SEARCH

SEARCH BY CITATION

Keywords:

  • claudin-5;
  • skinks;
  • tight junction;
  • uterine epithelium;
  • viviparity

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

Claudin-5, a tight junctional protein associated with ion and size selectivity, has been found in the uterus of skinks. This study has generated critical information about the molecular assembly of the tight junction at various stages of the reproductive cycle in the skink uterus. Recent studies looking at tight junctional proteins found occludin expression in the tight junction region of uterine epithelial cells in the skink uterus; however, occludin did not disclose any further information about the ions and size of ions permeating across the paracellular pathway. A ∼22-kDa claudin-5 band was detected in the uterus of the skinks present in this study and immunohistochemistry revealed that claudin-5 redistributes to the tight junction region of the lateral plasma membrane of uterine epithelial cells in late stage pregnancy/gravidity. This finding indicates that the tight junction becomes more assembled to precisely regulate ion and solute permeation in late stage pregnancy/gravidity. Claudin-5 with its functional role as a molecular sieve due to the formation of ion and size selective pores suggests that permeation of ions smaller than 0.8 kDa are restricted when claudin-5 is redistributed to the tight junction region of the later plasma membrane. This report is the first description of the molecular mechanisms that may be involved in regulating nutrient provision in the reptilian uterus. Anat Rec, 291:547–556, 2008. © 2008 Wiley-Liss, Inc.

Viviparity, or the birth of live young, occurs in many animal groups (Kaye,1971) and has evolved over 120 independent times in vertebrates. Over 100 of these origins have occurred in the Order Squamata (lizards and snakes) and viviparity is particularly common in the family Scincidae (Blackburn,2006), including the Eugongylus and Sphenomorphus group. In addition to a reduction in eggshell thickness and prolonged egg retention (Blackburn,1982; Shine,1983), viviparity involves the transfer of water, oxygen, and in many cases, nutrients (Guillette,1993). Thus, the uterus ultimately transforms from a passive environment to a nourishing chamber, in which the developing embryo relies on placentotrophy (Blackburn,1993; Stewart and Thompson,1993). Skinks are an ideal model to study the physiological and morphological changes taking place in the uterine epithelium and, thus, the evolution of placentation and viviparity, because some species exhibit oviparity (egg laying), whereas others are viviparous (live bearing) with chorioallantoic placentae ranging from simple (Weekes,1935) to complex (Thompson et al.,2002).

Viviparous skinks and some mammals (e.g., ungulates) have an epitheliochorial placenta (Luckett,1977). Epitheliochorial placentation involves close apposition of fetal and maternal tissue, but there is no breaching of the uterine epithelial layer (Grosser,1927; Amoroso,1952; Friess et al.,1980). The persisting barrier associated with epitheliochorial placentation illustrates the importance of the uterine epithelium to maximize nutrient provision and gas exchange between the mother and embryo. To cross the epithelium, molecules must pass through cells (transcellular pathway) or go between cells (paracellular pathway; Citi and Cordenonsi,1998; Anderson,2001). High molecular weight tracers can freely diffuse along the paracellular pathway until they reach the tight junction (TJ), the most apical structure of the epithelial junctional complex (Farquhar and Palade,1963), which seals the paracellular route (Anderson and Van Itallie,1995; Balda and Matter,1998). The TJ makes up a barrier that is an essential feature of epithelial and endothelial cells for the regulation of nonspecific passive diffusion of water, solutes, and immune cells driven by electro-osmotic gradients (Anderson and Van Itallie,1995; Nusrat et al.,2000; Van Itallie and Anderson,2004). In freeze fracture replicas, the TJ shows an anastomosing meshwork of strands (Winterhager et al.,1987; Claude and Goodenough,1973; Staehelin,1973; Murphy et al.,1982) which associate laterally with other TJ strands in the opposing membrane of adjacent cells (Tsukita et al.,2001). Recent studies carried out to identify TJ proteins in the uterine epithelium of skinks revealed that occludin, a 60-kDa membrane protein directly incorporated into individual TJ strands (Furuse et al.,1993; Ando-Akatsuka et al.,1996) was present in some lineages of Australian skinks but not in others (Biazik et al.,2007). The discovery of occludin suggests that the normally passive paracellular pathway becomes more regulated to prevent free diffusion of ions and solutes as the embryo develops in utero; however, this barrier is relatively nonspecific. Suggestions have also been made that the number of TJ strands does not determine the properties of the TJ barrier and changes in permeability result from changes in the quality of the TJ strands, such as their molecular composition, rather than the quantity of the TJ strands (Kojima,2002; Saitou et al.,1998). This finding led to the discovery of two ∼22-kDa novel TJ integral proteins named claudin-1 and claudin-2 (Furuse et al.,1998a).

Claudins have relevance in embryonic development and organogenesis by influencing epithelia–mesenchymal transitions (Bello et al.,2007) and are associated with forming ion selective pores (Tsukita and Furuse,2000; Heiskala et al.,2001). When opposing uterine epithelial cells express different claudins, a mismatch or a heteropolymeric interaction occurs (Furuse et al.,1999; Coyne et al.,2003; Wang et al.,2003a), thus, resulting in a pore formation that increases ion and solute flow (Furuse et al.,1999). Because the discovery of occludin in the uterus of skinks did not disclose information about the ion selectivity or size of solutes passing across the paracellular pathway, determining the presence or absence of claudins in the uterine epithelial cells in the skinks uterus may indeed reveal the possible mechanism for transepithelial nutrient permeation.

The claudin genome shows conservation between species. Mammals have 24 claudin protein members (Van Itallie and Anderson,2004), zebrafish 15 (Hardison et al.,2005), the puffer fish 56 (Loh et al.,2004), and six claudin species have also been described in Drosophila (Behr et al.,2003). Additionally, the expansion of the claudin gene family that exhibits tissue- and cell-type restricted expression in mammals (Furuse et al.,1993; Morita et al.,1999a; Heiskala et al.,2001) has resulted in the evolution of increasingly complex tissues and organs (Loh et al.,2004). Claudin-1, -2, -3, -4, and -5 are the most commonly studied claudins and have generated the most interest in research. Claudin-1 prevails in mouse epithelial liver and kidney TJ (Furuse et al.,1998b) and is a barrier to fluid loss (Furuse et al.,2002). Claudin-2 occurs in the lung (Mitic et al.,2000) and forms aqueous pores in epithelia (Furuse et al.,2001). Claudin-3 is present in lung and liver (Wolburg and Lippoldt,2002) and is a constituent of TJ strands (Tsukita et al.,2001) and claudin-4 occurs in mouse rat and kidney (Mitic et al.,2000) and influences paracellular ion selectivity (Van Itallie et al.,2001). Claudin-5 occurs in the TJ region of uterine epithelial cells in diestrus and proestrus rat uterus (Mendoza-Rodriguez et al.,2005), in retinal pigment epithelium (Kojima et al.,2002), airway epithelium (Coyne et al.,2003), and colonic epithelium (Amasheh et al.,2005). Claudin-5 is also a major component of the TJ in brain endothelial cells and fundamentally organizes the blood–brain barrier (Morita et al.,1999b). The deletion of the claudin-5 gene from brain endothelial cells led to an incremental increase in the size of tracers allowed to exit from the vascular space into the brain, resulting in the blood–brain barrier becoming a molecular sieve that is more permeable to small molecules (<0.8 kDa), but not larger molecules (Nitta et al.,2003).

Each claudin plays a unique role in epithelia. Examining the presence or absence and distribution of claudin-1, -2, -3, -4, and -5 in the uterine epithelium of skinks at different stages of the reproductive cycle and with differing parity modes will describe the molecular composition and function of the TJ in the uterus of skinks. Understanding the permeation property of the paracellular pathway in the skink uterus may determine whether this pathway is associated with possible nutrient provision in the skink uterus.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

Species

Oviparous and viviparous species of Australian skinks from two lineages, Eugongylus and Sphenomorphus groups (Greer,1989), were chosen for this investigation to avoid confounding effects attributable to evolutionary history and not the evolution of viviparity. From the Sphenomorphus lineage Eulamprus tympanum, a viviparous skink with a simple reptilian placenta (Hosie et al.,2003) as well as the bimodally reproductive Saiphos equalis were included, allowing for intraspecific comparison of an egg laying population and a live bearing population with a simple reptilian placenta (Smith and Shine,1997). Within the Eugongylus lineage, two closely related viviparous skinks, Pseudemoia entrecasteauxii and P. spenceri, were chosen as they have the most complex reptilian placentae to be described in Australia (Weekes,1935; Stewart and Thompson,1998), Lampropholis guichenoti because it is oviparous and the bimodally reproductive species Lerista bougainvillii species was also included.

Animal Collection and Husbandry

Pregnancy/gravidity of the skinks was determined by their large abdominal size and by palpitation of the lower abdomen to feel for embryos, but gestational stage remained unknown until the females were killed and the developing embryo was staged. On average, each female contained two to five embryos/eggs in the uterus. Vitellogenic skinks were collected early in the reproductive season (September).

Oviparous Saiphos equalis and Lampropholis guichenoti were collected from the grounds of The University of Sydney (33°53minS, 151°11minE), and viviparous S. equalis were collected from the Riamukka State Forest, New South Wales, Australia (31°20minS, 151°39minE). Pseudemoia entrecasteauxii, P. spenceri, and Eulamprus tympanum females were collected from Kanangra Boyd National Park, New South Wales, Australia (33°59minS, 150°03minE). Females were transported to The University of Sydney. Oviparous Lerista bougainvillii were collected from Burra, South Australia (33°38minS, 139°00minE), and viviparous L. bougainvillii were collected from Cape Willobough, Kangaroo Island, South Australia (35°51minS, 138°08minE). Pseudemoia entrecasteauxii, P. spenceri, E. tympanum, L. bougainvillii, and L. guichenoti were housed in glass aquaria lined with newspaper and rocks to provide basking sites. Aquaria were kept at 21°C with an 8:16 hr light/dark photoperiod. Saiphos equalis was housed under similar conditions, but aquaria contained a 20-mm-thick substrate of moist peat moss because the species is fossorial. Skinks were fed three times a week with meal worms (Tenebrio molitor) and crickets (Acheta domestica) dusted with Herptivite™ (Multivitamins) and phosphorus-free calcium (Per-Cal CA, USA). Water was provided ad libitum in open dishes. Skinks were housed until they reached the relevant reproductive stage and were killed by intrathoracic injection of 0.1 ml of sodium pentobarbitone 60 mg/ml (Nembutal, Boehringer, Australia) and cervical dislocation.

Embryonic Stages

Twenty-five skinks from each of the five species were examined. Embryonic development in reptiles is temperature dependent so the 40 embryonic stages are based on morphological development rather than time. For the purpose of this study, five individuals from each species were allocated to different reproductive stages by using the Dufaure and Hubert (1961) staging method. In oviparous skinks, two reproductive stages (mid-embryonic stages 25–27 and late stage 30, which coincides with oviposition) were chosen. In viviparous species, three reproductive stages were recognized (early, embryonic stages 20–25; mid, embryonic stages ranged from stage 30–37; and late, stages 38–40). Five vitellogenic and postparturient (2 weeks) skinks from each species were also examined.

Uterus Excision

After killing, a central ventral incision exposed the uterus. An incision was made between the incubation chambers of the uterus containing the embryos and the uterine tissue was peeled away carefully. Embryos were fixed in 10% buffered formalin. Control sections were rat lungs and kidneys and skink lungs and kidneys (Wolburg and Lippoldt,2002; Coyne et al.,2003, Wang et al.,2003a). All tissue for immunohistochemistry was coated in Tissue–Tek OCT cryoprotectant (Merck, Vic, Australia), immersed in super-cooled isopentane (Unilab, NSW, Australia), and stored in liquid nitrogen. Tissue used for Western blot analysis was frozen directly in liquid nitrogen.

Sodium Dodecyl Sulfate-Polyacrylamide Gel Electrophoresis and Western Blot Analysis

Frozen tissue was homogenized in lysis buffer (Sigma-Aldrich, St Louis, MO) using a bead mill homogenizer (Biospec Products, OK). After incubation on ice for 30 min, homogenate was centrifuged at 12,000 × g for 10 min and supernatant was collected. Quantity of total protein in the supernatant was determined using a Micro BCA Protein Assay kit (Quantum Scientific, Australia). Protein samples (20 μg) were denatured in Laemmli sample buffer at 90°C for 10 min and separated on a 12% sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) gel and transferred onto a polyvinylidene fluoride membrane (Millipore, MA). Nonspecific antibody binding was blocked with 0.5% skim milk powder in 10 mM Tris buffered saline (TBS) pH 7.4, 0.01% Tween-20 (TBST) for 1 hr at room temperature (RT). Membranes were incubated overnight in 1 μg/ml of appropriate primary antibodies; Rb/poly anti–claudin-1, Rb/poly anti–claudin-2, Rb/poly anti–claudin-3, Mo/mono anti–claudin-4 and Mo/mono anti–claudin-5 (Zymed Laboratories, CA). The membranes were then washed in TBST, and incubated in secondary antibody, horseradish peroxidase (HRP) -conjugated goat-anti rabbit (Dakocytomation, Glostrup, Denmark) 1:2,000 dilution for claudin-1, -2, and -3 and HRP-conjugated anti-mouse IgG (Amersham, UK) 1:2,000 dilution for claudin-4 and -5 and in 0.2% skim milk powder / TBST at RT for 1 hr. After final wash in TBST, membranes were incubated in enhanced chemiluminescent substrate (ECL plus; Amersham, UK) for 5 min and signals were captured using the Alpha Innotech imaging system and FluorChem SP software. A loading control for the Western method was carried out by stripping the membrane in 2% SDS, 100 mM 2-mercaptoethanol, 62.5 mM Tris HCl stripping buffer, blocking in 0.5% skim milk powder in TBST, and re-probing in 1:1,000 mouse monoclonal anti–β-actin (Sigma-Aldrich) in 0.5% skim milk TBST for 1 h. Secondary antibody incubation and chemiluminescent visualization was carried out as above.

Immunohistochemistry

Four sections (7 μm thick) were collected per gelatin-coated slide using a Leica CM3050 cryostat at −23°C. Four slides were cut per animal. All slides were fixed in 100% acetone at −20°C overnight. Two experimental and two control slides were randomly selected. After 30 min of air-drying, slides were incubated for 30 min in 1% bovine serum albumin (BSA; Sigma-Aldrich) in phosphate buffered saline (PBS). Depending on the investigation, slides were incubated as appropriate: Rb/poly anti–claudin-1, Rb/poly anti–claudin-2, Rb/poly anti–claudin-3, Mo/mono anti–claudin-4 and Mo/mono anti–claudin-5 antibodies (Zymed Laboratories, CA), 1 μg/ml in 1% BSA/PBS for 1 h at RT. After three rinses in PBS, sections were incubated in secondary fluorescein isothiocyanate (FITC) -conjugated goat anti-rabbit IgG (Zymed Laboratories) for claudin-1, -2, and -3 and FITC conjugated rabbit anti-mouse IgG for claudin-4 and -5 (Jackson ImmunoResearch Laboratories, West Grove, PA) 15 for 30 min at RT. The positive controls for immunohistochemistry (skink and rat kidney and lungs) were treated with the same primary and secondary incubation protocol. Negative controls carried out on skink tissue and rat tissue were treated by omitting the primary antibody and nonimmune controls were carried out by substituting mouse or rabbit matched IgG for the primary antibody. After rinsing with PBS, slides were mounted with Vectashield (Vector, Torrance, CA) and visualized using a Leitz-Diaplan microscope. Digital images were taken using a Leica DFC480 digital camera. Micrographs were imported into Adobe Photoshop 6.0, and the scale of figures was determined by a stage micrometer.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

Western blotting and indirect immunofluorescence studies were carried out to determine claudin-1, -2, -3, -4, and -5 expression in the uterus of oviparous and viviparous skinks and to determine whether changes in expression occurred in skinks at different stages of the reproductive cycle.

Claudins-1, -2, and -3

Claudins-1, -2, and -3 were not detected in the skink uterus of any species used in this study with immunohistochemical or Western blotting analysis. Claudin-1 and claudin-2 were not detected in skink kidney control. However, claudin-1 and -2 were present in the epithelium lining the tubules of the duct system in the rat kidney. Claudin-3 was not detected in the uterine epithelium of any skink uterus and was not detected in the skink lung control. Claudin-3 was present in the TJ region of alveolar epithelium and endothelial cells in the rat lung.

Claudin-4

Claudin-4 expression was detected using Western blotting in the skink kidney control tissue as a single band ∼22-kDa protein, and in rat kidney control tissue as a monomer at ∼22 kDa and a dimer at ∼44 kDa. Claudin-4 was not present in the uterine epithelium of any skink in the study at any gestational/gravid stage.

Claudin-5

Unlike other claudins, claudin-5 was expressed in the uterine epithelium of all skinks studied. Immunohistochemistry and Western blotting revealed changes in claudin-5 expression at various stages of the reproductive cycle in these skinks.

Immunohistochemically, in the nonreproductive uterus of L. bougainvillii (ovip) (Fig. 1a,b), P. spenceri (Fig. 1d), S. equalis (vivip; Fig. 2g), and P. entrecasteauxii (Fig. 2i), claudin-5 is ubiquitously expressed in the uterine and glandular epithelial cells. Claudin-5 is found in the apical, lateral, and basal region of the cytoplasm in uterine epithelial cells. In the pregnant/gravid uterus of L. bougainvillii (ovip; Fig. 1b), P. spenceri (Fig. 1e), S. equalis (vivip; Fig. 2h), and P. entrecasteauxii (Fig. 2j,k), claudin-5 expression is greatly reduced and has migrated to the apical region of the lateral plasma membrane, corresponding to the ultrastructural location of the TJ. Claudin-5 is not present elsewhere within the uterine epithelial cells, or stroma and myometrium.

thumbnail image

Figure 1. Immunofluorescence micrographs of skink uterus at various stages of pregnancy/gravidity. Uterine epithelium (arrows), uterine lumen (L), and uterine glands (G). A–C: Cross-section of L. bougainvillii uterus (ovip). A: Vitellogenic uterus showing claudin-5 expression in apical, lateral, and basal region of the cytoplasm of uterine and glandular epithelium. B: High power of uterine epithelium in vitellogenic uterus showing claudin-5 expression along apical, lateral, and basal regions of the cytoplasm of uterine epithelial cells. C: Gravid uterus showing claudin-5 is expressed only in the TJ region of the plasma membrane and is not expressed in glands. D,E: Cross-section of P. spenceri uterus. D: Closed down convoluted vitellogenic uterus with claudin-5 staining along apical, lateral, and basal regions of the cytoplasm of uterine epithelial cells. E: Late stage pregnant uterus with claudin-5 expression only in tight junction region of the lateral plasma membrane and opening to a uterine gland (asterisk) F: Negative nonimmune control using mouse matched IgG shows no immunofluorescence in vitellogenic uterus of P. spenceri. Scale bars = 30 μm in A,D,F, 10 μm in B, 20 μm in C,E.

Download figure to PowerPoint

thumbnail image

Figure 2. Immunofluorescence micrographs of skink uterus at various stages of gestation/gravidity. G,H: Cross-section through uterus of S. equalis (vivip). G: Closed down vitellogenic uterus showing claudin-5 expression along apical, lateral, and basal regions of the cytoplasm of uterine epithelium and glandular epithelium. H: Late stage pregnant uterus showing claudin-5 confinement to tight junction (TJ) region of lateral plasma membrane and no other staining present in stroma or glands. I–K: Cross-section through uterus of P. entrecasteauxii. I: Closed down vitellogenic uterus showing claudin-5 expression along apical, lateral, and basal regions of the cytoplasm as well as in glandular epithelium. J: Late stage pregnant uterus with claudin-5 expression in TJ region of the lateral plasma membrane. Uterine stroma has thinned out, and no glands are present. K: Villous folds (V) of the chorioallantoic placenta in late stage uterus showing claudin-5 expression in the TJ region of the lateral plasma membrane. L: Positive control of skink lungs from P. spenceri showing claudin-5 expression in the TJs between alveolar cells (a) and in the TJs between endothelial cells of alveolar blood vessel (BV). Scale bars = 30 μm in G–K, 10 μm in L.

Download figure to PowerPoint

Uterine gland numbers are greatly reduced in the pregnant/gravid uterus; consequently, claudin-5 glandular expression is reduced also. Claudin-5 expression in the positive control was detected in the region of the TJ in rat and skink lung alveolar epithelium and in the TJ between alveolar endothelial cells (Fig. 2l).

The same amount of total protein was extracted from each skink uterus and separated on a 12% SDS-PAGE followed by immunoblotting with anti–claudin-5 antibody, all species studied showed a closely separated doublet at ∼22 kDa (S. equalis, vivip, Fig. 3a; and S. equalis, ovip, Fig. 3d) or a ∼22-kDa single band associated with Cl-5 expression (P. spenceri, Fig. 3b; P. entrecasteauxii; Fig. 3e; and in L. guichenoti, Fig. 3f).

thumbnail image

Figure 3. Immunoblots of whole lysate of skink uterus at various stages of pregnancy/gravidity and positive skink lung controls. A total of 20 μl of protein was loaded into each well. Claudin-5 expression was a single ∼22-kDa band or a closely separated double band at ∼22 kDa. A: Viviparous S. equalis. Vitellogenic uterus (lane 1), mid stage pregnant uterus (lane 2), late stage pregnant uterus (lane 3), a 2 week postparturient uterus (lane 4), and a positive control S. equalis (vivip) lung (lane 5). B:P. entrecasteauxii. Vitellogenic uterus (lane 1), mid stage pregnant uterus (lane 2), late stage pregnant uterus (lane 3), and a positive P. entrecasteauxii lung control (lane 4). C: Positive lung controls with claudin-5 depicted as a single ∼22 kDa. P. spenceri (lane 1), P. entrecasteauxii (lane 2), E. tympanum (lane 3), and S. equalis (ovip) (lane 4). β-Actin loading control depicted as a single ∼42-kDa band shows even loading of protein into each well D:S. equalis uterus (ovip) shows a closely separated claudin-5 doublet at ∼22 kDa. Vitellogenic uterus (lane 1), late stage gravid uterus (lane 2), a 2 weeks postparturient uterus (lane 3), and positive control S. equalis (ovip) lung (lane 4). E:P. entrecasteauxii showing claudin-5 expression as a single ∼22-kDa band. Vitellogenic uterus (lane 1), mid stage pregnant uterus (lane 2), late stage pregnant uterus (lane 3), and positive control P. entrecasteauxii lung (lane 4). F:L. guichenoti showing a single claudin-5 band at ∼22 kDa. Vitellogenic uterus (lane 1), mid stage gravid uterus (lane 2), late stage gravid uterus (lane 3), and positive control L. guichenoti lung (lane 4).

Download figure to PowerPoint

In all the skinks studied, the uterus of vitellogenic and 2-week postparturient animals had a greater claudin-5 band compared with the uterus of skinks in mid/late pregnancy/gravidity.

Positive controls, of lung tissue from each animal showed claudin-5 expression at ∼22 kDa and β-actin, which is a ∼42-kDa ubiquitous protein used as a loading control expressed equal amounts of protein in skink uterus homogenate (Fig. 3c).

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

The discovery of claudin-5 in the skink uterus as seen in this study describes the specificity of the tight junctional regulation of the paracellular pathway in uterine epithelial cells. Additionally, differential expression of claudin-5 at various stages of the reproductive cycle suggests that possible changes in ion and solute permeation occurs depending on the stage of embryonic development in the skink uterus. The uterine epithelium of nonreproductive skinks showed claudin-5 expression along the apical, lateral, and basal regions of the cytoplasm. In the uterine epithelium of late stage pregnant/gravid skinks, claudin-5 redistributed to the TJ region of the cells and suggests that paracellular permeability is changed to precisely regulate ion and molecule diffusion. Without these changes to paracellular barriers in late stage pregnancy/gravidity, close regulation of paracellular diffusion would not be possible. Expression of claudin-5 in endothelial cells of the blood–brain barrier (Nitta et al.,2003) is indicative of a completely suppressed paracellular pathway and suggests that the barrier that is established in the uterus of skinks in late stage pregnancy/gravidity has similar attributes.

Claudin-5 was detected in the uterine epithelium of all the skinks studied, and was also confirmed in skink control tissue in the TJ region between alveolar and endothelial cells in the lung. In nonreproductive skinks, the uterine wall is thick and large numbers of uterine glands are present. Claudin-5 was expressed not only in the TJ region of the lateral plasma membrane, but also along the apical, lateral, and basal region of the cytoplasm of uterine epithelial and glandular epithelial cells. In various cells, it has been documented that claudins are distributed not only in the TJs, but also at the lateral membranes without forming TJ strands (Furuse et al.,2002; Li et al.,2004) or it can occur as diffuse labeling in endothelial cell cytoplasm as in the human brain (Virgintino et al.,2004). Similarly, in the estrus phase in rats, claudin-5 is detected in the basolateral region of the plasma membrane and in the cytosol (Mendoza-Rodriguez et al.,2005), and is thus not associated with a TJ seal. At this point, the TJ has not been assembled and fluid fills the uterine lumen to provide an important microenvironment for sperm capacitation (Wang et al.,2003b) and an energy source for the blastocyst (Magnuson et al.,1978). This finding suggests that, in the nonreproductive skink uterus, the TJ is not yet established and paracellular diffusion is not regulated, therefore allowing ions and solutes to freely enter the luminal space.

In the uterus of late stage pregnant/gravid skinks, the thickness of the uterine wall is reduced from that of the nonreproductive uterus and consequently the uterine glands are reduced or completely lost. Claudin-5 expression dramatically alters, and there is a redistribution and confinement of claudin-5 only to the TJ region of the lateral plasma membrane of uterine epithelial cells. No other immunofluorescence is detected elsewhere in the uterus. During the diestrus and proestrus phase in rats, claudin-5 and occludin shifts to the TJ region of the lateral plasma membrane, therefore assembling a strict paracellular barrier (Mendoza-Rodriguez et al.,2005) associated with regulation of the luminal fluid to prepare the uterus for implantation (de Jesus et al.,1972). The redistribution of claudin-5 from the cytosol to the region of the TJ in endothelial cells has also been detected during fetal development of the brain (Virgintino et al.,2004). This finding suggests that, as claudin-5 redistributes to the TJ region of the lateral plasma membrane in the uterine epithelium of skinks, free diffusion of solutes across the paracellular pathway is greatly reduced.

Claudin-5 expression is associated with the formation of size selective and ion selective pores, and claudin-5 knockout manipulation suggests that smaller molecules are able to diffuse across the normally “tight” barrier (Nitta et al.,2003). The expression of claudin-5 in the TJ of uterine epithelial cells in the skink uterus suggests that the size of the diffusional pore formed by claudin-5 expression allows for molecules that are larger than 0.8 kDa to pass across the paracellular space. This pathway is, therefore, impermeable to ions such as Ca2+ and Mg2+ because they are all smaller than 0.8 kDa and must utilize the transcellular pathway instead. Additionally, there is a limit to the size of the molecule that can diffuse across these pores. Histotrophy is another mechanism where macromolecules and lipids are transported to the developing embryo by means of vesicle secretion in reptiles (Corso et al.,2000; Blackburn and Lorenz,2003; Adams et al.,2005), and indeed occurs at the time when claudin-5 is present in the TJ region of the lateral plasma membrane. It may be that smaller ions and molecules are transported by means of the TJ pores, which are formed by claudin-5 and transcellular channels and larger molecules are transported by means of histotrophy.

The redistribution of claudin-5 in late stage pregnancy/gravidity to the TJ region of the lateral plasma membrane in oviparous, viviparous, and bimodally reproductive skinks suggests that TJ regulation is not only important in viviparous species, but also in egg layers. The significance of this finding reveals that, even in oviparous species in which most nutrients needed for development are confined to the egg yolk, a maternal nutrient contribution still exists before calcification of the egg in utero. This explanation is also suggested by studies carried out on egg yolk calcium levels, which revealed that calcium levels were insufficient to sustain development (Tuan et al.,1991) and our recent work carried out in the uterus of oviparous skinks, indeed show the presence of active Ca2+ATPase channels in the uterine epithelium of the skink uterus during the egg shelling phase of the reproductive cycle (Herbert et al.,2006). It may also be that, as transcellular and histotrophic activity imports fluid and ions into the luminal space between the maternal and fetal epithelium, the redistribution of claudin-5 and occludin to the tight junction may be associated with confining histotrophic and transcellular products to the fetal and maternal interface to allow for a more effective fetal absorption of those products.

Claudins -1, -2, and -3 were not detected in the uterine epithelium at any stage of the reproductive cycle in this present study and were not detected in the skink control tissue either. This finding suggests that these claudins are not associated with ion and solute regulation in the skink uterus or that the antibodies for claudins-1, -2, and -3 do not cross react in the skink. Claudin-4 was detected in the skink kidney control tissue with immunofluorescence in the TJ region of epithelium lining the duct system and with Western blotting detection as a single ∼22-kDa band from whole skink kidney homogenate. Claudin-4, however, was absent in the skink uterine epithelium at all stages of the reproductive cycle, and this finding suggests that claudin-4 is tissue specific in the skink.

In summary, this is the first documentation of the presence and distribution of claudin-5 in the TJ region of uterine epithelium in the skink uterus. Due to the uniqueness and ionic selectivity of claudin-5, the presence of this protein in the skink uterus provides evidence of a precise mechanism that regulates ion and solute permeation. The claudin-5 pore essentially forms an impermeable barrier to ions and molecules that are smaller than 0.8 kDa and, therefore, signifies the importance of transcellular and histotrophic activity in the uterus of late stage pregnant/gravid skinks. Because the paracellular pathway is slowly being understood in the uterus of skinks, future work on the transcellular pathway will further provide a molecular mechanism for the transepithelial nutrient provision and, thus, the evolution of placentotrophy and viviparity in reptiles.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

Skinks were collected with permits from the NSW National Parks and Wildlife Service (S10693), and the work was conducted under University of Sydney Animal Ethics Committee number L04/1-2005/3/4038. We thank the many people who have volunteered in the field and lab, especially Jacquie Herbert, Jim Stewart, Scott Parker, and Trevor Wilson. M.B.T. and C.R.M. were funded by the Australian Research Council.

LITERATURE CITED

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED