Principles of Limb Development

by Michael Rindler, Ph.D. and Cynthia Loomis, M.D., Ph.D.

Objectives
I. Introduction
II. Initiation of Limb Bud Formation
III. Apical Ectodermal Ridge and the Progress Zone Theory
IV. The Anterior-Posterior Axis and the Zone of Polarizing Activity
V. Dorsal-Ventral Polarity
VI. Positional Information
VII. Development of Tissues in the Limbs
VIII. Rotation of the Limbs
IX. Abnormalities of the Limbs
X. Formation of the Digits and Programmed Cell Death
STUDY QUESTIONS

Objectives

1. To learn about the early development of the limb bud and the regions that are most critical in the organization and outgrowth of the limbs, including the apical ectodermal ridge, the progress zone, and the zone of polarizing activity.

2. To understand the molecules which specify growth, differentiation, and pattern formation in the limb.

3. To describe the concept of a developmental field.

 

I. Introduction

Limb formation begins relatively late in gestation after the basic body plan has taken shape. The limb buds are first visible along the embryonic flanks in the middle (upper) and end (lower) of the fourth week (Fig. 8.12/9.12). The activation of mesenchymal cells derived from lateral mesoderm (to form the limb field) appears to be a crucial step. Indeed, the skeletal structures of the limbs are derived from this mesoderm as opposed to the somitic sclerotomes that form the rest of the skeletal system from the neck down. Limb buds have a specialized distal ectodermal region at the dorsal/ventral margin called the apical ectodermal ridge (AER) (Fig. 8.13/9.13), which signals to the underlying mesenchyme called the progress zone. As will be explained below, limb development occurs as a result of continuous interactions between the ectodermal and mesodermal components of the bud.

Limb development occurs in four dimensions, each controlled by a separate process and different growth factors (Fig. 8.16/9.17). First, there is the proximal-distal axis, initiated by signals from axial structures, including the secretion of fibroblast growth factors (FGF), and later maintained by the AER. Second, there is the anterior-posterior axis that specifies the position of the thumbs versus little finger, for example. A region located along the posterior side of the limb bud called the zone of polarizing activity (ZPA) controls this process. The ZPA produces sonic hedgehog (Shh) which is critical for its function. Third, there is the dorsal/ventral axis, which involves the ectoderm and a growth factor of the Wnt family. And finally, there is a time-dependent change in the cells of the progress zone that determines the type of structures that can be formed in the proximal-distal axis. Superimposed on all of these is the expression of specific transcription factors that mediate the cellular programs of differentiation, one of which is the retinoic acid receptor family of transcriptional regulators. Perhaps most important is the regional expression of HOX gene family members (paralogues). As was true in the axial region of the embryo, HOX genes encode transcription factors that control the formation of particular structures (e.g., type of digit) in specific regions of the limb (Fig. 8.16/9.17). Finally, the differentiation of the mesoderm into the various tissues inside the limb, bone, cartilage, muscle, connective tissue, and blood vessels, are controlled by another series of growth factors. Bone morphogenic proteins (BMP's) initially play a role downstream of Shh in patterning the A-P axis of the limb and then later function to induce cartilage and bone formation.

It is fairly obvious that the development of the limb has implications not only for congenital malformations of the limbs, but also for tissue regeneration and wound healing processes. Indeed, some of the knowledge gained from the study of the control of bone development has been applied to accelerate healing of bone fractures. Many compounds are being tested for their applicability in this regard.

 

II. Initiation of Limb Bud Formation

It is not currently known what is the primary signal for initiation of limb buds, but based on model systems that can be manipulated experimentally FGF10 is a good candidate. FGF10, a member of the FGF family of growth factors is secreted locally from axial structures in the specific regions adjacent to the limb buds, apparently dictated by the HOX gene expression pattern of those segments. FGF10 when implanted ectopically can induce the formation of accessory limbs. Moreover, targeted gene ablation of FGF10 results in amelia (no limbs). Once initiation has occurred and migration of mesoderm has begun, the limb field develops independently of any signals from the axial region. The AER is thought to take over the function of maintaining the limb bud. Transplantation of a limb bud to an ectopic site will lead to the development of a limb in this location. Rotation of the limb bud in chick or amphibian embryos will lead to a limb that is 180o rotated from its normal position.

 

III. Apical Ectodermal Ridge and the Progress Zone Theory

The AER is required for limb outgrowth. Removal of the AER results in a stunted limb whose development is arrested shortly thereafter. However, segments that have begun to form continue to do so reasonably normally. The AER produces FGF8 at an early stage, and thus seems to take over the function from axial structures that initiate limb bud formation. Later, production of several different FGFs (including FGF2, 4, 9,and 17) by the AER appears to be necessary for maintenance of limb development. Indeed, if the AER is removed, replacement with FGF-expressing cells or FGF-coated beads can restore its function and lead to normal limb development.

The mesoderm in the progress zone, just underneath the AER is thought to be the most important determinant of the actual limb structures that will form. These mesodermal cells determine whether, for example, leg or arm structures will form. In chick embryos, when the mesoderm in this region in the wing is replaced early on with mesoderm from the leg, leg structures develop. The converse experiment leads to a wing where a leg should be. The formation of arm versus leg structures appears to be set in the progress zone mesoderm very early by cues from the axial skeleton. A family of transcription factors, one specific for arm and a related one for leg, are turned on in the limb buds and appear to be important in the arm/leg distinctions. But in general, most of the growth factors and transcription factors produced are similar in the upper and lower limb buds.

The mesoderm in the progress zone also is thought to set the time clock for the proximal-distal axis. Cells leaving this region are fixed in their developmental potential and those leaving earlier have different capabilities than those leaving later on. These cells have received specific positional information that will determine the type of tissue they will become. This is best illustrated when an older limb bud is substituted for a younger one. Only distal structures are produced. It is too late for the proximal structures to form. Conversely, when a young bud is grafted onto an old stump, the proximal structures are duplicated ­ the clock is not reset. It is believed that the expression of transcription factors like Msx are responsible for the clock function. Msx is exclusively found in the progress zone and not in cells that have left the zone.

 

IV. The Anterior-Posterior Axis and the Zone of Polarizing Activity

As mentioned above, the ZPA, located in the posterior mesoderm, is responsible for establishing polarization along the anterior-posterior axis. Shh produced by the ZPA is necessary for its function. Shh production is sustained by FGF produced by the AER (Fig. 8.16/9.17).

The role of the ZPA was discovered in grafting experiments in chick embryos (Fig. 8.17/9.18). Grafting of an additional ZPA resulted in mirror image duplication of posterior limb structures, particular the digits, with the middle structures often fused or malformed. This result was interpreted as indicating that diffusible factors from the ZPA were responsible for inducing the anterior/posterior axis. Indeed, after Shh was discovered, its expression in the limb was shown to be exclusively in the ZPA. Ectopic expression of Shh in an anterior position led to the same mirror image duplications as grafting a ZPA itself. However, since Shh itself does not diffuse readily in the limb, other factors must be responsible.

Ultimately, the signals from the ZPA promote a shift from a proximal-to-distal ³striped² pattern of the HOX gene transcripts (paralogues 9-13) to an anterior-to-posterior ³striped² pattern (Fig. 8.16D/9.17D). Both in humans with limb malformations, like the synpolydactyly syndrome, and in mice and chicks whose HOX patterns have been altered through genetic means, limb malformations result. It is clear that the HOX gene expression pattern is critical for cells to understand the positional information they are receiving during development.

 

V. Dorsal-Ventral Polarity

Limbs are different on their dorsal and ventral surfaces. The skin on the palm of your hand, for example, is different from that on the back of your hand. This dorsal-ventral polarity of the limb is directed by the ectoderm. It is possible to detach the ectoderm from the mesoderm in a limb bud. If the ectoderm is rotated, the dorsal-ventral polarity is reversed while the other axes are maintained. The dorsal ectoderm produces the growth factor Wnt7, a member of the Wnt family, which has been shown to be necessary for dorsalization. Ectopic expression of Wnt7 can dorsalize a ventral surface and ablation of Wnt7a results in ventralization of the dorsal limb. Wnt7a signaling from the ectoderm is critical for the dorsal expression of the transcription factor Lmx1b in the distal mesenchyme. Lmx1b is the gene defective in the human nail-patella syndrome. This autosomal dominant disease is characterized by abnormal development of dorsal

 

Table 1:  Control of Axis Formation in the Limb Bud

______________________________________________________________________________

        Axis                                      Signaling Center                              Molecular Signal

 

Proximal/distal                        Apical ectodermal ridge              FGF2, FGF4, FGF8, FGF9, FGF17

                                                Limb mesenchyme                     FGF10

 

Anterior/posterior                    Zone of polarizing activity          Sonic hedgehog, BMPıs

 

Dorsal/ventral                          Dorsal ectoderm,                        WNT7a

                                                (Dorsal mesenchyme)                (LMX1b)

                                                Ventral ectoderm                        Engrailed-1

______________________________________________________________________________

limb structures, such as nails and patellae. Another transcription factor, engrailed-1, is expressed in the ventral ectoderm and is required for specifying ventral limb fates. It functions in part through repression of the dorsalizing genes Wnt7a and indirectly Lmx1b (Fig. 8.16).

 

VI. Positional Information

The differentiation of the cells within the limb is dependent on their precise position and the growth factor signals they receive once their fate is restricted. The limb has served as a classical example of how gradients of signals from centers such as the ZPA and AER help establish positional information in the limb field. In a simple model, the positional information is imparted by a source of a diffusable substance (morphogen). The distance a cell is from the source of the substance determines whether it will respond and differentiate in a particular way. In a more complex model (French flag model) the cells would have at least three developmental choices and select the appropriate one based on the concentration of the signal, if it is a diffusible molecule, that it received. For example, in vitro culture of mesodermal cells with activin beads leads to the expression of one set of transcription factors (including goosecoid) in cells nearest the beads, while a different set (e.g., brachyury) of factors are expressed by cells a little farther away. Those cells farthest away from the bead fail to respond at all. In reality, the scheme is more complex by virtue of the fact that the cells closest to the signal may, in response to the initial signal, send out their own additional signals. The combination of signals may alter the subsequent response from more distal cells. In addition, in a setting such as the limb, any given cell has the potential to respond to at least three signal gradients in each of the polar axes (FGF, Shh, Wnt, for example) and each of these can be modified in a similar fashion.

While it will take decades to work out the details of how positional information is imparted in the limbs, the general principles are likely to follow the pattern of superimposable gradients of factors and responses in several dimensions.

 

VII. Development of Tissues in the Limbs

Mesoderm in the limb has the potential to make blood vessels, muscle, bone, connective tissue and cartilage. In part, the fate of the cells is predetermined by its site of origin in the body of the embryo (Fig. 9.4/10.4). Lateral plate mesoderm migrating into the region can only make certain tissue (blood vessels, connective tissue, bone, and cartilage) while somitic myotomes make the muscle masses. Differentiation and selective growth of the different mesodermal components is further influenced by growth factors, including bone morphogenic proteins (BMP's), BMP's were discovered as molecules isolated from bone that could induce bone and cartilage formation when injected ectopically into developing limb buds. Other members of the family influence the development of joints. All of the bones in the limbs initially form as cartilage models and ossify as endochondral bones (Fig. 8.15/9.16).

Dorsal and ventral muscle masses form as the limb buds grow. Nerves migrate into the muscle masses from the developing spinal cord. Individual muscles in the limbs often are sup­plied by more than one spinal segment. While this might suggest that the myotomes themselves are mixed when they form the muscle masses, experiments on innervation pathways indicate that the pathways of neural migration are relatively fixed even when myotomes are prevented from forming. Thus, the innervation of the limb musculature does not correspond to somite origin.

Blood vessels in the limbs initially form by the fourth week with an anastomosing network with branches from the nearby aorta (these being the intersegmental arteries in the case of the upper limb) and the cardinal veins. In the upper limb, the vessels will coalesce to form the subclavian (7th intersegmental) artery and vein. In the developing limb itself, a large venous plexus forms along the outer rim of mesoderm outside and a central or axial artery forms in the center with extensive branches. As the limb grows further, some of the smaller branches become dominant. There is extensive anastomosis of new branches with old and considerable restructuring to form the mature vascular systems of the limbs.

Cutaneous sensory innervation is established as peripheral nerves and neural crest cells migrate into the limb buds during the fifth week. Initially they are in a regular segmented pattern as is true in the trunk. However, as the limbs undergo distal growth and extensive rotation, the dermatomes grow and rotate along with them, yielding a more complex sensory innervation pattern.

 

VIII. Rotation of the Limbs

Our limbs are derived evolutionarily from the limbs of primitive fishes. As animals migrated to land, the initially straight limbs evolved rotation in the shoulders/pelvic joints and further rotation at the elbows. As animals evolved to stand on two legs, the legs underwent further changes in their orientation such that the elbows and knees bend in opposite directions when viewed from the normal standing position.

 

IX. Abnormalities of the Limbs

There are many abnormalities associated with limb development. Most common are those that affect the fingers ­ syndactyly (fusion of fingers) and polydactyly (extra digits) (Fig. 8.19/9.20). These have been linked to disturbances in the HOX gene expression patterns and the Shh signaling pathway as well as to defects in AER formation. However, amelia (lack of limbs) or meromelia (a.k.a., phocomelia; lack of whole segments of limbs) can also occur. The latter was linked to thalidomide administration to pregnant women 40 years ago. It is thought that meromelia, where a normal distal structure (hand) can be attached to the shoulder, for example, is an interference with the clock of the progress zone during development (Fig. 8.18/9.19).

 

X. Formation of the Digits and Programmed Cell Death

Cell death plays a major role in limb development. A classic example of so-called programmed cell death occurs in the regions between the developing digits (Figs. 8.13/9.13, 8.14/9.14, 9.15). Here cells undergo apoptosis, that is, signal-induced death at the appropriate time. While the mechanisms of programmed cell death in the limbs are not well understood, in other model systems apoptosis involves the activation of cell surface receptors that instruct the cell to activate a program whereby proteases and nucleases destroy the cells proteins and nucleic acids in a defined way. The molecular details of this process will be discussed in your Cell Biology in Medicine course.

 

Further Reading: Gilbert, SF (2003) Developmental Biology. 7th Edition. Sinauer, Sunderland,

MA

 


 

STUDY QUESTIONS

 

What is a limb bud?  An AER? 

What growth factor family is necessary for initiation of limb bud formation and maintenance of the bud?

What is the ZPA and how does it function in anterior/posterior polarity?  What growth factor or factors are produced by the ZPA to mediate its effects?

What type of mesoderm gives rise to muscles in the limb?  Bone and cartilage?

What is a HOX gene?  How could the HOX gene patterns explain the anomaly of a child born with four additional toes arranged such that a single large big toe was present in the middle flanked by four toes on each side that are mirror images of one another?

How does the innervation pattern of the dermatomes of the limb reflect the origin and growth of the limb buds?