Formation of feathers on chick foot scales: a stage-dependent morphogenetic response to retinoic acid

It has been known for many years, from the pioneering work of Fell and Mellanby in 1953, and from many other reports (Fell, 1957; McLoughlin, 1961; Rothberg, 1967; Beckingham-Smith, 1973; Peck, Elias & Wetzel, 1977), that retinoids (vitamin A and derivatives) produce a mucous metaplasia in embryonic tarsometatarsal chick epidermis, and inhibit keratinization. In mammals, it has been found that retinoids not only act at the cellular level by transforming kératinocytes into mucus-secreting cells (Lawrence & Bern, 1963; New, 1965; Piewig, Wolff & Braun-Falco, 1971; Wong, 1975), but that in at least one developing tissue they are also able to elicit a morphogenetic response, namely the transformation of whisker follicles into skin glands, as shown by Hardy in 1968. Vitamin A was first applied to avian and mammalian skin in the form of retinol or retinyl acetate. More recently, it has become evident that retinoic acid (RA) is a more potent molecule than the above for many of the reported actions on mammalian and avian skin (Fulton, 1975; Peck et al. 1977).

All the transformations at the cellular level effected by retinoids seem to be basically reversible events, and they require the continued presence of the retinoids for their maintenance (Fell & Rinaldini, 1965). We report here another type of response of chick embryo tarsometatarsal skin to retinoic acid. This is a morphogenetic response, which was obtained in ovo and, like the whisker follicle transformation (Hardy & Bellows, 1978), it resulted in a permanent change in cutaneous appendages, namely the formation of feathers on the chick foot scales.

In a preliminary study (Dhouailly & Hardy, 1978), this experimental ptilopody was obtained in 85 % of embryos which had received three injections of 125 μg of RA at 10, 11 and 12 days. In this case of repeated treatments, the number and distribution of feathers on the feet were highly variable and the data were difficult to analyse.

More recently we found that a single dose of RA was sufficient to elicit feather formation and that a strict correspondence could be established between the time sequence of appearance of scales (Sawyer, 1972; Sawyer & Craig, 1977) and their sensitivity to the retinoid.

The chick embryos were from F₂ crosses of Wyandotte × Rhode Island Red breeds. Eggs were incubated at 38 °C and windows in the shell were prepared at 3 days of incubation.

Retinoic acid (all-trans retinoic acid, Sigma) was dissolved in absolute ethanol, and administered as single or repeated injections (0·05 ml) into the amniotic cavity. Nine-day embryos received a relatively small (31·2μg) single dose, as higher doses killed 100% of them within 24 h. Ten-, 11- and 12-day embryos received a single dose of 125 μg. Our preliminary experiments (Dhouailly & Hardy, 1978) had shown that, at these stages, 125 μg was the most suitable dose to obtain an exploitable percentage of both surviving embryos and morphogenetic aberrations. Embryos with repeated treatment either received 125 μg of RA at 10 and 11 days (i.e. a cumulative dose of 250 μg), or were injected with three doses of either 62·5 or 125 μg at days 12, 13 and 14 (i.e. a cumulative dose of 187·5 or 375μg). Control embryos were injected at corresponding times with 0·05 ml of absolute ethanol alone.

Most of the surviving embryos were recovered at 17 days of incubation and fixed in Bouin-Hollande’s fluid. A few others were sacrificed at hatching.

For precise determination of early developmental stages of scale primordia which were present at the time of injection, the entire skin from anterior and posterior surfaces of feet of 9- to 14-day chick embryos was dissected away from the underlying tissues, fixed in 2 % glutaraldehyde, post-fixed in 1 % osmium tetroxide, then dehydrated, embedded in Epon, and observed in transmitted and reflected light with a stereomicroscope. This technique allowed a precise observation of the earliest epidermal thickening during placode formation in scale rudiments, which appeared as opaque patches in an otherwise translucent skin.

Three types of scales are distinguished on the feet of chickens (Lucas & Stettenheim, 1972). Large, distally overlapping scales (scuta) are arranged in two longitudinal rows (ca. 2× 18 scales) on the anterior face of the shank (tarsometatarsus) and one row on the upper face of each toe (ca. 7, 15, 20 and 18 scales on toes I, II, III and IV respectively) (Figs. 1 and 2). Smaller, proximally overlapping scales (scutella) are arranged in two longitudinal rows (ca. 11 and 14 scales) on the posterior face of the shank (Fig. 3). The number of scuta and scutella is variable and cannot be determined with accuracy, because they may be split longitudinally and they gradually decrease in size towards the proximal (and, in the case of scutella, also the distal) end of the scale field, where the latter merges indistinctly with areas occupied by smaller scales. These small scales, called reticula, are non-overlapping and cover the remainder of the surface of the foot (Fig. 3). Furthermore, it appears appropriate to distinguish, among the reticula, larger ones which we propose to call marginal reticula and which are located in two or three rows along the lateral aspect, and in three to six rows along the medial aspect of the shank, between the scuta and scutella, and in one to three rows on the medial and lateral aspects of the digits, between the scuta and the small reticula which cover the digital pads and inter-pad spaces.

The primordia of scuta and scutella (epidermal placodes) and those of reticula (slight skin elevations) appear in a strict temporal sequence, as described by Sawyer (1972) and Sawyer & Craig (1977) and illustrated in detail from our specimens in Fig. 1.

A slight manifestation of ptilopody (feathers on the feet) may occasionally occur in breeds of fowl characterized by feet without feathers, such as the Wyandotte and Rhode Island Red breeds. Out of 90 non-treated 17-day embryos, one of them (1·1 %) presented one feather on each of two scuta of the lateral row of the left shank, and one feather on each of three scuta of the corresponding row of the right shank.

Sixty-five embryos, ranging from 9 to 14 days, received from one to three intra-amniotic injections of 0 · 05 ml of absolute ethanol. Thirty-eight (58 %) survived to 17 days and were normal, particularly with respect to foot integument. No case of even slight ptilopody was observed (Figs. 2 and 3). One embryo had abnormally short claws.

Four types of external anomalies were observed in the 17-day embryos surviving after RA injection (Table 1). Three of them were malformations of beak, feathers and toes and one was characterized by the formation of normally shaped cutaneous appendages in an abnormal location, namely feathers on the foot scales (Figs. 4 and 5). The supernumerary feathers did not appear to inhibit or interfere with the normal appearance and pattern of scales. The resulting morphology was similar to that which is characteristic of certain breeds of fowl (Lucas & Stettenheim, 1972) and referred to as ptilopody. Two of the anomalies were obtained with a relatively high frequency: malformed feathers within the feather tracts (club-shaped feathers) and ptilopody. The frequency of the anomalies was dependent on the dose of RA, whereas their type and their distribution (in the case of malformed feathers and of feathered scales) were dependent, as reported below, on the age of embryos at the time of treatment.

Malformations of toes and beak occurred when RA was administered to embryos before 11 and 12 days of incubation respectively. Anomalies of toes consisted in acropodial truncature, resulting in shortening or absence of claws (Fig. 6), associated or not with the loss of the last phalanx. Malformed beaks were characterized either by a shortening of the lower jaw, associated with a downward curvature and twisting of the upper jaw, or by the downward curvature of both upper and lower jaws (Fig. 6). It should be noted that many embryos observed one day after RA treatment showed swollen and blood-filled feather buds, and a few showed also swelling of digit extremities, accompanied by blood stasis.

Malformed feather filaments were obtained after RA treatment at 10, 11 or 12 days of incubation. They were club-shaped, with a proximal peduncle of normal diameter but variable length, and an apex which was dilated into a spherical shape. The distribution of these abnormal feathers within the different tracts varied with the age at treatment: in embryos treated at 10 days, the most frequently affected tracts were the spinal and femoral ones, and more specifically the earliest feathers in these tracts (Fig. 6). In embryos treated at 11 days, ventral, alar and capital tracts were preferentially affected (Fig. 7). After treatment at 12 days, most embryos presented club-shaped feathers around the eyes.

Retinoic acid-induced ptilopody occurred in 57% of embryos treated at 10 days, in 48 % of embryos treated at 11 days and in 83 % of embryos treated successively at 10 and 11 days. When RA was administered to 12-day embryos or repeatedly at 12, 13 and 14 days, only a small percentage (5 – 15) presented feathered feet. The distribution of feathers on scales, which was strongly dependent on the age of embryos at the time of treatment, is analysed in detail in the following section.

In contrast with those of the regular feather tracts, the majority of feather filaments formed on the scales were normally shaped (Figs. 8 – 11). In those embryos which were carried to hatching, these feather filaments gave rise to normal down feathers with barbs (16 – 26) and barbules.

As a rule, one or two feathers were borne by the distally oriented free edge of the affected scuta (Figs. 8 and 9). In a few cases, more than two feathers developed on each scutum, and were then more or less fused laterally (Fig. 12). The majority of the affected scutella bore only one feather (Fig. 13), but some formed two feathers, from their proximally oriented free edge. The affected reticula never formed more than one feather (Fig. 14). In most cases, the interscale furrows were normally developed and the distribution pattern of scales was normal. In a few embryos (2 %), the reticula of the plantar face had not formed at 17 days and the skin was folded into irregular wrinkles.

The regional distribution of feathered scales on the feet of treated embryos showed a high degree of bilateral symmetry. This distribution was analysed in detail by counting the number of feathers formed on the right foot of the embryos which had been treated with a single dose of RA at 9, 10, 11 or 12 days and of those which were treated repeatedly at 10 and 11 days or at 12, 13 and 14 days. It varied markedly with the age at which the embryos were treated (Table 2).

Only one embryo treated at 9 days (31 ·2 μ g) formed two feathers on each foot; they were located on the metatarsus, at the base of the fourth digit.

Embryos treated at 10 days (Figs. 8, 9 and 12) developed numerous (up to 210) feathers on most (sometimes all) tarsometatarsal scuta, as well as on a variable number of scutella and of digital scuta. A small number (four) of these embryos (16%) also formed a few feathers on the marginal reticula, and from one to three feathers (Fig. 9, inset) in the central metatarsal pad. It was noted that, in three of these four embryos, a group of two to nine scuta at the junction between tarsometatarsus and digits III and IV were normal and did not bear any feathers (Fig. 9).

Embryos treated at 11 days developed feathers on scuta in only 36% of cases, chiefly on the distal phalanges and at the base of digit II (Fig. 10). The tarsometatarsal scuta were unaffected or else bore only short feather filaments. In addition, these embryos formed up to 62 feathers on the marginal reticula, up to 50 on the scutella (Figs. 11 and 13), up to 70 on the digital pads (Fig. 11) and up to 35 feathers on the central metatarsal pad (Figs. 11 and 14).

When embryos received two successive injections of RA, at 10 and 11 days, three different distribution patterns of feathered scales were observed in the 83 % of chickens which were affected : either the pattern was similar to that of the embryos treated at 10 days (41 %) or at 11 days (10%), or it displayed the features of both patterns simultaneously (32 %) (Figs. 4 and 5).

Embryos treated at 12 days did not form any feathers on either scuta or scutella; but most if not all reticula formed feather buds (Fig. 15).

Embryos which received three successive injections at 12,13 and 14 days bore feathers on reticula only (Fig. 16), in the same location as in embryos which were treated with a single injection at 12 days.

Some of the effects of retinoic acid on chick embryos could be related to its well-known influences on chondroitin sulphate metabolism (Kochhar, 1967) and on developing blood vessels (Fraser & Travill, 1978). Impaired chondroitin sulphate synthesis could explain beak malformations, which appear to result from interference with maxillary and mandibular growth. Damage to forming capillaries could lead to fluid imbalance, oedema and haemorrhage in the apical parts of feather filaments and of toes. The correspondence between the affected feather tracts and the age of embryos at the time of treatment shows that feathers are RA-sensitive only during a short period of their development, at the stage when vascularization of the feather bud is being established (buds of 0 · 3 – 0 · 5 mm in length).

In addition, retinoic acid induces the formation, on the foot scales, of normal feather filaments, which at hatching give rise to normal down feathers. Thus it is able to elicit an abnormal morphogenetic response from prospective scaled foot integument. The specific distribution pattern of feathered scales for each of the three stages at which RA was administered, considered together with the time-sequence of appearance of the scales, leads us to conclude that scale primordia are sensitive to RA only during a limited period of their development. No information can be derived from the result of 9-day injections : although one out of ten surviving embryos presented a few feathers on the earliest appearing scuta, this slight ptilopody cannot be distinguished from that which can occasionally occur without treatment (see Results, Section II). Single injections at 10 days were found to produce feathers constantly on scuta which appear as placodes at 10 (Fig. b) and 10 · 5 days (Fig. 1c) and inconstantly on those which form at 9 days (Fig. 1 a) and 11 days (Fig. 1 b), and on a variable number of scutella which appear as placodes at 11 days (Fig. 1 e). Single injections at 11 days caused the formation of feathers on the scuta and scutella which appear at 11 days (Fig. 1 d and e), on the last scuta, which form at 12 days at the base of the second digit (Fig. 1f) and on scutella and reticula which appear at 12 days (Fig. 1g), and only inconstantly on scuta already present at 10 days. Single injections at 12 days induced the formation of feathers on reticula which appear at 12 and 13 days.

Thus, it was found that a single injection affected only those scuta and scutella which became discernible as opaque placodes, or reticula which appeared as elevations, on the day of injection and the following day. The scales which were due to appear (as placodes or elevations) 48 h after the injection of RA were seldom affected. Likewise the scales which had started their morphogenesis 24 h or more before the injection, and had thus attained the asymmetrical placode stage at the time of injection, were rarely affected, while those that had attained the hump stage were never affected.

It is not known for how long after administration the retinoic acid molecule remains biologically active in the tissues of the chick embryo. Injections at 10 days affect tarsometatarsal and digital scuta which arise as placodes at 10 and 11 days, but do not affect the scuta at the base of digit II (which appear at 12 days), only rarely (16 %) affect a few marginal reticula and central metatarsal reticula, and never affect any of the other 12-day reticula (such as those of the digital pads). These results can be interpreted in different ways according to whether RA remains active for less than 24 h or for more than 24 h. If RA is degraded within 24 h it follows that the pre-placodal (or pre-elevation) stage, as well as the placodal (or elevation) stage must be sensitive to the retinoid. If, however, RA persists in the tissues for more than 24 h, it is possible that only the placodal (or elevation) stage of each scale is the sensitive one. Results obtained by De Luca & Zile (1975) on the persistence of injected RA in rats show that RA is degraded in these animals within 48 h. If the chick embryo does not degrade RA any faster than the adult rat, the second of the above possibilities is favoured. In the present state of our knowledge, we can best interpret the findings by the following hypothesis : RA, reaching the skin within hours of injection, acts on those scales which are at the placode (or elevation) stage of development, and possibly acts also on some scales at a slightly earlier and a slightly later stage. It is degraded after 48 h or less, and perhaps after less than 24 h.

Two injections of 125 μg of RA, at 10 and 11 days, resulted in perturbation of foot cutaneous appendages which could be classified into three categories: (1) some surviving embryos had a distribution pattern of feathered scales similar to that of embryos treated at 10 days; (2) some had a distribution resembling that of embryos treated at 11 days; (3) some displayed the features of both distribution patterns simultaneously, and some had normally scaled feet. This variability may be easily interpreted if one considers the results of the 10-day and 11-day series, each of which yielded a certain proportion of feathered feet (0·57 and 0·48, respectively) and of normal feet (0·43 and 0·52, respectively). The combination of these proportions leads to the following ‘theoretical’ percentages of double-treated embryos: 30% (0·57 ×0·52) should display the 10-day phenotype; 21% (0·43×0·48) should display the 11-day phenotype; 27% (0·57×0·48) should show both phenotypes simultaneously; and 22 % (0·43 ×0·52) should be non-affected. As these calculated percentages are not significantly different from the observed ones (41, 10, 32 and 17%, respectively), it is reasonable to entertain the hypothesis that two successive injections of RA have additive effects on the foot integument. In addition, the results obtained with triple injections at 10, 11 and 12 days (Dhouailly & Hardy, 1978) can be interpreted and understood in a similar manner.

In conclusion, RA is able to elicit the formation of feathers on the scaled foot integument of the chick embryo. The mechanism of action of the retinoid is as yet unknown. It is clear, however, in view of the existence of domestic breeds of fowl with feathered feet (genetic ptilopody), that RA somehow interferes with scale morphogenesis and thereby reveals a latent ability of avian foot integument to produce feathers. Apparently the formation of scales requires additional and possibly inhibitory region-specific information on top of the trivial and ‘ubiquitous feather message’ (Dhouailly, 1978). Retinoic acid, by weakening the scale message, would leave the feather message free to be expressed.

The collaboration was made possible by a travel grant from the National Research Council of Canada to Margaret H. Hardy, who was on sabbatical leave from the University of Guelph when the work was undertaken. An operating grant from the National Research Council helped to defray the expenses; it was continued in 1979 by financial help from INSERM (Contrat de recherche libre n° 7915142). The authors are grateful to Mrs N. Cambonie, Mrs Y. Bouvat and Mrs J. Lana for their technical assistance.