Dinosaur’s feather and chicken’s tooth? Tissue engineering of the integument

Abstract

The integument forms the interface between animals and the environment. During evolution, diverse integument and integument appendages have evolved to adapt animals to different niches. The formation of these different integument forms is based on the acquisition of novel developmental mechanisms. This is the way Nature does her tissue/organ engineering and experiments. To do tissue engineering of the integument in the new century for medical applications, we need to learn more principles from developmental and evolutionary studies. A novel diagram showing the evolution and development of integument complexity is presented, and the molecular pathways involved discussed. We then discuss two examples in which the gain and loss of appendages are modulated: transformation of avian scale epidermis into feathers with mutated beta catenin, and induction of chicken tooth like appendages with FGF, BMP and feather mesenchyme.

This paper was first presented at the European Hair Research Society as the John Ebling lecture. One of the crucial issues in dermatology is how to make hairs and other skin appendages regenerate. Recent progress in molecular and developmental biology has shed new light on the development of feathers and hairs [1–4]. We have learned that a single piece of epidermis can be folded to form many different shaped epithelial organs. All of these appendages are the result of epithelial-mesenchymal interactions. These include hair, nails and teeth which protrude out of the body surface, and mammary glands and sweat glands, that invaginate into the body surface (Fig. 1A) [5]. The morphogenetic process is based on differential cell proliferation, apoptosis, and migration in time and space. These processes are regulated by a set of morphogenesis related signaling molecules and adhesion molecules [6]. Alterations of molecular pathway activities can lead to the formation of different types of appendages [7] that form as the result of variations on top of a common theme. Indeed integuments form the interface between animals and the environment and are subject to direct selection pressure. During evolution many types of epithelial appendages were produced and “experimented” on by Nature (Fig. 1B–D). These diverse integuments allow animals to adapt to different niches on the earth. Comparing the integument appendages of animals present today with those found through animal history provides insights into how different appendages evolved (Fig. 1E). Clinically, examples of homeotic aberrations in appendage phenotypes can be seen in patients with teeth growing from the neck and hair growing from the gum (Dr. Gorlin, personal communication). These examples and others suggest that epithelial stem cells have the capacity to become different organs and through erroneous regulatory mechanisms are shunted to other phenotypes. Ectodermal dysplasia, a genetic disease with abnormalities in hair, nails, sweat glands and teeth [8], presents further evidence that some components of the appendage-forming pathway are shared.

Figure 1.

(A) A prototype animal with various epithelial appendages. From Chuong, 1998 [5].

(B) Restoration of a Sinosauropteryx (Sinosauropteryx prima). Notice this theropod dinosaurs has elongated skin appendages which are hair like or some call proto-feathers. They are of the same type over the body and are more likely to be involved in keeping temperature rather than flight. The animal also has long legs, long tails, and sharp teeth. Adopted from National Geographic [48] based on Chen et al. [26].

(C) Restoration of a mesozoic bird Largirostronis sexdentoris. Notice this mesozoic bird has evolved different types of feathers over the body and should be a good flyer. It has a beak which still has teeth. Adopted from Hou et al. [49].

(D) Restoration of a Confuciusornis sanctus [30, 31]. Note the claws on the wing still remain but a beak without teeth has evolved. Adopted from Hou et al. [49].

(E) Conceptual model of integument appendage forms. Arrows imply the progression from simple to more complex integument appendage forms, with some possible underlying morphogenetic processes. Under each form, examples from current and ancient integument appendages are listed. Arrows do not indicate specific evolutionary steps, nor that one appendage is directly evolved from another. Sometimes evolution can occur in the opposite direction of the arrows when it is advantageous. For example, the loss of chicken teeth and wing claws. Glands, teeth and other epithelial organs can be complex too but are not elaborated because we focus on feathers here.

The human genome project is approaching an end. This accomplishment is equivalent to compiling the dictionary of Nature’s language. The next major challenge is to understand the grammar of this language of life. We wish to learn how Nature uses these words (genes) to write her articles and poems (organs in species that are present today and in history) in development and evolution. From the principles (grammar) we learn, we can then begin to write our own articles (tissue engineering), and guide skin stem cells to form skin, hairs, glands, teeth, etc. [9].

Phenotypic determination of epithelial appendages in embryonic development

During embryonic development, different epithelial appendages go through induction, morphogenesis and differentiation stages, and in some they also go through cycling/regeneration [10]. We have been using feathers as a model to study skin morphogenesis. We chose feathers because they are the most spectacular epithelial organs nature has ever produced. Feather formation initiates from a flat piece of epithelium [11]. Following induction, the feather primordia are formed, mainly composed of elongated feather placode epithelia and dermal condensations. From here, feathers start to protrude out and form feather buds. Feather buds are originally radially symmetric, but soon acquire anterior-posterior polarity through interactions with the epithelium. Feathers then start to elongate and develop the proximal-distal axis. Feather filaments then go through another layer of epithelial invagination and evagination to form the barb plates and marginal plates. Barb plate cells will be keratinized and become barbs, while marginal plate cells will die and become space between the barbs. Similar processes repeat in a fractal-like fashion to form barbules, and in this way branching morphogenesis occurs in feathers (Fig. 1E). Hair is formed in a different way. Following induction, the primordia form hair germs through invagination, but eventually also form hair follicles. Also in the hair filament, epithelial cells form the cortex and medulla without mesenchymal pulp in the center. The blood vessels in feather pulp may have allowed feathers to grow much bigger than hairs. There are other forms of integument appendages (Fig. 1E). Nails and scales form from protruding placodes, while glands form from invaginating epithelia.

Experimental embryology has led us to appreciate that the phenotypes of skin appendages can be inter-convertible. When epithelial-mesenchymal recombination is performed during the early developmental stages, the mesenchyme determines what type of appendages will form. First, in the induction stage (see above paragraph) the decision to make an appendage is made, setting a piece of epithelia to be different from the rest. Then, in the morphogenesis stage, the phenotypes are determined [1, 12, 13]. Abnormal retinoid pathway activities are known to cause epithelial metaplasia [14] as well as appendage phenotypes. When retinoic acid is added to the skin before phenotypes are irreversibly determined, scales are converted to feathers and vice versa, and hair germs are converted to gland-like structures. Regional differences of the Hox expression pattern in chicken skin prompted us to suggest that the Skin Hox code may determine the phentoypes of skin appendages [1]. In retinoic acid induced scale _ feather metaplasia, the expression of Hox-D13 in the scale region disappeared and became more similar to that of the feather dermis [15].

To appreciate the origin and progression of integument evolution and development, we have devised an integument complexity map (Fig. 1E) with invoked developmental pathways, or the “price” the species have to pay, on the X-axis. The appendage complexity, or the performance the species will gain, on the Y-axis. We will need to identify the molecular pathways behind the arrows if we want to learn how to engineer skin and skin appendages.

Dinosaur’s feather

During the more than 500 million years of vertebrate evolution, there has been major progress in integument evolution (evolutionary novelty) [16]. For example, the emergence of scales for protection, the establishment of a (water) barrier to prepare reptiles to live on land, the formation of follicles to put stem cells away for cyclic renewal, etc. (Fig. 1E). All these transitions represent evolutionary novelty that must come from the emergence of new genes and/or new wiring of developmental pathways. Some animals in evolution posses both the old and new appendage forms, representing an intermediate species (Fig. 1B–D). The evolution of birds from reptiles is one such case since the need to fly, and fly well, required a lot of “re-engineering” of the integument. Here we will use the reptile-bird transition as a paradigm for discussion.

The reptile integument is mainly made of scales [17]. In birds, there are scales on the foot and feathers on most of the rest of the body [18]. Feathers probably evolved because they provide novel functions in insulation, display (communication), and flight, characteristics that make the Aves unique [19–22]. The discovery of Archeopteryx (145 million years ago (MYA)) and other fossils led to the compelling Dinosaur-bird hypothesis, suggesting that birds evolved from the dinosaur [23]. There are some objections to the Dinosaur-bird hypothesis because differences in skeletal structures are not consistent with this hypothesis. These scientists consider that the bird and dinosaur share common ancestors, such as the basal archosaur, an ancient reptile [19, 20], but the “feathers” found on later dinosaurs result from later convergent evolution. In either case, one has to agree that feathers have evolved from the reptile integument, and there is a gradual transformation from the simple scale to the advanced forms of feathers (Fig. 1E, also [2, 24]).

Archeopteryx already has different types of feathers over the body, a toothed jaw, claws in the wing, and a bony tail. The flight feathers in the wing are asymmetric, suggesting that it could fly [25], although it may not have been an excellent flyer. Therefore in the reptile-bird transition, Archeopteryx represents a relatively advanced form with complex feathers. Can we find fossil records representing the earlier processes in feather evolution? More integument species of the mesozoic reptiles and birds will be needed. Unfortunately, in most fossils, only the bony skeleton is preserved. Recent new finds from several sites in China brought us exciting news because the preservation conditions for integuments are excellent. Sinosauropteryx [26] (Fig. 1B, about 120 MYA) has “fuzz fibers” on the body, especially along the dorsal midline. These filamentous “protofeathers” are about 20 (5–40) mm long and appear to be rather homogenous over the body without regional tract specificity. The filaments appear to be like down feathers without aerodynamic properties and were probably used for insulation. They are hollow, and appear to have a short shaft with “barbs”, but no barbules. The body shape is similar to the theropod. Theropods were carnivorous, bipedal, terrestrial dinosaurs with small forelimbs and special predatory features such as long hands with three digits for grasping prey [21, 27]. Beipiaosaurus also has similar primitive feather filaments that appear to be hollow, reflecting the cylindric developmental stages of the feather filament. They are longer, averaging about 50 mm long [28]. Caudipteryx [29] (145 MYA) has evolved different types of feathers over different regions of the body (Tracts have evolved). In both the wing and tail, they have spectacularly symmetric pennaceous feathers, probably used for display. It had vaned and barbed wing remiges, tail rectrices with tapering shafts, and plumulaceous feathers covering the body, most notably at the hips and the proximal region of the tail. They still have teeth. Protarchaeopteryx [29] (145 MYA) also has bilaterally symmetric pennaceous feathers. However, these feathers still lack the asymmetric vane required for flying. The tail rectrice feathers of Protarchaeopteryx are plumulaceous below the mid-shaft and pennaceous from the mid-shaft and above [29]. The vaned Protarchaeopteryx feathers appear to be structurally transitional between the hair-like structures of Sinosauropteryx and the modern feathers of Archaeopteryx. The continuous reptile-bird transition has led to the formation of many mesozoic birds. The crow sized Confuciusornis [30, 31] (around 140 MYA) has both down and flight feathers. The well developed asymmetric flight feathers and toothless beak suggests it flew well. The fossils even show they already have sexual dimorphism in the tail feathers. They have evolved beaks and have no teeth, and are considered to be more advanced than Archaeopteryx. These are representatives of reptiles/birds in evolution. Overall, the integument engineering during the reptile-bird transition included the formation of feathers, beaks, and the loss of teeth, wing claws, etc. Can we learn the molecular basis for these changes? This would have to be tested in a laboratory. Here we will use the chicken of today as a model to show one example of a gain of a pathway: growing feathers out from scale epidermis, and one example of the re-activation of a lost pathway: re-growing teeth from the chicken oral mucosa.

Studies in the recent 10 years by us and other laboratories have revealed the involvement of several major molecular signaling pathways in feather morphogenesis. The general order of appearance and functioning are FGF4, BMP4 => SHH, Wnt-7a => Notch-1, Serrate-1 and Delta-1 => Msx-1, -2 => Hox, NCAM [1, 11, 32–36], in which these pathways assume different functions including induction, mesenchymal condensation, localized cell proliferation, etc. While testing the effect of mis-expression of these morphogenesis related molecules, we were able to transform avian scale epidermis into feathers using constitutively active beta-catenin transduced by the avian retroviral vector, RCAS (Fig. 2B) [37]. Beta catenin is first expressed all over, and then segregates into individual primordia. During that process, they become stronger in the primordia area and absent in the interprimordial area. In scales, beta catenin is weak and diffuse. So we wondered whether enhanced expression of beta catenin in scale epidermis may make them progress into feathers. Experiments indicate that this is indeed the case (Fig. 2B). K14-beta catenin transgenic mice show new hair formation as well as hair follicle tumors [38]. Molecularly, beta catenin interacts with APC in the colon. Deletion of APC leads to colon polyps, which are also extra growths of epithelia. Thus activation of beta catenin is able to push the epithelia into a more active growth status (Fig. 2A). Along this line, activation of the delta pathway and suppression of the BMP pathway in scales also can induce some feathery scales [39]. These molecular pathways are likely to work in concert during scale _ feather metaplasia. We propose that parallel, but not necessarily identical, molecular processes may have taken place during avian evolution that took place since about 150–175 MYA, and initiated the formation of ancestral feathers.

Figure 2.

Transformation of avian scale epidermis into feathers with beta catenin. (A) Schematic drawing showing that beta catenin can cause epidermis to become more “activated” and the pathway may work in concert with other molecular pathways. (B) Whole mount view and tissue sections showing the feather follicles and filaments induced by RCAS beta catenin retroviral transduction. Adopted from Widelitz et al., 2000 [37].

Chicken teeth

Recently, we also tried to see if we can induce chicken tooth formation. The logic is that Mesozoic birds had teeth. In fact the loss of teeth is a relatively late trait. In the oral mucosa of the modern chicken, there is still formation of a dental lamina, but it soon degenerates (Fig. 3A). In situ hybridization shows that the chicken oral mucosa expresses Pitx2, Pax9, and FGF8, but is missing the expression of Bmp4, Msx1, and Msx2 genes. All these genes are expressed in the mouse tooth germ, and are critical for tooth formation. The tooth phenotype is similar to that of Msx1-Msx2 knockout mice [40]. In mouse tooth morphogenesis, there is a Bmp4-Msx1-Bmp4 pathway from epithelia to mesenchyme [41]. We therefore wondered whether this pathway became defective in the chicken, and whether we could rescue this process, at least partially, by supplying exogenous BMPs. BMP4-coated beads could indeed induce both Msx1 and Msx2 from the chicken oral mesenchyme. Morphologically, indeed, BMP4 could further enhance the morphogenesis of the chicken dental lamina. The addition of FGF has a more pronounced effect. Furthermore, the use of feather mesenchyme from dorsal skin has the most profound effect, leading to the formation of follicles (Fig. 3B). Are these formed epithelial appendages indeed teeth? Because of the lack of a chicken tooth marker (enamel), it is difficult to make a definite conclusion. However, it is obvious that the morphogenetic process of the dental lamina has advanced and formed follicles. Work with chimeric recombination has suggested that the mouse dental mesenchyme can induce the chicken enamel gene from the chicken oral mucosa [42], and mouse odontogenic epithelium can also induce chicken mandibular mesenchyme to exhibit some characteristics similar to dental mesenchyme [43].

Figure 3.

Induction of tooth-like appendages from chicken oral mucosa by feather mesenchyme. (A) A schematic drawing of normal tooth development. Normal chickens can form dental laminae (marked by 1), which can be promoted by FGF and BMP to the cap stage (marked 2), and by feather mesenchyme to form follicles (marked 3). (B) Recombination of a single piece of chicken oral mucosa (red) and aboral epithelium (green) is recombined with dermal mesenchyme from the trunk (blue). The results show oral mucosa, which otherwise will be smooth, form many tooth-like appendages. Sections show the epithelium invaginate to form cap like structures, and eventually form a nice epithelial appendage follicle. Adopted from Chen et al., 2000 [40].

Follicle formation is one major step in the evolution of epithelial appendages. Reptiles do not have follicles, and follicles are mainly seen in hair, feathers, and teeth. Another important message is that when the oral and chin epithelia are recombined with feather mesenchyme, one forms feather buds, while the other forms these tooth-like appendage follicles. This demonstrates that the epithelia are already different at this stage and will respond differently to the same mesenchymal signal. Epithelial cells are modulated in certain ways during development. In the beginning, they can be “molded” into any form of integument and integument appendages. This pluri-potentiality gradually becomes more limited. What are the molecules that accompany these changes in cellular potential? To answer these questions, we have to go back to the laboratory and try to find out what guides epithelial stem cells to form an epithelial organ.

Tissue engineering of the integument

Can epithelial stem cells be guided to form a new integument organ, such as hairs or teeth? Or asking the question in a different way, how different are these cells when they are derived from a dinosaur and a bird, i.e. can the cells from a dinosaur be incorporated to become part of a feather? Do the differences in integument phenotypes reside more in the different tissue interactions and less on the properties of single cells? As we learn more about how molecular cascades contribute to various morphogenetic processes, we come closer to determining how to build complex epithelial organs.

The long-term strategy of doing integument engineering and characterizing appendage precursor cells of mesenchymal and epithelial components is diagramed in Figure 4. Mesenchymal cells are dissociated and recombined with an intact piece of epithelium. After culturing, the cells were able to reorganize into evenly spaced feather primordia [44], and further could produce feather follicles. Mesenchymal cells can come from dissociated dermal papilla or the dermal sheath of adult vibrissa and be recombined with hair epithelia [45]. In contrast, dissociated epithelial cells can derive from the epithelia or hair follicles [46], and be recombined with mesenchymal components that contain inducing ability. With appropriate combinations and novel available molecular markers, we should be able to learn how these cells can be assembled to build complex integument appendages. One elegant example is the recent demonstration that Wnt signaling is important in maintaining the inducing ability of the dermal papilla [47]. More of these analyses will help us to establish the molecular cascade.

Figure 4.

Strategy of skin engineering using different stages of epithelial and mesenchymal precursor cells. Through recombination assay and gene alteration, we should be able to characterize the properties of different stages of epithelial and dermal stem cells, and learn to regenerate different integument structures (hair, teeth, glands, etc.) with different combinations. E: epithelial cells, also in blue color; M: mesenchymal cells, or dermal cells, also in orange color. Arrows represent progression and reversibility.

Tissue engineering is considered to be the prime science of the new century. We would like to learn how to guide epithelial cells to direct the formation of specific organs. In this overview, we pointed out that the best way is by learning how Nature does her tissue engineering in development and during evolution. By analyzing the molecular processes, we may now learn to apply these principles to integument appendages such as hairs, teeth, glands and other organs, so we will help to improve human health in the new century of biotechnology.