Bridget Murphy, School of Biological Sciences (A08), University of Sydney, Camperdown, NSW 2006, Australia. T: +612 9351 5118; F: +612 9351 4119; E:email@example.com
The eastern water skink (Eulamprus quoyii) has lecithotrophic embryos and was previously described as having a simple Type I chorioallantoic placenta. Indeed, it was the species upon which the definition of a Type I placenta was thought to be based, although we had cause to question that assumption. Hence we have described the morphology of the uterus of E. quoyii and found it to be more complex than previously supposed. The mesometrial pole of the uterus in E. quoyii displays a vessel-dense elliptical structure (the VDE) with columnar uterine epithelial cells. As pregnancy proceeds, the uterine epithelium near the mesometrial pole becomes folded and glands become hypertrophied, so that the morphology of VDE resembles that of a placentome, characteristic of Type III placentae. Unlike species with a Type III placenta, the apposing chorioallantoic membrane of E. quoyii is lined with squamous cells and interdigitates with the folded uterine epithelium. The remainder of the uterus is thin with a squamous uterine epithelium throughout pregnancy. Immunohistochemical localisation of blood vessels reveals a dense network of small capillaries directly beneath the folded epithelium of the VDE, while blood vessels are larger and sparser at the abembryonic pole of the uterus. Alkaline phosphatase (AP) activity is present in the uterine epithelium and sub-epithelial blood vessels in newly ovulated females. AP activity disappears from the epithelium between stages 27 and 29 of embryonic development and from the blood vessels after stage 34, but appears in the uterine glands at stage 35, where it remains until the end of pregnancy. Although the VDE is structurally similar to the placentomes found in other viviparous lizards, different distributions of AP activity in the uterus of E. quoyii and Pseudemoia spenceri suggest that the VDE may be functionally different from the placentome of the latter species. Our description of uterine morphology in E. quoyii provides evidence that, at least in some lineages, the evolution of a placentome may not occur in concert with the evolution of microlecithal eggs and obligate placentotrophy.
The morphology and function of the placenta in viviparous (live-bearing) lizards is extremely diverse (Weekes, 1935; Blackburn, 1993). Morphologically, lizard placentae are classified according to the structure of their chorioallantoic placenta (Weekes, 1935), formed by the apposition of the uterus and the chorioallantois of the embryo (Stewart and Blackburn 1988).
The majority of viviparous lizards have a relatively simple chorioallantoic placenta, in which there are no regional modifications of the chorioallantoic placenta and the epithelia of both the uterus and chorioallantois are squamous (Weekes, 1935; Blackburn, 1993). The chorioallantoic placentae of some scincid lizard species have more complex morphologies and regional modifications; some exhibit chorioallantoic placentae in which the uterine capillaries bulge into the uterine lumen (Weekes, 1930, 1935; Adams et al. 2007; Stewart & Thompson, 2009), while others comprise an elliptical region of the chorioallantoic placenta, known as the placentome, in which the uterine epithelium is thrown into folds. Each fold contains blood vessels and is covered by hypertrophied uterine epithelium; the apposing chorionic epithelium is also enlarged (Weekes, 1929, 1930, 1935; Stewart and Thompson 1996; Adams et al. 2005). The most complex chorioallantoic placenta described so far occurs in species from the scincid genus Mabuya, in which the chorionic epithelium interdigitates with the uterine epithelium of the placentome (Blackburn & Vitt, 2002; Jerez & Ramírez-Pinilla, 2003).
Functionally, lizard placentae range from those which facilitate the transfer of water, respiratory gases and some ions (calcium, potassium and sodium) (Thompson et al. 2000), to those which can also transport amino acids and lipids to the developing embryo (Thompson, 1981; Thompson et al. 2000). Evidence of histotrophic transport in several species with placentomes has been revealed by scanning and transmission electron microscopy (Corso et al. 2000; Adams et al. 2005; Biazik et al. 2009). Immunohistochemical studies have identified that nutrient transport via paracellular (Biazik et al. 2007, 2008) and transcellular pathways (Herbert et al. 2006) is regulated during pregnancy in skinks. Recently, alkaline phosphatase activity in the uterus of two skinks that have placentomes and are highly placentotrophic, Pseudemoia entrecasteauxii and Pseudemoia spenceri, was used to identify sites of transcellular nutrient transport (Biazik et al. 2009).
The classical model for the evolution of viviparity in lizards predicts that nutrient transport increases as the chorioallantoic placenta becomes more morphologically complex (Weekes, 1935; Blackburn, 2006). Previous evidence also suggests that morphological modifications of the uterus for nutrient transport occurs only in placentotrophic species (Blackburn et al. 2010), in which placental nutrition is crucial for embryonic development because ovulated eggs contain small amounts of yolk. However, placental morphology may not be the most reliable predictor of placental function as was first thought (Stewart & Thompson, 2009). Furthermore, modifications of the yolk sac placenta for histotrophic absorption and secretion in Sceloporus jarrovi, a lecithotrophic species, challenge the idea that modifications for nutrient transport are restricted to obligatory placentotrophic species (Blackburn et al. 2010).
The eastern water skink, Eulamprus quoyii, is a lecithotrophic species (Thompson, 1977a, 1981) and formed much of the basis of Weekes’ definition of a Type I chorioallantoic placenta [Lygosoma (Hinulia) quoyii; Weekes, 1927]. We recently identified a thickened, vessel-dense ellipse (VDE) in the mesometrial region of the uterus in E. quoyii (Murphy et al. 2010). The elliptical shape of the VDE bears a striking resemblance to Weekes’ illustration of a placentome in ‘Lygosoma (Liolepisma) weeksae’ (Pseudemoia spenceri; Stewart & Thompson, 1998), and preliminary histological examination of the VDE revealed shallow folds in the uterine epithelium, indicating a greater degree of chorioallantoic placental complexity than was previously described in this species. We have used light microscopy, immunohistochemistry and enzyme histochemistry to ask the following questions: (i) What is the morphology of the VDE in E. quoyii during pregnancy, and does it resemble the morphology of a placentome? (ii) What is the spatial arrangement of the dense network of blood vessels within the uterine stroma of the VDE of E. quoyii? (iii) What is the function of the VDE, based on morphological characters and enzyme histochemical analyses using an alkaline phosphatase assay? (iv) Are chorioallantoic placentomes always associated with obligatory placentotrophic embryos?
Materials and methods
Animal collection and husbandry
Pregnant E. quoyii (n = 29) were collected between October and December in 2008 and 2009 from locations in southern Sydney, Australia (Table 1). Lizards were housed individually at a constant temperature of 20 °C in 450 × 250 × 250 mm aquaria lined with newspaper. Lizards were provided with an incandescent light at one end of the enclosure so that females could thermoregulate. Flat rocks were provided for basking and shelter, and lizards were fed approximately five crickets dusted in calcium gluconate every 2 days. A large water dish supplied water ad libitum and a 12 : 12 photophase : scotophase was provided.
Table 1. The number of uterine samples examined at each stage of pregnancy in Eulamprus quoyii, using different staining techniques.
Number of specimens observed with each staining technique
Haematoxylin and eosin
Alkaline phosphatase activity
von Willebrand factor immunostain
Only one specimen per female was examined for each staining technique.
Less than 25
Tissue and embryo harvest
Pregnant lizards were killed at various times between October and December by an intrathoracic injection of 0.2 mL sodium bentabarbitone (6 mg mL−1). Lizards were dissected and each uterus containing embryos was removed and floated in a Petri dish of reptile Ringer’s solution (Frye 1973).
Each uterus was cut into separate uterine incubation chambers. One uterine chamber containing an intact embryo was fixed in 10% neutral buffered formalin for 24 h and then stored in 70% ethanol. Once the uterine tissue had been fixed while in its native shape and still stretched around the embryo, it was later removed from the embryo by cutting longitudinally along the abembryonic side of the uterus. No further dissection of the uterine tissue was performed before it was embedded in paraffin, and the tissue was sectioned (7 μm thick) transversely with regard to the long axis of the uterus. Serial sections were obtained from the entire length of each uterine incubation chamber and photographed sections indicate the morphology of the incubation chamber at its maximum width. Sections were stained with haematoxylin and eosin (Table 1) and mounted with Hydromount (GeneWorks, Adelaide, SA, Australia). Embryonic stages were determined using the staging scheme developed by Dufaure & Hubert (1961).
The uterine chamber was removed from around a second embryo by cutting longitudinally along the abembryonic side of the uterus. The chorioallantois was also removed intact from the embryo. Both the uterine chamber and chorioallantois were coated in Tissue-Tek OCT cryoprotectant, immersed in super-cooled isopentane and stored in liquid nitrogen. Uterine and chorioallantoic tissue was sectioned at 7 μm using a Leica CM3050 cryostat at −23 °C for immunohistochemistry and enzyme histochemistry.
Two sections were collected per slide, and two experimental and two negative control slides were cut per animal (Table 1). Sections were fixed in 4% paraformaldehyde for 10 min at room temperature and washed three times in phosphate-buffered saline (PBS). Slides were incubated for 30 min in 1% bovine serum albumin (BSA; Sigma-Aldrich, St. Louis, MO, USA) in PBS. Slides were then incubated in rabbit anti-human von Willebrand factor antibody (Dako, Carpenteria, CA, USA), 6.2 μg mL−1 in 1% BSA/PBS overnight at 4 °C. von Willebrand factor is a glycoprotein that is constitutively expressed in the endothelial cells of capillary walls (Sadler, 1998). After three rinses in PBS, slides were incubated in secondary Cy3-conjugated anti-rabbit IgG (0.3 μg mL−1) (Sigma-Aldrich) for 30 min at room temperature. The primary von Willebrand antibody was omitted from negative control slides and rabbit IgG was substituted for the primary antibody in non-immune control slides. After three rinses with PBS, slides were mounted with Vectashield with DAPI (Vector, Torrance, CA, USA).
Two sections were collected per slide, and two experimental and two negative control slides were cut per animal (Table 1). Alkaline phosphatase localisation was performed using a modified Gomori (1952) method (Kiernan, 2008). The incubation medium, containing 5 mL of 3% sodium β-glycerophosphate, 10 mL of 0.25% lead nitrate and 20 mL of 0.05 m barbital buffer (pH 9), was filtered and allowed to stand at 29 °C for 15 min. Slides were incubated in medium for 1 h at 29 °C, the mean selected body temperature for E. quoyii (Schwarzkopf, 1998). Sodium β-glycerophosphate was omitted from the incubation medium in negative control slides. After washing in barbital buffer, slides were incubated in filtered 1% ammonium sulphide for 10 s until a brown precipitate was visible and then washed again in barbital buffer. Sections were counterstained with toluidine blue and mounted with Hydromount (GeneWorks).
All slides were visualised using a Zeiss Axioplan 2 upright microscope. Images were captured using an AxioCAM HR digital monochrome CCD camera and axiovision image acquisition software (Carl Zeiss, Germany). Micrographs were imported into imagej (National Institute of Health, Bethesda, MD, USA) and scale bars were added.
The uterine epithelium consists of a single layer of mononucleated cells throughout pregnancy. In pregnant females with embryos of stage 28 or earlier, uterine epithelial cells vary in morphology from squamous to cuboidal along most of the uterine luminal surface (Fig. 1A). Close to the mesometrial pole of the uterus, uterine epithelial cells vary in morphology from cuboidal to columnar (Fig. 1B). Deep folds of the uterine epithelium occur in the inter-embryonic segments of the uterus, forming rectangular-shaped plateaus, but there are no folds present elsewhere in the uterine epithelium (Fig. 1C). Small, compressed uterine glands are visible in the underlying lamina propria (Fig. 1B,D). Round white circles in the lamina propria of the stage 28 specimen are not blood vessels but instead may indicate lipid inclusions inside adipose cells (Fig. 1D), as distinct cell borders and basal nuclei are visible around the white spaces at high magnification.
Between embryonic stages 32 and 34, the uterine epithelium near the mesometrial pole shows various degrees of folding. A stage 32 specimen exhibits folds in the uterine epithelium near, but not centered across, the mesometrial pole of the uterus (Fig. 2A). Uterine epithelial cells in each fold are columnar or cuboidal (Fig. 2B,C) but become squamous near the edges of the folded region (Fig. 2C). Uterine glands directly beneath the crest of each fold are hypertrophied (Fig. 2B). Uterine epithelial cells near the mesometrial pole in a stage 33 specimen are also columnar but do not exhibit folding, and secretory vesicles are visible in the epithelium in this region (Fig. 2D). The columnar epithelial cells near the mesometrial pole in a stage 34 sample overlay uterine glands that are larger than those in the uterus of newly ovulated females. The epithelium is rippled but does not exhibit the epithelial folding seen in the stage 32 specimen examined (Fig. 2E). Uterine epithelial cells in the lateral and abembryonic regions of the uterus between stages 32 and 34 vary in morphology from squamous to cuboidal and do not exhibit any rippling or folding (Fig. 2F).
During the remainder of pregnancy, the uterine glands continue to hypertrophy until the glandular lumina are clearly distinguishable and uterine epithelial cells on the crest of each fold remain columnar (Fig. 3A,B). The folds in the uterine epithelium are lobe-like in one stage 37 specimen (Fig. 3C,D). In places, the lobes are separated from each other by squamous epithelial cells that are highly attenuated over an underlying blood vessel (Fig. 3C,D). A dense network of small capillaries lies directly beneath the columnar epithelial cells (Fig. 3E). The ectodermal layer of the chorioallantoic membrane has a squamous morphology (Fig. 3F) and interdigitates with the folds in the opposing uterine epithelium (not shown). Glands are also hypertrophied in the abembryonic region of the uterus, but the uterine epithelium remains squamous.
Uterine blood vessels are situated directly beneath the uterine epithelium throughout pregnancy (Fig. 4A–E). Many small blood vessels are present in the region of the uterus near the mesometrial pole, forming an almost continuous line of fluorescent staining beneath the uterine epithelium (Fig. 4A,C). Fewer, larger blood vessels are present in the abembryonic region of the uterus, resulting in more discrete patches of staining beneath the uterine epithelium (Fig. 4B,D). The negative control shows no staining (Fig. 4E) and the non-immune control shows a light, non-specific stain across the entire uterus (Fig. 4F).
Distribution of uterine alkaline phosphatase activity during pregnancy
The distribution of alkaline phosphatase activity in the uterus changes during pregnancy (Fig. 5). At any given stage of embryonic development, the distribution of alkaline phosphatase activity is the same in the abembryonic, lateral and mesometrial regions of the uterus. In five of the six specimens examined from newly ovulated females, as well as in a stage 26, stage 28 and two stage 29 specimens, alkaline phosphatase activity is present in the uterine epithelium and in the sub-epithelial blood vessels (Fig. 6A,B). In one sample, staining for alkaline phosphatase activity is visible on the cilia of some epithelial cells (Fig. 6C). In one newly ovulated specimen, two stage 28 specimens, and all four specimens examined between stages 30 and 34, alkaline phosphatase activity was present in the sub-epithelial blood vessels only and not in epithelial cells (Fig. 6D). Alkaline phosphatase activity is present in both the uterine glands and blood vessels of a stage 35 specimen (Fig. 6E), and all four specimens examined between stages 37 and 40 display alkaline phosphatase activity in the uterine glands only (Fig. 6F,G). Negative control slides showed no brown staining (Fig. 6H).
Eulamprus quoyii displays specializations in its chorioallantoic placenta that are incongruous with Weekes’ original description of a Type I placental morphology of this species (Weekes, 1927). In Weekes’s description, uterine epithelial cells in the chorioallantoic placenta were squamous, maternal capillaries were close to the epithelial surface, and there were no villous folds of the uterus. Our histological examination revealed that the recently described VDE in this species (Murphy et al. 2010) is morphologically different from the rest of the chorioallantoic placenta. In females with newly ovulated eggs, the VDE is characterized by a short stretch of unfolded columnar uterine epithelium near the mesometrial pole of the uterus. This region expands and hypertrophied glands create folds in the epithelium later in pregnancy. Uterine epithelial cells do not exhibit uniform morphology in the folded region; uterine epithelium is tall and columnar on the crest of folds but is cuboidal or squamous in the trough of each fold. The chorioallantois exhibits a squamous ectoderm that interdigitates with the folds in the VDE.
Numerous small blood vessels are located directly beneath the folded columnar epithelium of the VDE in the mesometrial region of the uterus. Blood vessels in the abembryonic uterine region are larger but less numerous than in the mesometrial region. These results confirm our previous observations of uterine vasculature in the VDE of E. quoyii using immunofluorescent confocal microscopy (Murphy et al. 2010). We had previously proposed that stimuli of embryonic origins may be responsible for the position and expansion of the VDE. We extend this hypothesis and suggest that stimuli from the growing embryo may not only cause regional differentiation in the vasculature of the uterus, but also contribute to the hypertrophy and folding of the uterine epithelium during pregnancy.
It is likely that the ‘Lygosoma (Hinulia) quoyii’ that Weekes collected for the original placental description comprised a combination of currently recognized species of Eulamprus. Weekes’ description was based on specimens collected from relatively high elevation environments (Barrington Tops, Mount Kosciusko and the Blue Mountains in NSW) and low elevation coastal environments (Sydney and Kiama in NSW). She referred to specimens collected at higher elevations as the ‘mountain type’ and specimens collected at lower elevations as the ‘coastal type’. Weekes observed that the ‘mountain type’ was much smaller than the ‘coastal type’, had smaller litter sizes and smaller neonates, but all specimens were identified as ‘L. quoyi’. Several inconsistencies in Weekes’s comprehensive descriptions are probably because she was describing placentation of at least two and potentially up to five species; according to current nomenclature and species distributions, it is likely that the ‘coastal type’ is E. quoyii and perhaps some Eulamprus heatwole while the ‘mountain type’ could be a combination E. heatwole, Eulamprus kosciuskoi, Eulamprus leuraensis, E. quoyii and Eulamprus tympanum (Swan et al. 2004).
However, Weekes described an area near the mesometrial pole of the uterus of one specimen that exhibited ‘slight folding of the maternal and foetal tissues’ and ‘the foetal tissue is seen fitting loosely into the shallow maternal crypts’ (Weekes, 1927). In this specimen ‘(uterine epithelial) cells (were) not as uniform in shape as in the non-pregnant condition, being crowded together in some places until they (were) cone-shaped’, and Weekes’ noted that ‘the (epithelial) cells appear to be secretory, being non-vacuolated, and somewhat resemble those lining the villous folds of L. (Pseudemoia) entrecasteauxii’. We suspect that what Weekes referred to as a ‘freak in placental development’ may have been the only specimen of Eulamprus quoyii sensu stricto that she described, as her observations of this single embryo match very closely with our own. However, two elements of her description of this specimen do not match with our observations; we observed that the chorionic cells opposite the uterine folds were squamous, not cuboidal as in her illustration (Weekes, 1927). Secondly, Weekes notes that the folded region is not as well vascularised as the non-folded region. Our observations, from both whole-mount specimens and sectioned material, indicate that the folded region is highly vascular, but blood vessels are so small that they are often difficult to locate from histological preparations alone.
Our observations indicate that E. quoyii possesses a chorioallantoic placenta that is more complex than a Type I placenta. But do the characteristics of the VDE match the definition of a placentome in Type III placentae? The VDE in E. quoyii fulfills three of the four requisites in Weekes’ definition of a placentome (Weekes, 1935); The VDE (i) is opaque and elliptical in shape; (ii) comprises a folded uterine epithelium; and (iii) exhibits enlarged uterine epithelial cells. However, chorionic cells opposite the VDE are not enlarged but are instead squamous. A similar morphology in the chorioallantoic placenta occurs in Chalcides ocellatus tiligugu (Corso et al. 2000), which, like E. quoyii, is also a lecithotrophic skink with a chorioallantoic placenta that was previously described as Type I (Giacomini, 1906; Weekes, 1935). We suggest that the placental morphology of E. quoyii and C. ocellatus tiligugu are forms of complex chorioallantoic placentae that share many morphological features with the placentomes of C. chalcides and Pseudemoia spp.
We described the distribution of alkaline phosphatase activity in the uterus of E. quoyii to test whether the VDE might have a nutritive function as would be predicted by its morphology, according to the classical model for the evolution of viviparity (Weekes, 1935). Alkaline phosphatase is a membrane-bound glycoprotein involved in the hydrolysis of extracellular phosphate esters (McComb et al. 1979) and exhibits strong activity in the epithelium of tissues that engage in active transport, such as the small intestine and the kidney (Danielli, 1954; Harris, 1989). In non-pregnant and vitellogenic P. spenceri (Type III chorioallantoic placenta), AP activity is present only in the uterine glands (Biazik et al. 2009). During pregnancy, AP activity is localized apically in the uterine epithelium of the omphaloplacental region. During the final stages of pregnancy (stages 39 and 40), some activity also localized to the tips of the villous folds in the placentome (Biazik et al. 2009), suggesting that the placentome is involved in the transcellular transport of nutrients to the embryo during late pregnancy.
In contrast, AP activity is absent from the villous folds of the VDE in E. quoyii during late pregnancy and there are no consistent differences in the intensity or distribution of AP activity between the mesometrial, lateral and abembryonic regions of the uterus. This suggests that the VDE is functionally different from the placentome of P. spenceri, and that the VDE has the same capacity for transcellular transport as other regions of the uterus. High AP activity associated with the sub-epithelial blood vessels suggests that transcellular exchange of substances between capillaries and uterine epithelium, and presumably transport by other means from the uterine epithelium into the uterine lumen, occurs until stage 34 of embryonic development. AP activity in the blood vessels after stage 34 is no longer visible (Fig. 6). From stage 35 until the end of pregnancy, the now hypertrophied uterine glands display strong AP activity (Fig. 6), which suggests that uterine transport changes at about stage 35 from one mediated by transcellular transport from the blood vessels to another form that comprises secretions from uterine glands. The functional significance of this change in AP must be elucidated using molecular markers that indicate transcellular transport and electron microscopy to look for evidence of secretory activity.
Alkaline phosphatase activity is lost from the uterine epithelium between stages 27 and 29 of embryonic development (Fig. 6). It is unlikely that AP activity so early in pregnancy is indicative of a nutritive role for uterine epithelium, as the eggshell membrane is still intact for most of this period (B. F. Murphy, personal observation). It is more likely that the loss of AP activity in the uterine epithelium of E. quoyii is an indicator of the period during which the eggshell membrane disintegrates, allowing the embryo to attach to the uterine wall. Similar changes in the AP activity of uterine epithelium occur during blastocyst attachment in rabbits (Classen-Linke et al. 1987), rats (Pritchard, 1947; Bansode et al. 1998; Bucci & Murphy, 2001), cows (Leiser & Wille, 1975) and sheep (Boshier, 1969). The apical plasma membrane of uterine epithelial cells undergoes major morphological and biochemical changes during the ‘plasma membrane transformation’ that occurs during blastocyst implantation in both mammals and reptiles (Murphy et al. 2000). We suggest that the loss of AP activity in the uterine epithelium of E. quoyii between stages 27 and 29 is an indicator of the plasma membrane transformation in this species.
Embryos of E. quoyii are lecithotrophic; embryos can be excised from the uterus and cultured in vitro to full term with no organic supplement (Thompson, 1977a). The ratio of neonatal dry mass to egg dry mass, often used to quantify the relative degrees of placentotrophy in squamate reptiles (Thompson et al. 2000), indicates that there is no net uptake of dry mass by E. quoyii embryos (Thompson, 1981), or by embryos of a closely related species, E. tympanum (Thompson et al. 2001) during gestation. However, the VDE of E. quoyii has striking structural similarities to the uterine contribution of a placentome, which supplies copious amounts of organic nutrients to the embryo in obligately placentotrophic species. So what then is the function of the VDE in E. quoyii? Radiolabelled amino acids are able to cross the placenta of E. quoyii (Thompson, 1977b), which suggests that this species is facultatively placentotrophic (Stewart, 1989), and the VDE may be involved in transporting these nutrients across the placenta. Facultative placentotrophy may be adaptive in unpredictable environments because it allows a degree of flexibility in the amount of nutrients provided across the placenta by the female (Stewart, 1989; Stewart & Thompson, 1993; Thompson et al. 1999a,b). However, even though the VDE is structurally similar to the uterine contribution to the placentome of Pseudemoia spp., the distribution of AP activity in the uterus of E. quoyii and P. spenceri is different, indicating that if the VDE does have a nutritive function, the mechanisms of nutrient transport in the VDE may be different from those in the uterine component of a placentome.
Until recently, all species with a placentome were also obligate placentotrophs (Thompson et al. 2000). Eulamprus quoyii and C. ocellatus tiligugu are both lecithotrophic species, yet they display a chorioallantoic placenta that resembles a placentome. This study provides evidence consistent with that of recent studies (Stewart & Thompson, 2009; Blackburn et al. 2010) contradicting the long-held idea that placental morphology is a reliable predictor of placental function. Furthermore, our results indicate that the evolution of a placentome is not necessarily accompanied by the evolution of obligate placentotrophy and microlecithal eggs. In addition to their role in some lineages in supplying nutrients to obligate placentotrophic embryos, we suggest that placentomes can also be involved in facultative placental transport, which supplements rather than substitutes nutrients deposited in the ovulated egg of viviparous lizards.
This research was funded by an Australian Research Council Discovery grant to M.B.T. and C.R.M. Animals were collected with New South Wales Scientific Collecting number S10693 and research was approved by the University of Sydney Animal Ethics Committtee (number L04/9-2008/3/4883). We thank J. Herbert for assistance in dissections, and S. Clayman, K. Murphy, M. Murphy, D. Nelson and N. Pezaro for assistance in the field.