Desmosomes are adhesive multimolecular intercellular junctions composed of interacting proteins and transmembrane glycoprotein molecules. Desmosomes facilitate adhesion between adjacent cells and link intermediate filaments of adjacent cells, producing an integrated scaffolding of cytoskeletal elements throughout the tissue sheet. The main molecular components of desmosomes are members of three gene superfamilies, the desmosomal cadherins, the armadillo proteins and the plakins. Desmosomal cadherins are members of the cadherin superfamily of cell adhesion molecules and are the adhesion molecules of the desmosomal plaque. The two principal desmosomal cadherin subclasses are the desmocollins (Dsc) and the desmogleins (Dsg). Dsc exist as three distinct isoforms (Dsc 1–3) encoded by separate genes, each of which exist as alternatively spliced ‘a’ and ‘b’ forms. To date, 6 isoforms of Dsg have been discovered. Different Dsg isoforms exhibit tissue and differentiation specific expression with Dsg1 restricted to stratified epithelia whereas Dsg 2 appears to be more ubiquitous.

Successful implantation of the blastocyst in the uterus can only occur during a narrow time range and under specific hormonal conditions, the so-called “window of implantation”. In the rat, implantation is achieved by displacement penetration, during which trophoblastic projections penetrate between adjacent uterine epithelial cells and then extend between the uterine epithelium and the basal lamina. Decidual cells subsequently breach the basal lamina and the embryo invades the endometrial stroma. Uterine epithelium readily detaches from the underlying basal lamina and is sloughed off both as sheets of cells and as individual cells. Uterine epithelium is considered to be a barrier to implantation at times of non-receptivity. Therefore, it is likely that adhesive properties of uterine epithelium undergo alterations when the uterus shifts from the non-receptive state to the receptive state in order to allow these key processes of implantation to occur.

Previous studies have described that the integrity of lateral junctional complexes of uterine epithelial cells is indeed compromised during the peri-implantation period. In the rat, several tight junction molecules are redistributed and adherens junctions are lost during early pregnancy. Ovarian hormones have been shown to regulate these aspects of lateral junctional complexes, but to date hormonal regulation of desmosomal function is yet to be established. Proteins associated with adherens junctions, such as E-cadherin, -catenin and ?-catenin, have been shown to be downregulated by oestrogen and upregulated by progesterone in the human endometrium. Freeze-fracture studies have demonstrated that progesterone alone or in conjunction with oestradiol leads to an increase in the depth of tight junctions, similar to that observed at the time of implantation in the rat.

Various desmosomal proteins have been investigated in a variety of species. In the rabbit, it was found that desmoplakin I & II are strongly expressed in apical and lateral domains of the cell membrane prior to implantation and are redistributed along the lateral plasma membrane at the time of implantation. Human uterine epithelial cell lines also exhibit punctate regions of desmoplakin I staining. Illingworth et al. (2000) demonstrated a decrease in desmoplakin I & II mRNA and protein levels at the time of implantation in mouse uterine epithelial cells.

Clearly, there are major changes in adhesive properties of uterine epithelial cells when the uterus shifts from the non-receptive to the receptive state and these changes are likely to be under the control of ovarian hormones. Since luminal epithelial cells are the first site of contact with the implanting blastocyst, these cells are the primary focus in the present study. As far as we know, this is the first study to use an antibody against Dsg 1&2 to investigate distributional patterns of desmosomes, and their control by ovarian hormones in rat uterine epithelial cells during the peri-implantation period.

All animals used in the present study were housed at 21 °C with a 12 h light:dark cycle and were given food and water ad libitum. To investigate the localisation of Dsg1&2 within the uterus during normal pregnancy, 20 female adult virgin Wistar rats were used. Vaginal smears were taken during the late afternoon and any females in proestrus were caged overnight with a male of proven fertility. Sperm in a morning vaginal smear indicated that mating had been successful and this was designated as day 1 of pregnancy. Five rats per group were sacrificed at days 1, 3, 6 and 7 of pregnancy.

To evaluate the effects of ovarian hormones on the localisation of Dsg1&2 in uterine epithelial cells, female rats were bilaterally ovariectomised using a ventral approach under isoflurane anaesthesia. After allowing a recovery period of 4 weeks, rats were randomly divided into four groups, each consisting of five animals. All hormone doses given were within the normal physiological range. Hormones were dissolved in benzyl alcohol (Sigma, St. Louis, MO, USA) and diluted in peanut oil to achieve levels as described below. Subcutaneous injections were given in the scruff of the neck. Animals in the control group were injected with 0.1 ml peanut oil only for three consecutive days. Animals in groups 2 and 3 were injected with either 0.5 ?g 17?-oestradiol (Sigma) in 0.1 ml peanut oil or 5 mg progesterone (Sigma) in 0.2 ml peanut oil, respectively, for three consecutive days. Animals in the fourth group were injected with 5 mg progesterone for 2 days and 0.5 ?g 17?-oestradiol as well as 5 mg progesterone on day 3 on opposite sides of the neck. All rats were killed at 24 h after the last injection with an intraperitoneal injection of sodium pentobarbitone (Nembutal; Merial Australia, Parramatta, NSW, Australia).

Uterine tissue was collected from all animals, immediately placed in ice-cold 0.1 M phosphate buffer and cut into 5-mm blocks which were immersed in OCT compound (Tissue Tek, Torrance, CA, USA). Blocks were then placed in supercooled isopentane (BDH, Poole, UK) before being stored in liquid nitrogen. Two blocks per animal were cut using a CM3050 cryostat (Leica, Heerbrugg, Switzerland) and 8-?m-thick sections were placed onto gelatin-coated slides. After briefly air drying, sections were fixed in acetone for at least 1 h at ?20 °C. Sections were air dried and non-specific binding was blocked with PBS containing 1% bovine serum albumin (BSA; Sigma) for 10 min at room temperature. PBS containing 1% BSA was also used as diluent for all primary and secondary antibodies. All subsequent incubation steps were carried out at room temperature. Sections were incubated in anti-Dsg1&2 primary antibodies (Progen Biotechnik, Heidelberg, Germany) at a concentration of 5 ?g/ml for 1 h. This mouse monoclonal antibody recognises an epitope within the repeating unit domain of the cytoplasmic C-terminal and is yet the only commercially available antibody of this type. After washing in PBS, sections were incubated in goat anti-mouse secondary antibodies conjugated with FITC (Jackson, Baltimore, MD, USA) at a concentration of 14 ?g/ml for 30 min. Sections were then washed in PBS, mounted in Vectashield (Vector, Burlingame, CA, USA), coverslipped and the edges sealed with clear nail polish. Sections were viewed using a Diaplan fluorescence microscope (Leica). Digital images were acquired using a DC 200 digital camera (Leica) and micrographs were produced using Adobe Photoshop (Adobe, San Jose, CA, USA). Final magnifications were calculated using a stage micrometer.

At least four randomly selected sections per animal were processed as negative controls and run concurrently with experimental sections. Negative controls were used to identify any non-specific binding of the secondary antibody and were treated in the same way as experimental sections except for omission of the primary antibody. Non-immune controls were carried out by replacement of primary antibodies with normal purified mouse IgGs (Sigma).