Development of body, head and brain features in the Australian fat-tailed dunnart (Sminthopsis crassicaudata; Marsupialia: Dasyuridae); A postnatal model of forebrain formation

Most of our understanding of forebrain development comes from research of eutherian mammals, such as rodents, primates, and carnivores. However, as the cerebral cortex forms largely prenatally, observation and manipulation of its development has required invasive and/or ex vivo procedures. Marsupials, on the other hand, are born at comparatively earlier stages of development and most events of forebrain formation occur once attached to the teat, thereby permitting continuous and non-invasive experimental access. Here, we take advantage of this aspect of marsupial biology to establish and characterise a resourceful laboratory model of forebrain development: the fat-tailed dunnart (Sminthopsis crassicaudata), a mouse-sized carnivorous Australian marsupial. We present an anatomical description of the postnatal development of the body, head and brain in dunnarts, and provide a staging system compatible with human and mouse developmental stages. As compared to eutherians, the orofacial region develops earlier in dunnarts, while forebrain development is largely protracted, extending for more than 40 days versus ca. 15 days in mice. We discuss the benefits of fat-tailed dunnarts as laboratory animals in studies of developmental biology, with an emphasis on how their accessibility in the pouch can help address new experimental questions, especially regarding mechanisms of brain development and evolution.

Genomic imprinting in marsupial placentation

Genomic imprinting is a widespread epigenetic phenomenon in eutherian mammals, which regulates many aspects of growth and development. Parental conflict over the degree of maternal nutrient transfer is the favoured hypothesis for the evolution of imprinting. Marsupials, like eutherian mammals, are viviparous but deliver an altricial young after a short gestation supported by a fully functional placenta, so can shed light on the evolution and time of acquisition of genomic imprinting. All orthologues of eutherian imprinted genes examined have a conserved expression in the marsupial placenta regardless of their imprint status. Differentially methylated regions (DMRs) are the most common mechanism controlling genomic imprinting in eutherian mammals, but none were found in the marsupial imprinted orthologues of IGF2 receptor (IGF2R), INS or mesoderm-specific transcript (MEST). Instead, histone modification appears to be the mechanism used to silence these genes. At least three genes in marsupials have DMRs: H19, IGF2 and PEG10. PEG10 is particularly interesting as it is derived from a retrotransposon, providing the first direct evidence that retrotransposon insertion can drive the evolution of an imprinted region and of a DMR in mammals. The insertion occurred after the prototherian–therian mammal divergence, suggesting that there may have been strong selection for the retention of imprinted regions that arose during the evolution of placentation. There is currently no evidence for genomic imprinting in the egg-laying monotreme mammals. However, since these mammals do have a short-lived placenta, imprinting appears to be correlated with viviparity but not placentation.

Insulin is imprinted in the placenta of the marsupial, Macropus eugenii

Therian mammals (marsupials and eutherians) rely on a placenta for embryo survival. All mammals have a yolk sac, but while both chorio-allantoic and chorio-vitelline (yolk sac) placentation can occur, most marsupials only develop a yolk sac placenta. Insulin (INS) is unusual in that it is the only gene that is imprinted exclusively in the yolk sac placenta. Marsupials, therefore, provide a unique opportunity to examine the conservation of INS imprinting in mammalian yolk sac placentation. Marsupial INS was cloned and its imprint status in the yolk sac placenta of the tammar wallaby, Macropus eugenii, examined. In two informative individuals of the eight that showed imprinting, INS was paternally expressed. INS protein was restricted to the yolk sac endoderm, while insulin receptor, IR, protein was additionally expressed in the trophoblast. INS protein increased during late gestation up to 2 days before birth, but was low the day before and on the day of birth. The conservation of imprinted expression of insulin in the yolk sac placenta of divergent mammalian species suggests that it is of critical importance in the yolk sac placenta. The restriction of imprinting to the yolk sac suggests that imprinting of INS evolved in the chorio-vitelline placenta independently of other tissues in the therian ancestor of marsupials and eutherians.

Society for Reproductive Biology Founders' Lecture 2006 - life in the pouch: womb with a view

Marsupials give birth to an undeveloped altricial young after a relatively short gestation period, but have a long and sophisticated lactation with the young usually developing in a pouch. Their viviparous mode of reproduction trades placentation for lactation, exchanging the umbilical cord for the teat. The special adaptations that marsupials have developed provide us with unique insights into the evolution of all mammalian reproduction. Marsupials hold many mammalian reproductive 'records', for example they have the shortest known gestation but the longest embryonic diapause, the smallest neonate but the longest sperm. They have contributed to our knowledge of many mammalian reproductive events including embryonic diapause and development, birth behaviour, sex determination, sexual differentiation, lactation and seasonal breeding. Because marsupials have been genetically isolated from eutherian mammals for over 125 million years, sequencing of the genome of two marsupial species has made comparative genomic biology an exciting and important new area of investigation. This review will show how the study of marsupials has widened our understanding of mammalian reproduction and development, highlighting some mechanisms that are so fundamental that they are shared by all today's marsupial and eutherian mammals.

Marsupial Hypoblast: Formation and Differentiation of the Bilaminar Blastocyst in Sminthopsis macroura

Hypoblast formation in Sminthopsis macroura starts in blastocysts with a size between 1.0 and 1.4 mm, in which cells appear to be similar to each other, and finishes at the complete 2.6- or 2.7-mm bilaminar blastocyst, which is fully lined with hypoblast cells. When hypoblast cells begin allocation, the pluriblast region progressively differentiates from the trophoblast. Some pluriblast cells, which are otherwise undistinguished, lying on one side near the boundary of the circular pluriblast, move to the inside as hypoblast cells by mitosis or migration. They initially line the pluriblast and then the trophoblast. Hypoblast cells continue to leave the pluriblast/epiblast and intercalate into the underlying hypoblast layer until the advanced stages of bilaminar blastocysts. Associated with the origin of the hypoblast cells, the residual surface epiblast cells become less flatted and more cuboidal or rounded in shape. Characteristics are increased density of ribosomes, granular endoplasmic reticulum and a marked apical-basal polarity related to apical microvilli and endocytosis and more vesicles with flocculent content and a loss of the crystalloid deposits that were typical for earlier stages. Trophoblast cells become flat and elongated with only few vesicles, and they transform into extra-embryonic ectoderm cells, which are broader, rather square and with a higher density of ribosomes. Hypoblast cells are characterized by a relatively high level of ribosomes and endoplasmic reticulum, fewer small vesicles and no noticeable endocytotic processes and initially form a reticulum because the cells preferentially migrate along cell-cell boundaries by extension of long filopodia. Once hypoblast cells reach the boundary of the embryonic area and extend to line the trophoblast, they progressively consolidate into a squamous epithelium. It is suggested that the origin of the hypoblast from one side of the pluriblast and its invasion under the trophoblast from proliferating centres at the edge of the embryonic area provide mechanisms for patterning epiblast, hypoblast, trophoblast and extra-embryonic ectoderm.

Craniofacial development in marsupial mammals: Developmental origins of evolutionary change

Biologists have long studied the evolutionary consequences of the differences in reproductive and life history strategies of marsupial and eutherian mammals. Over the past few decades, the impact of these strategies on the development of the marsupial embryo and neonate has received attention. In this review, the differences in development in the craniofacial region in marsupial and eutherian mammals will be discussed. The review will highlight differences at the organogenic and cellular levels, and discuss hypotheses for shifts in the expression of important regulatory genes. The major difference in the organogenic period is a whole-scale shift in the relative timing of central nervous system structures, in particular those of the forebrain, which are delayed in marsupials, relative to the structures of the oral-facial apparatus. Correlated with the delay in development of nervous system structures, the ossification of the bones of the neurocranium are delayed, while those of the face are accelerated. This study will also review work showing that the neural crest, which provides much of the cellular material to the facial skeleton and may also carry important patterning information, is notably accelerated in its development in marsupials. Potential consequences of these observations for hypotheses on constraint, evolutionary integration, and the existence of developmental modules is discussed. Finally, the implications of these results for hypotheses on the genetic modulation of craniofacial patterning are presented.

Constraints on the morphological evolution of marsupial shoulder girdles

Throughout their evolutionary histories, marsupial mammals have been taxonomically and morphologically less diverse than their sister taxa the placentals. Because of this, it has been proposed that the evolution of marsupials has been constrained by the functional requirements of their mode of reproduction. Marsupials give birth after short gestation times to immature neonates that immediately crawl, under the power of their precociously developed shoulder girdles, to the teat where they attach and complete their early development. Using a novel approach incorporating adult and embryological morphological data, this study is the first to both: (1) statistically support adult patterns of morphological divergence consistent with the constraint hypothesis, and (2) identify ontogenetic patterns of morphological change that demonstrate that the constraint was responsible, at least in part, for their formation. As predicted by the marsupial constraint, the shoulder girdles of adult marsupials are less diverse than those of adult placentals, and adult marsupial scapulae are less morphologically diverse than adult marsupial pelves. Furthermore, marsupials that complete an extensive crawl to the teat are restricted to a common pattern of ontogenetic scapular shape change, strongly supporting the hypothesis that the morphological development of the marsupial scapula has been limited evolutionarily by its obligate role in the crawl to the teat. Because this study establishes that ontogenetic and evolutionary morphological change is correlated within mammalian scapulae, it is probable that the marsupial constraint also restricted the morphological divergence of the scapula over evolutionary time by limiting ontogenetic change in the scapula. These findings, coupled with the importance of the shoulder girdle in mammalian locomotor specialization, support the conclusion that the low morphological diversity of marsupial forms over evolutionary time could be directly due to the constraint on marsupial morphological evolution caused by the functional requirements of the crawl to the teat.

Precedence of Cell-Zona Adhesion over Cell-Cell Adhesion during Marsupial Blastocyst Formation Prohibits Morula Formation and Ensures that both the Pluriblast and Trophoblast Are Superficial

This study, based on 38 samples taken between the 16-cell stage on day 2.5 of gestation and the expanded 1.0-mm-diameter unilaminar blastocyst on day 6, describes the ultrastructural changes that occur in the conceptus of the marsupial Sminthopsis macroura in relation to cell-zonal adhesion initiated at the zygote stage and cell-cell adhesion initiated at the 16-cell stage, lineage allocation, extracellular matrix (ECM) secretion and embryo coat changes. In S. macroura, rather flattened pluriblast and rounded trophoblast cells appear as different cell types during the fourth division when nucleolar reticulation suggests activation of the zygotic genome in both cell types. The differences disappear nearly completely in blastocysts of 0.6-0.8 mm in diameter, but the two cell types then reappear as two distinct populations. The ECM varies depending on its location within the conceptus up to the stage of the expanding blastocyst. It is of rather granular appearance between the cell lining and zona pellucida and consists of patches of homogeneous material embedded in an electron-lucent substance in the cleavage cavity. Homogeneous ECM coats trophoblast but not pluriblast cells on blastocoelic surfaces. Transient structures such as 'myosin-like' fibrillar arrays, probably associated with exocytosis of ECM, and pearl string-like whorls are still present, but both disappear during further expansion of the 0.6- to 0.8-mm blastocyst. During blastocyst expansion, the patchy homogeneous ECM in the blastocoel changes structure and appears flocculent, while the continuous ECM coating trophoblast cells disappears. Pluriblast cells and yolk mass identify the embryonic pole and hemisphere, and the opposite hemisphere becomes abembryonic and is eventually fully lined by trophoblast cells. An increase in endocytotic, mainly coated vesicles at the apical, zona-orientated surface of both cell types is noticed and is probably responsible for uptake of the mucoid coat. In 1-mm blastocysts, numerous vesicles contain rod-shaped crystalline inclusions.

Conceptus Polarity and Cell-Zona Adhesion during Early Cleavage (Fertilized Tubal Egg to 8-Cell Stage) in the Marsupial Sminthopsis macroura

This study outlines the ultrastructural changes that occur in Sminthopsis macroura tubal zygotes to the 8-cell stage in relation to observations of development in vitro, oocyte polarity and cell-zona adhesion. The extremely polarized mature oocytes and zygotes have nuclear material at one pole and accumulated vesicular bodies at the other. The first division is associated with extrusion of vesicular bodies and some cytoplasm as a membrane-bound yolk mass into the perivitelline space. Early cleavage is accompanied by the appearance of an extensive, highly structured extracellular matrix (ECM) comprised of amorphous substance, granules and filaments. At the 2- and 4-cell stage the decrease in density of the ECM in the vicinity of the blastomeres may facilitate cell-zona contact. At the 8-cell stage, discharge of vesicular bodies, which mostly appear to be empty, may contribute to the ECM by increasing the area of plasma membrane for synthesis of a hyaluronan-like ECM. As in other marsupials, the precedence of cell-zona adhesion over cell-cell contacts prevents morula formation. The earliest cell-zona contacts appear when microvilli contact the zona in the uterine zygote 12-16 h after uterine entry and continue at later stages. This early contact is possible because of the absence of a dense subzonal ECM in this species. Between late zygote and late 4-cell stage the cytoplasm also contains, beside a large amount of vesicular bodies, demarcated areas where smooth endoplasmic reticulum encloses mitochondria, vesicles, granular material and fibrillar arrays. The latter develop in the late zygote stage and are found outside demarcated areas as well, often closely surrounding large vesicles, probably helping vesicle extrusion. A putative germ plasm was identified at the 4-cell stage.

Peri-gastrulation development of the dasyurid marsupial Sminthopsis macroura (stripe-faced dunnart) in vitro and evidence for patterning of the epiblast prior to gastrulation

Marsupials are potentially excellent models for the study of gastrulation because of their superficial embryonic area (EA), post-gastrulation implantation and their potential to provide information about the evolution of gastrulation. Very few studies have examined this developmental period in marsupials. Using an established developmental timetable, peri-gastrula stage Sminthopsis macroura blastocysts were collected and described in detail by observations on live blastocysts and by the use of histological and immunohistochemical techniques on fixed blastocysts. Gastrulation in S. macroura shares several aspects common to that of both eutherian mammals and birds, but in terms of tissue arrangement and conceptus form, is more similar to the chick than other mammals. Two methods of culturing EA explants flat without their shell were devised. The techniques will markedly increase the number of possible experimental manipulations, which previously were limited by the presence of blastocyst investments. Exposure of fractions of explants of round, morphologically uniform pre-gastrula stage EA to growth factors or signaling molecules implicated in vertebrate gastrulation suggests that like the chick and mouse, the marsupial epiblast is patterned prior to gastrulation. Of all factors tested, basic fibroblast growth factor (bFGF) had the most prominent effect, promoting cell differentiation, and possible mesoderm formation. Data from explant culture suggests that similar to the chick and mouse, limited specification precedes the onset of gastrulation.

Ontogenesis of the scapula in marsupial mammals, with special emphasis on perinatal stages of didelphids and remarks on the origin of the therian scapula

The development of the scapula was studied in embryonic and postnatal specimens of Monodelphis domestica and perinatal specimens of Philander opossum, Caluromys philander, and Sminthopsis virginiae using histological sections and 3D reconstructions. Additionally, macerated skeletons of postnatal M. domestica were examined. This study focused on the detachment of the scapulocoracoid from the sternum and on the acquisition of a supraspinous fossa, a supraspinatus muscle, and a scapular spine, all these events associated with the origin of the therian shoulder girdle. In none of the specimens is there a continuity of the cartilaginous scapulocoracoid with the sternum, even though the structures are in close proximity, especially in S. virginiae. At birth, the first rib laterally presents a pronounced boss that probably contacts the humerus during certain movements. Only the acromial portion of the scapular spine, which originates from the anterior margin of the scapular blade, is preformed in cartilage. The other portion is formed by appositional bone ("Zuwachsknochen"), which expands from the perichondral ossification of the scapula into an intermuscular aponeurosis between the supra- and infraspinous muscles. This intermuscular aponeurosis inserts more or less in the middle of the lateral surface of the developing scapula. Thus, the floor of the supraspinous fossa is present from the beginning of scapular development, simultaneously with the infraspinous fossa. The homology of the therian spine with the anterior border of the sauropsid and monotreme scapula is questioned. We consider the dorsal portion (as opposed to the ventral or acromial portion) of the scapular spine a neomorphic structure of therian mammals.

Sex down under: the differentiation of sexual dimorphisms during marsupial development

Marsupials have many characteristic features that make them ideal models to study the control of sexual differentiation and development. They are distinguished from eutherian mammals in their mode of reproduction and their greater dependence on the teat and mammary gland than on the placenta for development. They give birth to a highly altricial young which completes its development while firmly attached to a teat, usually within the confines of a pouch. At birth, the marsupial neonate has a well-developed digestive, respiratory and circulatory system, but retains its fetal excretory system with a fully functional mesonephric kidney and undifferentiated gonads and genitalia.

Gestational length in the koala, Phascolarctos cinereus

We report a possible case of extended gestation in the koala, Phascolarctos cinereus. Birth of a pouch young was first observed 127 days after the removal of the male from a multi-female colony at Taronga Zoo. No other males were present at that time or had access to the facility. Head measurements and other growth data collected at the time of detection and over the period of pouch life indicates the time from removal of the male and the date of birth to be between 50 and 77 days. DNA fingerprinting using microsatellite loci unambiguously assigned paternity of the pouch young to this male. These observations suggest either an extended period of gestation of at least 50 days, or activation of a dormant blastocyst from the previous breeding season, as the female entered the period of seasonal oestrus.

Ontogenetic and phylogenetic transformations of the ear ossicles in marsupial mammals

This study is based on the examination of histological sections of specimens of different ages and of adult ossicles from macerated skulls representing a wide range of taxa and aims at addressing several issues concerning the evolution of the ear ossicles in marsupials. Three-dimensional reconstructions of the ear ossicles based on histological series were done for one or more stages of Monodelphis domestica, Caluromys philander, Sminthopsis virginiae, Trichosurus vulpecula, and Macropus rufogriseus. Several common trends were found. Portions of the ossicles that are phylogenetically older develop earlier than portions representing more recent evolutionary inventions (manubrium of the malleus, crus longum of the incus). The onset of endochondral ossification in the taxa in which this was examined followed the sequence; first malleus, then incus, and finally stapes. In M. domestica and C. philander at birth the yet precartilaginous ossicles form a supportive strut between the lower jaw and the braincase. The cartilage of Paauw develops relatively late in comparison with the ear ossicles and in close association to the tendon of the stapedial muscle. A feeble artery traverses the stapedial foramen of the stapes in the youngest stages of M. domestica, C. philander, and Sminthopsis virginiae examined. Presence of a large stapedial foramen is reconstructed in the groundplan of the Didelphidae and of Marsupialia. The stapedial foramen is absent in all adult caenolestids, dasyurids, Myrmecobius, Notoryctes, peramelids, vombatids, and phascolarctids. Pouch young of Perameles sp. and Dasyurus viverrinus show a bicrurate stapes with a sizeable stapedial foramen. Some didelphids examined to date show a double insertion of the Tensor tympani muscle. Some differences exist between M. domestica and C. philander in adult ossicle form, including the relative length of the incudal crus breve and of the stapes. Several differences exist between the malleus of didelphids and that of some phalangeriforms, the latter showing a short neck, absence of the lamina, and a ventrally directed manubrium. Hearing starts in M. domestica at an age in which the external auditory meatus has not yet fully developed, the ossicles are not fully ossified, and the middle ear space is partially filled with loose mesenchyme. The ontogenetic changes in hearing abilities in M. domestica between postnatal days 30 and 40 may be at least partially related to changes in middle ear structures.

Germ cells, gonads and sex reversal in marsupials

The formation of the testis or ovary is a critical step in development. Alterations in gonadal development during fetal or postnatal life can lead to intersexuality or infertility. Several model systems have been particularly useful in studying gonadal differentiation, the eutherian mammal and amphibia, fish, and birds. However, marsupials provide a unique opportunity to investigate gonadal development and the interactions of genes and hormones in gonadal differentiation and germ cell development in all mammals. On the one hand the genetic mechanisms appear to be identical to those in eutherian mammals, including the testis-determining SRY gene. On the other hand, marsupials retain in part the plasticity of the amphibian gonad to hormonal manipulation. It is possible to induce female to male and also male to female gonadal sex reversal in marsupials by hormonal manipulation, and oestradiol can induce male germ cells to enter meiosis at the time the oogonia do. In addition, in marsupials the development of the scrotum and mammary glands are independent of testicular androgens and instead are controlled by a gene or genes on the X-chromosome. Thus marsupials provide a number of opportunities for manipulating the sexual differentiation of the gonads that are not possible in eutherian mammals and so provide a unique perspective for understanding the common mechanisms controlling sexual development.

Significance of serum early pregnancy factor concentrations during pregnancy and embryonic development in Sminthopsis macroura (Spencer) (Marsupialia: Dasyuridae)

Marsupial pregnancy differs from that in eutherians in duration, placentation and hormonal profile so much so that maternal recognition of pregnancy may not occur in polyovular marsupials. However, a comparison of gravid and non-gravid uteri reveals differences indicative of histological and physiological adaptations to pregnancy. In the present study, the hypothesis that embryo-maternal signalling occurs in polyovular marsupials was tested by examining serum from non-pregnant and pregnant Sminthopsis macroura for the presence of early pregnancy factor (EPF), a serum protein secreted by the ovary in response to the presence of a newly fertilized egg in the oviduct. EPF is detectable in the serum of pregnant, but not in non-pregnant, females in all eutherians studied to date. In the present study, EPF was detected in S. macroura serum by the rosette inhibition test during the first 9 days of the 10.7 day gestation period in this marsupial. However, EPF was not detected on day 10, just before parturition, or in non-pregnant or preovulatory animals. Immunohistochemical analysis of ovaries from gravid and non-gravid animals demonstrates that EPF is found in the capillaries, interstitial spaces and secretory cells of the corpus luteum. It is concluded that the spatiotemporal pattern of EPF activity described strongly indicates that maternal recognition of pregnancy in marsupials is mediated, at least in part, by EPF. Because the endocrinological milieu is the same in pregnant and non-pregnant marsupials, the possibility of using marsupials as an experimental system for studying EPF function unconfounded by hormonal effects is presented.

Mechanisms for pattern formation leading to axis formation and lineage allocation in mammals: a marsupial perspective

Developing patterns in early embryogenesis are analysed in conceptuses from several families, including Dasyuridae, Phalangeridae, Macropodidae and Didelphidae, in which cleavage has been examined in some detail. Features common to cleavage and blastocyst formation, and in some cases to hypoblast formation, are used to develop an outline of possible mechanisms leading to axis formation and lineage allocation. Relevant features that have been described only in some species are also included. It is suggested that certain features of marsupial cleavage establish patterns in the developing blastocyst epithelia, pluriblast, trophoblast and hypoblast that contribute to axis formation and lineage allocation. All marsupials examined had a polarized oocyte or conceptus, the polarity of which was related to the conceptus embryonic-abembryonic axis and, eventually, the conceptus dorsal-ventral axis and the formation of the pluriblast (future embryo) and trophoblast. The embryonic dorsal-ventral and anterior-posterior axes were established after the allocation of hypoblast and epiblast. Mechanisms that appear to result in patterning of the developing epithelia leading to axis formation and lineage allocation are discussed, and include sperm entry point, gravity, conceptus polarity, differentials in cell-zona, cell-cell and cell-type (boundary effects) contacts, cell division order during cleavage and signals external to the conceptus. A model of the patterning effects is included. The applicability of these mechanisms to other amniotes, including eutherian mammals, is also examined.

Identification of a homologue of POU5F1 (OCT3/4) in a marsupial, the brushtail possum

Several morphological features of early development differ between marsupials and eutherians. These include early overt conceptus polarity, the mode of segregation of pluriblast and trophoblast lineages, and the superficial position of the pluriblast within the unilaminar blastocyst epithelium. Knowledge of the molecular mechanisms of early development is lacking in marsupials. We have cloned from the brushtail possum, Trichosurus vulpecula, the partial cDNA of a marsupial homologue of mammalian POU5F1, which encodes octamer-binding transcription factor-4 (Oct-4). The 579 bp cloned coding sequence of possum POU5F1 (tvPOU5F1) shares 74, 78, and 79% identity with murine Pou5f1, human POU5F1, and bovine bPOU5F1, respectively, at the protein level. The mRNA for tvPOU5F1 was detected by in situ hybridization in oogonia and in oocytes of growing follicles, but not in early meiotic oocytes. In a gastrulation-stage conceptus, high tvPOU5F1 expression was found in the embryonic ectoderm, with low or undetectable expression in other cell types. These results are comparable with those of murine and bovine expression studies, and suggest that the putative role of POU5F1 in totipotent and pluripotent cells is conserved among mammals.

Development of the thymus, spleen, lymph nodes and liver in the marsupial, Isoodon macrourus (Northern brown bandicoot, Peramelidae)

We report for the first time the development and morphological characteristics of the spleen thymus, lymph nodes and liver of the northern brown bandicoot, Isoodon macrourus. To date few marsupial species have been studied. The development and morphological characteristics of the organs shared the typical features of those few other marsupials studied as well as those of eutherians. These suggest comparative functional properties with the eutherian immune system. The thymus differentiated within the first week of pouch life and showed evidence of immunolymphopoiesis up to juvenile stages. The spleen, though seeded by lymphocytes within the first week of pouch life, was slower to mature, but differentiated and showed signs of immunocompetency by the time young left the pouch. The mature spleen displayed the same anatomical blood filtering and immunosurveillance properties as that of the eutherian spleen, with evidence of erythrocyte destruction, thrombopoietic activity, activation and differentiation of immunocompetent lymphocytes. However, the absence of sheathed capillary structures in the spleen may indicate differences in the humoral response to circulating antigens. Similarly, lymph nodes also mature by this stage with anterior nodes appearing before posterior nodes. The mature lymph nodes displayed structural features of secondary immuno-lymphoid organs consistent with production of immune responses. Finally, the liver displayed haemopoietic activity for the first four weeks of pouch life. The pattern of development in the bandicoot appears to parallel the pattern reported for other marsupials, yet the thymus matures considerably earlier than previously reported and may be of significance in the development and onset of cell-mediated immunity. Current studies to characterise cellular components, such as T/B lymphocyes and accessory cells of these organs will help to define the mechanisms of immune recognition, activation and hence outline the basis of the marsupial humoral and cellular immunity.

Development of the male urogenital system of the koala phascolarctos cinereus

This paper described several developmental stages of the male urogenital system in the koala Phascolarctos cinereus, employing both light and scanning electron microscopy. There are few studies of the development of the urogenital system in male marsupials. Findings by White and Timms (1994) that male koalas can be infected with Chlamydia psittaci emphasise the importance of studies on male animals and in particular their reproductive system. Specimens in our study ranged in age from 15 days postnatal to adults. Due to the rarity of such specimens, details of each specimen are linked to the changes of the structures at each available stage. Light microscopy revealed that differentiation of the gonads had commenced by 15 days postnatal and that the cytological arrangements of the urogenital system are essentially the same as those of other mammals. Scanning electron microscopy revealed stereocilia and microvilli along the lumen of each ductus epididymis and cilia and microvilli along the lumina of the vasa deferentia and urethra. The development of these structures coincided with the onset of sexual maturation, sperm production and differentiation at about three years of age.

Development in vitro of Marsupials: a comparative review of species and a timetable of cleavage and early blastocyst stages of development in Monodelphis domestica

The development of marsupial oocytes and embryos in vitro is reviewed. Most stages of development have been cultured successfully, usually in a complex medium with added fetal calf serum. Simpler media without added serum have been developed for fertilization and cleavage in vitro. Culture systems have been established for oocyte maturation and fertilization in the grey short-tailed opossum and for cleavage from the zygote to the early expanding unilaminar blastocyst in a number of other marsupials. Survival in vitro of the unilaminar and early bilaminar blastocyst stages is limited in all species examined. In contrast, late bilaminar, trilaminar, embryonic and fetal stages develop at rates approximating those in vivo. More stages have been cultured successfully in Sminthopsis macroura than in any other species. It has been cultured from the late bilaminar blastocyst to within 18 h of birth. Stages of cleavage and unilaminar blastocyst formation of Monodelphis domestica timed by videotaping mating animals, proceeded at similar rates in vivo and in vitro. As in other marsupials, cleavage in this opossum is characterized by a polarized conceptus. This polarity is expressed in the distribution of organelles in the zygote and the localization of secretion of the extracellular matrix material into the cleavage cavity and of the initial cell-zona attachment. Because cell-cell adhesion follows cell-zona adhesion, a unilaminar blastocyst forms without the development of an intervening morula stage.

Early cleavage to formation of the unilaminar blastocyst in the marsupial Antechinus stuartii: ultrastructure

The development of Antechinus stuartii from the 2-cell stage to the blastocyst stage in vivo was examined by routine transmission electron microscopy. The 2-8-cell stages had a similar organization of organelles, whereas the 16- to 32-cell stages had pluriblast cells and trophoblast cells forming an epithelium closely apposed to the zona pellucida. Specialized cell-zona plugs were formed at the 8-cell stage, and primitive cell junctions appeared in later conceptuses. The cytoplasmic organelles included mitochondria, lysosomes, aggregates of smooth endoplasmic reticulum, lipid and protein yolk bodies and fibrillar arrays, possibly contractile in function. Nuclei had uniformly-dispersed dense chromatin. Nucleoli of 2-4-cell conceptuses were dense, compact and fibrillar, and those of 8-cell conceptuses and later conceptuses were finely granular and became progressively reticulated. The embryonic genome is probably not switched on before the 8-cell stage. Sperm tails were detected in cells in several early conceptuses. The yolk mass had the same organelles as cells. Centrioles were discovered for the first time in marsupial conceptuses. These were prominently situated at a spindle pole in a 32-cell blastomere and were associated with a nucleus and sperm tail at the 4-cell stage. It is very likely that the paternal centrosome is inherited at fertilization and perpetuated in Antechinus embryos during cleavage.

Development of the ovary in the brushtail possum Trichosurus vulpecula (Marsupialia)

The urogenital region of 25 fetuses and 75 pouch young, ranging in age from newborn to 103 days (d) in development, was examined in serial histological sections. The rete ridges formed the anterior extensions of the gonadal ridges and gave rise to the rete system and gonads respectively. Sexual differentiation of the ovary commenced 2.5 d after birth, when 2 cell types appeared: the larger of these then clumped together to form the medullary cords, while the smaller cells gave rise to the stroma. Primordial germ cells were still migrating, dividing and populating the peripheral gonadal regions on d 8. Cortex and medulla were distinguishable by d 12, when a thick fibrous zone separated them. The cortex was augmented by cells from the mesothelium. The rete ovarii developed from cell condensations within the rete ridges, made secondary contact with the mesonephroi and penetrated the ovaries but did not contribute to the granulosa cells. It is concluded that, contrary to the situation in most eutherians, in Trichosurus, as in other marsupials examined, the mesonephros does not contribute to rete formation, or to the granulosa cells, which appear to arise from the medullary cords.

A molecular and evolutionary study of the beta-globin gene family of the Australian marsupial Sminthopsis crassicaudata

Beta-globin gene families in eutherians (placental mammals) consist of a set of four or more developmentally regulated genes which are closely linked and, in general, arranged in the order 5'-embryonic/fetal genes-adult genes-3'. This cluster of genes is proposed to have arisen by tandem duplication of ancestral beta-globin genes, with the first duplication occurring 200 to 155 MYBP just prior to a period in mammalian evolution when eutherians and marsupials diverged from a common ancestor. In this paper we trace the evolutionary history of the beta-globin gene family back to the origins of these mammals by molecular characterization of the beta-globin gene family of the Australian marsupial Sminthopsis crassicaudata. Using Southern and restriction analysis of total genomic DNA and bacteriophage clones of beta-like globin genes, we provide evidence that just two functional beta-like globin genes exist in this marsupial, including one embryonic-expressed gene (S.c-epsilon) and one adult-expressed gene (S.c-beta), linked in the order 5'-epsilon-beta-3'. The entire DNA sequence of the adult beta-globin gene is reported and shown to be orthologous to the adult beta-globin genes of the North American marsupial Didelphis virginiana and eutherian mammals. These results, together with results from a phylogenetic analysis of mammalian beta-like globin genes, confirm the hypothesis that a two-gene cluster, containing an embryonic- and an adult-expressed beta-like globin gene, existed in the most recent common ancester of marsupials and eutherians. Northern analysis of total RNA isolated from embryos and neonatals indicates that a switch from embryonic to adult gene expression occurs at the time of birth, coinciding with the transfer of the marsupial from a uterus to a pouch environment.

The marsupial shell membrane: an ultrastructural and immunogold localization study

In the dasyurid marsupial, Sminthopsis crassicaudata, as the oocytes/embryos travel down the female reproductive tract two extracellular coats, the mucoid and shell membrane, come to surround them. Embryos recovered from the oviduct have a mucoid coat but no shell membrane which is only found surrounding uterine embryos. Initially, the shell membrane has a compact granular consistency but it later thins and becomes fibrous in texture with fibres oriented mainly in the plane of the membrane. Immunogold labelling with polyclonal antibodies raised against the extracellular coats was employed to determine the location and ultrastructural appearance of the secretory granules which contain mucoid and shell membrane precursors. Secretory granules in the luminal epithelium of the ampulla of the oviduct are of irregular electron density, while those in the isthmus are electron-dense and homogeneous. Both types give rise to the mucoid coat. Secretory granules in the epithelia of the utero-tubal junction and some endometrial glands are electron-lucent and contain some flocculent material which, after exocytosis, gives rise to the shell membrane.

Variation in ultrastructure of mucoid coat and shell membrane secretion of a dasyurid marsupial

In the dasyurid marsupial Sminthopsis crassicaudata, the shell membrane of cleaving embryos has a compact granular structure but becomes fibrous around blastocysts. Polyclonal antibodies were raised against the extracellular coats, mucoid and shell membrane, of oocytes and early embryos. Immunogold cytochemistry resulted in labelling of secretory granules in the epithelia of both the ampulla and isthmus of the oviduct, although the secretory granules of these two regions differed in their ultrastructural appearance. Those in the ampulla were heterogeneous with areas of varying electron density, whereas those in the isthmus were electron dense and homogeneous. Shell membrane precursors were found in secretory granules in the epithelia of the uterotubal junction and endometrial glands and were electron lucent.

Nutrient uptake and culture of Sminthopsis macroura (stripe-faced dunnart) embryos

Glucose and pyruvate uptake by individual embryos were measured in a marsupial species (stripe-faced dunnart) and a eutherian species (mouse). At each stage of development, nutrient uptake by the dunnart embryo was around an order of magnitude greater than that of the mouse embryo. The pattern of glucose uptake by the dunnart embryo was not like that for any eutherian embryo, all of which have a low glucose uptake before the blastocyst stage. Rather, in the dunnart embryo there was a significant increase in glucose uptake after the third cleavage division, increasing from 13.6 pmol embryo h-1 at the 4-cell stage to 34.9 pmol embryo h-1 by the 8-cell stage. This increase in glucose uptake before blastocyst formation may be attributed to an increased energy demand associated with the movement of cells within the dunnart embryo. Using a new culture system, it was possible to culture 66% of dunnart embryos at the 2-4-cell stage and 80% of those at the 8-16-cell stage to the unilaminar blastocyst stage. Embryos cultured from the 2-cell to the 4-cell stage were retarded by around 12 h when they reached the blastocyst stage. Developmental retardation was also reflected in the pattern of nutrient uptake, which lagged behind that of embryos developed in vivo. The present study has shown that it is possible to culture the early marsupial embryo to the blastocyst stage in a serum-free culture system, while concomitantly quantifying embryonic nutrient requirements. Such an approach is essential for species where there is a paucity of material for study.

The blastocyst epithelium is not a protoderm in dasyurid marsupials: a review of the evidence

Evidence from studies of cleavage and blastocysts in dasyurids in reviewed to show that the unilaminar blastocyst is not a protoderm but consists of two cell types. Cleavage and blastocyst formation in marsupials has been most comprehensively studied in dasyurids, in which the secondary oocyte and zygote are polarized with respect to the position of the nucleus, cytoskeletal elements and cytoplasmic vesicles. Polarity is reinforced by fertilization. Early cleavage divisions are associated with the polarized elimination of a yolk mass and many vesicular structures into the perivitelline space. Because secretion of the vesicular structures, of which several types are found, facilitate blastomere-zona then blastomere-blastomere associations during cleavage, a unilaminar blastocyst forms without an intervening morula stage. Polarization of the cleavage cavity is related to the appearance of two cell lineages, pluriblast and trophoblast, at the 16-cell stage. In species in which the yolk mass persists, the tier of eight cells lying nearest the yolk mass forms the pluriblast (future embryonic and extra-embryonic lineages), and the other tier forms the trophoblast (future extra-embryonic ectoderm of the yolk sac). Thus, the unilaminar blastocyst epithelium is not a protoderm. Blastocyst expansion is associated with increased cell numbers mainly in the trophoblast. Pluriblast cells begin to increase just before the appearance of the hypoblast. The two cell populations can be distinguished by ultrastructural and histological features at the end of cleavage. During blastocyst expansion they differ in appearance and behaviour in vitro and in cell-doubling time. Two populations of cells have also been found in cleavage and unilaminar blastocyst stages in other marsupials, such as opossums, brushtail possums and bandicoots.

Fate-map analysis of the epiblast of the dasyurid marsupial Sminthopsis macroura (Gould)

Allocation of cells in the marsupial epiblast to embryonic and extra-embryonic domains has to date been studied only histologically. An unresolved issue in marsupial embryology has been the existence of a medullary plate. We re-examined the hypotheses that the medullary plate, or neurectoderm, arises before notochord formation and that the integumentary ectoderm is segregated from the ectoderm after the formation of the medullary plate. By marking epiblast cells in 65 Day-8 embryos of the dasyurid marsupial Sminthopsis macroura, with the lipophilic cell-surface marker, DiI, we demonstrated that the so-called medullary plate contains progenitors of the integumentary ectoderm of the neck, chest, limbs and flank of the embryo. Thus, cell-allocation processes in the epiblast must have segregated the entire complement of embryonic ectoderm in one event, not separate events. It is concluded that the embryonic structure called 'medullary plate' in marsupial embryos is misnamed since, in fact, it consists of both integumentary ectoderm and neurectoderm.

The type and differentiation of cells in vitro from unilaminar and bilaminar blastocysts of two marsupials, Antechinus stuartii and Sminthopsis macroura

The type and ability to differentiate in vitro of cells found in blastocysts of two marsupials were examined. Thirteen unilaminar blastocysts on Day 7 to Day 12 of gestation and 24 bilaminar blastocysts on Day 16 and Day 18 of gestation were collected from 11 brown antechinus, Antechinus stuartii. A total of of 77 unilaminar blastocysts on Day 5 and Day 6 of gestation and a total of 61 bilaminar blastocysts on Day 6 and Day 7 of gestation were collected from 40 stripe-faced dunnarts, Sminthopsis macroura. Pluriblast and trophoblast cells, confined to separate hemispheres, were found in unilaminar and bilaminar blastocysts, establishing that the blastocyst epithelium was not a protoderm. Hypoblast cells were found only in bilaminar blastocysts. Pluriblast and hypoblast cells did not differentiate or proliferate in up to eight weeks in culture. A small number of trophoblast cells transformed to a multinucleate state but the remainder did not proliferate or differentiate further. The presence of murine leukaemia inhibitory factor or medium conditioned by exposure to marsupial fibroblast feeder layer was not required for the maintenance of an undifferentiated state. Differentiation, proliferation and attachment of cells were not influenced by the presence of the yolk mass or the egg coats in culture. The time taken for attachment of dissociated cells varied significantly between cultures with no substrate and with collagen, fibronectin or laminin (P = 0.001, ANOVA) but did not vary significantly between substrates. The substrates did not influence the state of differentiation of the cells.

Changes in structure of the trophectoderm of a marsupial in Mid-pregnancy up to the time of implantation

Pre- and peri-implantation embryos of the dasyurid marsupial Sminthopsis crassicaudata were examined for morphological differentiation of the trophectoderm. The cells of unilaminar blastocysts were all squamous and stained intensely with toluidine blue. In bilaminar blastocysts and embryos at the early embryonic-disc stage, the trophectoderm was similar in appearance to, but stained more lightly than, the underlying endoderm. Trophoblast differentiation did not appear to occur until the mesoderm had begun to migrate between the trophoblast and endoderm beyond the embryonic disc. At this stage, trophoblasts had three distinct morphologies: (1) vacuolated, tall and columnar cells in the trilaminar region; (2) large cuboidal cells in the adjacent bilaminar region; and (3) squamous cells in the abembryonic pole of the bilaminar region. These variations in cell structure correlate with differences in subsequent functional activity in these three regions of the yolk sac placenta.

The marsupial male: a role model for sexual development

Sexual differentiation in male marsupials has many similarities with that of eutherians. Marsupials have an XX-XY sex determining mechanism, and have a homologue of the testis-determining SRY gene on their Y-chromosome. However, the development pattern of SRY gene expression is different from the mouse in that it is expressed for a much longer period. SRY is expressed in a range of non-gonadal tissues in male pouch young and adults which is similar to the human pattern, and raises questions as to its particular role(s) in sexual differentiation. Similarly Müllerian inhibiting substance (MIS) is produced in the developing testis over a longer period than in the mouse. Since ovaries cultured with MIS or transplanted into male recipient pouch young develop tubular structures, MIS may induce Sertoli cell formation. Testosterone is produced by the neonatal testis, and this stimulates Wolffian duct development to form the vas deferens and epididymis. Virilization of urogenital sinus is also androgen-dependent. However, virilization of the prostate and phallus occurs more than three weeks after the onset of testosterone production, suggesting that the timing of this may be regulated by delayed activation of the androgen receptor pathway. Unlike in eutherians, differentiation of the scrotum and mammary glands is not dependent on testicular hormones, but is independently regulated by an X-linked genetic mechanism. Clearly marsupials provide a unique perspective to help us clarify the mechanisms underlying sexual development in all mammals.

Widespread expression of the testis-determining gene SRY in a marsupial

There is compelling evidence from mutation analysis and transgenesis that the SRY gene isolated from human and mouse encodes the testis-determining factor on the mammalian Y chromosome. However, how SRY achieves this function is unclear. Although marsupials have been separated from eutherian mammals for approximately 100 million years, homologues of SRY have been localised to the Y chromosome of two unrelated marsupial species, the tammar wallaby and the Darling Downs dunnart. Gonadal development is fundamentally similar in eutherian and marsupial mammals, but the timing of morphological events is different. Fetal Sry transcripts are confined to somatic cells of the male mouse genital ridge between 10.5-12.5 days post coitum, corresponding with the onset of testis differentiation. Analysis of Sry gene expression in the genital ridge of normal and germ cell-deficient fetal mice has established that this gene acts in the somatic cell lineage, and is presumed to induce the formation of Sertoli cells. This assumption can be tested more critically in the tammar, where the equivalent stages of testis differentiation are observed over a 7-day period. We have examined the relationship of SRY expression to testis differentiation in the tammar wallaby. We show the marsupial SRY gene cannot be exclusively coupled to Sertoli cell differentiation, as this gene is expressed in the male fetus from several days before genital ridge formation until 40 days after birth. SRY transcripts are also present in a variety of extra-gonadal tissues in the developing young and adult male, a pattern of SRY expression similar to that observed in humans. These data indicate that, in addition to a role in testis determination, SRY may have other functions [corrected].

Central cardiovascular shunts in the perinatal marsupial

BACKGROUND

Marsupials are born at an early stage of development after a short period of gestation. In this study the nature and timing of closure of the central cardiovascular shunts was investigated.

METHODS

Light and scanning electron microscopy were used to determine changes in central cardiovascular shunts in eight marsupial species with gestation periods of between 12.5 and 36.5 days and birth weights ranging from 12.5 mg to 740 mg. Laboratory mice with a birth weight of about 1,000 mg and a gestation period of 21 days were included for comparison.

RESULTS

Marsupials have a ductus arteriosus and an interatrial communication. The former closes rapidly after birth in the marsupial; however the interatrial communication is in the form of a fenestrated septum, which closes as a result of tissue proliferation over a period of days after birth. An additional central shunt, an interventricular foramen, was found to persist in three species for a short time after birth. In one species, the eastern native cat, Dasyurus viverrinus, which has a gestation period of about 19 days and low birth weight of about 12.5 mg, in addition to the two common shunts there was a large interventricular communication and septation of the outflow tract was incomplete.

CONCLUSION

In adapting from intra-uterine life, it seems that marsupials have adopted different, but equally effective strategies, with regard to the circulatory system.

Developmental determinants at the mammalian optic chiasm

The mammalian optic chiasm is widely and properly regarded as a region where axons from the temporal retina take an uncrossed course and separate from axons arising in the nasal retina that take a crossed course. However, this is but a rough approximation of the adult situation, and developmental studies must take account of several distinctive stages and axon rearrangements that characterize the region of the chiasm. At the early and late stages of development of nonprimate species the axons do not segregate in accordance with a strict naso-temporal rule at all, and their behavior at the chiasm is not relevant to the formation of the naso-temporal division. As the axons pass from the eye to the chiasm they tend to lose their retinotopic order, to gain a chronotopic order, and then, in the region of the chiasm, to regain some aspects of the retinotopic order before reaching their terminal sites. Molecular or cellular cues that allow the several distinct organizational steps to occur must be expected in the retina, on the axons themselves, and also along the pathway of the axons, prechiasmatically and at the chiasm. Some of these cues will be associated with local nerve cells, some with specialized glial elements and some with the retinofugal axons themselves. Several candidate molecules have been identified in the retina and along the path of the axons, but to date no clearly defined role in the specific events of the pathway determination have been identified. The sequence of developmental processes that characterizes the formation of the optic chiasm provides an interesting and useful challenge to experimentalists, because the advancing axons can now be observed in vitro and in the living brain. The pattern of growth changes as development proceeds, it shows distinctive properties in different species and in their genetic mutants, and it can be readily modified by simple experimental procedures. These all provide opportunities for investigating the function of proposed molecular cues that act in the development of the chiasm.

Cytoskeletal organization in the oocyte, zygote, and early cleaving embryo of the stripe-faced dunnart (Sminthopsis macroura)

Ovulation occurs in Sminthopsis macroura approximately 160 hr after administration of 1.3 IU PMSG, and yields significantly more oocytes than does spontaneous ovulation (P = 0.001). Germinal vesicle (GV)-stage oocytes have a thin cortical rim of microfilaments, which is disrupted by exposure to cytochalasin D. After GV breakdown, the first meiotic spindle forms subcortically and parallel to the oolemma. It rotates during anaphase and telophase to extrude the first polar body. This rotation is associated with a local cortical concentration of microfilaments, which is extruded in the first polar body. The second meiotic spindle is orthogonal to the surface, and extrusion of the second polar body is not associated with obvious local changes in cortical actin, resulting in a polar body containing little polymerized actin. The sites of second polar body emission and sperm entry are always in the half of the oocyte opposite the concentrating yolk mass, and are within 60 degrees of each other in most oocytes. During the concentration and eccentric movement of the yolk, microfilaments condense around it. During yolk expulsion, these microfilaments become continuous with those located subcortically. During early cleavage, the cytocortex of the zygote, but not of the extruded yolk mass, stains heavily for polymerised actin. Multiple sites of pericentriolar material are detectable in the cytoplasm of some secondary unfertilized oocytes which, in the presence of taxol, generate large cytasters and pseudospindle structures. After fertilization, a large aster is formed in association with the sperm entry point and serves as the center of an extensive cytoplasmic network of microtubules which surrounds but does not enter the yolk mass. Taxol treatment generates small cytasters within this meshwork and promotes selective stabilization of some periyolk microtubules opposite to the sperm aster.

Monotreme and marsupial reproduction

Marsupials were regarded as curiosities by their early European discoverers, animals to be wondered at. Monotremes were even more surprising; the platypus was such an amalgam of characters that it was thought to be a hoax. They were recognized very early as mammals that could make a major contribution to our understanding of reproductive processes, and work on marsupials at the turn of the century was much in evidence. It is, however, only in the past two decades, and especially in the past few years that marsupial research has regained this position. There is no doubt that future research will strengthen this contribution, but we are faced with serious conservation questions that must be solved if we are to maintain these wonderful animals as a resource for future generations.

Development of craniofacial musculature in Monodelphis domestica (Marsupialia, Didelphidae)

Development of craniofacial muscles of Monodelphis domestica (Marsupialia, Didelphidae) is described. In a period of 4-6 days all craniofacial muscles in M. domestica progress from myoblast condensation, to striated myofibers that are aligned in the direction of adult muscles and possess multiple, lateral nuclei. This process begins 1 to 2 days before birth and continues during the first few days after birth. Compared to other aspects of cranial development, muscle development in M. domestica is rapid. This rapid and more or less simultaneous emergence of craniofacial muscles differs from the previously described pattern of development of the cranial skeleton in marsupials, which displays a mosaic of acceleration and deceleration of regions and individual elements. Unlike the skeletal system, craniofacial muscles show no evidence of regional specialization during development. M. domestica resembles eutherian mammals in the relatively rapid and more or less simultaneous differentiation of all craniofacial muscles. It differs from eutherian taxa in that most stages of myogenesis occur postnatally, following the onset of function. The timing of the development of muscular and skeletal structures is compared and it is concluded that the relatively early development of muscle is not reflected by any particular acceleration of the differentiation or growth of skeletal structures. Finally, the difficulties in accounting for complex internal arrangements of muscles such as the tongue, given current models of myogenesis are summarized.

Embryonic-maternal cell interactions at implantation in the fat-tailed dunnart, a dasyurid marsupial

BACKGROUND

In marsupials implantation occurs about two-thirds the way through the short gestation before which time the embryo is surrounded by the permeable shell membrane which prevents physical contact between the trophoblast and uterine epithelium. Although the trophoblast has been shown to be invasive to varying degrees in several species of marsupials, the ultrastructure of the embryonic-uterine cell interactions at the time of implantation has not been described in this group.

METHODS

Thick plastic sections and transmission electron microscopy were employed to investigate the cellular interactions at implantation in the fat-tailed dunnart (Sminthopsis crassicaudata), a dasyurid Australian marsupial.

RESULTS

Our results show that epithelial penetration begins when the embryo is at the late presomite/early somite stage. In the trilaminar region of the yolk sac (TYS), trophoblast cells adjacent to the embryo form desmosomes with uterine epithelial cells and also appear to fuse with them to form hybrid cells, the cytoplasm of which resembles that of trophoblast. Later in the TYS, as the placenta develops, trophoblast microvilli and larger cell processes invaginate, and interdigitate with, the highly folded maternal epithelium but do not invade it. At this time in the bilaminar, or avascular, yolk sac (BYS), multinucleate trophoblast giant cells (TGCs) from an annular region adjacent to the sinus terminalis intrude between, and possibly fuse with, the maternal epithelium. The invading TGCs spread laterally above the residual basal lamina before migrating into the stroma.

CONCLUSIONS

In this species of marsupial at least, the cell interactions at the time of implantation are similar to those seen in some eutherian species despite the fact that the fetal chorion is of yolk sac rather than allantoic origin.

Microtubule configurations in oocytes, zygotes, and early embryos of a marsupial, Monodelphis domestica

The marsupials represent a separate evolutionary lineage from eutherians from which they diverged over 100 million years ago. In order to explore the origin and mode of centrosome inheritance amongst this group of mammals, this study investigates the microtubule organization during fertilization, parthenogenesis, and polyspermy in the didelphid, Monodelphis domestica. Microtubules and DNA were visualized in maturing ovarian oocytes, parthenogenetically activated oocytes, monospermic and polyspermic zygotes, and early embryos. Ovarian oocytes had a central region of yolky cytoplasm that, after fertilization, became polarized; much of the yolk was then extruded into the perivitelline space as an enucleated cytoplasmic mass. Immunofluorescence microscopy, using a monoclonal antibody to beta-tubulin, demonstrated microtubules in the meiotic spindle in unfertilized oocytes, but cytasters were not detected. After fertilization, a cluster of microtubules forming into a sperm aster was evident around the male pronucleus. The sperm aster remained largely restricted to the nonyolky region of the egg cytoplasm, resulting in a cytoplasmic heterogeneity between a microtubule-rich region and one in which microtubules were largely absent. Once the two pronuclei came close together, abundant microtubules were found surrounding both pronuclei. In the early embryo, microtubules were found in the outer cortical region of the blastomeres and, in addition, there was an extensive and elaborate network of microtubules throughout the yolk mass. Disruption of the meiotic spindle microtubules with nocodazole or cold treatment did not result in chromosome dispersion in the cortex and recovery from drug or cold depolymerization demonstrated that microtubules might not be as dynamic as those in eutherian mammals. Taxol stabilization resulted in an increase in cortical microtubules. In this marsupial species, therefore, the centrosome appears to be of paternal origin, and the radiating microtubules that form may well be involved both in bringing the pronuclei together and in the cytoplasmic polarization that results in extrusion of the yolk mass.

An SRY-related sequence on the marsupial X chromosome: implications for the evolution of the mammalian testis-determining gene

The SRY gene on the human, mouse, and marsupial Y chromosomes is the testis-determining gene that initiates male development in mammals. The SRY protein has a DNA-binding domain (high mobility group or HMG box) similar to those found in the high-mobility-group proteins. SRY is specific for the Y chromosome, but many autosomal genes have been identified that possess a similar HMG box region; those with the most closely SRY-related box regions form a gene family now referred to as SOX genes. We have identified a sequence on the marsupial X chromosome that shares homology with SRY. Sequence comparisons show near-identity with the mouse and human SOX3 gene (formerly called a3), the SOX gene which is the most closely related to SRY. We suggest here that the highly conserved X chromosome-linked SOX3 represents the ancestral SOX gene from which the sex-determining gene SRY was derived. In this model SOX3/SRY divergence and the acquisition of a testis-determining role by SRY might have preceded (and initiated) sex chromosome differentiation or, alternatively, might have been a consequence of X chromosome-Y chromosome differentiation initiated at the locus of an original sex-determining gene(s), later superseded by SRY.

Development of early cell lineages in marsupial embryos: an overview

All major embryonic and extra-embryonic cell lineages are established before implantation in marsupials. In reptiles, birds, monotremes and most marsupials, the zygote is polarized, sometimes markedly so, and the cleavage pattern in relation to the polarized state provides the mechanism for the generation of positional signals. These ensure that the embryonic cell lineages develop in the centre of the developing blastoderm or blastocyst epithelium and the extra-embryonic lineages at the periphery. The evolution of the vertebrate yolky egg was accompanied by a decreasing dependence on maternal determinants and increasing dependence on positional signals to determine cell fate. It is proposed that when a less yolky egg evolved, the mechanisms for determination of cell fate in a developing epithelium were retained. It is proposed that in marsupials, positional signals are involved in the determination of cell fate of embryonic and trophectoderm cells but do so in a two-dimensional epithelium not a three-dimensional morula. The next lineage formed is the primary endoderm which separates off from the primitive ectoderm in the embryoblast and eventually lines the blastocyst cavity. Positional signals are responsible for the determination of primary endoderm in eutherian mammals, birds and probably also marsupials. Order of cell division during cleavage provides a mechanism whereby some cells in the embryoblast of marsupials have earlier and greater contact with their neighbouring cells. The mechanism for determination of primary endoderm cells in the blastocyst epithelium is examined in the Virginia opossum and the stripe-faced dunnart.

Uterine histology of the dasyurid marsupial, Antechinus stuartii: relationship with differentiation of the embryo

Uterine samples from Antechinus stuartii on days 1, 4, 6, 8, 11, 13, 15, 18, 21 and 23 after ovulation were examined histologically. Animals were pregnant, nonpregnant and unmated, or nonpregnant and mated but found to have only unfertilized eggs on autopsy. The histological parameters used were thickness of the myometrium, endometrial stroma, and endometrial epithelium, and density of uterine stromal glands and of lymphocytes at the endometrial basal lamina. Overall, the fluctuation patterns of these parameters were superficially similar between pregnant and nonpregnant animals (mated or unmated). However, statistically significant differences were detected between pregnant and unmated nonpregnant animals in every parameter examined at nearly every time point except day 13. Comparison of these results with known data on embryonic stages, corpus luteum development and plasma progesterone concentrations revealed that the gravid uterus underwent histological changes co-incident with changes in both progesterone concentration and developmental delay or embryonic arrest. It was concluded that the uterus mediates progesterone-driven changes in embryonic developmental rate. Although determination of number of lymphocytes provided inconclusive evidence of cellular immunity against embryos, the possibility that embryonic signalling to the uterus occurs is discussed.

Monotreme development with particular reference to the extraembryonic membranes

This paper considers many of the salient features of monotreme development, particularly morphogenesis of the extraembryonic membranes. The uterine endometrium of both monotremes and marsupials exhibits a progesterone driven luteal phase where accelerated utilization of endometrial nutrients is evidenced by a rapid post-primitive streak expansion in the dimensions of the extraembryonic membranes. Monotremes share with marsupials, birds, and reptiles an unspecialized vertebrate mode of genesis of the embryonic disc on the peripheral surface of the yolk-sac. The fused vascularized respiratory chorioallantois is estimated to have a functional life of not more than the terminal 4 to 5 days of the monotreme incubation period. This time interval is of a slightly greater order of magnitude than that found in marsupials with a fused chorioallantois, so far described, but in the context of the proportional elapse of post-primitive streak organogenesis would fall within the marsupial grade. The dominant extraembryonic membrane for nutritive and respiratory function in both monotremes and marsupials is the yolk-sac. This contribution shows that monotremes and marsupials share a much larger suite of developmental anatomical features than previously reported. The evolutionary biologist is confronted with the challenge of how to assign an appropriate weighting to these features as marsupials are considered by many researchers to be allied phylogenetically more closely with eutherian mammals than with monotremes.

Gonadal sex differentiation in embryos and neonates of the marsupial, Monodelphis domestica: arrest of testis development in postterm embryos

Growth and histological differentiation were studied in 8 litters of embryos and 4 litters of neonate grey short-tailed opossums, Monodelphis domestica. The embryonic litters included 2 that had passed their expected birth date, and whose weights exceeded the usual birthweights; we refer to these litters as 'postmature'. There was an abrupt increase in the growth rate of XY gonads after birth, but this was not seen in XX gonads. Although there was evidence of testicular differentiation in XY gonads on the day before the expected birth, testicular differentiation was found to be blocked in postmature litters. The growth of XX gonads in postmature embryos was not affected. In view of evidence that exogenous oestrogens feminise the gonads of genetic males in some species of marsupials including Monodelphis domestica, the question arises whether oestrogen is responsible for the failure of testes to continue their development in utero. We suggest that the ability of functional testes to develop in the presence of oestrogen may be a fundamental requirement distinguishing eutherian mammals from other vertebrates, including marsupials.

Cranial osteogenesis in Monodelphis domestica (Didelphidae) and Macropus eugenii (Macropodidae)

The pattern of onset and general rate of cranial ossification are compared in two marsupials, Monodelphis domestica (Didelphidae) and Macropus eugenii (Macropodidae). In both species a similar suite of bones is present at birth, specifically those surrounding the oral cavity and the exoccipital, and in both postnatal events follow a similar course. The facial skeleton matures more rapidly than the neurocranium, which is characterized by an extended period of ossification. Most dermal bones begin ossification before most endochondral bones. Endochondral bones of the neurocranium are particularly extended in both the period of onset of ossification and the rate of ossification. These data confirm suggestions that morphology at birth is conservative in marsupials and we hypothesize that the pattern of cranial osteogenesis is related to two distinct demands. Bones that are accelerated in marsupials are correlated with a number of functional adaptations including head movements during migration, attachment to the teat, and suckling. However, the very slow osteogenesis of the neurocranium is probably correlated with the very extended period of neurogenesis. Marsupials appear to be derived relative to both monotreme and placental mammals in the precocious ossification of the bones surrounding the oral cavity, but share with monotremes an extended period of neurocranial osteogenesis.

Embryonic development in culture of the marsupials Antechinus stuartii (Macleay) and Sminthopsis macroura (Spencer) during preimplantation stages

Forty-nine blastocysts from 11 brown antechinus, Antechinus stuartii, and 96 blastocysts from 17 stripe-faced dunnarts, Sminthopsis macroura, were used to develop a culture system for embryos during preimplantation stages. Blastocysts of brown antechinus were collected on Days 6-9 for unilaminar stages, Days 16-21 for bilaminar stages and Days 20 and 21 for trilaminar stages. Blastocysts of stripe-faced dunnarts were collected on Day 6 for unilaminar stages, Days 6-8 for bilaminar stages and Day 8 for trilaminar stages. Culture media were Dulbecco's modified Eagle's medium (DMEM) with 4.5% glucose and Whittingham's T6 medium both of which were supplemented with 5, 10, 12.5 and 20% fetal calf serum (FCS). Antechinus serum (5%) and bovine serum albumin (0.1%, 0.2%) were also added to some media. Human amniotic fluid (HAF) and Monomed media were also tested. Blastocysts were cultured at 35 degrees C in 5% CO2 in air. DMEM + 10% FCS and HAF supported normal development for the longest periods and over the greatest range of stages. Developmental failure of blastocysts in vitro during expansion of the unilaminar blastocyst and formation of the bilaminar blastocyst suggests that these stages may be dependent on uterine signals. When cultured in DMEM + 10% FCS, the rate of development of bilaminar and trilaminar blastocysts into organogenesis was 4 h slower than in vivo in the stripe-faced dunnart and about 6 h slower than in vivo in the brown antechinus. Embryos of stripe-faced dunnarts were cultured to within 18 h of birth.

Developmental staging in a marsupial Dasyurus hallucatus

In a marsupial, Dasyurus hallucatus, pouch-young of various ages from newborn to 55 days were embedded in wax and serially sectioned. On the basis of the relative development of external and internal characteristics, they were placed in the Carnegie staging system developed by Streeter and elaborated by O'Rahilly and associates. Birth occurred at stage 15, and the end of embryogenesis (stage 23) was reached about day 33. Whereas stage 23 is characterised in all eutherians by the closure of the secondary palate, this occurs before stage 15 in D. hallucatus. Since most other characters of the newborn are at a stage 15 level of development, there has been a relative acceleration of development of the secondary palate (and forelimb) in D. hallucatus that allows it to suckle and breathe at the same time. Between D. hallucatus and eutherians, there is general agreement in the sequence of development and in the relative degree of most structures at each stage. Further marsupials should be examined to see if the minor differences noted are peculiar to D. hallucatus or apply to marsupials generally.