Trent D. Stephens

Mechanism of action in thalidomide teratogenesis.

In this commentary, we describe a model to explain the mechanism of the embryopathy of thalidomide. We propose that thalidomide affects the following pathway during development: insulin-like growth factor I (IGF-I) and fibroblast growth factor 2 (FGF-2) stimulation of the transcription of alphav and beta3 integrin subunit genes. The resulting alphavbeta3 integrin dimer stimulates angiogenesis in the developing limb bud, which promotes outgrowth of the bud. The promoters of the IGF-I and FGF-2 genes, the genes for their binding proteins and receptors, as well as the alphav and beta3 genes, lack typical TATA boxes, but instead contain multiple GC boxes (GGGCGG). Thalidomide, or a breakdown product of thalidomide, specifically binds to these GC promoter sites, decreasing transcription efficiency of the associated genes. A cumulative decrease interferes with normal angiogenesis, which results in truncation of the limb. Intercalation into G-rich promoter regions of DNA may explain why certain thalidomide analogs are not teratogenic while retaining their anti-tumor necrosis factor-alpha (TNF-alpha) activity, and suggests that we look elsewhere to explain the action of thalidomide on TNF-alpha. On the other hand, the anti-cancer action of thalidomide may be based on its antiangiogenic action, resulting from specific DNA intercalation. The tissue specificity of thalidomide and its effect against only certain neoplasias may be explained by the fact that various developing tissues and neoplasias depend on different angiogenesis or vasculogenesis pathways, only some of which are thalidomide-sensitive.

Hypothesis: thalidomide embryopathy-proposed mechanism of action.

We propose that thalidomide affects the following pathway during limb development: Growth factors (FGF-2 and IGF-I) attach to receptors on limb bud mesenchymal cells and initiate some second messenger system (perhaps SP-1), which activates alphav and beta3 integrin subunit genes. The resulting alphav beta3 integrin proteins stimulate angiogenesis in the developing limb bud. Several steps in this pathway depend on the activation of genes with primarily GC promoters (GGGCGG). Thalidomide, or a hydrolysis or metabolic breakdown product, specifically binds to GC promoter sites and inhibits the transcription of those genes. Inhibition of the genes interferes with normal angiogenesis, which results in truncation of the limb.

Axial and paraxial influences on limb morphogenesis

Previous studies by Stephens and McNulty and Strecker and Stephens have demonstrated that foil barriers placed between the mesonephros and lateral plate at stages 12 to 15 inhibited limb development, but foil barriers placed between the neural tube and somites at stages 11 to 12 resulted in limbs with normal skeletal patterns. It was concluded that some influence present in the paraxial region of the embryo at stages 11 to 15 is necessary for normal limb development. The present study was undertaken to localize that influence more precisely. Foil barriers were placed in the lateral edge of the somites or segmental plate of stage 10 to 15 chick embryos. Barriers placed into stage 13 to 15 embryos resulted in chicks with normal limbs, but barriers placed into stage 10 to 11 embryos resulted in chicks with defective limbs. Barriers inserted just lateral to Hensens node at stages 6 to 8 resulted in embryos with defective or absent wings. We also grafted stage 4 to 9 presumptive limb territories with and without Hensens node. Explants without Hensens node formed limb‐like structures in 1% of the cases. Explants with Hensens node formed limb‐like structures in 27% of the cases. When barriers were implanted and a node was placed on the lateral side of the barrier, limbs formed in 40% of the cases. These data suggest a medial to lateral progression of some as yet unknown morphogenetic influence necessary for normal limb development and we hypothesized that the influence may initially emanate from Hensens node.

Mesonephros has a role in limb development and is related to thalidomide embryopathy

Recent studies have demonstrated a link between limb reduction defects and mesonephros removal [Geduspan and Solursh, 1992) Dev. Biol., 151:242-250]. However, there is some question as to whether the limb-reduction defects seen in that study resulted from the removal of mesonephros or from the formation of scar tissue medial to the limb territory. The current study was conducted to test the hypothesis that elimination of the mesonephros without producing scar tissue adjacent to the limb will adversely affect limb morphogenesis. The hypothesis was tested by the insertion of tantalum foil barriers into various levels of the intermediate mesoderm of developing chick embryos to prevent the caudal elongation of the mesonephros. Limb reduction defects were obtained when the mesonephros was prevented from forming caudal to somite 14. No limb defects were seen when a foil barrier was placed into the intermediate mesoderm at the level of somite 21 or 25. Our results support the notion that a signal from the mesonephros is necessary for normal limb development. In addition, it appears that a craniocaudal factor emanating from the mesonephros plays a role in limb development. The limb reduction defects obtained in this study were also compared to the pattern of thalidomide embryopathy in humans. There is a close correspondence between the types of limb reduction anomalies seen with thalidomide and mesonephric blocks and between the severity of defects vs. the timing of thalidomide intake or mesonephric blockage. A model for possible thalidomide embryopathy is presented.

Determinants in the morphogenesis of muscle tendon insertions

The factors that normally determine the location and insertion of a muscle were explored in human experiments of nature with early problems in morphogenesis. Monozygotic conjoined twins for whom there could be no genetic determinants for muscle attachments at the sites of juncture served as one model; these attachments had to follow general principles of morphogenesis. A second type of problem involved absence of bone that presumably antedated muscle and tendon development (e.g., genetically determined radial aplasia). A third category included mechanical alteration of early limb position that may have occurred prior to the development of muscle attachments (e.g., early amnion rupture sequence). The dissection findings from all three types imply a general hierachy of muscle tendon attachments. Tendons appear to attach preferentially to bone. In the absence of the bone to which they would normally attach, they will attach to the next closest bone. If no such bone is available, tendons may attach to other tendons; and if no tendon is available, occasionally they will attach to the fascia of another muscle. If there is no connective tissue attachment site, there will be no muscle, implying a need for function in the development and preservation of muscle.

Paraxial and lateral plate influences on reinitiation of wing development in chicken embryos

Stephens et al. ([1993] Dev. Dynam. 197:157-168) hypothesized that the chick embryo wing territory is not an equipotential system. They proposed that following extirpation of the wing territory, limb formation is reinitiated, stimulated by more paraxial tissues. It was the purpose of this study to test this hypothesis. Four sets of procedures were undertaken in stage 11-12 chick embryos: 1) the wing territory lateral plate was removed (negative controls); 2) medial (between the central axis and somites), lateral (between the mesonephros and lateral plate), and intermediate (between the somites and mesonephros) foil barriers were placed adjacent to the wing territory (positive controls); 3) barriers were placed as above, with the removal of the lateral plate (experimental); and 4) barriers were placed lateral to the area of lateral plate extirpation (experimental). Normal limbs developed in 89% of the embryos without foil barriers with the lateral plate removed. In 93% of the embryos that contained medial barriers, normal wings developed whether or not the limb territory was extirpated, and in 100% of the embryos with intermediate or lateral barriers, with or without extirpation, deficient limbs occurred. When foil barriers were placed lateral to the wound, 87% of the chick wings were either reduced or absent. Closure of the wound following extirpation appeared to progress from lateral to medial. The data from this study appear to support the hypothesis that the limb territory is not equipotential but is reformed from cells closing the wound from lateral to medial, and is reinitiated from paraxial tissues (medial to lateral) following wound healing.

1204 DETERMINANTS IN THE MORPHOGENESIS OF MUSCLE TENDON INSERTIONS

Recent studies suggest that muscle originates from somites and tendons from lateral plate mesoderm. This study explored the factors which normally determine the location and insertion of a muscle, a previously unanswered question. Human experiments of nature with early problems in morphogenesis were used to determine how muscle development proceeds when known critical developmental factors are varied. In a variety of monozygotic conjoined twins for whom there could be no genetic determinants for muscle attachments at the sites of juncture, these attachments must follow general principles of morphogenesis. A second type involves absence of bone that antedated muscle and tendon development (e.g. radial aplasia). A third category includes mechanical alteration of early limb position prior to development of muscle attachments (e.g. early amnion rupture sequence). The dissection findings from all 3 types strongly imply a general hierarchy of muscle tendon attachments. Tendons appear to attach preferentially to bone. If the bone they would normally attach to is absent, they will attach to the next closest bone. If no such bone is available, they will attach to tendons, and if no tendon is available, occasionally they will attach to the aponeurosis of another muscle. If there is no connective tissue attachment site, there will be no muscle, implying a need for function in the development and preservation of muscle.

Radial condensation in the axis of the evolving limb

The origin of the tetrapod limb has been a topic of discussion and debate in the scientific literature for well over a century. At least two disparate approaches have been employed in an attempt to establish the origin of the limb. One is the evaluation of adult fossil remains and the other is the observation of the developing embryo. While both approaches obviously have their place in evolutionary biology, the latter has proved to be the most successful in elucidating apparent phylogenetic pathways. Darwin argued that embryology offered by far the strongest single class of facts in favor of evolution (cf., Gould, 1977; Oster and Alberch, 1982). Experimental embryology offers an even more valuable tool than does purely observational embryology (cf., Morgan, 1907). However, in spite of the importance of experimental embryology in evolutionary theory, Neo-Darwinian thought has largely ignored this discipline (Maderson et al., 1982). Recently there has been a trend among some investigators to apply the data and models derived from experimental developmental biology to approach some difficult problems of evolutionary theory. Concurrent with this trend is the tendency to discuss evolution in terms of changes in developmental programs rather than changes in visible (usually adult) structures (Waddington, 1975; Maderson, 1975; Oster and Alberch, 1982; Maderson, 1983). Hinchliffe and Johnson (1980), in their excellent book on the evolution and development of the vertebrate limb, have epitomized this approach. Their work, we believe, quite clearly and accurately articulates the current majority view among evolutionary biologists as to the evolution of the tetrapod limb. While we agree with the vast majority of what Hinchliffe and Johnson (1980) have to say, we propose, in this paper, to challenge the current views of the evolution of the limbs skeletal axis as reflected in their book.