois (Irx) genes have various functions including the specification of the identity of wing veins. Vertebrate Iroquois (Irx) genes have been reported to be expressed in the developing digits of mouse limbs. Here we carry out a phylogenetic analysis of vertebrate Irx genes and compare expression in developing limbs of mouse, chick and human embryos and in zebrafish pectoral fin buds. We confirm that the six Irx gene families in vertebrates are well defined and that Clusters A and B are duplicates; in contrast, Irx1 and 3, Irx2 and 5, and Irx4 and 6 are paralogs. All Irx genes in mouse and chick are expressed in developing limbs. Detailed comparison of the expression patterns in mouse and chick shows that expression patterns of genes in the same cluster are generally similar but paralogous genes have different expression patterns. Mouse and chick Irx1 are expressed in digit condensations, whereas mouse and chick Irx6 are expressed interdigitally. The timing of Irx1 expression in individual digits in mouse and chick is different. Irx1 is also expressed in digit condensations in developing human limbs, thus showing conservation of expression of this gene in higher vertebrates. In zebrafish, Irx genes of all but six of the families are expressed in early stage pectoral fin buds but not at later stages, suggesting that these genes are not involved in patterning distal structures in zebrafish fins.

Micro-magnetic resonance imaging and embryological analysis of wild-type and pma mutant mice with clubfoot.

Gross similarities between the external appearance of the hind limbs of the peroneal muscle atrophy (pma) mouse mutant and congenital talipes equinovarus (CTEV), a human disorder historically referred to as 'clubfoot', suggested that this mutant could be a useful model. We used micro-magnetic resonance imaging to visualize the detailed anatomy of the hind limb defect in mutant pma mice and performed 3D comparisons between mutant and wild-type hind limbs. We found that the pma foot demonstrates supination (i.e. adduction and inversion of the mid foot and fore foot together with plantar flexion of the ankle and toes) and that the tibiale and distal tarsals display 3D abnormalities in positioning. The size and shape of the tibia, fibula, tarsal and metatarsal bones are similar to the wild-type. Hypoplasia of the muscles in the antero-lateral (peroneal) compartment was also demonstrated. The resemblance of these features to those seen in CTEV suggests that the pma mouse is a possibly useful model for the human condition. To understand how the observed deformities in the pma mouse hind foot arise during embryonic development, we followed the process of foot rotation in both wild-type and pma mutant mice. Rotation of the hind foot in mouse embryos of wild-type strains (CD-1 and C57/Black) occurs from embryonic day 14.5 onwards with rotation in C57/Black taking longer. In embryos from both strains, rotation of the right hind foot more commonly precedes rotation of the left. In pma mutants, the initiation of rotation is often delayed and rotation is slower and does not reach completion. If the usefulness of the pma mutant as a model is confirmed, then these findings on pma mouse embryos, when extrapolated to humans, would support a long-standing hypothesis that CTEV is due to the failure of completion of the normal process of rotation and angulation, historically known as the 'arrested development hypothesis'.

Integrating technologies for comparing 3D gene expression domains in the developing chick limb.

Chick embryos are good models for vertebrate development due to their accessibility and manipulability. Recent large increases in available genomic data from both whole genome sequencing and EST projects provide opportunities for identifying many new developmentally important chicken genes. Traditional methods of documenting when and where specific genes are expressed in embryos using whole amount and section in-situ hybridisation do not readily allow appreciation of 3-dimensional (3D) patterns of expression, but this can be accomplished by the recently developed microscopy technique, Optical Projection Tomography (OPT). Here we show that OPT data on the developing chick wing from different labs can be reliably integrated into a common database, that OPT is efficient in capturing 3D gene expression domains and that such domains can be meaningfully compared. Novel protocols are used to compare 3D expression domains of 7 genes known to be involved in chick wing development. This reveals previously unappreciated relationships and demonstrates the potential, using modern genomic resources, for building a large scale 3D atlas of gene expression. Such an atlas could be extended to include other types of data, such as fate maps, and the approach is also more generally applicable to embryos, organs and tissues.

The expression of Fat-1 cadherin during chick limb development.

Cellular adhesion is fundamental to the behaviour of cell populations during embryonic development and serves to establish correct tissue pattern and architecture. The cadherin superfamily of cell adhesion proteins regulates cellular organization and additionally influences intracellular signalling cascades. Here we present for the first time a detailed account of chick Fat-1 gene expression during embryogenesis visualised by whole-mount in situ hybridisation. In part, we focus on the expression pattern in limb buds that has not been accurately documented. While Fat-1 is generally expressed in epithelial tissues and its Drosophila counterpart Fat-like regulates formation of ectodermally-derived organs, in limb buds we have found that chick Fat-1 is uniquely restricted to mesenchyme. This Fat-1 expression pattern is remarkably dynamic throughout tissue differentiation, limb maturation and pattern formation. Diffuse expression of Fat-1 begins at stage HH17 as the limb bud is forming. It then becomes more proximal as the limb bud grows and is expressed within both tendon and muscle progenitors in the dorsal and ventral subectodermal mesenchyme. Later, Fat-1 transcripts were more abundant in anterior and posterior domains of the limb bud. During hand plate formation, Fat-1 transcripts were expressed in the mesenchyme adjacent to the wrist joint zone and in the interdigit mesenchyme.

The chick limb: embryology, genetics and teratology.

The chick embryo has a long history in investigations of vertebrate limb development because of the ease with which its limbs can be experimentally manipulated. Early studies elucidated the fundamental embryology of the limb and identified the key signalling regions that govern its development. The chick limb became a leading model for exploring the concept of positional information and understanding how patterns of differentiated cells and tissues develop in vertebrate embryos. When developmentally important molecules began to be identified, experiments in chick limbs were crucial for bridging embryology and molecular biology. The embryological mechanisms and molecular basis of limb development are largely conserved in mammals, including humans, and uncovering these molecular networks provides links to clinical genetics. We emphasise the important contributions of naturally occurring chick mutants to elucidating limb embryology and identifying novel developmentally important genes. In addition, we consider how the chick limb has been used to study mechanisms involved in teratogenesis with a focus on thalidomide. These studies on chick embryos have given insights into how limb defects can be caused by both genetic changes and chemical insults and therefore are of great medical significance.

Morphological diversity: taking the spine out of three-spine stickleback.

A gene already known to play a crucial developmental role in chick and mouse embryos has been fingered as a candidate for naturally occurring variation in three-spine stickleback anatomy.

Fgf signaling controls the number of phalanges and tip formation in developing digits.

Tetrapods have two pairs of limbs, each typically with five digits, each of which has a defined number of phalanges derived from an archetypal formula. Much progress has been made in understanding vertebrate limb initiation and the patterning processes that determine digit number in developing limb buds, but little is known about how phalange number is controlled. We and others previously showed that an additional phalange can be induced in a chick toe if sonic hedgehog protein is applied in between developing digit primordia. Here we show that formation of an additional phalange is associated with prolonged Fgf8 expression in the overlying apical ridge and that an Fgf Receptor inhibitor blocks its formation. The additional phalange is produced by elongation and segmentation of the penultimate phalange, suggesting that the digit tip forms when Fgf signaling ceases by a special mechanism, possibly involving Wnt signaling. Consistent with this, Fgfs inhibit tip formation whereas attenuation of Fgf signaling induces tip formation prematurely. We propose that duration of Fgf signaling from the ridge, responsible for elongation of digit primordia, coupled with a characteristic periodicity of joint formation, generates the appropriate number of phalanges in each digit. We also propose that the process that generates the digit tips is independent of that which generates more proximal phalanges. This has implications for understanding human limb congenital malformations and evolution of digit diversity.

A comprehensive collection of chicken cDNAs.

Birds have played a central role in many biological disciplines, particularly ecology, evolution, and behavior. The chicken, as a model vertebrate, also represents an important experimental system for developmental biologists, immunologists, cell biologists, and geneticists. However, genomic resources for the chicken have lagged behind those for other model organisms, with only 1845 nonredundant full-length chicken cDNA sequences currently deposited in the EMBL databank. We describe a large-scale expressed-sequence-tag (EST) project aimed at gene discovery in chickens (http://www.chick.umist.ac.uk). In total, 339,314 ESTs have been sequenced from 64 cDNA libraries generated from 21 different embryonic and adult tissues. These were clustered and assembled into 85,486 contiguous sequences (contigs). We find that a minimum of 38% of the contigs have orthologs in other organisms and define an upper limit of 13,000 new chicken genes. The remaining contigs may include novel avian specific or rapidly evolving genes. Comparison of the contigs with known chicken genes and orthologs indicates that 30% include cDNAs that contain the start codon and 20% of the contigs represent full-length cDNA sequences. Using this dataset, we estimate that chickens have approximately 35,000 genes in total, suggesting that this number may be a characteristic feature of vertebrates.