The mammary gland-specific marsupial ELP and eutherian CTI share a common ancestral gene
© Pharo et al.; licensee BioMed Central Ltd. 2012
Received: 29 July 2011
Accepted: 8 June 2012
Published: 8 June 2012
The marsupial early lactation protein (ELP) gene is expressed in the mammary gland and the protein is secreted into milk during early lactation (Phase 2A). Mature ELP shares approximately 55.4% similarity with the colostrum-specific bovine colostrum trypsin inhibitor (CTI) protein. Although ELP and CTI both have a single bovine pancreatic trypsin inhibitor (BPTI)-Kunitz domain and are secreted only during the early lactation phases, their evolutionary history is yet to be investigated.
Tammar ELP was isolated from a genomic library and the fat-tailed dunnart and Southern koala ELP genes cloned from genomic DNA. The tammar ELP gene was expressed only in the mammary gland during late pregnancy (Phase 1) and early lactation (Phase 2A). The opossum and fat-tailed dunnart ELP and cow CTI transcripts were cloned from RNA isolated from the mammary gland and dog CTI from cells in colostrum. The putative mature ELP and CTI peptides shared 44.6%-62.2% similarity. In silico analyses identified the ELP and CTI genes in the other species examined and provided compelling evidence that they evolved from a common ancestral gene. In addition, whilst the eutherian CTI gene was conserved in the Laurasiatherian orders Carnivora and Cetartiodactyla, it had become a pseudogene in others. These data suggest that bovine CTI may be the ancestral gene of the Artiodactyla-specific, rapidly evolving chromosome 13 pancreatic trypsin inhibitor (PTI), spleen trypsin inhibitor (STI) and the five placenta-specific trophoblast Kunitz domain protein (TKDP1-5) genes.
Marsupial ELP and eutherian CTI evolved from an ancestral therian mammal gene before the divergence of marsupials and eutherians between 130 and 160 million years ago. The retention of the ELP gene in marsupials suggests that this early lactation-specific milk protein may have an important role in the immunologically naïve young of these species.
Marsupials and eutherians diverged between 130 and 160 million years ago [1–3] and evolved very different reproductive strategies [4–6]. Marsupials have an ultra-short gestation ranging from 10.7 days for the stripe-faced dunnart (Smithopsis macroura)  to 38 days for the long-nosed potoroo (Potorous tridactylus)  and deliver an altricial young .
Organogenesis is completed after birth supported by a long and physiologically complex lactation, during which there is an increase in maternal mammary gland size and milk production, and there are dramatic changes in milk composition [5, 9–13]. In contrast, eutherians have a long pregnancy during which maternal investment is high [14, 15]. During eutherian lactation, milk composition remains relatively constant apart from the initial production of colostrum 24–36 hr postpartum (pp) .
The tammar wallaby (Macropus eugenii) has a 26.5-day pregnancy after embryonic diapause . After giving birth, the tammar produces milk for ~300 days until the young is weaned. Phase 1 of lactation is comprised of mammary development during pregnancy and lactogenesis around parturition. At birth, the altricial young (~400 mg) attaches to one of the four teats [5, 9, 13, 18]. Lactation proceeds only in the sucked gland, whilst the remaining three glands regress [5, 9]. The young remains permanently attached to the teat from the day of birth until day 100 pp (Phase 2A) followed by detachment from the teat and a period of intermittent sucking while confined in the pouch between days 100–200 pp (Phase 2B) [5, 13, 18]. The final phase is from day 200 to at least day 300 when the young suckles variably and begins to graze as well as maintaining a milk intake (Phase 3) . These phases are highly correlated with changes in milk composition and mammary gland gene expression [10, 13, 19]. Milk protein genes such as α-lactalbumin β-lactoglobulin (LGB), α-casein β-casein and κ-casein are induced at parturition and expressed throughout lactation, whilst others are expressed and secreted in a phase-specific manner . Early lactation protein (ELP) is expressed during Phase 2A only [13, 20, 21], whey acidic protein (WAP) is Phase 2B-specific  and late lactation protein A and B are characteristic to late Phase 2B/Phase 3 and Phase 3 respectively [23, 24].
The ELP gene was first identified in an Australian marsupial, the brushtail possum (Trichosurus vulpecula) . ELP encodes a small precursor protein with a single bovine pancreatic trypsin inhibitor (BPTI)-Kunitz domain characteristic to serine protease inhibitors. ELP is secreted in milk in multiple isoforms, which include an ~8 kDa peptide and a heavily N-glycosylated protein (~16 kDa) . ELP was later identified in the tammar [13, 20, 21, 26], the stripe-faced and fat-tailed dunnarts (Sminthopsis macroura and Sminthopsis crassicaudata respectively) and the South American grey short-tailed opossum (Monodelphis domestica)  (Refer to Additional file 1: Table S1 for the species in which the putative functional ELP/CTI gene, transcript and protein have been identified). Marsupial ELP expression is limited to the early phase of lactation [13, 20, 21, 27, 28] at the time the mother produces milk for an immunologically naïve young [29, 30]. During this period, the tammar young is permanently attached to the teat and protected by humoral (passive) immunity acquired from its mother’s milk and its own innate immunity [18, 30].
Whilst an ELP orthologue is yet to be identified in eutherians, tammar and possum ELP share ~37% similarity with bovine colostrum trypsin inhibitor (CTI) [20, 25]. CTI was discovered by chance in bovine colostrum over 60 years ago . Putative CTI proteins with trypsin inhibitor activity were subsequently isolated from colostrum of the pig , cat, sheep, goat, dog, reindeer, ferret and Blue fox , but were not found in equine colostrum . These glycosylated proteins inhibited serine endopeptidases such as trypsin, pepsin and chymotrypsin [31, 32, 35]. However, of these putative CTI proteins, only bovine CTI has been sequenced (Additional file 1: Table S1) and found to contain a Kunitz domain which generally indicates serine protease inhibitor activity (see below) . Laskowski and Laskowski  hypothesised that bovine CTI protected immunoglobulins against proteolysis during the crucial period of immunoglobulin transfer from cow to calf via colostrum. However, its function is yet to be determined. Although CTI and ELP are expressed in early milk, bovine CTI secretion is brief (~1-2 days) [31, 37], but marsupial ELP expression is prolonged (up to 100 days pp) [20, 21, 25, 28]. However, their secretion in milk is correlated with the period of immuno-incompetence in the young [29, 31].
The Kunitz domain was thought to have evolved over 500 million years ago  and is now ubiquitous in mammals, reptiles, birds, plants, insects, nematodes, venoms from snakes, spiders, cone snails and sea anemones and in viruses and bacteria [39–42]. The archetypal protein of the Kunitz domain and the BPTI-Kunitz family I2, clan IB of serine endopeptidase inhibitors in the MEROPS database [43, 44] is the much studied bovine pancreatic trypsin inhibitor, also known as aprotinin (reviewed in ). The Kunitz domain is characterised by six conserved cysteine residues which form three disulphide bonds, producing a compact, globular protein of α + β folds [43, 46, 47]. Serine endopeptidase inhibition occurs through the binding of the P1 reactive site residue within the ‘binding loop’ of the Kunitz domain to a serine residue within the catalytic cleft of the protease [47, 48]. This is a reversible, tight-binding, 1:1 interaction [44, 48]. Furthermore, the Kunitz domain P1 residue determines protease-specificity [39, 47].
Since its evolution, the Kunitz domain has been incorporated into many different genes [43, 44]. In general, each domain is encoded by a single exon [43, 49]. Some genes encode proteins with a single Kunitz domain, e.g. ELP CTI PTI spleen trypsin inhibitor (STI), the five trophoblast Kunitz domain protein genes (TKDP1-5) and serine protease inhibitor Kunitz-type-3 (SPINT3) and SPINT4. These genes, apart from the TKDPs, have 3 exons. The first exon encodes the signal- and pro-peptide, the second, a single Kunitz domain and the third, a short C-terminus. However, the TKDPs have a variable number of unique N domains inserted between the signal peptide and the Kunitz domain-encoding exon [50, 51]. Genes that encode multiple Kunitz domains include: hepatocyte growth factor activator inhibitor 1 and 2, also known as SPINT1 and SPINT2 respectively (two domains), tissue factor pathway inhibitor 1 and 2 (three domains); with up to 12 domains in the Ac-KPI-1 I nematode (Ancylostoma caninum) protein [38, 43, 44]. In addition, the Kunitz domain has been integrated into multi-domain proteins, some of which include: the collagen α3(VI), α1(VII) and α1(XXVIII) chains, WFDC6 and WFDC8, amyloid beta A4 protein, α1-microglobulin/bikunin precursor (AMBP), SPINLW1 [serine peptidase inhibitor-like, with Kunitz and WAP domains 1 (eppin)] and the WAP, follistatin/kazal, immunoglobulin, Kunitz and netrin domain containing (WFIKKN)1 and 2 proteins . Furthermore, each domain within a multi-Kunitz domain protein, may exhibit different protease activity, such as for the three tandemly repeated domains within both tissue factor pathway inhibitor 1 and 2 [43, 44, 52].
The early lactation/colostrum-specific expression of ELP/CTI suggests these Kunitz domain-encoding genes may play an important role in the neonate. The sequencing of the tammar genome , in addition to the availability of numerous vertebrate genomes including one other marsupial, the opossum, a monotreme, the platypus, many eutherians, birds (chicken, Zebra finch), fish (Zebrafish, Japanese medaka, Three-spine stickleback, Tiger and Green spotted puffers), amphibian (African clawed frog) and reptile (Green anole lizard), provides an invaluable resource with which to investigate the evolution of these genes. We used a comparative genomics approach based upon bioinformatics and PCR-based cloning of cDNA and genomic DNA to characterise the marsupial ELP and eutherian CTI genes and investigate their evolutionary history.
ELP/CTIevolved from a common ancestral gene
To determine whether the marsupial ELP gene was present in other species, we used multiple approaches. We cloned the ELP genes of the koala and fat-tailed dunnart and isolated tammar ELP from a genomic library. ELP/CTI transcripts were cloned from the mammary gland of the cow, opossum and fat-tailed dunnart and the dog CTI transcript was cloned from epithelial cells isolated from canine colostrum. We performed BLAST searches of genomic databases (Ensembl, Release 62, April 2011 , NCBI GenBank nr and WGS  and UCSC ), using a cut-off of E-value ≤ 1e-8 (nucleotides) and E-value ≤ 1e-17 (proteins). To further refine the identification of ELP/CTI orthologues based upon protein sequence, we also compared gene structures (where possible) to identify genes with a similar three-exon structure to ELP/CTI. Based upon these methods, no genes orthologous to marsupial ELP/eutherian CTI were present in fish (Zebrafish, Tiger and green spotted puffers, Three-spined stickleback), birds (chicken, zebra finch), amphibian (African clawed frog), reptile (Green anole lizard), monotreme (platypus), nor sea squirts, fruit fly, nematode (Caenorhabditis elegans) or yeast. However, many of the current genomes available provide only low sequence coverage (e.g. anole lizard, 2x; green spotted pufferfish, 2.5x; chicken, zebra finch and platypus, 6x; elephant, 7x). Many assemblies are also incomplete (contain gaps) and may contain incorrect assemblies. Hence it is possible that ELP/CTI orthologues may be identified within these genomes with future improvements in sequence coverage and assemblies.
The CTI gene was present in the Laurasiatherian orders Cetartiodactyla (cow, pig, common bottle-nosed dolphin) and Carnivora (dog, cat, Giant panda). However, based upon current genome assemblies, it is a pseudogene in Afrotheria, Xenarthra, Euarchontoglires and the Laurasiatherian orders Chiroptera and Perissodactyla.
Homology between and within the marsupial ELP and eutherian CTI peptides 1
Kunitz motif2 (51 aa)
Kunitz motif1 (19 aa)
85 - 95%
67.5 - 100%
59.1 - 100%
76.5 - 100%
84.2 - 100%
20 - 100%
57.1 - 90.5%
70.7 - 88.6%
59.1 - 90.9%
76.5 - 94.1%
40 - 83.3%
Marsupial ELP vs Eutherian CTI
57.1 - 81.0%
44.6 - 62.2%
54.9 - 68.6%
63.2 - 73.7%
10 - 60%
Conserved amino acid residues within a protein provide an indication of sites essential for its structure and biological function. Comparison of the marsupial ELP and eutherian CTI precursor proteins showed that the signal peptide (57.1%-81.0% similarity), the 51 aa BPTI KUNITZ 2 motif (54.9%-68.6%), plus the shorter 19 aa BPTI KUNITZ 1 motif within it (63.2%-73.7%) were conserved. However, the 20–22 residue linear chain of the mature ELP/CTI N-terminus had marsupial-specific and eutherian-specific homology (59.1%-100%, Table 1; Additional file 4: Tables S3B, S3C, S3D, S3E). Conservation of the short (3–10 residue) C-terminus was variable (Additional file 4: Table S3F). This was in part due to the use of different stop codons in ELP/CTI transcripts across divergent species. The opossum and dunnart ELP proteins were truncated at the end of exon 2, with the stop codon encoded by one nucleotide in exon 2 and two in exon 3 (nt 323–325 inclusive; Additional file 2: Figure S1). For all other species, two different stop codons within exon 3 were used. For the panda, cat and dog, the TAA stop codon (nt 333–335) was used. However, for the pig, cow, dolphin and the remainder of the marsupials, the equivalent TGA stop codon (nt 344–346 inclusive) was used.
Surprisingly, there was little conservation of the amino acid residue type (physiochemical properties) at the P1 reactive site within the Kunitz domain (residue 33, Figure 2). Although the P1 residue type (basic amino acid with a positively charged side chain) was conserved amongst eutherians: K (lysine) for the pig, cow and dolphin and R (arginine) for the cat, dog and panda, this was not so for marsupials. The opossum and possum ELP P1 residue was acidic with a negatively charged side chain (D, aspartate). However, the P1 residue for tammar (S, serine) and the koala and dunnarts (N, asparagine) was polar with uncharged side chains.
Although P1 residues differed, all ELP/CTI peptides were predicted to be N-glycosylated at asparagine-42, consistent for bovine CTI  and therefore should be larger than their predicted masses (8.6 to 9.6 kDa, data not shown).
Selective pressure acting upon marsupial ELP and eutherian CTI
Average rates of synonymous (dS) and non-synonymous (dN) substitutions occurring in marsupial ELP and eutherian CTI
(a) Neutral selection test (dN ≠ dS)+*
selection test (dN < dS)+*
(c) Positive selection test (dN > dS)+*
BTPI KUNITZ 2 #
BPTI KUNITZ 1 ~
Trypsin interaction site ^
The protein-coding marsupial ELP and eutherian CTI transcripts and regions within them generally exhibited a trend towards purifying selection, with a dN/dS ratio <1 (Table 2). However, based upon codon-based Z-tests, only the eutherian CTI BPTI KUNITZ 1 motif (57 nt encoding 19 amino acids) was found to be undergoing purifying selection (p < 0.05). Although the regions encoding the marsupial BPTI KUNITZ 1 motif (p = 0.103) and the marsupial and eutherian BPTI KUNITZ 2 motifs (p = 0.101 and p = 0.105 respectively) exhibited a strong trend towards purifying selection, the test values (dN < dS) were not significant. This tendency was also consistent for the putative trypsin interaction site. In contrast, three regions of the ELP/CTI transcripts showed a trend towards positive selection (dN/dS > 1). These included the regions encoding the ELP/CTI N-terminus and the eutherian CTI signal peptide. However, based upon codon-based Z-tests (dN > dS), only the eutherian CTI signal peptide (p < 0.05) was undergoing positive selection.
Marsupial ELP and eutherian CTIshare common flanking genes
The PIGT WFDC2 region of bovine chromosome 13 (~74.51-75.14 Mb) was unique. Bovine CTI was adjacent to PIGT, but there was an insertion of ~602 kb between the CTI and WFDC2 genes [49, 55] (data not shown). This region included 7 Artiodactyla-specific Kunitz domain-encoding genes including PTI STI, plus the five placenta-specific TKDP1-TKDP5 genes inclusive [50, 63]. Furthermore, the SPINLW1 gene which contains both a Kunitz and a WAP domain and the eutherian-specific SPINT4 gene were located a further ~38 kb and ~90 kb respectively downstream from WFDC2[49, 55] (data not shown). As mentioned previously, these genes, with the exception of SPINLW1 and the TKDPs, share a similar 3-exon structure. However, the TKDPs differ due to the likely “exonisation” of an intron and its subsequent duplication to produce a variable number of tripartite N-domains between the exon encoding the signal peptide and the Kunitz domain [50, 51].
CTIhas been lost in some eutherians
A closer examination of the nucleotide sequence between PIGT and WFDC2 in these and other species using the Ensembl and UCSC genome databases revealed that different mutations had most likely disrupted the CTI gene. Exon 1 was disrupted in the elephant, Hoffmann's two-toed sloth (Choloepus hoffmanni), armadillo (Dasypus novemcinctus), human and other primates and horse, with exon 2 (Kunitz domain) also excised for these species, apart from the horse. Additional file 5: Figure S2A (i) depicts a nucleotide alignment of the functional/protein-coding dog CTI exon 1 compared with the putative disrupted CTI exon 1 of the elephant, sloth, human and horse. Additional file 5: Figure S2A (ii) shows the translated sequences to highlight mutations and/or deletions within the signal peptide region of CTI. The deletion of two nucleotides within human CTI exon 1 would produce a frame-shift (as depicted by the +1 and +2 reading frames). CTI exon 2 of the mouse, rat, large flying fox (Pteropus vampyrus) and horse also appeared to have been disrupted by deletions resulting in frame-shifts when compared to the functional/protein-coding dog CTI exon 2. The disruption of the protein-coding region of equine CTI exons 1 and 2 by at least one mutation and one deletion respectively would produce a frame-shift, suggested these were a recent occurrence (Additional file 5: Figure S2B (ii)).
Transposable elements within the ELP/CTIgenes
Transposable elements integrate randomly into the genome, so the probability of the same element(s) integrating independently into orthologous positions in different species is extremely low. They therefore act as genetic markers and can be used to determine the phylogenetic relationship between genes and species . Further evidence that marsupial ELP and eutherian CTI evolved from a common ancestral gene was provided by CENSOR retrotransposon analysis  (Additional file 6: Figure S3). Retroelements of conserved fragment size and orientation were located within the PIGT ELP/CTI region. However, the elephant and human which appear to have lost CTI exons 2 and 3, had also lost retrotransposons in the corresponding region, but gained a MER5A element.
Bovine CTI, PTI, STI and the TKDPsshare a common ancestral gene
Tammar ELPexpression is up-regulated at parturition and is mammary-specific
LGB expression peaked in the mammary gland during Phase 3, consistent with .
Mammalian ELP/CTI and the evolution of bovine PTI, STI and the TKDPs
The Kunitz-type inhibitor domain has been duplicated many times throughout evolutionary history . This was no more evident than for the region of bovine chromosome 13 on which CTI and the 7 CTI-like genes were located. The PTI STI and TKDP1-5 genes were specific to the order Cetartiodactyla, sub-order Ruminantia [50, 51, 63, 72], strong evidence they evolved from CTI after the divergence of the Ruminantia ~25-35 MYA . The CTI PTI and STI genes had a similar 3-exon structure and conserved regions within both coding and non-coding segments. The PTI and STI genes and proteins were homologous and almost certainly arose by gene duplication . However, the TKDP1-5 genes had one or more additional exons inserted between the signal- and pro-peptide-encoding and Kunitz domain-encoding exons (equivalent to intron 1 of CTI PTI and STI) resulting in an expansion to 4 (TKDP5), 6 (TKDP2 3 and 4) and 12 exons (TKDP1) [50, 51, 72]. These added exons encode tripartite N-domains which had no similarity to database sequences or motifs and evolved recently due to the “exonization” of an intron within an active MER retrotransposon and its subsequent duplication [50, 63]. These elements have been associated with genetic rearrangements and deletions . This may explain the excision of CTI exons 2 (Kunitz domain) and 3 (C terminus) for the elephant and primates, based upon current genome sequencing and assemblies.
Lack of conservation of the ELP/CTI putative P1reactive site residue
All putative ELP/CTI peptides were predicted to be secreted and shared a conserved single 51 amino acid Kunitz domain. The conserved location of the 6 cysteine residues which form three disulphide bonds suggested ELP/CTI would, like bovine CTI  and PTI  form a globular protein. However, neither the identity, physiochemical properties of the ELP/CTI P1 reactive site residue, the trypsin interaction site, nor the N- and C-terminus of the proteins were conserved. The P1 “warhead” residue plays an essential role in the interaction of a Kunitz inhibitor domain with a serine protease and a P1 mutation may alter the protease specificity of the Kunitz domain to a particular substrate and the reaction kinetics [48, 76]. Kunitz inhibitors with a basic residue, K (Cetartiodactyla) or R (Carnivora) at P1 generally inhibit trypsin or trypsin-like serine endopeptidases such as chymotrypsin, pepsin, plasmin and kallikrein in vitro (e.g. bovine CTI and PTI) [31, 38, 77]. However, Kunitz domains with smaller, uncharged residues at P1, such as serine, generally inhibit elastase-like proteases (eg. neutrophil elastase) [43, 47, 76]. In contrast, Kunitz domains with an acidic, negatively-charged P1 residue (e.g. TKDP2) exhibit minimal antiprotease activity in vitro. Comparison of BPTI Kunitz domains suggested that the marsupial ELP P1 amino acids were quite rare [43, 49, 55]. Furthermore, the absence of purifying selection within the putative ELP/CTI trypsin interaction site and the lack of conservation of P1 residues provides intriguing questions as to the role(s) of the marsupial ELP and eutherian CTI proteins in vivo.
Not all Kunitz domains act as protease inhibitors . As mentioned previously, snake and spider venoms contain proteins with Kunitz domains . Some domains inhibit trypsin or chymotrypsin via P1, whilst others lack anti-protease activity but have neurotoxic effects by acting as potassium channel blockers . Peigneur and colleagues  recently reported a sea anemone Kunitz domain protein, APEKTx1 (Anthopleura elegantissima potassium channel toxin 1) which had dual functions. It exhibited both trypsin-inhibitor activity and selectively blocked the Kv1.1 type of voltage-gated potassium channels. Furthermore, not all Kunitz protease inhibitors act via the P1 residue. The tick anticoagulant peptide (TAP) inhibits Factor X, Factor Xa and thrombin but the reactive site is located towards the N-terminus of the protein, rather than at the P1 residue of the Kunitz domain .
ELP/CTI – a conserved N-glycosylation site predicted within the Kunitz domain
All ELP/CTI proteins shared a putative conserved N-glycosylation site within the Kunitz domain at asparagine-42 (asparagine-40 for koala ELP), consistent with the site identified for bovine CTI in vitro. The proportion of sugars attached to glycosylated bovine CTI, possum ELP and tammar ELP varies, 25-40% [58, 80], 60%  and ~47-55% [20, 21, 26], respectively. However, as the N-glycosylation site occurs at the base of the pear-shaped protein and at the opposite end to the P1 site, it is unlikely to affect protease-binding activity . Unlike bovine CTI, the Kunitz domains of neither bovine PTI, STI, nor for the placenta-specific TKDPs are predicted to be N-glycosylated. In fact, very few Kunitz domains are N-glycosylated, or predicted to be so [43, 49, 55]. The exceptions are SPINT4, SPINLW1, the first Kunitz domains of bikunin and hepatocyte growth factor activator inhibitor, the second domain of tissue factor pathway inhibitor 1, as well as selected sea anemone peptides. The precise effect of N-glycosylation is uncertain, but it may enhance protein hydrophilicity and solubility, reduce proteolysis, influence cell surface signalling and adhesion and affect protein folding, turnover and quality control [81–83]. Furthermore, oligosaccharides may act as soluble receptor analogues for bacterial and viral pathogens, preventing them from attaching to the wall of the intestines, thereby stopping their passage through the gastrointestinal and urinary tracts of the young [84, 85].
The lack of conservation of the ELP/CTI N- and C-terminus was intriguing, particularly the positive Darwinian selection (p < 0.05) acting upon the coil-like marsupial ELP N-terminus. In contrast, the eutherian CTI N-terminus tended towards neutral selection. The N- and C-termini of proteins have been associated with sub-cellular targeting, protein-protein and protein-lipid interactions and macromolecular complex formation . The marsupial- and eutherian-specific homology of the mature ELP/CTI N-terminus suggested these regions may have different activities. However, the lack of conservation of the ELP/CTI C-terminus suggested these areas may have species-specific effects. Interestingly, the conservation of the TGA codon used by the tammar, koala, pig, dolphin and cow for all species but the cat (CGA) suggested it was the ancestral ELP/CTI stop codon, with more recent mutations producing a shortened ELP/CTI C-terminus in some species. Furthermore, a conserved marsupial-specific region within the 3' UTR may regulate ELP gene transcription.
ELP/CTI is expressed and secreted in milk during the early lactation/colostrogenesis period only [this study, [20, 21, 25–28, 31, 36, 37]]. Furthermore, all mammalian neonates have an innate immune system but an immature adaptive immune system and a gut which is yet to undergo maturation or ‘closure’ and is therefore permeable to macromolecules [16, 29, 87–89]. For the calf, gut maturation occurs 24–36 hr pp , whereas for the tammar, this process does not occur until ~200 days pp . Therefore, maternal milk immunoglobulins such as IgG can be passively transferred via colostrum and Phase 2A/2B milk to the gut of the young calf and tammar, respectively, where they are absorbed by the intestines and enter the circulatory system [16, 89]. Hence ELP/CTI may enhance the survival of the young by preventing the proteolytic degradation of maternal immunoglobulins , or by protecting the young against pathogens . Although sequence comparisons predict the ELP/CTI peptides are likely to inhibit serine endopeptidases, their true function(s) will only be determined through in vitro and/or in vivo studies.
The importance of local control mechanisms in the regulation of the tammar mammary glands and ELP were highlighted in this study. Whilst ELP expression proceeds in the sucked gland, the gene is down-regulated and milk production ceases in the non-sucked glands, as for the possum . However, this partitioning of mammary glands and lactation does not occur in eutherians . Marsupial ELP/eutherian CTI expression was specific to the mammary gland and lactation (Figure 8), unlike the genes that most likely evolved from bovine CTI. PTI and STI are produced in mast cells, which have a protective role and are distributed throughout the body to tissues such as the duodenum, pancreas, lung, pituitary gland, spleen and chondrocytes . In contrast, the five bovine TKDPs are differentially expressed in trophoblast cells of the ruminant placenta only during the peri-implantation period, suggesting they have an important role in the maintenance of the conceptus and pregnancy [51, 63, 72]. Hence, the bovine PTI STI and TKDP1-5 genes have undergone positive (adaptive) selection, changes in tissue-specific expression and function compared to the putative CTI ancestral gene, consistent with gene duplication and neofunctionalisation [91, 92].
The location of the CTI gene in a rapidly evolving region of the eutherian chromosome [51, 62] may explain the conversion of CTI into a putative pseudogene in Afrotheria (elephant), Xenarthra (sloth, armadillo), Euarchontoglires (humans, primates, rodents) and in selected Laurasiatherians such as the horse and flying fox.
This region included many additional genes with Kunitz and WAP 4-DSC domains , unlike for marsupials. It is possible that the role of CTI is fulfilled by one of these genes and hence the loss of the CTI gene is tolerated. Alternatively, CTI function may have become non-essential due to physiological changes in selected species. Notably, milk protein gene loss is not common amongst mammals, as genes involved in milk production are generally under negative selection . However, the conservation of the ELP/CTI gene in marsupials and Laurasiatherian orders Carnivora (dog, cat, dolphin, panda) and Cetartiodactyla (cow, pig) suggests ELP/CTI has an important role in these species.
Marsupial ELP and eutherian CTI evolved from a common ancestral gene and encode a milk protein with a single BPTI-Kunitz serine protease inhibitor domain. Although CTI was identified as the putative ancestral gene of PTI, STI and the placenta-specific trophoblast TKDP1-5 gene family, the origin of the ELP/CTI gene is inconclusive. ELP/CTI expression in the postpartum mammary gland is brief (~24-48 hrs) in eutherians but prolonged in the tammar and other marsupials (up to 100 days). However, this period correlates with the provision of milk to an immuno-incompetent young, suggesting ELP/CTI may play a vital role in immune protection of the young at this time.
Tammar wallabies (Macropus eugenii) were provided from two different marsupial colonies: VIAS (Victorian Institute of Animal Science), DPI (Department of Primary Industries), Attwood, Victoria and The University of Melbourne, Victoria. Animals were kept in open grassy yards with ad libitum access to food, water and shelter, using standard animal husbandry conditions in accordance with the National Health and Medical Research Council guidelines . All experiments were approved by the Animal Experimentation Ethics Committees of the Department of Primary Industries and The University of Melbourne.
Tissues (salivary gland, adrenal gland, pituitary gland, lymph node, spleen, liver, kidney, lung, pancreas, brain, small intestines, hind gut, muscle, heart, ovaries) were collected from adult female tammars (n = 2). Mammary glands were also collected from adult females at different stages of pregnancy and lactation (n = 60). Mammary glands from virgin females were collected from tammar pouch young (~220 days of age, n = 3). Testes and epididymides were collected from adult tammar males (n = 2). Tissue samples derived from ear-tagging of a population of koalas (Phascolarctos cinereus) located on French Island, Victoria, were donated by Dr. Kath Handasyde and Dr. Emily Hynes from the Department of Zoology, The University of Melbourne. Total RNA extracted from a grey short-tailed opossum (Monodelphis domestica) mammary gland from day 15 of lactation (early-lactation) was provided by Dr Denijal Topcic (The University of Melbourne) from animals provided by Professor Norman Saunders (The University of Melbourne). Dr Peter Frappell (Latrobe University) provided fat-tailed dunnart mammary gland tissue from day 37 of lactation (Phase 2) and liver tissue. Dr Amelia Brennan (The University of Melbourne) provided total RNA isolated from the mammary gland of a late-pregnant (~8 months) Holstein-Friesian cow. A small quantity of dog colostrum (~20 μL) from a late-pregnant (~2 weeks prepartum) Labrador in its first pregnancy was also kindly donated by Cate Pooley (The University of Melbourne). All samples were snap frozen in liquid nitrogen and stored at −80°C until use, with the exception of the koala ear punches, which were stored at 4°C.
RNA extraction and northern analysis
Total RNA was extracted from tissues using the Qiagen RNeasy Midi Kit (Qiagen) and from cells isolated from colostrum using RNAWIZ (Ambion). RNA extracted from cells shed into milk during the lactation process provides a good representation of gene expression in the mammary gland  and therefore eliminates the need for destructive tissue sampling. RNA was electrophoresed through a 1% agarose, low-formaldehyde (1.1%) gel with 1X MOPS [3(N-Morpholino) Propane Sulfonic Acid] buffer at 4°C and then transferred to Zeta-Probe GT Blotting Membrane (BioRad) in 20X SSC (3.0 M sodium chloride, 0.3 M trisodium citrate, pH 7.0) overnight.
Membranes were rinsed in 2X SSC, UV crosslinked at 1200 J (Stratagene UV Stratalinker1800) and hybridized in 25 mL [30% deionised formamide, 5 X SSC, 50 mM sodium acetate, herring sperm DNA (100 μg/μL), 5 mL Denhart’s 50X stock solution, 0.1% SDS] with an [α-32P] dCTP-labelled probe [DECAprime II Random Priming DNA Labelling Kit (Ambion)] and incubated for ~16 hr at 42°C. The tammar ELP RsaI digested LGB (to detect both LGB transcripts ) and CST3 probes were either amplified by RT-RCR from tammar mammary gland total RNA or sourced from clones in a tammar mammary gland EST library held by the Cooperative Research Centre for Innovative Dairy Products , with plasmid DNA isolated and the cDNA insert amplified by PCR. Membranes were washed (0.1X SSC, 0.1% SDS) twice for 15 min at 60°C, wrapped in cling film, sealed into plastic pockets and exposed to a General Purpose Storage Phosphor screen and scanned on a Typhoon 8600 Scanner (Molecular Dynamics/GE Healthcare). Membranes were stripped of probes by incubation with boiling (100°C) 1X SSC, 0.1% SDS on a shaking platform for two 15 min periods, then rinsed with RT 1X SSC, 0.1% SDS.
RT-PCR and cloning of ELP/CTI
Primer sequences and conditions used to amplify ELP/CTI genes and transcripts
PCR Product Size (bp)
FT dunnart transcript
94°C for 2 min; 35 cycles of 94°C for 30 sec; 59°C for 30 sec; 68°C for 1 min; 68°C for 10 min
94°C for 2 min; 35 cycles of 94°C for 30 sec; 60°C for 30 sec; 68°C for 30 sec; 68°C for 10 min
94°C for 2 min; 35 cycles of 94°C for 30 sec; 58°C for 30 sec; 68°C for 30 sec; 68°C for 10 min
94°C for 2 min; 35 cycles of 94°C for 30 sec; 55°C for 30 sec; 68°C for 1 min; 68°C for 10 min
94°C for 2 min; 35 cycles of 94°C for 30 sec; 58°C for 30 sec; 68°C for 30 sec; 68°C for 10 min
FT dunnart gene
94°C for 2 min; 35 cycles of 94°C for 30 sec; 55°C for 30 sec; 68°C for 6 min; 68°C for 10 min
94°C for 2 min; 35 cycles of 94°C for 30 sec; 52°C for 30 sec; 68°C for 4 min; 68°C for 10 min
Tammar gene (6.2 kbpromoter)
94°C for 2 min; 35 cycles of 94°C for 30 sec; 57°C for 30 sec; 68°C for 8 min; 68°C for 10 min
Tammar gene (7.9 kbpromoter)
94°C for 2 min; 35 cycles of 94°C for 30 sec; 57°C for 30 sec; 68°C for 8 min; 68°C for 10 min
Genomic DNA isolation and cloning
Genomic DNA was isolated from koala and fat-tailed dunnart tissues as described . The ELP/CTI genes were amplified by PCR (Table 3) using Platinum Taq DNA Polymerase and ~200 ng of genomic DNA template, cloned into pGEM-T Easy and sequenced.
Isolation of the tammar ELPgene from a genomic library
A tammar genomic library (liver) in the E. coli phage vector lambda EMBL3 T7/SP6 was screened with tammar ELP cDNA and a positive clone isolated. The clone was SalI digested and the ~14.7 kb genomic DNA fragment cloned into a modified pBeloBACII plasmid vector. Digestion of pBeloBACII-14.7kbtELP with SalI and HindIII yielded three fragments, 6.2 kb SalI/HindIII, 5.2 kb HindIII/HindIII and 3.3 kb SalI/HindIII. These fragments were sub-cloned into pBluescript SK and the latter two clones sequenced by the Australian Research Genome Facility (Australia). The remaining 6.2 kb was sequenced (Department of Pathology, The University of Melbourne), providing the full sequence of the genomic clone (14.704 kb). BLAST  searches of the NCBI Macropus eugenii WGS (Whole Genome Shotgun) trace archives and assembly of hits with CAP3 [99, 100] produced a contig of 54,363 bp which included ELP and the first 2 exons of WFDC2.
Fluorescence in situhybridisation (FISH)
Metaphase spreads were prepared from the tammar and FISH performed as described . The 14.7 kb tammar ELP genomic clone was used as a probe. Slides were examined using a Zeiss Axioplan microscope and images captured using the Spot Advance software package. Pictures were processed with Confocal Assistant, Image J, Adobe Illustrator and Adobe Photoshop. Chromosomal location of ELP was verified by at least ten metaphase spreads that had at least three or four signals out of a maximum of four.
cDNA microarray analysis of tammar ELPgene expression
ELP gene expression in the tammar mammary gland was investigated by analysing a microarray database [69, 102–104] produced from custom-made cDNA microarray slides and total RNA collected from glands at each phase of the lactation cycle [69, 102–104]. Glass microarray slides were printed by the Peter MacCallum Cancer Centre Microarray Core Facility, Melbourne, Australia and contained 10,368 tammar cDNA spots which were derived from a commercially prepared (Life Technologies, Rockville, MD, USA), normalised 15,001 tammar mammary gland EST (expressed sequence tag) library. The library was prepared using tammar mammary gland total RNA pooled from various time points in pregnancy (P), lactation (L) and involution (I). These included: day 26P, d55L, d87L, d130L, d180L, d220L, d260L and d5I (tissue from a d45L female 5 days after removal of the pouch young (RPY)) . Gene expression changes in the tammar mammary gland during the reproductive cycle were investigated by a large-scale microarray experiment involving 36 comparisons (72 slides including dye swaps, 144 channels in total) [69, 102–104].
Sixteen different time points were used in the experiment: virgin female ~ 300 days old (n = 3), pregnancy (Phase 1: d5P, d25P, d26P; n = 1 per time point), lactation (Phase 2A: d1L, d5L, d80L; Phase 2B: d130L, d168L, d180L; Phase 3: d213L, d220L, d260L; n = 1 per time point) and involution (pouch young were removed at d264L and mammary tissue sampled 1, 5 and 10 days after RPY; n = 1 per time point). Microarray probes were prepared from total RNA (50 μg per sample) using a two-step procedure which involved incorporation of aminoallyl-modified dUTP and then coupling with either Cy3 or Cy5 fluorescent dye [102, 104]. Slides were hybridised overnight (14–16 hr) in a humidified chamber [102, 104], scanned (Agilent scanner) and the images analysed with Versarray software (Bio-Rad).
Quantile-quantile normalisation within and between microarray slides was implemented using the Limma Package of Bioconductor . The complete data set was analysed simultaneously using a large-scale, linear mixed-model, which included random effects to account for the microarray experiment design, plus gene effects and gene-contrast effects [102, 106]. For each time point during pregnancy and lactation, there were a total of 4 different microarray comparisons made; 8 including the Cy3/Cy5 dye swap experiments. For the virgin tissues, there were a total of 12 comparisons, with these values combined for each gene and the average determined. The relative gene expression levels were determined by exponentiation of the gene effects values. The expression levels of the ELP and LGB milk protein genes and the housekeeping gene glyceraldehyde 3-phosphate dehydrogenase (GAPDH) were based upon the average expression of n = 3, 7 and 2 non-identical clones on each microarray respectively ± SEM. Microarray experiment data (E-MTAB-1057) was submitted to the EBI Array Express Archive .
ELP/CTI genes and pseudogenes were identified by BLAST searches of the NCBI GenBank nr and WGS trace archives and BLAST searches of the Ensembl Release 62, April 2011  and UCSC  genome databases. We used an Expect-value ≤ 1e-8 as a cut-off for orthologue identification for nucleotide comparisons and gene structure comparison and an E-value ≤ 1e-17 for protein comparisons. Contigs were assembled with CAP3. The following ELP/CTI genes and transcripts were submitted to GenBank: the ELP gene of the tammar (14.704 kb) [GenBank: JN191335], Southern koala [GenBank: JN191337] and fat-tailed dunnart [GenBank: JN191336], the ELP transcripts of the tammar [GenBank: JN191338], fat-tailed dunnart [GenBank: JN191339] and South American opossum [GenBank: JN191340] and CTI transcripts of the cow (Holstein-Friesian breed) [GenBank: JN191341] and dog (Labrador breed) [GenBank: JN191342]. Third party annotations of the ELP/CTI gene were also submitted to GenBank for the cat: [GenBank: BK008083], dog: [GenBank: BK008082], dolphin [GenBank: BK008086], opossum [GenBank: BK008085] and panda [GenBank: BK008084].
The genomic regions encompassing the PIGT ELP/CTI and WFDC2 genes in different species were sourced from either the Ensembl or UCSC genome databases for sequence comparisons using mVISTA . These included: dog build CanFam2 chr24: 35680293–35758485, elephant build loxAfr3:
SuperContig_scaffold_19:44809970–44903157, horse build EquCab2 chr22: 34,465,586-34568786, human build hg19 chr20: 436717–510935, mouse build mm9/NCBI37 chr2: 164320020–164401749, opossum build MonDom5 chr1: 501309327–501453154 and cow build Btau_4.0 chr13: 74506302–74550554 (included the PIGT and CTI genes) and 75064658–75139756 (included the WFDC2 gene). The tammar genome sequences used for comparisons included the incomplete PIGT gene in tammar build Meug_1.0 GeneScaffold_3597: 2268–20682, and a 54,363 bp contig which included tammar ELP and the first 2 exons of WFDC2. The contig was compiled by BLAST searches of the NCBI Macropus eugenii WGS trace archives with the tammar ELP gene and assembly with CAP3. The following bovine chromosome 13 genes were also extracted for comparisons: CTI (74530701–74533686), PTI (75011365–75016221), STI (75065067–75069211), TKDP1 (74843274–74860062), TKDP2 (74913592–74923363), TKDP3 (74567402–74577188), TKDP4 (74874966–74883256), and TKDP5 (74976879–74983345). The web-based CENSOR tool  was used to mask sequences and identify transposable elements by comparison to the Repbase database of repeat elements . Putative exons, transcripts and proteins within genomic sequences were predicted using GENSCAN . However, the third exon of ELP/CTI was incorrectly predicted by GENSCAN and was therefore determined by manual comparison to known ELP/CTI splice sites. Splice site location was confirmed by comparison of transcripts and putative proteins. Masked sequences were analysed with mVISTA . Specifications used for each analysis are described in the relevant figure legends.
The ELP/CTI, PTI, STI, SPINT4 (bovine SPINT3 has not been detected) and TKDP family of proteins were subjected to a Prosite database scan  to identify putative conserved motifs and post-translational modifications. Putative leader sequences (indicative of secreted proteins) and N-glycosylation sites based upon the NX(S/T) motif were predicted by SignalP 3.0 and NetNGlyc 1.0 Server, respectively, using the Center for Biological Sequence analysis Prediction Servers . Sequences were aligned with CLUSTALW2  and homology within ELP/CTI transcripts and proteins assessed with MatGAT (Matrix Global Alignment Tool) 2.01 software . MatGAT produces pairwise alignments only and determines homology between each sequence pair based upon the BLOSUM50, BLOSUM62 (used for this study) or PAM250 matrix.
Selection pressures acting upon different regions of the marsupial ELP and eutherian CTI precursor proteins were determined by dN/dS analysis with MEGA5 software . The protein-coding regions of the marsupial and eutherian transcripts were analysed separately. For each region, the average transition/transversion ratio was calculated using the Maximum Composite Likelihood estimate of the pattern of nucleotide substitution based upon the Tamura-Nei model  and then used in the subsequent dN/dS analysis. All codon positions were used, but positions within the alignment containing gaps were eliminated from the analysis. In pairwise comparisons, dN (number of non-synonymous changes per non-synonymous site) and dS (number of synonymous changes per synonymous site) were estimated using the Nei-Gojobori method  with modified Jukes-Cantor correction  and their variances determined by boostrapping (1000 replications). Codon-based Z-tests for positive (dN > dS), purifying (dN < dS) and neutral (dN = dS) selection were carried out using the Modified Nei-Gojobori method with Jukes-Cantor correction in MEGA5.
The phylogenetic relationship between the protein-coding regions of the marsupial ELP, eutherian CTI, bovine TKDP1-5 PTI and STI transcripts was investigated using PHYLIP software version 3.69 . Bovine secretory leukocyte protease inhibitor (SLPI, GenBank: NM_001098865) was used as an outgroup for the analysis.
Transcripts were aligned with MUSCLE  and then 100 bootstrapped alignments generated with SEQBOOT (PHYLIP). The phylogenetic relationship between the sequences was determined using different methods including the character-based maximum likelihood and maximum parsimony methods, as well as distance-based methods. Maximum likelihood trees were generated with DNAMLK which uses a molecular clock assumption. A transition/transversion ratio of 1.34 and a coefficient of variation for the rate of substitution among sites of 0.848 (based upon a gamma distribution with a shape of 1.39) were also specified for the analysis. These values were derived from a Maximum Likelihood test of best fit for 24 different nucleotide substitution models with MEGA5. A Hidden Markov Model using 5 categories, global rearrangements and a randomized input order jumbled once were also used for the DNAMLK analysis. A consensus tree was generated with CONSENSE specifying SLPI as an outgroup root, redrawn with RETREE and plotted with DRAWGRAM. Bootstrapped trees were also generated without the molecular clock assumption (DNAML) and using maximum parsimony (DNAPARS). Distance-based analysis on bootstrapped alignments was carried out with DNADIST using the Kimura  model of nucleotide substitution. The values used for transition/transversion ratio and gamma distribution were the same as for the maximum likelihood analysis. Trees were generated with the FITCH joining method  using global rearrangements, a randomized input order jumbled 10 times and SLPI as an outgroup root. The bovine CTI TKDP1-5 PTI STI SPINT4 and SLPI protein-coding transcripts were also analysed with PHYLIP as described above. However, a transition/transversion ratio of 1.39 and a coefficient of variation for the rate of substitution among sites of 0.913 were used.
After the submission of this manuscript, we identified the ELP gene in the Tasmanian devil (Sarcophilus harrisii) by in silico analysis of the DEVIL7.0 assembly.
Four disulphide core
Expressed sequence tag
Long terminal repeat
Medium Reiterated frequency repeat
Million years ago
Removal of pouch young
Serine peptidase inhibitor-like with Kunitz and WAP domains 1
Serine protease inhibitor Kunitz type
Tissue factor pathway inhibitor
Whey acidic protein
Wap four disulphide core
WAP, follistatin/kazal, immunoglobulin, Kunitz and netrin domain containing protein
Whole genome shotgun.
We thank Dr Kaylene Simpson, Michael Wilson, Jenni Carfi and Dr Jane Whitley for their work in the cloning of the tammar ELP gene, Prof Geoff Shaw for his assistance in tammar tissue dissections and Scott Brownlees, Kerry Martin and Jenni Carfi for assistance with animal handling. We also thank Dr Matthew Digby and Sonia Mailer for their work on the microarray experiments. We thank Prof Geoff Shaw and Keng Yih Chew for the provision of tammar photos and Dr Andrew Pask for helpful comments on this manuscript. This research was funded by the Cooperative Research Centre for Innovative Dairy Products (CRC-IDP) and the Department of Zoology, The University of Melbourne. EAP was supported by an Australian Postgraduate Award, AAD by a Melbourne Research Scholarship and both EAP and AAD were recipients of a scholarship top-up from the CRC-IDP.
- Bininda-Emonds OR, Cardillo M, Jones KE, MacPhee RD, Beck RM, Grenyer R, Price SA, Vos RA, Gittleman JL, Purvis A: The delayed rise of present-day mammals. Nature. 2007, 446: 507-512.PubMedGoogle Scholar
- Luo ZX, Yuan CX, Meng QJ, Ji Q: A Jurassic eutherian mammal and divergence of marsupials and placentals. Nature. 2011, 476: 442-445.PubMedGoogle Scholar
- Luo Z-X, Ji Q, Wible JR, Yuan C-X: An early Cretaceous Tribosphenic mammal and Metatherian evolution. Science. 2003, 302: 1934-1940.PubMedGoogle Scholar
- Oftedal OT: Lactation: land mammals, species comparisons. Encyclopedia of Animal Science. Edited by: Ullrey DE, Baer CK, Pond WG. 2011, Taylor & Francis: Published online:19 Nov 2010, New York, 664-666. 2Google Scholar
- Tyndale-Biscoe CH, Renfree MB: Reproductive Physiology of Marsupials. 1987, Cambridge University Press, Cambridge, UKGoogle Scholar
- Lefèvre CM, Sharp JA, Nicholas KR: Evolution of lactation: ancient origin and extreme adaptations of the lactation system. Annu Rev Genomics Hum Genet. 2010, 11: 219-238.PubMedGoogle Scholar
- Selwood L, Woolley PA: A timetable of embryonic development, and ovarian and uterine changes during pregnancy, in the stripe-faced dunnart, Sminthopsis macroura (Marsupialia: Dasyuridae). J Reprod Fertil. 1991, 91: 213-227.PubMedGoogle Scholar
- Hughes RL: Reproduction in the macropod marsupial Potorous tridactylus (Kerr). Aust J Zool. 1962, 10: 193-224.Google Scholar
- Stewart F: Mammogenesis and changing prolactin receptor concentrations in the mammary glands of the tammar wallaby (Macropus eugenii). J Reprod Fertil. 1984, 71: 141-148.PubMedGoogle Scholar
- Green B, Newgrain K, Merchant J: Changes in milk composition during lactation in the tammar wallaby (Macropus eugenii). Aust J Biol Sci. 1980, 33: 35-42.PubMedGoogle Scholar
- Joss JL, Molloy MP, Hinds L, Deane E: A longitudinal study of the protein components of marsupial milk from birth to weaning in the tammar wallaby (Macropus eugenii). Dev Comp Immunol. 2009, 33: 152-161.PubMedGoogle Scholar
- Tyndale-Biscoe CH: Life of Marsupials. 2005, CSIRO Publishing, Collingwood, VIC, AustraliaGoogle Scholar
- Nicholas K, Simpson K, Wilson M, Trott J, Shaw D: The tammar wallaby: a model to study putative autocrine-induced changes in milk composition. J Mammary Gland Biol Neoplasia. 1997, 2: 299-310.PubMedGoogle Scholar
- Hayssen V, Lacy RC, Parker PJ: Metatherian reproduction - transitional or transcending. Am Nat. 1985, 126: 617-632.Google Scholar
- Renfree MB: Marsupial reproduction the choice between placentation and lactation. Oxf Rev Reprod Biol. 1983, 5: 1-29.Google Scholar
- Kruse PE: The importance of colostral immunoglobulins and their absorption from the intestine of the newborn animals. Ann Rech Vet. 1983, 14: 349-353.PubMedGoogle Scholar
- Renfree MB, Fletcher TP, Blanden DR, Lewis DR, Shaw G, Gordon K, Short RV, Parer-Cook E, Parer D: Physiological and behavioural events around the time of birth in macropodid marsupials. Kangaroos, Wallabies and Rat-Kangaroos. Edited by: Grigg G, Jarman PJ, Hume ID. 1989, Surrey Beatty and Sons Pty. Ltd, Sydney, 323-337.Google Scholar
- Tyndale-Biscoe CH, Janssens PA: The Developing Marsupial - Models for Biomedical Research. 1988, Springer, Berlin, GermanyGoogle Scholar
- Lefèvre CM, Digby MR, Whitley JC, Strahm Y, Nicholas KR: Lactation transcriptomics in the Australian marsupial, Macropus eugenii: transcript sequencing and quantification. BMC Genomics. 2007, 8: 417-PubMedPubMed CentralGoogle Scholar
- Simpson K, Shaw D, Nicholas K: Developmentally-regulated expression of a putative protease inhibitor gene in the lactating mammary gland of the tammar wallaby, Macropus eugenii. Comp Biochem Phys B. 1998, 120: 535-541.Google Scholar
- Simpson K: The tammar wallaby, Macropus eugenii, a model to study autocrine and endocrine regulation of lactation. PhD Thesis. Latrobe University: Department of Agriculture; 1998
- Simpson KJ, Ranganathan S, Fisher JA, Janssens PA, Shaw DC, Nicholas KR: The gene for a novel member of the whey acidic protein family encodes three four-disulfide core domains and is asynchronously expressed during lactation. J Biol Chem. 2000, 275: 23074-23081.PubMedGoogle Scholar
- Trott JF, Wilson MJ, Hovey RC, Shaw DC, Nicholas KR: Expression of novel lipocalin-like milk protein gene is developmentally-regulated during lactation in the tammar wallaby, Macropus eugenii. Gene. 2002, 283: 287-297.PubMedGoogle Scholar
- Nicholas KR, Messer M, Elliott C, Maher F, Shaw DC: A novel whey protein synthesized only in late lactation by the mammary gland from the tammar (Macropus eugenii). Biochem J. 1987, 241: 899-904.PubMedPubMed CentralGoogle Scholar
- Piotte CP, Grigor MR: A novel marsupial protein expressed by the mammary gland only during the early lactation and related to the Kunitz proteinase inhibitors. Arch Biochem Biophys. 1996, 330: 59-64.PubMedGoogle Scholar
- Joss J, Molloy M, Hinds L, Deane E: Proteomic analysis of early lactation milk of the tammar wallaby (Macropus eugenii). Comp Biochem Physiol Part D Genomics Proteomics. 2007, 2: 150-164.PubMedGoogle Scholar
- De Leo AA, Lefevre C, Topcic D, Pharo E, Cheng JF, Frappell P, Westerman M, Graves JAM, Nicholas KR: Characterization of two whey protein genes in the Australian dasyurid marsupial, the stripe-faced dunnart (Sminthopsis macroura). Cytogenet Genome Res. 2006, 115: 62-69.PubMedGoogle Scholar
- Demmer J, Ross IK, Ginger MR, Piotte CK, Grigor MR: Differential expression of milk protein genes during lactation in the common brushtail possum (Trichosurus vulpecula). J Mol Endocrinol. 1998, 20: 37-44.PubMedGoogle Scholar
- Old JM, Deane EM: Development of the immune system and immunological protection in marsupial pouch young. Dev Comp Immunol. 2000, 24: 445-454.PubMedGoogle Scholar
- Basden K, Cooper DW, Deane EM: Development of the lymphoid tissues of the tammar wallaby Macropus eugenii. Reprod Fertil Dev. 1997, 9: 243-254.PubMedGoogle Scholar
- Laskowski M, Laskowski M: Crystalline trypsin inhibitor from colostrum. J Biol Chem. 1951, 190: 563-573.PubMedGoogle Scholar
- Laskowski M, Kassell B, Hagerty G: A crystalline trypsin inhibitor from swine colostrum. Biochim Biophys Acta. 1957, 24: 300-305.PubMedGoogle Scholar
- Baintner K: Occurrence of trypsin inhibitors in colostrum, meconium, and faeces of different species of ungulates and carnivores. Acta Vet Hung. 1984, 32: 91-95.PubMedGoogle Scholar
- Baintner K, Csapo J: Lack of acid-resistant trypsin inhibitor in mare's colostrum: short communication. Acta Vet Hung. 1996, 44: 95-97.PubMedGoogle Scholar
- Kassell B, Laskowski M: The comparative resistance to pepsin of six naturally occurring trypsin inhibitors. J Biol Chem. 1956, 219: 203-210.PubMedGoogle Scholar
- Cechova D, Svestkova VV, Keil B, Sorm F: Similarities in primary structures of cow colostrum trypsin inhibitor and bovine basic pancreatic trypsin inhibitor. FEBS Lett. 1969, 4: 155-156.PubMedGoogle Scholar
- Veselsky L, Cechova D, Jonakova V: Secretion and immunochemical properties of the trypsin inhibitor from bovine colostrum. Hoppe Seylers Z Physiol Chem. 1978, 359: 873-878.PubMedGoogle Scholar
- Ikeo K, Takahashi K, Gojobori T: Evolutionary origin of a Kunitz-type trypsin inhibitor domain inserted in the amyloid beta precursor protein of Alzheimer's disease. J Mol Evol. 1992, 34: 536-543.PubMedGoogle Scholar
- Rawlings ND, Tolle DP, Barrett AJ: Evolutionary families of peptidase inhibitors. Biochem J. 2004, 378: 705-716.PubMedPubMed CentralGoogle Scholar
- Yuan CH, He QY, Peng K, Diao JB, Jiang LP, Tang X, Liang SP: Discovery of a distinct superfamily of Kunitz-type toxin (KTT) from tarantulas. PLoS One. 2008, 3: e3414-PubMedPubMed CentralGoogle Scholar
- Fry BG, Roelants K, Champagne DE, Scheib H, Tyndall JD, King GF, Nevalainen TJ, Norman JA, Lewis RJ, Norton RS, Renjifo C, de la Vega RC: The toxicogenomic multiverse: convergent recruitment of proteins into animal venoms. Annu Rev Genomics Hum Genet. 2009, 10: 483-511.PubMedGoogle Scholar
- Schweitz H, Bruhn T, Guillemare E, Moinier D, Lancelin JM, Beress L, Lazdunski M: Kalicludines and kaliseptine. Two different classes of sea anemone toxins for voltage sensitive K + channels. J Biol Chem. 1995, 270: 25121-25126.PubMedGoogle Scholar
- MEROPS: . the peptidase database http://merops.sanger.ac.uk/
- Rawlings ND: Peptidase inhibitors in the MEROPS database. Biochimie. 2010, 92: 1463-1683.PubMedGoogle Scholar
- Ascenzi P, Bocedi A, Bolognesi M, Spallarossa A, Coletta M, De Cristofaro R, Menegatti E: The bovine basic pancreatic trypsin inhibitor (Kunitz inhibitor): a milestone protein. Curr Protein Pept Sci. 2003, 4: 231-251.PubMedGoogle Scholar
- Huber R, Kukla D, Ruhlmann A, Steigemann W: Pancreatic trypsin inhibitor (Kunitz). I. Structure and function. Cold Spring Harb Symp Quant Biol. 1972, 36: 141-148.PubMedGoogle Scholar
- Laskowski M, Kato I: Protein inhibitors of proteinases. Annu Rev Biochem. 1980, 49: 593-626.PubMedGoogle Scholar
- Laskowski M: Protein inhibitors of serine proteinases - mechanism and classification. Adv Exp Med Biol. 1986, 199: 1-17.PubMedGoogle Scholar
- Ensembl Genome Browser : . http://www.ensembl.org/,
- Chakrabarty A, Green JA, Roberts RM: Origin and evolution of the TKDP gene family. Gene. 2006, 373: 35-43.PubMedGoogle Scholar
- Chakrabarty A, MacLean JA, Hughes AL, Roberts RM, Green JA: Rapid evolution of the trophoblast Kunitz domain proteins (TKDPs) - A multigene family in ruminant ungulates. J Mol Evol. 2006, 63: 274-282.PubMedGoogle Scholar
- Rawlings ND, Barrett AJ, Bateman A: MEROPS: the peptidase database. Nucleic Acids Res. 2010, 38: D227-233.PubMedPubMed CentralGoogle Scholar
- Renfree MB, Papenfuss AT, Deakin JE, Lindsay J, Heider T, Belov K, Rens W, Waters PD, Pharo EA, Shaw G, Wong ES, Lefèvre CM, Nicholas KR, Kuroki Y, Wakefield MJ, Zenger KR, Wang C, Ferguson-Smith M, Nicholas FW, Hickford D, Yu H, Short KR, Siddle HV, Frankenberg SR, Chew KY, Menzies BR, Stringer JM, Suzuki S, Hore TA, Delbridge ML, et al: Genome sequence of an Australian kangaroo, Macropus eugenii, provides insight into the evolution of mammalian reproduction and development. Genome Biol. 2011, 12: R81-PubMedPubMed CentralGoogle Scholar
- NCBI BLAST. http://www.ncbi.nlm.nih.gov/BLAST/,
- UCSC Genome Browser. http://genome.ucsc.edu/,
- Center for Biological Sequence analysis prediction servers (SignalP 3.0 and NetNGlyc 1.0). http://www.cbs.dtu.dk/services/,
- Perona JJ, Tsu CA, Craik CS, Fletterick RJ: Crystal Structure of Rat Anionic Trypsin Complexed with the Protein Inhibitors APPI and BPTI. J Mol Biol. 1993, 230: 919-933.PubMedGoogle Scholar
- Klauser R, Cechova D, Tschesche H: The carbohydrate linkage site of cow colostrum trypsin inhibitor. Hoppe Seylers Z Physiol Chem. 1978, 359: 173-180.PubMedGoogle Scholar
- Nei M, Kumar S: Molecular Evolution and Phylogenetics. 2000, Oxford University Press, New YorkGoogle Scholar
- Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar S: MEGA5: Molecular Evolutionary Genetics Analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Mol Biol Evol. 2011, 28: 2731-2739.PubMedPubMed CentralGoogle Scholar
- Hughes AL: The evolution of functionally novel proteins after gene duplication. Proc Biol Sci. 1994, 256: 119-124.PubMedGoogle Scholar
- Clauss A, Lilja H, Lundwall A: A locus on human chromosome 20 contains several genes expressing protease inhibitor domains with homology to whey acidic protein. Biochem J. 2002, 368: 233-242.PubMedPubMed CentralGoogle Scholar
- MacLean JA, Chakrabarty A, Xie S, Bixby JA, Roberts RM, Green JA: Family of Kunitz proteins from trophoblast: expression of the trophoblast Kunitz domain proteins (TKDP) in cattle and sheep. Mol Reprod Dev. 2003, 65: 30-40.PubMedGoogle Scholar
- Frazer KA, Pachter L, Poliakov A, Rubin EM, Dubchak I: VISTA: computational tools for comparative genomics. Nucleic acids research. 2004, 32: W273-W279.PubMedPubMed CentralGoogle Scholar
- Kriegs JO, Churakov G, Kiefmann M, Jordan U, Brosius J, Schmitz J: Retroposed elements as archives for the evolutionary history of placental mammals. PLoS Biol. 2006, 4: e91-PubMedPubMed CentralGoogle Scholar
- Kohany O, Gentles AJ, Hankus L, Jurka J: Annotation, submission and screening of repetitive elements in Repbase: RepbaseSubmitter and Censor. BMC Bioinformatics. 2006, 7: 474-PubMedPubMed CentralGoogle Scholar
- Huttley GA, Wakefield MJ, Easteal S: Rates of genome evolution and branching order from whole genome analysis. Mol Biol Evol. 2007, 24: 1722-1730.PubMedGoogle Scholar
- Bird PH, Hendry KA, Shaw DC, Wilde CJ, Nicholas KR: Progressive changes in milk protein gene expression and prolactin binding during lactation in the tammar wallaby (Macropus eugenii). J Mol Endocrinol. 1994, 13: 117-125.PubMedGoogle Scholar
- Sharp JA, Digby M, Lefevre C, Mailer S, Khalil E, Topcic D, Auguste A, Kwek JH, Brennan AJ, Familari M, Nicholas KR: The comparative genomics of tammar wallaby and Cape fur seal lacation models to examine functions of milk proteins. Milk Proteins: from Expression to Food. Edited by: Thompson A, Boland M, Singh H. 2009, Academic Press/Elsevier, San Diego, 55-79. Food science and technology, international seriesGoogle Scholar
- Luo Z-X: Transformation and diversification in early mammal evolution. Nature. 2007, 450: 1011-1019.PubMedGoogle Scholar
- Murphy WJ, Eizirik E, Johnson WE, Zhang YP, Ryder OA, O'Brien SJ: Molecular phylogenetics and the origins of placental mammals. Nature. 2001, 409: 614-618.PubMedGoogle Scholar
- MacLean JA, Roberts RM, Green JA: Atypical Kunitz-type serine proteinase inhibitors produced by the ruminant placenta. Biol Reprod. 2004, 71: 455-463.PubMedGoogle Scholar
- Kingston IB, Anderson S: Sequences encoding two trypsin inhibitors occur in strikingly similar genomic environments. Biochem J. 1986, 233: 443-450.PubMedPubMed CentralGoogle Scholar
- Jurka J, Kaplan DJ, Duncan CH, Walichiewicz J, Milosavljevic A, Murali G, Solus JF: Identification and characterization of new human medium reiteration frequency repeats. Nucleic Acids Res. 1993, 21: 1273-1279.PubMedPubMed CentralGoogle Scholar
- Wagner G, Wutherich K, Tschesche H: A 1 H nuclear-magnetic-resonance study of the conformation and the molecular dynamics of the glycoprotein cow-colostrum trypsin inhibitor. Eur J Biochem. 1978, 86: 67-76.PubMedGoogle Scholar
- Laskowski M, Qasim MA: What can the structures of enzyme-inhibitor complexes tell us about the structures of enzyme substrate complexes?. Biochim Biophys Acta. 2000, 1477: 324-337.PubMedGoogle Scholar
- Cechova D, Muszynska G: Role of lysine 18 in active center of cow colostrum trypsin inhibitor. FEBS Lett. 1970, 8: 84-86.PubMedGoogle Scholar
- Peigneur S, Billen B, Derua R, Waelkens E, Debaveye S, Beress L, Tytgat J: A bifunctional sea anemone peptide with Kunitz type protease and potassium channel inhibiting properties. Biochem Pharmacol. 2011, 82: 81-90.PubMedGoogle Scholar
- Jordan SP, Mao SS, Lewis SD, Shafer JA: Reaction pathway for inhibition of blood coagulation factor Xa by tick anticoagulant peptide. Biochemistry. 1992, 31: 5374-5380.PubMedGoogle Scholar
- Tschesche H, Klauser R, Cechova D, Jonakova V: On the carbohydrate composition of bovine colostrum trypsin inhibitor. Hoppe Seylers Z Physiol Chem. 1975, 356: 1759-1764.PubMedGoogle Scholar
- Walsh CT: Posttranslational modification of proteins: expanding nature's inventory. 2006, Roberts and Company Publishers, Greenwood Village, ColoradoGoogle Scholar
- Varki A, Cummings R, Esko J, Freeze H, Hart G, Marth J: (Eds): Essentials of Glycobiology. 2009, Cold Spring Harbor Laboratories Press, Cold Spring Harbor, 2Google Scholar
- Roth J, Zuber C, Park S, Jang I, Lee Y, Kysela KG, Le Fourn V, Santimaria R, Guhl B, Cho JW: Protein N-glycosylation, protein folding, and protein quality control. Mol Cells. 2010, 30: 497-506.PubMedGoogle Scholar
- Shoaf-Sweeney KD, Hutkins RW: Adherence, anti-adherence, and oligosaccharides preventing pathogens from sticking to the host. Adv Food Nutr Res. 2009, 55: 101-161.PubMedGoogle Scholar
- Kunz C, Rudloff S: Potential anti-inflammatory and anti-infectious effects of human milk oligosaccharides. Adv Exp Med Biol. 2008, 606: 455-465.PubMedGoogle Scholar
- Chung JJ, Shikano S, Hanyu Y, Li M: Functional diversity of protein C- termini: more than zipcoding?. Trends Cell Biol. 2002, 12: 146-150.PubMedGoogle Scholar
- Kwek J, De Iongh R, Nicholas K, Familari M: Molecular insights into evolution of the vertebrate gut: focus on stomach and parietal cells in the marsupial, Macropus eugenii. J Exp Zool B Mol Dev Evol. 2009, 312: 613-624.PubMedGoogle Scholar
- Yadav M: The transmission of antibodies across the gut of pouch young marsupials. Immunology. 1971, 21: 839-851.PubMedPubMed CentralGoogle Scholar
- Brambell FWR: The Transmission of Passive Immunity from Mother to Young. 1970, North-Holland Publishing Company, AmsterdamGoogle Scholar
- Fioretti E, Angeletti M, Fiorucci L, Barra D, Bossa F, Ascoli F: Aprotinin-like isoinhibitors in bovine organs. Biol Chem Hoppe Seyler. 1988, 369 (Suppl): 37-42.PubMedGoogle Scholar
- Rastogi S, Liberles D: Subfunctionalization of duplicated genes as a transition state to neofunctionalization. BMC Evolutionary Biology. 2005, 5: 28-PubMedPubMed CentralGoogle Scholar
- Force A, Lynch M, Pickett FB, Amores A, Yan YL, Postlethwait J: Preservation of duplicate genes by complementary, degenerative mutations. Genetics. 1999, 151: 1531-1545.PubMedPubMed CentralGoogle Scholar
- Lemay D, Lynn D, Martin W, Neville M, Casey T, Rincon G, Kriventseva E, Barris W, Hinrichs A, Molenaar A, Pollard K, Maqbool N, Singh K, Murney R, Zdobnov E, Tellam R, Medrano J, German JB, Rijnkels M: The bovine lactation genome: insights into the evolution of mammalian milk. Genome Biology. 2009, 10: R43-PubMedPubMed CentralGoogle Scholar
- NHMRC: Australian Code of Practice for the Care and Use of Animals for Scientific Purposes. 2004, Australian Government, Canberra, 7Google Scholar
- Boutinaud M, Jammes H: Potential uses of milk epithelial cells: a review. Reprod Nutr Dev. 2002, 42: 133-147.PubMedGoogle Scholar
- Collet C, Joseph R, Nicholas K: A marsupial β-lactoglobulin gene: characterization and prolactin-dependent expression. J Mol Endocrinol. 1991, 6: 9-16.PubMedGoogle Scholar
- Sambrook J: Russell DW: Molecular Cloning: A Laboratory Manual. 2001, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, 3Google Scholar
- Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ: Basic local alignment search tool. J Mol Biol. 1990, 215: 403-410.PubMedGoogle Scholar
- Huang X, Madan A: CAP3: A DNA sequence assembly program. Genome Res. 1999, 9: 868-877.PubMedPubMed CentralGoogle Scholar
- CAP3. http://pbil.univ-lyon1.fr/cap3.php,
- Toder R, Wilcox SA, Smithwick M, Graves JA: The human/mouse imprinted genes IGF2, H19, SNRPN and ZNF127 map to two conserved autosomal clusters in a marsupial. Chromosome Res. 1996, 4: 295-300.PubMedGoogle Scholar
- Daly KA, Digby M, Lefèvre C, Mailer S, Thomson P, Nicholas K, Williamson P: Analysis of the expression of immunoglobulins throughout lactation suggests two periods of immune transfer in the tammar wallaby (Macropus eugenii). Vet Immunol Immunopathol. 2007, 120: 187-200.PubMedGoogle Scholar
- Khalil E: Identification of novel milk proteins and their functions by exploiting the lactation strategy of the tamar wallaby (Macropus eugenii). PhD Thesis. University of Melbourne: Department of Zoology; 2007
- Khalil E, Digby MR, Thomson PC, Lefèvre C, Mailer SL, Pooley C, Nicholas KR: Acute involution in the tammar wallaby: Identification of genes and putative novel milk proteins implicated in mammary gland function. Genomics. 2011, 97: 372-378.PubMedGoogle Scholar
- Bioconductor. http://www.bioconductor.org,
- Thomson PC: Analysis of microarray data: a mixed model finite-mixture approach [Abstract]. Proceedings of the XXIIIrd International Biometric Conference. 2006, McGill University, Montréal, International Biometric SocietyGoogle Scholar
- ArrayExpress Archive. http://www.ebi.ac.uk/arrayexpress/,
- CENSOR. http://www.girinst.org/censor/,
- Burge C, Karlin S: Prediction of complete gene structures in human genomic DNA. J Mol Biol. 1997, 268: 78-94.PubMedGoogle Scholar
- Prosite. http://prosite.expasy.org/,
- Larkin MA, Blackshields G, Brown NP, Chenna R, McGettigan PA, McWilliam H, Valentin F, Wallace IM, Wilm A, Lopez R, Thompson JD, Gibson TJ, Higgins DG: Clustal W and Clustal X version 2.0. Bioinformatics. 2007, 23: 2947-2948.PubMedGoogle Scholar
- Campanella JJ, Bitincka L, Smalley J: MatGAT: an application that generates similarity/identity matrices using protein or DNA sequences. BMC Bioinformatics. 2003, 4: 29-PubMedPubMed CentralGoogle Scholar
- Tamura K, Nei M, Kumar S: Prospects for inferring very large phylogenies by using the neighbor-joining method. Proc Natl Acad Sci U S A. 2004, 101: 11030-11035.PubMedPubMed CentralGoogle Scholar
- Nei M, Gojobori T: Simple methods for estimating the numbers of synonymous and nonsynonymous nucleotide substitutions. Mol Biol Evol. 1986, 3: 418-426.PubMedGoogle Scholar
- Zhang J, Rosenberg HF, Nei M: Positive Darwinian selection after gene duplication in primate ribonuclease genes. Proc Natl Acad Sci U S A. 1998, 95: 3708-3713.PubMedPubMed CentralGoogle Scholar
- Felsenstein J: Phylip (Phylogeny Inference Package) version 3.69. http://evolution.genetics.washington.edu/phylip.html,
- Edgar RC: MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res. 2004, 32: 1792-1797.PubMedPubMed CentralGoogle Scholar
- Kimura M: A simple method for estimating evolutionary rates of base substitutions through comparative studies of nucleotide sequences. J Mol Evol. 1980, 16: 111-120.PubMedGoogle Scholar
- Fitch WM, M E: Construction of phylogenetic trees. Science. 1967, 155: 279-284.PubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.