SEARCH

SEARCH BY CITATION

Keywords:

  • integument;
  • feathers;
  • scales;
  • BMP7;
  • competence;
  • induction;
  • Silkie;
  • scaleless;
  • heterochrony

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

The induction and specification of a large number of vertebrate organs require reciprocal signaling between an epithelium and subjacent mesenchyme. In the formation of integumentary organs, the initial inductive signaling events leading to the formation of the organ primordia stem from the mesenchyme. However, the epithelium must have the capacity to respond to these signals. We demonstrate that bone morphogenetic protein 7 (Bmp7) is an early molecular marker for epidermal organ development during development of feathers and scales of the chick. Bmp7 is expressed broadly in the preplacode epidermis and subsequently becomes localized to the forming placodes of feathers and scales. An examination of Bmp7 expression in the scaleless mutant chicken integument indicates that Bmp7 expression in the epidermis is associated with the ability to form epidermal organs. We show that BMP7 function is necessary for the formation of epidermal placodes in both feather and scale forming epidermis. In addition, precocious expression of Bmp7 in the metatarsal epidermis of the Silkie mutant or treatment of the metatarsus with ectopic BMP7 protein results in feather development from scale forming integument. From these data, we propose that Bmp7 is necessary and sufficient, in a developmental context, to mediate the competence of an epithelium to respond to inductive signals from the underlying mesenchyme to form epidermal organs in the chick. We propose that regulation of Bmp7 in localized areas of the embryonic epidermis facilitates the development of regional formation of integumentary organs. Developmental Dynamics 231:22–32, 2004. © 2004 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

The integument, or skin, is an important developmental model to investigate the signaling associated with organ induction and differentiation, in part, due to the varied organ types that form from similar tissues and common early inductive processes. Within the integument of amniotes, epidermal organs (e.g. feather, scale, hair, and gland) are formed from the response of an epithelium to inductive signals from the underlying dermal mesenchyme. However, in parallel, the epithelium must be regionally competent to react to these signals (for review see Sengel, 1984). Heterospecific recombination studies, in which epidermis from one species is combined with the mesenchyme of another species (Dhouailly, 1975), indicate that epidermal organs of amniotes share developmental and molecular mechanisms of competence and induction within the integument. The molecular nature of the signaling mechanisms involved in these processes and their integration during induction of epidermal organs is generally unknown.

The study of epidermal organ induction in birds has been aided by the use of specific mutations that affect the extent of feather or scale formation. In particular, the scaleless mutant that exhibits a loss of function of the metatarsal epithelium to respond to dermal signals has played an important role in dissecting the inductive roles of early appendage formation (e.g., Goetinck and Abbott, 1963; Sengel and Abbott, 1963; Brotman, 1977b; McAleese and Sawyer, 1981). In conjunction with gene expression studies, scaleless has provided insights into how molecular mechanisms are linked to developmental events in the formation of feathers and scales (e.g., Shames et al., 1991; Zeltinger and Sawyer, 1992; Song et al., 1994, 1996; Vaillet et al., 1998; Widelitz et al., 2000). There are several chicken breeds that exhibit extensive feather formation along the shank and digits (e.g., Bantams, Cochins, and Silkies) that also provide useful models for study of the initial induction of epidermal organs of chick. The allele(s) involved in these traits affect both the mesenchyme and epidermis as shown in recombination studies (Goetinck, 1967; Somes, 1992). The molecular nature of these alleles is not known nor are the signaling systems or developmental processes that they disrupt.

It is known, however, that Wnt signaling plays a common role in early molecular specification of epidermal organs. β-catenin is expressed in early placode stages of hair, feathers, and avian scales. Inappropriate activation of β-catenin signaling in the epidermis leads to ectopic development of hair and feathers (Gat et al., 1998; Noramly et al., 1999; Widelitz et al., 2000), while targeted inactivation of β-catenin in the epidermis of mice causes agenesis of hair (Huelsken et al., 2001). Consistent with these data, WNT- and LEF-1–mediated signaling are also required to maintain development of hair (van Genderen et al., 1994; Kishimoto et al., 2000) and are involved in the early induction and patterning of feathers (Noramly et al., 1999; Widelitz et al., 1999). Lef-1/TCF and β-catenin form a transcriptional complex that is sensitive to signaling by WNT ligand–receptor complexes (Eastman and Grosschedl, 1999). In the mouse, a Lef-1 deficiency results in loss of whiskers, teeth, hair, and glands (van Genderen et al., 1994). In addition, transgenic overexpression of Lef-1 is sufficient to cause ectopic hair growth in oral epithelia (Zhou et al., 1995). These data demonstrate an active role of β-catenin and Lef-1 in the development of epidermal organs. The maintenance of hair growth by WNT signaling (Kishimoto et al., 2000) and the attenuation of hair development by reduced WNT signaling (Andl et al., 2002) both complement and support the results on β-catenin/Lef-1 in hair development. The emerging model is that WNT signaling is essential for the proper development of early-stage epidermal organs and that this signaling integrates β-catenin/Lef-1–mediated transcription. However, the role of these signaling systems, alone or in combination, in the particular events of induction and organ specification is still poorly understood.

We examined early gene expression in the developing feathers and scales of the chick to define early components of epithelial–mesenchymal inductive mechanisms for epidermal organ specification. We found that bone morphogenetic protein 7 (Bmp7) is expressed in the epidermis before, or concomitant with, genes involved in placode induction. We propose that BMP7 signaling mediates the competence of the epidermis to respond to inductive signals from the dermis to form feather and scales. The function of BMP7 in mediating competence of an epidermis may be a component of a conserved mechanism of establishing competence in a whole class of epithelial-derived organs that form from epithelial placodes.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Bmp7 Expression Regionally Defines Preplacode Feather and Scale Forming Epidermis

Feathers and scales form in an invariant pattern in predictable domains on the embryonic chicken integument (Lucas and Stettenheim, 1972; Dhouailly et al., 1980; Tanaka and Kato, 1983; Mayerson and Fallon, 1985). Through reciprocal signaling between the epidermis and the mesenchyme, a hexagonal pattern of feather placodes becomes specified (see Fig. 1G). Scale development shows a similar process of induction and patterning within the metatarsal integument. The first scale primordia form at the digit 3 and 4 phalangeal–metatarsal boundary (Sawyer, 1972; Dhouailly et al., 1980). The formation of scutate scales progresses proximally along metatarsal three and four and laterally to digit 2 (Fig. 1D). Scale induction on the phalanges occurs as a secondary inductive event separate from initial scale induction along the metatarsal epidermis (Dhouailly et al., 1980).

thumbnail image

Figure 1. A–H: Developmental progression of Bmp7 expression in developing scales (A–D) and feathers (E–H) of the chicken. A,E: Hamburger and Hamilton stage (HHs) 32 limb and HHs 27 thigh do not show specific staining for Bmp7 in the metatarsal or femoral tracts that form scales and feathers, respectively. B,F:Bmp7 is expressed in the epidermis of forming scale (embryonic day [E] 8) and feather fields (HHs 29) in a diffuse manner (open arrowheads). The expression of Bmp7 becomes refined to the forming placodes (arrows) and down-regulated in interplacodal epidermis. Bmp7 expression occurs before histological indication of an epithelial placode structure. C: Secondary sites of scale induction occur on the digits as seen (E9). D and H outline the developmental progression of Bmp7 expression on the metatarsal epidermis (D) and thigh feather tract (H). Divergence of feather and scale development at the placode stage of induction is marked by Bmp7 expression. I,J: Histological sections of scale (E9) or feather placodes (HHs 30) at comparable developmental stages show that Bmp7 is expressed in the epithelium of scales (I), whereas feather placodes show expression in both the epithelium and the subjacent mesenchyme (J). ep, epidermis; me, mesenchyme.

Download figure to PowerPoint

To investigate the early inductive events in the formation of feather and scales, we analyzed the expression of genes known to be involved with the induction of other epidermal organs. Bmp7 showed an interesting expression profile that matched the timing of early genes, such as β-catenin, that are thought to act in the specification of the placode epithelium (Noramly et al., 1999; Widelitz et al., 2000; see below).

Bmp7 had two distinct expression modes in the development of both feathers and scales. There was an initial diffuse epidermal expression in early feather and scale appendage forming regions of the integument (arrowheads, Fig. 1B,F). Subsequently, this diffuse expression formed into refined foci of Bmp7 expression surrounded by regions of reduced expression (Fig. 1C,G; Patel et al., 1999). The regulation of epidermal Bmp7 expression, from diffuse to patterned foci, was similar in both feather and scale appendage fields and had an invariant pattern of expression that presaged the temporal appearance of the placodes (Fig. 1D,H). Thus, on the metatarsus of an embryonic day (E) 8 chick, Bmp7 expression first was detected just proximal to the digit 3 and 4 metatarsal boundary where the first scale placodes form (Sawyer, 1972; Dhouailly et al., 1980). Similarly, Bmp7 was diffusely expressed along the primary inductive row of feathers in each tract and becomes specified into placodes as expression moves laterally over the epidermis (Fig. 1F; see Patel et al., 1999).

Histological sections of feather forming integument showed epithelial expression of Bmp7 that correspond to the preplacode expression seen in whole-mount (data not shown). Bmp7 expression became localized in the epithelium at sites of placode formation. In addition, the expression of Bmp7 was seen underlying the feather placode in the subjacent mesenchyme (Fig. 1J). In contrast, Bmp7 was expressed in the epidermis but not detected in the mesenchyme of scale placodes (Fig. 1I). This difference in expression during the initiation of placode formation between feather and scales indicates an early divergence of the developmental specification of these epidermal organs.

Hierarchy of Genes Involved in Epidermal Organ Development of the Chick

Establishing gene expression hierarchy using hemisected embryos.

The hierarchical basis of early signaling in epidermal organ induction is only beginning to be understood. To address this question further, we analyzed the timing of expression of genes thought to act early in the induction of feather tracts of the chick. Expression of candidate genes was compared with the expression of Bmp7 within contralateral feather fields of hemisected embryos. Embryos of the same developmental stage can show variation in the extent of feather formation in a particular tract. However, contralateral tracts on the same embryo develop comparably. By comparing gene expression in hemisected embryos, we were able to devise temporal hierarchical relationships between genes and determine at what stage Bmp7 acts during the induction of epidermal organs (see Fig. 2 and text below).

thumbnail image

Figure 2. Analysis of the temporal expression of genes involved in early feather development. The timing of the expression of candidate genes involved in early epidermal signaling was compared by analyzing the spatial extent of their expression within contralateral feather fields of hemisected embryos. A,A′:Bmp7 expression (A) compared with Wnt7a expression (A′) in Hamburger and Hamilton stage (HHs) 29 pectoral feather tract shows that Bmp7 is expressed in expanded domains within the feather field. Because feather development is initiated at a discrete row within each feather field and expands laterally through the field as development proceeds, Bmp7 is expressed earlier in the development of feathers than Wnt7a. B–C′: By using a similar comparative analysis in feather tracts of HHs 30 hemisected embryos, Bmp7 is expressed concomitantly with β-catenin in both paired pectoral (B,B′) and thigh (C,C′) feather tracts.

Download figure to PowerPoint

Hierarchy of gene expressions in placode induction.

Initial observations indicated that Bmp7 was expressed before genes involved in placode formation such as Bmp2 and Sonic Hedgehog (Shh; data not shown). Wnt7a, previously shown to be a mediator of placode development in feathers (Widelitz et al., 1999), exhibited a similar early preplacode expression profile as Bmp7 within feather fields. However, a comparison of Bmp7 expression to that of Wnt7a in right and left forming pectoral feather tracts within the same embryo indicated a significant difference in the timing of their expression as shown by the extent of expression within the feather field (Fig. 2A). In contrast, a similar analysis of β-catenin expression in both the humeral and thigh feather tracts showed little difference from that of Bmp7 (Fig. 2B,C), suggesting that these two genes act contemporarily in preplacode specification of feather forming integument. These data suggest a chronology of gene expression within feather-forming integument in which Bmp7 and β-catenin are expressed in preplacode epidermis before Wnt7a- and placode-specific genes such as Bmp2 and Shh.

Experimental Dissection of the Paracrine Regulation of Bmp7 Expression

Initial expression data reported above indicate that Bmp7 is temporally and spatially regulated during early feather and scale formation. Epidermal expression of Bmp7 preceded the initiation of Bmp7 expression in the mesenchyme of the forming feather placode. Histological sections also showed a difference in the expression of Bmp7 in the mesenchyme between feather and scale placodes (Fig. 1). This finding suggests that signaling from the epidermis at sites of feather formation may regulate mesenchymal Bmp7 expression. Because Wnt7a and genes associated with WNT signaling (e.g., Lef-1, β-catenin, and APC; Widelitz et al., 2000) are present in both preplacode and placode epidermis, we centered our work on the putative role of WNT signaling in regulating Bmp7 in the mesenchyme.

By using feather skin explants grown in culture, we were able to assess the role of WNT signaling on Bmp7 expression using pharmacological mediators of WNT signaling and assaying for Bmp7 expression by whole-mount in situ hybridization (WMISH). Control explant cultures formed patterned arrays of feather-like placodes that expressed Bmp7 (Fig. 3A). Histological sections showed that Bmp7 was expressed in the epidermis and subjacent mesenchyme of the placodes (Fig. 3A′). Treatment with CKI7, an inhibitor of WNT signaling (Peters et al., 1999; Gao et al., 2001), had a minor effect on the patterning of placodes across the explant causing apparent fusions of placodes (arrowheads, Fig. 3B). However, histological sections of placodes revealed a loss of detectable mesenchymal expression of Bmp7 in the placodes (100%, n = 2; see Experimental Procedures section) compared with controls (0%, n = 4; compare Fig. 3A′ with B′). To complement the loss of WNT signaling experiment, we next treated explants with LiCl, which activates WNT signaling (Stambolic et al., 1996). Explants treated with LiCl caused global mesenchymal expression of Bmp7 throughout the explant and a failure to form placode foci in these cultures (75%, n = 4; Fig. 3C and C′) compared with untreated controls (0%, n = 4; Fig. 3A and A′). There was a general decrease in Bmp7 expression observed in the epidermis after LiCl treatment. It is unknown if this is a specific response due to LiCl in the epidermis or a consequence of the change in mesenchymal signaling. The complementary response of LiCl and CKI7 treated explants in the regulation of Bmp7 mesenchymal expression supports the function of these treatments in mediating WNT signaling in the integument during feather induction.

thumbnail image

Figure 3. Control of mesenchymal Bmp7 expression in developing feather tracts. The regulation of Bmp7 expression in feather integument was tested using explant cultures of thigh feather tracts in the presence of pharmacological modifiers of WNT signaling. A: Control explants form patterned arrays of placodes marked by Bmp7 expression. A′: Histological sections show that Bmp7 is primarily in the mesenchyme and is present in the epidermis as well (filter: ft; the dotted line marks the border of the explant on the filter). B: Treatment with CKI7 results in placode formation similar to wild-type, however, with a high percentage of placodal fusions (arrowheads). B′: However, histological sections show that Bmp7 expression is limited to the epidermis. C,C′: In contrast, treatment with LiCl results in Bmp7 expression throughout the mesenchyme (C′); however, this expression is not patterned (C). The expression of Bmp7 in the epidermis and mesenchyme is genetically separable. D,D′: Expression of Bmp7 in wild-type thigh feather tracts is in both the epidermal and mesenchymal components of a forming placode. E,E′: Thigh tracts of the scaleless allele (sc/sc) show expression and regulation of Bmp7 expression in the forming feather tract (E), however, the expression is limited to the epidermis (E′). F–G′: Treatment of scaleless thigh tracts with LiCl is sufficient to induce Bmp7 throughout the explant mesenchyme (G,G′) in contrast to KCl treated controls (F,F′). Thus, the expression of Bmp7 in feather formation can be genetically and functionally separated into two distinct regulated domains, of which the mesenchymal expression is mediated by WNT-dependent signaling and inhibited by the scaleless allele. ep, epidermis; me, mesenchyme.

Download figure to PowerPoint

Lessons From scaleless: Bmp7 and Epidermal Competence

In addition to pharmacological dissection of the regulation of Bmp7 expression seen in feather placode induction, we also found genetic evidence that the regulation of Bmp7 expression is essential for feather and scale development. The scaleless mutant is a Mendelian recessive allele (Abbott and Asmundson, 1957) that affects the function of the epidermis as shown by recombination studies in which normal epidermis rescues the mutant phenotype (Goetinck and Abbott, 1963; Sengel and Abbott, 1963). Histological sections of scaleless thigh feather tracts showed no detectable Bmp7 expression in the mesenchyme underlying Bmp7-positive epidermis (Fig. 3E and E′). This expression of Bmp7 in scaleless mimicked the loss of Bmp7 expression caused by blocking WNT signaling (Fig. 3B and B′). Following up on this, we treated scaleless thigh tract integument in culture with LiCl and found this led to a restoration of Bmp7 expression throughout the mesenchyme (100%, n = 3; Fig. 3G and G′) when compared with controls (0%, n = 4; Fig. 3F and F′). This finding indicated that the effect of the scaleless gene product on Bmp7 mesenchymal expression is rescued by activation of WNT signaling. These data suggest that the scaleless gene product mediates WNT paracrine signaling from the feather epidermis to regulate mesenchymal Bmp7 expression.

The scaleless mutation affects the morphogenesis of epithelial organs in a pleiotropic manner. A small number of feathers are formed in all tracts with varied abnormal morphologies, but no scales are formed (Abbott and Asmundson, 1957). Thus, the scaleless gene affects feather and scale development at different stages in their induction. Bmp7 is expressed diffusely in each scaleless forming feather tract, but its expression never refined into discrete placodes of expression as found in wild-type epidermis (compare Fig. 4A,B; with Fig. 1 and Patel et al., 1999). This expression pattern of Bmp7 during feather field specification in scaleless was similar to that seen for β-catenin (Noramly et al., 1999; Widelitz et al., 2000). The forming scaleless feather fields showed that the feather tracts are defined by Bmp7 expression, whereas the apteryic (nonfeather forming) regions did not show Bmp7 expression. This regional expression of Bmp7 is in contrast to scaleless metatarsal epidermis in which Bmp7 is expressed at only low levels in the epidermis, and which correlates with the lack of induction of scales (Fig. 4C,D). Thus, in scales, the scaleless allele inhibits the early epidermal expression of Bmp7, and the mutation results in the lack of scale formation; whereas scaleless feather tracts express Bmp7 in the epidermis and are able to form a small number of feathers, although mispatterned within the tract. We conclude that the expression of Bmp7 in the epidermis correlates with the ability to form epidermal organs. Of interest, β-catenin was expressed in scaleless shanks in a manner similar to wild-type (Fig. 4E), providing evidence that β-catenin functions upstream of, or parallel to, Bmp7 in the epidermis during preplacode specification and that the scaleless gene product is not necessary for preplacode β-catenin expression within the epidermis.

thumbnail image

Figure 4. The scaleless gene and initial specification of secondary fields in the chicken integument. A,B: Hamburger and Hamilton stage 31 scaleless embryos show expression of Bmp7 in the epidermis throughout the forming feather tracts: fe, femoral; hu, humeral; pc, pectoral; sp, spinal; rt, rectrices. Of interest, the expression looks bounded and demarcates the extent of forming feather tracts (cf. Mayerson and Fallon, 1985). C,D: Expression of Bmp7 in wild-type (arrows, C) and scaleless (D) shanks demonstrates a lack of early Bmp7 expression in forming scale fields in scaleless epidermis. Insets in C and D show epidermal expression of Bmp7. E: Notably, β-catenin expression is seen in comparably staged scaleless shanks, suggesting a function before Bmp7 and independent of the function of the scaleless gene product. Expression of β-catenin in wild-type shanks was identical to the pattern in scaleless and is not shown. Scale bar = 0.1 mm in insets.

Download figure to PowerPoint

Bmp7 Is Necessary for Placode Formation in Feathers and Scales

The above data suggest that epidermal Bmp7 expression may be necessary for the initial induction of epidermal placodes. To test this hypothesis, we treated explant cultures of thigh and metatarsal epidermis with a blocking antibody to BMP7 to abrogate BMP7 function in these tissues (see Wawersik et al., 1999, and references therein). Control explants showed feather and scale placode development as shown by WMISH detection of patterned Bmp7 expression (Fig. 5A,D, each 100%, n = 4). BMP7 antibody treatment led to an overall inhibition of placode formation in both feather and scale forming integument (Fig. 5B, 0% n = 8; Fig. 5E, 25% n = 4), while explants treated with a control antibody developed normally (feathers: Fig. 5C, 100%, n = 8; scales: Fig. 5F, 75%, n = 4). Thus, BMP7 function in the epidermis is necessary for the induction of both feather and scale placodes in developing skin.

thumbnail image

Figure 5. Bmp7 is necessary for placode formation. Hamburger and Hamilton stage (HHs) 30 thigh and HHs35 dorsal metatarsal skin explants were grown in the presence of BMP7 blocking antibody or LCAM control antibody and assayed for Bmp7 expression by whole-mount in situ hybridization as a measure of placode specification. A,D: Untreated explants show the formation of placodes and small feather buds (A) and scale rudiments (D). B,E: The treatment of explant medium with a BMP7 antibody, known to block the function of the ligand, inhibited feather (B) and scale placode formation (E) as measured by Bmp7 expression. C,F: Antibodies to LCAM did not show inhibition of Bmp7 expression and placode formation in feathers (C) or scales (F) and serve as an immunoglobulin control.

Download figure to PowerPoint

The scaleless phenotype can be rescued by wild-type epidermis indicating that normal inductive signals are present in the mutant mesenchyme (Goetinck and Abbott, 1963; Sengel and Abbott, 1963). The scaleless scale phenotype suggests that the lack of Bmp7 expression in the metatarsus results in the loss of the mutant epithelium's ability to respond to mesenchymal inductive signals. This hypothesis is supported by the failure of placode formation in both feather and scale wild-type skin explants after inhibition of BMP7 protein function. The function of Bmp7 in mediating competence to the epidermis is further supported by studies of the feathered shank (ptilopodous) phenotype of the chick mutant Silkie discussed below.

Lessons From Silkie: Bmp7 Is Sufficient to Mediate Competence of the Metatarsal Epidermis

The ptilopodous phenotype of the Silkie breed is due to a dominant allele (Pti) that causes feathering of the shanks and toes (Somes, 1992). Feathering on the shank in Silkie occurs in a developmental sequence starting from digit 4 and the proximal metatarsal and subsequently across the foot toward digit 3 (Fig. 6A). We examined Bmp7 expression in the shanks of Silkie embryos and found precocious epidermal expression of Bmp7 along digit 4 and proximally down the associated metatarsal (Fig. 6B). This expression occurred before the endogenous expression of Bmp7 seen at the metatarsal/phalanx border of digits 3 and 4 during normal scale induction (see Fig. 1B). As development proceeded, circular foci of Bmp7 expression, indicative of initial feather placodes, formed along the lateral shank and digit 4 (Fig. 6C). Further developmental specification of the placode pattern along the shank showed a merging of circular and oval, scale-like, placodes along metatarsal 3 (Fig. 6D). This transition from feather to scale placodes is the region where the feather to feather–scale morphology is seen in older Silkie shanks (Fig. 6A).

thumbnail image

Figure 6. The Silkie Pti allele and the early induction of Bmp7 and epidermal competence. A: Analysis of the formation of feathers on the metatarsus of embryonic day (E) 14 Silkie shows feathers (I, II, III, and IV mark the digits and associated metatarsi). Medially, the phenotype is less severe, showing shorter feathers and more obvious feather–scale combinations. B: The formation of the first feathers on the Silkie metatarsal epidermis is correlated with precocious Bmp7 expression in a stripe along digit 4 and lateral metatarsus 4 (E8; arrows). This expression of Bmp7 is before that associated with scale development (see Fig. 1B). Histological sections reveal that the expression of Bmp7 is in the epidermis at this stage (B inset, III and IV mark the respective metatarsals in the section). C,D: As development proceeds, feather-like placodes form along the lateral metatarsal (C) and start to fuse into more “scale-like” ellipsoid placodes along the medial metatarsal (D). To test whether BMP7 is sufficient to promote feather development in metatarsal epidermis, recombinant BMP7 protein was applied to the dorsal metatarsal surface of wild-type chick embryos in ovo. E–H: Control embryos show characteristic scale shape and pattern at E14, whereas BMP7 protein-treated shanks show formation of feathers from scale primordia along digit 4 and lateral metatarsus (see arrows in H and inset). Note the formation of barb ridges in the ectopic feathers (histological sections H inset). The effect of BMP7 protein is stage-dependent. H: Stage 31–32 (100%, n = 6). I: Stage 33 (0%, n = 3). Control shanks do not show any feathering at any stage (G, 0%, n = 4). Note novel scale morphologies in stage Hamburger and Hamilton stage 32 BMP7 protein-treated shanks (compare E with F). Scale bars = 0.1 mm in insets.

Download figure to PowerPoint

We tested the hypothesis that the Silkie phenotype is an effect of the alteration of epidermal competence overlying the lateral metatarsus by precocious Bmp7 expression before the onset of scale induction. A corollary to this hypothesis is that the underlying shank mesenchyme has inductive capacity to induce feathers but requires, in tandem, a competent epithelium to initiate appendage development (see Rawles, 1963; Goetinck, 1967). To examine the role of Bmp7 in mediating competence in the metatarsal epidermis, we applied BMP7 recombinant protein to the shank of wild-type (White Leghorn) embryos in vivo by using a polymer, pluronic acid F127 (see Experimental Procedures section), for slow release and localization of the protein. Application of BMP7 protein across the shank using this method resulted in the formation of feathers from lateral scutate (dorsal) and scutellate (dorsal–lateral) scales, indicating that epidermal BMP7-mediated signaling is sufficient to initiate precocious appendage formation along the wild-type shank, forming feathers (Fig. 6E–I, see figure legend for percentages). This effect was localized to the lateral margin of the shank, indicating a restricted effect dependent on the region of the metatarsus and not only on BMP7 treatment. Recombinant BMP7 protein also caused interesting transformations of the dorsal scales of the shank into novel morphologies (Fig. 6E,F). BMP7 protein-treated scales showed growth along the distal and lateral margins of the scale and a general change in scale shape. The ability of BMP7 protein to permit precocious feather formation on the shank was stage-dependent (Fig. 6G–I). This finding may represent a shift in the nature of the inductive signals present in the metatarsal mesenchyme from feather- to scale-specific cues, as the normal scale developmental program begins (Rawles, 1963).

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Developmental Cascades and Early Divergence Between Epidermal Organs

Among the genes expressed in early, naive chick epidermis are those involved in BMP signaling (including Bmp7, follistatin, gremlin; Patel et al., 1999; Ohyama et al., 2001) and twisted gastrulation (S. Maas, personal communication), as well as those active in WNT signaling (β-catenin and Lef-1; Widelitz et al., 1999, 2000). The timing of the expression of these genes is similar between feather and scale forming epidermal fields, thus supporting the conservation of early molecular mechanisms of induction between these homologous appendages (Widelitz et al., 2000, and data not shown). Here, we describe the expression of Bmp7 in the preplacode epidermis of both forming feather and scale fields and necessity of BMP7 function in the formation of placodes. We further provide evidence that BMP7 is functioning to mediate competence of the epidermis to respond to dermal inductive signals.

Our data on feather and scale development in the chick suggest a model of early epithelial specification by regulation of Bmp7 expression (Fig. 7). The early, preplacode expression of Bmp7 is dependent on epithelial responsiveness to an unknown signal (UNK, Fig. 7) as shown by the dependency on the function of the scaleless (sc) gene product in the metatarsal epidermis. Regulation of the preplacode expression of Bmp7 in the metatarsal epidermis is also affected by the ptilopody (Pti) allele of Silkie, leading to precocious expression along the lateral shank (Fig. 7). In both feather and scales, the expression of Bmp7 is coincident with that of β-catenin; however, it is unknown whether they function in parallel or in combination during feather and scale induction. The function of β-catenin at this stage remains enigmatic because of the limited data on WNT ligands expressed in the early appendage field. Nevertheless, we establish that the mechanisms that control regional expression of β-catenin in the skin are activated concomitantly with Bmp7. Subsequently, during feather formation, Bmp7 is expressed in the mesenchyme underlying the placode by canonical WNT signaling. This paracrine regulation of Bmp7 in the feather mesenchyme is also dependent on the scaleless gene product. The expression of Bmp7 in mesenchyme during feather formation is coincident with formation of the dermal condensation, and it is likely that it plays a role in mediating dermal signaling to the epidermis during further specification of the placode and patterning within the feather tract, but this hypothesis awaits further analysis.

thumbnail image

Figure 7. A model of the early molecular events of induction and competence in early epidermal appendage development. The schematic outlines the regulation of Bmp7 expression in the epidermis and subjacent mesenchyme of early feather and scale development. Processes that are specific to feather and scale development are noted in gray and light gray text, respectively. The competence of an epidermis to respond to inductive signals is mediated by Bmp7. Bmp7 expression is closely linked with signaling mediated by β-catenin in the preplacode epidermis of both feather and scales. It is not clear if Bmp7 and β-catenin act in parallel or in concert to specify the character of the preplacode epidermis. Initial epidermal expression of Bmp7 in scales requires an induction that is mediated by the Silkie Ptilopody (Pti) gene and dependent on the scaleless gene product (sc); the molecular basis of this induction is unknown (UNK). During placode phase of feather formation, WNT-dependent, paracrine signaling mediated by the scaleless gene product regulates Bmp7 expression in the subjacent mesenchyme of the feather. In scales, which exhibit similar early gene regulation as feather fields, mesenchymal Bmp7 expression is not detected. Arrows with dotted lines connecting Bmp7 and Wnt in epidermis indicate a temporal hierarchy based on expression in preplacode epidermis and inferred relationship between the two genes. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com]

Download figure to PowerPoint

It is noteworthy that our model is consistent with recent analysis of β-catenin mediated signaling in the initiation and patterning of hair (Huelsken et al., 2001; Lauikkala et al., 2002), as well as recent data suggesting a mechanistic link between BMP and WNT signaling in hair and mammary gland formation (Jamora et al., 2003). The timing of Bmp7 expression in the formation of epidermal organs of the mouse (M.P. Harris, unpublished observations) suggests it may have an early role in placode induction in the mouse as well. Bmp7 is expressed in the early formation of the otic placode (Groves and Bronner-Fraser, 2000) and is necessary for induction of epibranchial (Begbie et al., 1999), and lens placode formation (Wawersik et al., 1999). The similarity of the role of Bmp7 in the induction of ectodermal placodes is suggestive of a common molecular mechanism of competence in these structures.

Epidermal Competence in Secondary Fields

Previous reports have proposed a repressive role of BMP7 during feather formation (Patel et al., 1999). Our data specifically demonstrate that BMP7 function is necessary for placode formation of both feathers and scales, thus supporting the conclusion that BMP7 function acts to promote the development of these integumentary structures at the earliest stages. In addition, the comparison of expression of Bmp7 between scaleless feather tract and metatarsal epidermis is consistent with a model of Bmp7 mediating the response of the epidermis to a mesenchymal inductive signal during epidermal organ development. It is likely that the repressive function of ectopic BMP7 protein on feather formation as described by Patel et al. (Patel et al., 1999) may be due to high levels of BMP7 (333–660 μg/ml) used having dysmorphogenic and/or nonspecific effects, because BMP2 and 4 showed similar results. Alternatively, the BMP7 effect on placode formation in the Patel study may be secondary, regulating spacing within the feather field and not regulation of early placode formation per se (Patel et al., 1999). The loss-of-function experiments described in our study clearly indicate a necessary role for BMP7 in the early specification of feather and scale placodes. The finding that Bmp7 is expressed precociously along the region that feathers are induced in Silkie supports the conclusion that Bmp7 is acting to promote appendage formation in the shank. Furthermore, our experimental work applying exogenous BMP7 to wild-type shanks showing only localized feather induction along the shank suggests that the effect of Bmp7 is permissive for induction. Thus, we conclude that Bmp7 is acting to promote placode formation in the epidermis.

The molecular nature of competence has been a central question in the study of epithelial–mesenchymal signaling of both primary and secondary fields (Gurdon, 1987). Competence of the epidermis during placode formation is likely to be determined by the combined function of several signaling cascades that endow the tissue with temporal and spatial control of this property. Our data support the role of BMP7 signaling in the regulation of competence of the epidermis. We demonstrate that Bmp7 is necessary for initiation of placode formation in the epidermis of both feather and scales and sufficient to permit appendage growth precociously in response to dermal inductive stimuli. BMP7 is an extracellular signaling factor; thus, its effects would be predicted to act in a juxtacrine manner, in essence, priming the tissue to respond to paracrine inductive signaling from the mesenchyme (cf. Goetinck, 1967). The regulation of Bmp7 expression in the epidermis, thus, uncovers a molecular mechanism to regulate epidermal organ growth into defined regional tracts of the bird (feather tracts [pteryla] and localized scale fields) by temporal and spatial regulation of the competence of the epidermis (see Figs. 1, 4).

Ptilopody and Plumage Evolution

The formation of feathers from scales on the shank remains of great interest, largely because of the proposed ancestral relationship of the two appendages. Feathers and scales share early developmental mechanisms of induction and differentiation, and the cells of the early metatarsal epidermis retain the ability to form feathers given appropriate signals from the underlying dermis (see Rawles, 1963; Chuong et al., 2003; Sawyer and Knapp, 2003). Our data suggest that the early expression of Bmp7 in the metatarsal epidermis as seen in Silkie and recapitulated in experimental BMP7 protein-treated shanks results in the ability of the epidermis to form feathers from native inductive cues from the metatarsal mesenchyme. At this early stage of development of the metatarsal integument (stage 32; E8), the inductive signal is to make feathers given a competent epidermis, while scale-specific inductive signals arise much later in the metatarsus (E12–E13; Rawles, 1963). Our finding in Silkie reveals a shift in the timing (heterochrony) of the developmental mechanisms of normal epidermal appendage field specification, specifically in the metatarsal epidermis, which results in the formation of feathers from the lateral metatarsal integument.

A model has been proposed in which feather formation on the metatarsus as seen in ptilopodous breeds is the consequence of the repression of scale induction in this region (Sawyer and Knapp, 2003; Sawyer et al., 2003). This model is based on the early effects of bromodeoxyuridine (BrdU) treatment that specifically affects the action of the metatarsal dermis (Tanaka et al., 1987) and is consistent with ptilopody that occurs after abrogation of Bmp signaling in the limb mesenchyme (Zou and Niswander, 1996). Given the lack of a specific mechanism of action for the effects of BrdU and the pleiotropic roles of several Bmp family members in feather and scale development, it is difficult to resolve these data in light of the known molecular and developmental contexts of epidermal appendage formation. We demonstrate that an effect of the Pti allele is to cause early misexpression of Bmp7 in the epidermis and provide genetic and experimental evidence that the function of Bmp7 is necessary for placode formation. The timing of the effect of the Pti allele in Silkie before endogenous scale programs are initiated (as shown by Bmp7 expression), and the shift from feather to scale development in the mutant metatarsus as development proceeds, supports the notion that feathering is a response to temporal changes in induction not simply repression of scale formation.

A fossil theropod was discovered recently that exhibited feathers along the lateral surface of the shanks (Prum, 2003; Xu et al., 2003). The regional formation of feathers on the shank in this group of theropods is proposed to be a novel and crucial adaptation in the evolution of avian flight (Xu et al., 2003). The ancestors of this fossil theropod exhibit well-formed feathers elsewhere on the body but apparently lack evidence of feathering on the metatarsus. Hence, the presence of feathers on the metatarsus is thought to be a derived condition (reviewed in Prum and Brush, 2002). The growth of feathering on the leg in primitive dromaeosaurs could represent a similar example of heterochrony as seen in extant ptilopodous breeds. Both Silkie shanks and wild-type chicken shanks treated with BMP7 protein reveal that the lateral metatarsus exhibits a latent early inductive capacity that induces feathers when the epidermis is responsive. This latent ability of the lateral metatarsus has been described previously using recombination studies between wild-type and ptilopodous breeds (Goetinck, 1967) as well as scale and feather recombinants (Rawles, 1963). Thus, regional and temporal regulation of competence by Bmp7 is a plausible developmental mechanism of regulation of the variation in plumage across the integument seen in extant birds (this report), and revealed in ancestral theropods (Xu et al., 2003).

Our results, moreover, highlight an important fact in the discussion of feather formation from scales as a response to embryological perturbations or mutation. The chicken has the genetic program to make both feathers and scales. Thus, homeotic shifts in phenotype can be brought about by an evolutionary or experimental change in timing of the conditions that are necessary for induction in organisms having the genetic capacity to form the particular structures. Although, the formation of feathers from scale epidermis represents an important system to address the evolution and mechanisms of regional specification of integumentary appendages, as well as the phenotypic plasticity of a labile morphogenetic system, it does not have any direct bearing on the potential evolutionary changes that occurred in the formation of a feather from a scale.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Specimen Collection and Preparation

Embryos of the Babcock strain of White Leghorn chickens (Gallus gallus; Madison WI), scaleless (Storrs, Connecticut), and Silkie breeds (University of Wisconsin Poultry Center) were incubated at 39°C until needed. Embryos were fixed overnight in 4% paraformaldehyde/phosphate buffered saline (PFA/PBS), dehydrated in a methanol series, and stored at −20°C until use.

Whole-Mount In Situ and Immunohistochemistry

WMISH labeling was performed as described (Nieto et al., 1996) with the addition of 10% polyvinyl alcohol to the color reaction, post-fixation in PFA, and subsequent clearing in methanol. The signal was detected by using alkaline phosphatase–conjugated antibodies and NBT/BCIP as a substrate. Sections of tissue after WMISH analysis were accomplished by using isopropanol as an antimedium for paraffin embedment. Sections were cut at 10 μm thickness. Probes used in WMISH analysis were derived from chicken Bmp7 (E. Lewis), Wnt7a (C. Tabin), and chick β-catenin cDNA (cloned in the Fallon laboratory by J. Lancman).

Feather Explant Culture and Reagents

Skin was dissected from Hamburger and Hamilton stage 30 (HHs 30, Hamburger and Hamilton, 1951) thigh tracts or HHs 35 metatarsus by physical removal of the epidermis and associated dermis in PBS without enzymatic treatment. The skin was placed on Immobilon-P membranes and allowed to adhere overnight in 2% fetal bovine serum, 1% penicillin–streptomycin, Dulbecco's modified minimal essential medium (DMEM) at 37°C. The explants were treated as described in the text: CKI7 (Peters et al., 1999; 30 μm, US Biologicals); BMP7 blocking antibody (mouse IgG, 100 μg/ml; gift of Genetics Institute); LCAM blocking antibody (rabbit IgG, 100 μg/ml; kind gift of W. Gallin, Gallin et al., 1986); lithium chloride and potassium chloride (20 mM) and returned to the incubator for another 48–72 hr.

The effect of CKI7 on tissue specific Bmp7 expression was assessed in serial sections for every placode formed on each explant. Over 65 placodes were analyzed for each treatment group. Data are presented as the number of explants exhibiting a significant effect of CKI7 treatment compared with untreated controls.

In Ovo Treatment of Metatarsal Epidermis

Recombinant BMP7 protein (rBMP7; 0.9 μg/ml, gift of Genetics Institute) was applied in ovo to the epidermis on the shank of HHs31–33 White Leghorn chickens using a ∼18% pluronic acid F127; PBS; 0.5% dimethyl sulfoxide solution (Simons et al., 1992). The metatarsus of embryos at these stages was treated by isolating the shank of the chick through a small hole made in the amnion. After treatment, the shank and extraembryonic membranes were restored and the embryo returned to the incubator.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

We thank S&R Egg Farm (Whitewater, WI) for providing the Babcock White Leghorn Flock. M.P.H. thanks K. and M. Harris for their support. The authors thank L. Mathies, K. Siegfreid, R. Prum, D. Rudel, and members of the Fallon lab for helpful comments on the manuscript. A special thanks to Joseph Lancman for help with Figure 4. J.F.F. was funded by a grant from the NIH.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES