Mammary Gland Evolution

By at January 4, 2012 | 12:43 am | Print

In 1758, mammary glands, a defining characteristic of mammals, were first acknowledged for their uniqueness. It was that year that Linnaeus combined terrestrial organisms, formerly considered quadrupeds, with a group of aquatic cetaceans, considered fish, into a new group he called Mammalia. All of these organisms shared the ability to feed their young with a nutritive cutaneous gland secretion. Although the anatomy and physiology of the mammary gland is fairly well understood in extant animals, the evolution of this distinct adaptation is still mysterious and controversial.

In 1872, Darwin, in his sixth edition of On the Origin of Species, first proposed a mechanism for mammary evolution. He postulated that mammals descended from animals that used a pouch for rearing young, and individuals that were capable of producing nutritive secretions from specialized cutaneous glands to nurture their young would certainly give their offspring an advantage (1). However, ten years later when it was identified that the lactating platypus was egg-laying and lacked a pouch, Darwin’s explanation broke down. Since Darwin, many attempts (Table I) have been made to explain mammary evolution in egg-laying mammalian predecessors (2). In 1886 Gegenbauer looked at sweat and sebaceous glands as a possible origin for the mammary gland (3). In early 1900, Bresslau proposed that mammary glands derive from cutaneous glands associated with hair follicles (4). In the 1960’s, Haldane suggested that endothermic mammalian ancestors might have needed to keep their eggs cool, and did so by keeping fur moist through bathing, as seen in some Asian birds that dampen their feathers for this purpose, and sweat secretions (5).

Haldane postulated that these sweat secretions were possibly the earliest mammary secretions. Later that decade, Long suggested that since monotreme eggs absorb uterine secretions before eggs are laid, nutritive cutaneous secretions would have provided great benefits to already laid eggs (6). In 1973, Hopson stated that late Triassic mammals were so small that egg size had to be small. Also, these early endothermic animals, with a high metabolism, would require a nutrient dense diet. Based on small egg size and endothermic requirements, the altricial hatchings would depend heavily on the parent for nutrition. Hopson further proposed that early cutaneous secretions prevented egg/neonatal dessication but later provided the source of nutrition for the altricial young (7). Graves and Duvall (1983) suggested that pheromones were released by cutaneous glands that directed young to nuzzle and lick glands that eventually led to the evolution of the glands nutritive function (8). Fossil evidence of vomeronasal organs was used to support their ideas. Hayssen and Blackburn, several years later, used molecular evidence to support that the earliest cutaneous secretions were used for anti-microbial purposes, and only subsequently led to a nutritive function (9). The molecular similarities between α-lactalbumin and lysozyme, a very common anti-microbial agent, do support this idea. Blackburn also argued that mammary glands are actually a mosaic of multiple gland types (10). This was a novel idea, because most authorities have always argued that mammary glands evolved from the sweat (eccrine or merocrine) gland, the apocrine gland, or the sebaceous gland; certainly not as a hybrid.
It is very possible that none of the preceding hypotheses is correct by itself, but in actuality the truth may lie in some combination of them. Two papers published in July, 2002, in the Journal of Mammary Gland Biology and Neoplasia by Olav T. Oftedal expands on and synthesizes several established hypotheses in attempt to elucidate the evolution of the mammary gland (2,11). The remainder of this paper will be dedicated to providing a synopsis of Oftedal’s ideas.

First it is necessary to establish a time period and the types of early organisms in which the mammary gland evolved. The first evidence of bone appeared in the Cambrian period roughly 500 mya. About 300 million years later vertebrates known as amniotes, which produced eggs with extraembryonic membranes to support gas/nutrient exchange, waste storage, and water retention, split into the clades Sauropsida and Synapsida (Figure I). The sauropsids gave rise to all extant reptiles and birds. The synapsids are ancestral to extant mammals. The sauropsids and synapsids have been evolving independently for over 300 million years, so it is not necessarily correct to look at extant reptiles when trying to determine mammalian evolution. Mammals did not evolve from reptiles, but both evolved from amniotes. The synapsids evolved mammal-like characteristics and became the dominant fauna in the Permian and Triassic. In the late Triassic the dinosaurs displaced the synapsids as the dominant animals, however synapsid evolution toward the mammalian character continued. New structural adaptations favored elevated metabolic rates, increased aerobic abilities, enhanced processing of nutrients, and rapid growth rates. These changes favored endothermic metabolism. It must also be presumed that lactation, an efficient means of nutrient transfer which is required for endothermy, was also evolving during this time.

Mammals of the Triassic and Jurassic ranged in size from small shrew-like (2-3 g, Hadrocodium) through mice-like (30-90 g, morganucodontids) to large rat-like size (up to 500 g, Sinocododon). It is also generally agreed that they were egg-laying and endothermic. These small mammals are ancestral to the first monotremes, marsupials, and eutherians of the early Cretaceous. It is reasonable to assume that these Cretaceous animals all lactated and therefore, their common ancestor, probably Jurassic form, also must have lactated. The first organisms that lactated, therefore, must have been small, egg-laying, endothermic animals.
Hopson (7) argued that in order for newly hatched precocial young to be capable of endothermy they would have needed to be very large relative to mother size. This means that the necessary large size of the egg to produce precocial young would have been precluded based on maternal size. Therefore, these animals must have laid small-yolked eggs that would have produce altricial young. For reasons to be elaborated on later, these small yolked-eggs must have been incubated in a warm, humid environment (pouch). Also, the altricial hatchlings could only have survived by complete dependence on mother for food, either directly by mouth, as seen in birds, or by means of nutritive cutaneous secretions (milk). Because no known mammal has evidence of specialized lingual, esophageal, or gastric glandular secretions for feeding their young, as seen in birds, it is possible to conclude that a functional cutaneous gland secretory system was already in place. This system may have been functional in therapsids for the purpose of supplying water and nutrients to parchment-shelled eggs in order to prevent desiccation and supplement nutrition. Deposit of calcium over eggshell membranes, which does prevent desiccation, is a derived character by sauropsids and not a primitive trait of primitive amniotes (12). This view is consistent with the absence of fossil eggs from the Carboniferous, Permian, and early Triassic (13).

As synapsids evolved endothermy, the advantages of egg incubation at higher temperatures to allow for rapid growth and decreased incubation time also transpired. An obvious problem with incubation at a high temperature, however, would be a large water loss across the egg due to the high thermal gradient. If early synapsids did lay parchment-shelled eggs, as extant monotremes, then egg mortality due to rapid flux of water vapor loss would be at a premium. It has been calculated that for a given egg size and surface and a particular ambient vapor pressure, parchment-shelled eggs lose water at a rate of 50-150 times that of a rigid-shelled crocodilian or turtle egg (14). On the other hand, parchment-shelled eggs could gain moisture in a relative humid environment. Without a calcified shell, such a moist environment would be mandated to prevent desiccation.

Early synapsids could have buried eggs in the moist ground to prevent water loss, however this would be incompatible with the need for incubation at high temperatures for endothermy. A pouch or pouch-like enclosure, however, does allow for incubation at a high temperature and does provide necessary humidity to prevent desiccation. Echidnas incubate eggs in a well developed pouch at 32C, normal body temperature, and the air inside approaches saturation (15). The platypus is pouchless, however an “incubatorium” is produced as she lies on her side or back and covers eggs with her tail as the eggs rest on her abdomen. As with the echidna, body temperature and high humidity are established and maintained (15). In summary, it was necessary for parchment-shelled synapsid eggs to be incubated in a warm, moist environment for endothermy to evolve. The next case to be made is that the source of moisture was produce by cutaneous secretions. These secretions, initially functioning to provide moisture to the eggs/neonates, may be the origin of milk.

The mammary glands of extant mammals are large, complex glandular structures capable of producing large amounts of nutritive secretions. However, these glands must have evolved from some simple cutaneous glands present in Permian synapsids. The integument inherited from early amniotes by synapsids and sauropsids was likely modified to prevent transcutaneous water loss, thereby permitting a greater independence from water. It probably was very similar to the integument of extant amphibians, which have abundant mucous and granular cutaneous glands. It appears two different evolutionary approaches were used to enhance the prevention of transcutaneous water loss in these two major lineages. Sauropsids developed a thick outer stratum corneum that eventually formed thick keratinized scales. Also, a hydrophobic lipid layer was incorporated into the intermediate layers. The intricate cyclical pattern of shedding seen in lizards and snakes is much more complex than anything seen in mammals, and it is certainly a derived condition. The impervious structure of the sauropsid integument does not depend upon lipid secreting cutaneous glands to prevent water loss. The reduced number of glands seen in these animals is a derived characterisic. These glands have also been restricted to specific areas, as well as restricted to specific functions, such as sociosexual communication.

Despite an extensive fossil record, no scale impressions, like those of sauropsids, have ever been identified in basal synapsids or therapsids (16, 17). Synapsids, as well as extant mammals, instead evolved lipid secreting cutaneous glands to prevent water loss. The lipids coat the skin and hair forming a protective barrier. Not only do the secretions provide water-resistant lipids to protect and moisturize the integument and hair, but also anti-microbial agents are secreted to help prevent infection. In extant mammals these cutaneous glands are closely associated with hair follicles and are termed apo-pilo-sebaceous units. During development, the ectoderm grows into the dermis and produces the hair follicle, as well as lateral buds that form the apocrine and sebaceous glands. The eccrine (merocine) sweat gland, a third type of gland, unlike the lipid secreting apocrine and sebaceous glands, produces a watery secretion and are not associated with hair follicles. Eccrine glands are not profusely distributed in most mammals. However humans are an exception, as these glands cover most of our bodies and function in thermoregulation.

The question now becomes which of these glands, if any, is most closely related to the mammary gland. Comparing extant mammalian skin glands may help in attempting to answer this question (Table II). The morphology and function of sebaceous glands are very different than mammary glands. Sebaceous glands are highly specialized for holocrine secretion (Fig II). The mode of holocrine secretion involves the cell bursting to release its secretory product. This destroys the cell and requires that basal cells of the gland continually divide for replacement. This certainly does not resemble mammary glands, nor any other cutaneous gland. Eccrine (merocrine) glands, also, do not resemble mammary glands in many characteristics. First, eccrine glands are the only cutaneous glands that are not associated with hair follicles, and therefore probably evolved independently of hair, a seemingly critical structure in the evolution of lactation. Also, eccrine glands do not produce lipids, proteins, or complex carbohydrates, all necessary constituents of milk. Lastly, these glands do not respond to sex steroids, nor are they correlated with puberty as seen in other glands.
The glands that appear to resemble mammary glands the most are the apocrine glands. Unlike sebaceous glands, apocrine and mammary gland secretory cells are surrounded by a layer of myoepithelial cells. The myoepithelial cells contract in response to neural and/or hormonal stimulation, thereby releasing the substance from the secretory cells. The active secretory cell, in both, is also of the same type, simple cuboidal to columnar shape. Also, the development of both glands is associated with the subcutaneous fat layer, whereas sebaceous glands are only of dermal origin. Sebaceous and mammary glands do have a similar lobular appearance. However, in monotremes at times of egg-laying and hatching, the mammary gland begins secretory function when the gland is in an apocrine, tubular-like appearance (18). The mode of secretion, which perhaps reveals the most important and greatest similarity between the apocrine and mammary gland, is via the exocytosis/aprocrine process (Fig. II). Overall, it appears that a likely scenario for mammary gland evolution involves an apocrine-like skin gland giving rise to both the apocrine and mammary gland.

Synapsid radiations in the Carboniferous through Cretaceous times gradually led to the mammalian characteristics as we know them today. It is reasonable to conclude that lactation involved a similar step-like approach. As stated previously, early synapsids probably had a glandular skin capable of producing protective secretions to prevent transcutaneous water loss, bacterial infection, and abrasion. These early synapsids also produced parchment-shelled eggs vulnerable to desiccation. This combination would favor using moist, perhaps lipid protecting, secretions for egg- tending. 

In addition, the association formed between apocrine gland and hair follicle may lie in the function of the hairs ability to facilitate spreading secretions onto the eggs and reducing ambient air around the incubating eggs. This may be the origin of hair, and its thermoregulatory function may be secondary.
In summary, apocrine-like glandular secretions provided egg protection in early synapsids. These secretions then doubled in function as they evolved into hatchling food for more advanced egg-laying animals, and may still play the same dual role in monotremes. Although this has never been studied, if it can be demostrated that the monotreme mammary areolar patch transfers water and/or nutrients to unhatched eggs, as occurs in the uterus, this entire evolutionary scenario will be given greater credence.



Be Sociable, Share!

Student Papers

Related Posts

Leave a Reply