Evolution of Lactation in Mammalian Species

at least 500 antimicrobial peptides [6]. Some milk constituents, such as α-lactalbumin, β-lactoglobulin, whey acidic protein, and proteins in the mammary fat globule membrane, may have originated from antimicrobial components in carboniferous tetrapod skin secretions [4].

Another potentially significant tetrapod feature is the parental care of eggs, including the provision of secretions to keep them moist. Among all 3 living amphibian lineages (salamanders, frogs, and caecilians), a terrestrial system of egg development has evolved, with large-yolked eggs and direct development into adult-type hatchlings without a larval stage. In such species, parental care is nearly universal [7] and may include provision of supplemental moisture to the eggs via transcutaneous water movement on direct contact or via skin secretions [8]. Among some caecilians, a lactation-like pattern of care occurs in that maternal epidermis swells with lipid-containing material after the young hatch from direct developing eggs, and the hatchlings ingest skin and/or secretions using specialized teeth [9].

The amniotes, ancestral to both synapsids (including mammals) and sauropsids
(including dinosaurs/birds and crocodilians, for example), first appeared in the late
Carboniferous (i.e., mid-Pennsylvanian, about 310 mya; Fig. 1). The amniotic egg was a major evolutionary novelty characterized by a fibrous eggshell and a set of specialized extra-embryonic membranes that partitioned and enhanced egg physiologic functions [10], permitting eggs to become larger and to contain more yolk to support development. The parchment-like eggshell of these early amniotes lacked a calcified layer and thus remained highly permeable to water [11]. It is possible that these amniote eggs were dependent on moisture provided by the parents, just as in some terrestrial amphibians. I have argued that lactation originated as a secretion that provided water, and other secretory constituents, to eggs [11].

The synapsids subsequently underwent a series of extensive radiations and massive
extinction events, so that most lineages disappeared in the late Triassic, but the
mammaliaforms persisted and radiated in the late Triassic and Jurassic, and were
ancestral to mammals about 160 mya. The mammaliaforms became more mammal-
like, progressively smaller in size, with increased metabolic expenditure, growth rates, and activity [12]. The associated miniaturization of eggs implies hatching of altricial young, possibly fed on milk. There is evidence in the regional specialization of the extra-embryonic membranes in monotreme eggs that nutrient absorption could occur at the abembryonic end [11]. The mammaliaforms also delayed tooth eruption, developing simple “milk teeth” first, followed by adult dentition, suggesting the young were fed early in postnatal life on milk.

Thus, a plausible scenario is that lactation initially provided a “proto-lacteal” secretion containing moisture, antimicrobial constituents, and perhaps a few nutrients (such as calcium) to eggs. At some point, hatchlings began to ingest milk secretions including constituents with nutritional value that supplanted egg nutrients. Vitellogenins began to disappear about 170 mya [13], indicating the replacement of egg nutrients by milk nutrients.
 

It was the similarity in development, structure, and secretory processes that led to the hypothesis that apocrine glands may resemble the ancestral condition of mammary glands [1]. Both apocrine and mammary glands secrete by exocytosis secretory vesicles and by budding out and pinching off cellular contents with loss of cytoplasm [1]. In mammary glands, budding out and pinching off occurs during secretion of the milk fat globule (MFG); cytoplasmic crescents may be present but are minimal [14]. In most mammals, an apocrine gland on the general skin surface is typically associated with both a hair follicle and a sebaceous gland in a triad termed an apopilosebaceous unit. The development of the apopilosebaceous unit occurs in coordinated fashion, typically leaving the apocrine gland duct opening into the infundibulum of the hair follicle, such that secretion contacts the hair shaft. A parallel is found in monotremes (Fig. 1), in which mammary glands, hair follicles, and sebaceous glands form what can be termed a mammopilosebaceous unit (MPSU) [1]. The lactiferous ducts also open up into the infundibulum of an enlarged, specialized mammary hair [15]. The mammary glands in monotremes are organized into bilateral oval mammary patches consisting of 100–
200 MPSUs [15, 16]; there is no nipple. In earliest lactation, when monotreme eggs are incubated and hatched, the secreting mammary gland is still relatively small and tubular, and thus it has a superficial resemblance to an apocrine gland. There is also a developmental association of mammary glands with hair follicles and sebaceous glands in marsupials, such as koalas and kangaroos (Fig. 1). An oval primary primordium separates into nipple primordia which deepen into knobs and generate hair follicles (primary sprout), mammary glands (secondary sprout), and sebaceous glands (tertiary sprout). In the opossum, the nipple primordium develops into 8 MPSUs. Developing hair follicles penetrate the nipple epithelium and are then shed, and the mammary gland
ducts come to occupy these nipple-penetrating cavities. In the adult marsupial, the “mammary hairs” are no longer evident, but the galactophores reflect their prior existence.

In eutherian mammals, apocrine glands retain an association with hair follicles
(apopilosebaceous units), but the association of mammary glands with hair follicles, the presumed ancestral condition, is lost in many taxa. That this is a secondary condition is suggested by two phenomena. First in some taxa (e.g., the horse), the mammary gland develops as an MPSU association, and the mammary hairs and sebaceous glands remain in lactating mares [16]. Second, bone morphogenetic proteins inhibit hair follicle formation during nipple development in the mouse, and transgenic overexpression of a bone morphogenetic protein antagonist converts nipple epithelium into pilosebaceous units. This may suggest an ancestral state involving MPSUs that was altered by bone morphogenetic protein production. 
 

Caseins are unique to milk, convey a large proportion of the amino acids required
by offspring, and form large micelles containing calcium phosphate nanoclusters. Multiple caseins (characterized as αs1-, αs2-, β-, and κ-caseins) participate in these micelles, with κ-casein stabilizing the micelle in secreted milk. Caseins are a primary transport vehicle for amino acids, calcium, and phosphorus. After ingestion, κ-casein is vulnerable to proteases such as chymosin, causing the release of a macropeptide from κ-casein and precipitation of caseins into a gastric curd which entraps fat. Fat so entrapped is attacked by multiple lipases; the curd caseins themselves are hydrolyzed by proteases. A mechanism to retain milk as a semisolid (gastric curd) was no doubt an important evolutionary step.

All mammalian milks that have been studied contain the 4 primary types of caseins:
αs1-, αs2-, β-, and κ-caseins, indicating a premammalian origin, although casein genes have been variously duplicated in diverse mammalian groups. The caseins are members of a much larger family of proteins of unfolded nature that are secreted from cells, usually in association with tissue mineralization or regulation of calcium at target tissues. These proteins, termed secretory calcium- binding phosphoproteins (SCPP), are secreted by secretory epithelial cells or cells derived from underlying ectomesenchymal cells, and have an ancient history in the evolution of mineralized vertebrate tissue [17]. The SCPPs include extracellular matrix proteins secreted by ameloblasts, odontoblasts, and osteoblasts as well as salivary proteins that bind and transport calcium. As unfolded proteins, all SCPPs are low in cystine disulfide bridges, and a subclass of proteins (P/Q-rich SCPPs), which includes caseins, are particularly rich in proline and glutamine [17].

None of the milk casein genes have been found in sauropsids, but 2 members of this SCPP gene cluster, ODAM (odontogenic, ameloblast-associated protein) and FDCSP (follicular dendritic cell secreted peptide), have been found in the genomes of a frog (Xenopus) and a lizard (Anolis), respectively [17]. Based on the relative locations and structures of exons of P/Q-rich SCPPs, as well as their phylogenetic distribution, Kawasaki et al. [17] proposed that the αs- and β-caseins derive via gene duplication and exon changes from an ancestral gene (CSN1/2) that derives from another SCPP gene, either ODAM or SCPPPQ1 (which itself derived from ODAM), while κ-casein derives from the SCPP gene FDCSP (which also derived from ODAM). If this scenario is correct, the ODAM gene is ultimately the grandmother gene of all caseins and played a central role in the evolution of synapsid reproduction. The initial function of an ancestral SCPP in a protolacteal secretion may have been to regulate calcium delivery to the surface of an egg and to prevent precipitation of calcium phosphate on the eggshell [17]. This hypothesized role of ancestral casein(s) in delivering calcium to eggs is consistent with the view that early amniotes produced eggs with a fibrous calcium-free eggshell, that such eggs take up environmental calcium especially at the abembryonal end, and that the limited calcium supply in yolk would have made this beneficial [1, 11]. Subsequently, there was an expansion in the number and types of casein genes and their interaction in producing the large and complex casein micelle, which was essential to neonatal nutrition. This may have been one of the most important evolutionary novelties of lactation.
 

Milk varies tremendously in lipid content among mammals (from < 1% in rhinos and some lemurs to > 60% in some seals [18]). A unique feature of milk lipids is that they are packaged into membrane-enclosed structures known as MFGs. The core triacylglycerols in MFGs are bounded sequentially by a phospholipid monolayer, an inner protein coat, a bilayered phospholipid membrane, and a glycosylated surface [19]. Transmembrane proteins such as mucins, butyrophilin, and CD36 interact both with proteins of the inner coat, such as xanthine oxidoreductase (XOR), fatty acid-binding protein, and adipophilin, and with molecules at the milk-facing surface [19]. The glycosylated surface is primarily due to oligosaccharide chains attached to outwardly projecting domains of transmembrane proteins. Of particular interest is that cytoplasmic crescents become trapped in the MFG during the process of secretion, possibly akin to apocrine- type secretion involving terminal protrusion in a narrow (a bleb) or wide
(an endpiece) extension, which detaches via pinching off (narrow blebs), or via merging of exocytotic vesicles to create a gap (wide endpieces). From an evolutionar perspective, mammary glands had to minimize cytoplasmic loss since the volume of secretion was increased from micro- to milliliters or liters. The secretion of the MFG and its associated MFG membrane (MFGM) envelope is a highly regulated process. In particular 2 proteins, butyrophilin and XOR, play an obligatory structural role in MFGM synthesis, and if they are reduced or eliminated from mouse mammary cells via knockout of the coding genes, mice fail to produce normal milk [19]. These 2 proteins have apparently been coopted from other cellular functions during the evolution of the mammary gland. The butyrophilin in milk (butyrophilin 1A1) is one of the butyrophilin
family of proteins. In butyrophilin 1A1 in the MFGM, one domain projects inward into the underlying protein coat where it binds XOR with high affinity. XOR is an unusual partner for butyrophilin, as it is best known for its role in catalysis of the last 2 steps in uric acid formation and other enzymatic functions. XOR has particularly high expression in epithelial surfaces of the gastrointestinal tract, liver, kidney, lung, skin, and mammary gland [20]. Yet, upregulation and apical membrane localization of XOR in mammary epithelial cells during mammary gland development indicates a novel function for XOR in the MFGM. Given the antiquity of XOR and its long conservative evolutionary history,
evolution of this new mammary function must be considered a radical departure. MFG secretion presumably evolved from some prior less productive forms of fat secretion, perhaps by tetrapod or synapsid skin glands. If these were apocrinelike glands, they would have had the biochemical pathways for apocrine secretion.

One can imagine a scenario in which an ancestral apocrine secretion entailed the
pinching off of apical blebs containing cytoplasm, secretory vesicles, and perhaps cytoplasmic lipid droplets, similar to the process described for some specialized apocrine glands such as human axillary apocrine glands, glands of Moll, ceruminous glands in the outer ear canal, and rodent harderian glands. When the blebs disintegrate in the gland lumen, the various constituents are released.
 

All mammalian milks contain at least traces of sugar [18]; in many eutherians the predominant sugar is lactose (galactose [β1–4] glucose), while in monotremes, marsupials, and caniform carnivores oligosaccharides predominate, most of which contain a galactose (β1–4) glucose unit at the reducing end [21, 22]. Both lactose and oligosaccharides with lactose at the reducing end appear to be unique to mammary secretion and thus entail the evolution of a novel synthetic pathway. Some marine mammals produce milk devoid of lactose and lactose- based oligosaccharides, but this is a secondary loss.

The evolution of lactose synthesis is a remarkable example of a protein (or in this case 2 proteins) adopting completely new functions with minimal changes in structure. During lactose synthesis, a transmembrane protein in the trans- Golgi, β-1,4-galactosyltransferase 1 (β4gal-T1) binds UDP-galactose, producing a conformational change in β4gal-T1 that allows α-lactalbumin to be bound [23]. α-Lactalbumin binding to β4gal-T1 alters its specificity, allowing glucose to become the acceptor sugar for galactose transfer, resulting in the synthesis of lactose. α-Lactalbumin does not have other known functions, but β4gal-T1 does facilitate transfer of galactose from UDP-galactose to N-acetyl glucosamine at the terminus of N-linked oligosaccharide chains, which was presumably its ancestral function.

Based on amino acid sequence similarity, three-dimensional structure, and the structure of the exons that code for α-lactalbumin, it is apparent that α-lactalbumin is most closely related to c-type lysozyme and is derived from it via gene duplication and base pair substitution [24, 25]. Lysozymes are hydrolytic enzymes that cleave the glycosidic bonds in peptidoglycans, the major bacterial cell wall polymer, and thus play a key role in innate immune defense systems in both vertebrates and invertebrates. Amino acid substitutions have led to the loss of hydrolytic function in α-lactalbumin. The estimated date of origin of α-lactalbumin from c-lysozyme is ancient, prior to the time of the split of synapsids from sauropsids about 310 mya [24]. Given that c-lysozyme is widespread,
being a normal antimicrobial constituent of egg white and epithelial secretions, including amphibian skin secretions, it seems likely that an ancestral c-lysozyme, already present in secretions provided to eggs, was coopted as a modifier of carbohydrate secretion, generating lactose.

One uncertainty is whether the original milk carbohydrate was lactose or an oligosaccharide. A low rate of lactose synthesis, coupled with high activity of glycosyl transferases that could glycosylate lactose, may have produced diverse oligosaccharides rather than free lactose, similar to what is observed in many extant monotremes and marsupials. In a study of echidnas, Oftedal et al. [26] suggested O-acetylated sialyllactose may serve to provide protection for pouch young against microbial attack, and that this may have been an ancestral function of early oligosaccharides. However, the evolutionary factors that have produced such a diverse array of oligosaccharides in monotremes, marsupials, caniform carnivores, and primates (especially humans) remain uncertain.

Disclosure Statement
The author declares no conflicts of interest.

References
1. Oftedal OT: The mammary gland and its origin during synapsid evolution. J Mammary Gland Biol 2002; 7: 225–252.
2. Lemay DG, Lynn DJ, Martin WF, et al: The bovine lactation genome: insights into the evolution of mammalian milk. Genome Biol 2009; 10:R43– R43.
3. Oftedal OT: The evolution of milk secretion and its ancient origins. Animal 2012; 6: 355–368.
4. Oftedal OT: Origin and evolution of the major constituents of milk; in Fox PF, McSweeney PLH (eds): Advanced Dairy Chemistry-1A. Proteins: Basic Aspects, ed 4. New York, Springer, 2013, pp 1–42.
5. Carroll R: The Rise of Amphibians. 365 Million Years of Evolution. Baltimore, Johns Hopkins University Press,2009.
6. Jenssen H, Hamill P, Hancock REW: Peptide antimicrobial agents. Clin Microbiol Rev 2006; 19: 491–511.
7. Duellman W, Trueb L: Biology of Amphibians. Baltimore, Johns Hopkins University Press, 1994.
8. Taigen TL, Pough FH, Stewart MM: Water balance of terrestrial anuran (Eleutherodactylus coqui) eggs: importance of parental care. Ecology 1984; 65: 248–255.
9. Kupfer A, Müller H, Antoniazzi MM, et al: Parental investment by skin feeding in a caecilian amphibian. Nature 2006; 440: 926–929.
10. Packard MJ, Seymour RS: Evolution of the amniote egg; in Sumida SS, Martin KLM (eds): Amniote Origins: Completing the Transition to Land. San Diego, Academic Press, 1997, pp 265–290.
11. Oftedal OT: The origin of lactation as a water source for parchment-shelled eggs. J Mammary Gland Biol 2002; 7: 253–266.
12. Kemp TS: The Origin and Evolution of Mammals. New York, Oxford University Press, 2005.
13. Brawand D, Wahli W, Kaessmann H: Loss of egg yolk genes in mammals and the origin of lactation and placentation. PLoS Biol 2008; 6:e63.
14. Mather IH, Keenan TW: Origin and secretion of milk lipids. J Mammary Gland Biol 1998; 3: 259– 273.
15. Griffiths M: Biology of the Monotremes. New York, Academic Press, 1978.
16. Oftedal OT, Dhouailly D: Evo-devo of the mammary gland. J Mammary Gland Biol 2013; 18: 105–120.
17. Kawasaki K, Lafont A, Sire J: The evolution of milk casein genes from tooth genes before the origin of mammals. Mol Biol Evol 2011; 28: 2053– 2061.
18. Oftedal OT, Iverson SJ: Comparative analysis of non-human milks. A. Phylogenetic variation in the gross composition of milks; in Jensen RG (ed): Handbook of Milk Composition. San Diego, Academic Press, 1995, pp 749–789.
19. Mather IH: Milk fat globule membrane; in Fuquay J, Fox PF, McSweeney PL (eds): Encyclopedia of Dairy Sciences. San Diego, Academic Press, 2011, vol 3, pp 680–690.
20. Garattini E, Mendel R, Romão MJ, et al: Mammalian molybdo-flavoenzymes, an expanding family of proteins: structure, genetics, regulation, function and pathophysiology. Biochem J 2003; 372: 15–32.
21. Urashima T, Asakuma S, Leo F, et al: The predominance of type 1 oligosaccharides is a feature specific to human breast milk. Adv Nutr 2012; 3: 473S–482S.
22. Urashima T, Messer M, Oftedal OT: Comparative biochemistry and evolution of milk oligosaccharides of monotremes, marsupials and eutherians; in Pontarotti P (ed): Evolutionary Biology: Genome Evolution, Speciation, Coevolution and
Origin of Life. Cham, Springer, 2014, pp 3–33.
23. Ramakrishnan B, Qasba PK: Crystal structure of lactose synthase reveals a large conformational change in its catalytic component, the β-1,4- galactosyltransferase-I. J Mol Biol 2001; 310: 205– 218.
24. Prager EM, Wilson AC: Ancient origin of lactalbumin from lysozyme: analysis of DNA and amino acid sequences. J Mol Evol 1988; 27: 326–335.
25. Qasba PK, Kumar S: Molecular divergence of lysozymes and α-lactalbumin. Crit Rev Biochem Mol 1997; 32: 255–306.
26. Oftedal OT, Nicol SC, Davies NW, et al: Can an ancestral condition for milk oligosaccharides be determined? Evidence from the Tasmanian echidna (Tachyglossus aculeatus setosus). Glycobiology 2014; 24: 826–839.