Observations on Fur Development in Echidna (Monotremata, Mammalia) Indicate that Spines Precede Hairs in Ontogeny
ABSTRACT
In the primitive mammal echidna, the initial 2–3 generations of skin appendages produced from birth forms spines and only later true hairs appear. Microscopy on preserved museum specimens reveals that the morphogenesis of spines and hairs is similar but that a larger dermal papilla is formed in spines. The growing shaft comprises a medulla surrounded by a cortex and by an external cuticle. A thick inner root sheath made of cornified cells surrounds the growing shaft inside the spine canal that eventually exits with a pointed tip. Hairs develop later with the same modality of spines but have a smaller papilla and give rise to a fur coat among spines. Therefore the integument of developing echidnas initially produces spines from large dermal papillae but the reduction in size of the papillae later determines the formation of hairs. Although the morphogenesis of spines and hairs can represent a case of specialization in this species, the primitive mammalian characteristics of echidnas has also inspired new speculations on the evolution of the mammalian hair from mammalian-like reptiles with a spiny coat. The resemblance in the morphogenesis between spines and hairs has suggested some hypothesis on hair evolution, in particular that hairs might be derived from the reduction of protective large spines present in ancient mammalian-like reptiles possibly derived from the reduction of pre-existing pointed scales. The hypothesis suggests that spines became reduced and internalized in the skin forming hairs. Anat Rec, 298:761–770, 2015. © 2014 Wiley Periodicals, Inc.
The echidna (Tachyglossus aculeatus) is a monotreme mammals typically covered by a coat fur made of long spines or quills of variable size mixed to hairs or spiny-hairs (Hausman, 1920; Griffith, 1968, 1978; Rismiller, 1999). While the dorsal and lateral sides of the body contain numerous spines mixed with hairs, spines are reduced in number and disappear in the ventral side where softer hairs are present, especially around the incubatorium, the ventral pouch of females. The size and diameter of spines however varies and also a gradation in size of spines and stiff hairs is present.
Spines indicate medium–small (0.5–1.5 mm × 10–25 mm) pointed thick hairs, while quills indicate larger (3–4 mm × 40–70 mm) hard, and often curved spines. The microstructure of quills varies in different mammals (Sokolov, 1982; Zherebtsova, 2000; Chernova, 2006).
The observation of the mixed pelage of echidnas reveals a gradation of morphology between quills (large spines in adults) to spines (smaller), and true hairs, supporting the idea from most authors that spines are specialized and enlarged hairs. The microscopic study of echidna spines show that they are made of an external thin cuticle with irregular edges and a thick cortex, but that most of the thickness is occupied by a medulla in proximal-medium regions (Hausman, 1920; Sokolov, 1980; Zherebtsova, 2000). Cuticular scales are visible in the proximal part of spines but they gradually disappear in more distal regions so that no trace of scaling is visible in the surface of the spine (Hausman, 1920).
The few available histological descriptions of developing and mature echidna spines indicate that spines develop similarly to the remaining hairs, and that hairs become modified into spines (Spencer and Sweet, 1898; Griffith, 1968; Chernova, 2006). Based on this idea it is believed that spines derive from modified hairs, assuming that hairs are the primary skin appendages characteristics of mammals from which all the other appendages are derived.
In order to better understand the morphogenesis and the nature of spines and hairs in echidna in relation to the other mammals, the following histological analysis has been carried out on some museum specimens of baby echidnas. The preservation of the utilized samples was sufficient to reveal the microscopic cell organization of spines and hairs during development. The coexistence of spines and hairs of different dimension in growing echidnas, the appearing of true hairs after spines, and the fact that hairs derive from the reduction in size of the dermal papilla and of the bulb of spines has suggested a new hypothesis on the evolution of hairs from reptilian ancestors.
MATERIALS AND METHODS
The skin was sampled from one frozen specimen of echidna, of grossly estimated age around 40 days post-hatching according to images reported in literature (Griffith, 1968; Rismiller, 1999). This specimen was collected dead in the field and carried to the Australian Museum, Sydney to be store frozen. No information was available on the post-mortem span before freezing. The specimen measured about 110 mm and its surface on the neck and head appeared smooth but pointed with developing hairy-like or spine-like spots visible on the skin surface. The remaining skin areas were also smooth but the lateral trunk and the dorsal sides between fore limbs and hind limbs appeared wrinkled, giving a scale-like surface aspect to this area. Samples of 2–3 mm long by 2–3 mm large by 3–4 mm deep mm were collected from the neck, dorsal, and lateral trunk sides and fixed.
Other skin samples derived from five baby echidnas maintained in 70% vol/vol ethanol at the South Australian Museum in Adelaide, of larger dimension than the first specimen (Fig. 1). The 2–3 mm large skins pieces of the first specimen were fixed in 10% formaldehyde wt/vol in 0.1% vol/vol phosphate buffer at pH 7.4 for 8 hr, rinsed in buffer, dehydrated in ethanol, and embedded in the transparent KM-4 Lowcryl Resin under UV light at 0–4°C for 3 days. The samples collected from the other five specimens were instead fixed in 2.5% glutaraldehyde–4% paraformaldehyde wt/vol in phosphate buffer saline (PBS) for 24 hr, dehydrated in ethanol, and embedded in the semi-transparent Araldite Resin.
Sequence of baby specimens of Echidna aculeatus (the length in mm is indicated) showing the progressive formation of a fur from an apparently naked cutis in early stages from the left to the right (indicated as progressive numbers).
The samples embedded in plastic resins were sectioned with an ultramicrotome LKB-Nova at 3–4 μm thickness and stained with 1% wt/vol toluidine blue. The transparency of the resins allowed to directly photograph the embedded tissues under a stereomicroscope and also to orient the spines-hairs for sectioning. Several sections were collected sectioning the material in a plane along the axis of the spines or hairs or in a cross-section plane. The histological images were photographed under a stereomicroscope and a bright field optical microscope.
In order to get a more detailed view of the skin surface, other preserved samples were fixed and prepared for scanning electron microscopy. On this purpose, the preserved specimens in ethanol 70% vol/vol were transferred in ethanol 100%, air dried, mounted on stubs, coated with carbon-gold, and observed under a scanning electron microscope Philips XL-40 operating at 10 kV.
OBSERVATIONS
Gross Appearance
Four of the six babies here studied are shown in Fig. 1. The skin of the younger baby (far left Fig. 1) was still glabrous but presented over most of its surface numerous black spots, indicating the eruption of spines or hairs. The second younger baby (2nd from left in Fig. 1) also featured a pale skin with numerous black spots. In the third specimen, shown in Fig. 1, the skin appeared darker and covered with an initial fur coat. In the oldest specimen (far right in Fig. 1), spines and hairs were numerous and turned the skin surface brown. By comparison of these samples with those illustrated by Griffith (1965) and Rismiller (1999), we estimated that the ages of these specimens were comprised within 1–3 months post-hatching.
Although naked-eye observation of the skin surface indicated pointed hairy-like or stick-like appendages, it was not possible to determine with a superficial examination whether the different stages of development of the “sticks” represented only spines or whether developing hairs were also present. Embedding in a transparent resin allowed observation of the actual length of each developing and pigmented skin appendages, either spines or hairs that were later sectioned and analyzed for the histological structure.
The surface of the skin observed under the stereomicroscope revealed that all the hairy-like spots observed in the younger specimens were pointed and with a large base, indicating that they all represented different stages or successive generations of developing spines (Fig. 2A,B). The stereoscopic and scanning electron microscopic observation on the skin from the younger specimens (1 and 2 in Fig. 1) embedded in resin, confirmed that only spines were present in the skin at these earlier stages (Figs. 2, 3 and 4A–C). In the two older specimens shown in Fig. 1 (3 and 4), the first hairs appeared between the large spines at this stage, covering most of the dorsal–lateral areas of the skin surface. Typical hairs showed blunted tips, and were much thinner and often curved in comparison to the surrounding spines (Fig. 2D,E).
Aspect of the surface of the skin at progressive stages of development as viewed through the transparent resin in which the specimens were embedded before sectioning. Scale bars in all figures = 200 μm. A: Youngest specimen. The arrow indicates smaller, second generation of spines while the arrowhead indicates the first generation of larger spines. B: Second younger specimen. The double arrowhead indicates the first generation of spines, the arrowhead the second generation, and the arrow the third generation. C: Intermediate specimen. Developed spines(s) with their large base (arrow) viewed to exit with their tips (arrowheads) through the epidermal surface (e). D: Oldest specimen. Detail to show the presence of hairs (arrows) together spines (outlined by arrowheads). E: Wider area featuring various hairs (arrows) among spines (arrowheads).
Scanning electron microscopic view of the cutis at progressive stages of development. Scale Bars in all figures = 200 μm. A: Youngest specimen. Three spines with broken tips are seen (arrowheads). B: Second younger specimen. Detail on the base of three spines at different stages of development (the arrowhead points to spines of the first generation and the arrows to second generation spines). The central medullary cavity can be seen. C: Intermediate specimen. Only pointed spines (arrowhead) cover the skin surface. D: Oldest specimen. Only spines are surely observed (arrowhead) while hairs (arrows) are less recognizable. Double arrowheads show the bumps formed at the base of the emerging spines.
Stereomicroscopic view (A, B) and histological aspects (C, D) of developing spines in the 110 mm specimen. A: All the developing sticks are pointed spines at different stages of development, from large spines (arrowheads) to small spine tips (arrows) of more immature generations of spines. Bar = 100 μm. B: Detail of a developed spine with the typical triangular shape. Bar = 100 μm. C: Bulbs (bu) of developing spines at a relatively early stage of development beneath the epidermis (e), when the spine canal (arrows) is partially occupied by the spine shaft. Bar = 50 μm. D: Detail of the bulb of an early spine. The deep papilla (p) encounters numerous melanocytes (arrow) in the upper matrix that contribute to the pigment deposition in the cortex (co). Bar = 30 μm.
Scanning Electron Microscopic Aspect
The scanning electron microscopic study confirmed that the initial skin appendages present in the skin were developing and enlarging spines (Fig. 3). In the younger specimen, the tips often appeared broken revealing the medullated cavity (Fig. 3A). In the intermediate stages the emerging and sparse spines possessed a pointed tip and the cuticular surface was mostly smooth (Fig. 3B,C). As previously observed under the stereomicroscope (arrowheads in Fig. 1C), the spines in their point of emergence on the skin surface distorted the epidermis forming a bump (Fig. 3D, double arrowheads), an aspect not seen in emerging hairs. Only in the older specimens sparse hairs were seen among the prevalent spines (Fig. 3D, arrows).
Histological Observations
The analyzed skins were preserved for some years frozen or in 70% ethanol and were probably derived from animals that were collected within a relatively short time after death, possibly few hours. In fact, the fixation and embedding in plastic have allowed the detection of a sufficient degree of cell preservation to show useful histological features on some stages of the morphogenesis of spines and hairs. Since only spines were initially present in the skin of the younger specimens we considered most of the sectioned bulbs as “spine bulbs” rather than “hair bulbs.” Because the spines were forming at different stages (or successive generations) of growth, the random histological study over a number of spines of different lengths provided reliable information on the morphogenetic process of spine formation and later of hairs formation.
The random sectioning of intercepted spines at different stages of development (Fig. 4A,B) revealed small (still intra-epidermal) to large spines with their pointed shaft growing inside the spine-canal, emerging or already emerged from the epidermal surface (Fig. 4C). The small follicles with bulbs that produce the spines initially appeared as smaller dermal papillae in early stages of spine development and appeared as larger papillae at later stages when the spines enlarged. The spine canal was obliquely oriented and the bulb appeared small in relatively small spines still located inside the epidermis although the dermal papilla extended deeply in the bulb (Fig. 4C). Despite the different dimension of spines at different stages of development, the histological study revealed that their dermal papillae have basically the morphology of typical hairs.
The dermal papilla appeared deeply inserted inside the bulb and frequent melanocytes were seen in the bulb matrix (Fig. 4D). The spines at different stage of development showed small bulbs in early spines and larger bulbs in more developed and larger spines where the spine shaft was emerging on the surface of the epidermis (Fig. 5A–D). The spine shaft included a cortical and a medullary core but the dermal papilla extended for more that 1 mm inside the core of the shaft in larger spines. The papilla was formed by a loose connective tissue where numerous large blood vessels were likely present in the bulb (Fig. 5E). The external part of the multilayered inner root sheath was the first to form a corneous layer, equivalent to the Henle layer of typical hairs (Fig. 5E,F). The inner regions of the inner root sheath remained non-cornified in the distal regions of the bulb in the upper part of the spine canal. In this region, the spine mainly consisted of cortical cells forming a cortical spine cone still growing inside the spinal canal in spines at an intermediate stage of development (Fig. 5C,F).
Microscopic images of intermediate size spine follicles (A, B) of larger spine follicles erupting on the surface (C, D) and of details of the papilla (E) and hair shaft (F). A: Large spine canal (sc) contacting the epidermis (e). Bar = 50 μm. B: Spine cone (arrow) inside the still compact upper part of the spine canal (sc). Below the spine shaft, a long dermal appilla (P) is present. Bar = 100 μm. C: Developed shaft inside the spine canal (sc) with a cortex (arrowhead) and a medulla (arrow). The asterisk indicates either the tip of the dermal pulp or the spine medulla. Bar = 50 μm. D: Detail of the emerging spine shaft (s) from the spine canal (sc) surrounded by the outer root sheath (ORS) in continuation with the epidermis (e). Bar, 50 μm. E: Detail on the large dermal papilla (p) of a spine featuring the loose connective likely containing blood vessels (v) in life. The arrows point to the cornified part of the IRS while the arrowhead indicates the external ORS. Bar = 25 μm. F: Cross-sectioned spine showing the prevalent cortex (co) at this level of the shaft, the thin cuticle (arrow), the multilayered and largely cornified IRS, and the surrounding ORS. Bar = 50 μm.
In more developed and large spines with a triangular- or a barrel-like-base, the dermal papilla occupied more than 1 mm inside the spine bulb. The papilla comprised a fibrous connective tissue not surrounded by epithelial cells of a possible matrix. The papilla contained large and vacuolated cells that were in continuation with the medulla inside the hair shaft (Fig. 6A). Due to the relative poor preservation of the tissue after the death of the animal and before fixation, it is uncertain whether the dermal papilla was actually in contact with these medullated cells and whether a flat epithelium was present in spines during the life of the animal. The localization of the epithelium in the spine matrix was seen on the sides located at the base of the papilla (Fig. 6B). Therefore, because of the relatively poor preservation of the tissues, in particular of the epithelium forming the spine matrix, it was not clear the region of separation between the loose connective tissue of the papilla and the adjacent medulla of the spine. The inner root sheath appeared multilayered around the papilla (Fig. 6B) but was cornified in more distal regions around the hair shaft (Fig. 6C,D). Beneath a thin cuticle a thick cortex was formed, and a medulla was also present and reached the distal regions except the tip of spines.
Microscopic image of developing large spine (A–D) and hair (E, F). A: Typical triangular shape of well developed spine illustrating the dermal papilla (p) surrounded (arrowhead) by the medulla that colonizes the central part (arrow) of the hair shaft (s). e, epidermis. Bar = 100 μm. B: Detail of the dermal papilla (p) surrounded by a medullary tissue made of large vacuolated cells in the central region (asterisk) and by narrow and cornified cells forming a cortex (co). At the base of the papilla is present the area occupied by the spine matrix (ma). IRS, inner root sheath. The arrowhead indicates the scaled spine cuticle. Bar = 50 μm. C: Detail on a spine shaft with medulla (arrow), cortex (co), cuticle (arrowhead), surrounded by the corneous IRS and ORS. Bar = 50 μm. D: Detail of apical region of the spine with indicated cuticle (arrowhead), cortex (co), and medulla (me). The thick and corneous IRS and ORS are seen. 50 μm. E: Cylindrical shape of hair bulb featuring the dermal papilla (p), medulla (me), cortex (co), and ORS outside the corneous IRS (arrowhead). Bar, 50 μm. F: Detail of hair bulb with cells forming the papilla (p) and medullary cells (me). The arrowhead points to the corneous IRS. Bar, 25 μm.
Finally, sections of older stages of development (last two stages shown on the right of Fig. 1), revealed small bulbs that belong to true hairs, as it was indicated by their thinner and more constant diameter along the hair canal (Fig. 6E,F). These bulbs appeared narrower than the barrel-shaped spine bulbs, and possessed a much shorter dermal papilla surrounded by medullary cells. Also in this case, the more external part of the inner root sheath (IRS) appeared cornified above the bulb.
DISCUSSION
Reduction of Spines Gives Rise to Hairs
All studies on echidna hairs and spines agree that there is a gradation of dimension between spines and hairs (Spencer and Sweet, 1899; Hausman 1920; Griffith, 1968), although the pointed and sharp tip is only formed in spines. These sharp tips are used in monotremes for defense (Zherebtsova, 2000), and they were probably similarly used in their reptilian and therapsid ancestors as well as tactile appendages (Maderson, 1972). Spines appear roughly similar to reptilian frills but the latter possess neither the papilla complexity nor the hardness of spines (Chang et al., 2009).
In their histological description of hair and spine development in echidna (Spencer and Sweet, 1899) as well as in descriptions from other authors (Hausman, 1920; Griffith, 1968; Chernova, 2006), it is indicated that spines are modified hairs, so that developing hairs become later enlarged and produce spines. In particular Spencer and Sweet (1899) worked on two 55 mm long babies of similar age, indicated as embryos by these authors. They reported that initially the skin formed “tips of large hairs that subsequently become modified into spines.” These authors continue to use this terminology despite the evidence that the initial tips develop from large bulbs that are characteristic of spines. Small hairs do not transform into spines but are derived from characteristically different hair-bulbs. We have observed that spines are the first hairy-like skin appendages formed in echidna since sharp and large pointed tips appear initially on the surface of the epidermis of baby echidnas and evolve into 2–3 generations of spines (Fig. 7). Hairs appear among spines, and by comparison hairs have a narrow cylindrical shaft, a blunt tip and derive from a bulb in which the dermal papilla is smaller than in spines.
Schematic summary of the sequence of spines and hair generations formed on the echidna integument from birth. From A to B only spines are produced and they emerge on the skin surface. From Stage C also smaller hair follicles are formed but emerge at Stage D among the spines. Hairs determine the beginning of fur formation and the darkening of the integument (see also Fig. 1, Stages 3–4).
Clearly spines and hairs have a similar origin but hairs, at least in echidna, are derived from the smaller size of later generations of spine bulbs and not vice-versa, an opposite view from that of previous authors. In fact, it appears that it is the size of the dermal papilla that determines the output of the epidermal derivative and that large papillae give rise to spines or to large hairs, with the remaining small follicles growing into hairs. Therefore naming a bulb as a hair bulb or a spine bulb is determined by the final appendix produced from the follicle since the papilla appears initially small but progressively enlarges probably reflecting the amount of cells produced in the basal matrix of the bulb. No data are available on the presence and localization of likely stem cells in echidna spine follicles. The initial hairy-like appendages produced in the echidna tegument are therefore spines, and it is from a modification of typical spines that spiny-hairs or a thinner and softer type of hairs are later formed. Although a specific study is not apparent from the literature it is likely that during the lifetime of echidnas, spine replacement in adults gives rise to new spines through the formation of larger dermal papillae while the formation of hairs takes place from smaller papillae.
The formation of spines suggests a case of specialization for echidna skin since spines are utilized for protection (Zherebtsova, 2000; Chernova, 2006). The reduction of size from spine bulbs to hair bulbs has contributed to formulate a new hypothesis on the evolution of hairs derived from the idea that “ontogeny recapitulates phylogeny, with some modifications.”
Speculations on Hair Evolution From Spiny Mammalian-Like Reptiles
Spines and hairs are related appendages and follow a similar developmental pathway in the echidna. Despite no similar studies for other spiny mammals are available to the best of the authors' knowledge, the systematic position of the echidna in relation to placentals spiny mammals and the various reptilian characteristics present in this species (Griffith, 1968, 1978) allows for a new hypothesis on hair evolution. This hypothesis offers another view to the traditional concept indicating that spines are modified hairs (Spencer and Sweet, 1899; Hausman, 1920; Zherebtsova, 2000; Chernova, 2006). Only the study on fossilized skin from synapsids will show whether a spiny coat was also present in ancient mammalian-reptiles.
The presence of a scaly integument in synapsids of the Permian is indicated by the presence of osteoderms (Botha-Brink and Modesto, 2007) but whether this integument was spiny or not remains undetermined. However it is likely that the furred integument shown in some Mesozoic fossils of monotremes (Ji et al., 2012, estimated from 165 million years old) and eutherians (Ji et al., 2002, estimated from 125 million years ago), also included numerous spines among the furred skin (unfortunately not analyzed in detail in these cited papers). The spiny integument of possible synapsids–therapsids was mainly utilized for defense against predators, like in extant echidnas, porcupines, and hedgehogs, but no fossilized skin showing a spiny skin is available or documented so far. This type of integument however presents several limitations for the functionality of the skin in mammals that is richly innervated for sensing, a function that would be limited in presence of large spines (Spearman, 1964; Maderson, 1972). It may be possible that the transition between spines and hairs observed from the dorsal–lateral to the ventral side of echidna evolved as an adaptation to viviparity, from egg laying to the feeding of the juvenile in the incubatorium area where spines would have been hazardous for feeding babies.
The relatively scarce presence of spines in mammals speaks for a main trend to lose these robust and sharp appendages, especially on the ventral part of the body, in favor of a hairy coat (Sokolov, 1982). The reduction of spines and their replacement with smaller hairs (Fig. 8) would have maintained skin sensitivity and improved thermoregulation (Ruben and Jones, 2000). When spines became reduced following miniaturization of the dermal papilla they produced thinner stick-like appendages with blunted tips, the true hairs and later a pelage, a hypothesis presented in Fig. 8. Therefore this new hypothesis would sustain the scaling origin of hairs from a scaled reptilian integument (Spearman, 1964; Maderson, 1972; Alibardi, 2004a2004b), an alternative to the glandular hypothesis of hair evolution (Stenn et al., 2007; Dhouailly, 2009; Alibardi, 2012).
Schematic drawing illustrating a hypothesis on hair evolution from spines. A–F and G–J show a hypothetical sequence of transformation from reptilian scales through conical scales and large spines that, after deepening in the dermis, were transformed into spines with large bulbs. The reduction of these spines forms single hairs and the multiplication of them gave rise to the pelage. In K–N the sebogenic hypothesis of hair evolution (Stenn et al., 2007) is sketched, which also converge into the formation of a bulb spine (G). The latter hypothesis suggests that hairs were derived from epidermal pegs (K) that originated sebaceous glands (L). Later, the association to a dermal papilla (M) gave later origin to a hairy bulb (bu) with a central, solid stick instead of the previous sebaceous secretion.
It may be possible that future studies will find in the fossil records types of integument of synapsids or early therapsids where a spiny cover made of large pointed tubercles, or even large spines was present in some Paleozoic or Mesozoic species. The lens-like structures found in the synapsid skin may represent the base of sectioned horny spines, not glands (Chudinov, 1967). Spiny or frilled integument appendages have been found in several archosaurian reptiles (Martin and Czerkas, 2000), but the histological structure, in particular the presence of inner and basal cavities harboring dermal papillae remain unknown in fossilized skin casts.
ACKNOWLEDGMENTS
Research largely self-supported (Comparative Histolab). Dr. Sandy Ingleby of the Australian Museum, Sydney, allowed sampling of the echidna skin while Mr. David Stemmer from the South Museum in Adelaide kindly made available preserved Echidna specimens for sampling. Ms. Lyn Waterhouse of the Adelaide Microscopy Center assisted with the SEM operations. Mr. Mauro Cesarini (University of Bologna) made part of the drawings.
