Molecular mechanisms in uterine epithelium during trophoblast binding: the role of small GTPase RhoA in human uterine Ishikawa cells

Abstract

BACKGROUND: Embryo implantation requires that uterine epithelium develops competence to bind trophoblast to its apical (free) poles. This essential element of uterine receptivity seems to depend on a destabilisation of the apico-basal polarity of endometrial epithelium. Accordingly, a reorganisation of the actin cytoskeleton regulated by the small GTPase RhoA plays an important role in human uterine epithelial RL95-2 cells for binding of human trophoblastoid JAR cells. We now obtained new insight into trophoblast binding using human uterine epithelial Ishikawa cells. METHODS: Polarity of Ishikawa cells was investigated by electron microscopy, apical adhesiveness was tested by adhesion assay. Analyses of subcellular distribution of filamentous actin (F-actin) and RhoA in apical and basal cell poles were performed by confocal laser scanning microscopy (CLSM) with and without binding of JAR spheroids as well as with and without inhibition of small Rho GTPases by Clostridium difficile toxin A (toxin A). In the latter case, subcellular distribution of RhoA was additionally investigated by Western blotting. RESULTS: Ishikawa cells express apical adhesiveness for JAR spheroids and moderate apico-basal polarity. Without contact to JAR spheroids, significantly higher signalling intensities of F-actin and RhoA were found at the basal as compared to the apical poles in Ishikawa cells. RhoA was equally distributed between the membrane fraction and the cytosol fraction. Levels of F-actin and RhoA signals became equalised in the apical and basal regions upon contact to JAR spheroids. After inhibition of Rho GTPases, Ishikawa cells remained adhesive for JAR spheroids, the gradient of fluorescence signals of F-actin and RhoA was maintained while the amount of RhoA was reduced in the cytosolic fraction with a comparable increase in the membrane fraction. CONCLUSION: Ishikawa cells respond to JAR contact as well as to treatment with toxin A with rearrangement of F-actin and small GTPase RhoA but seem to be able to modify signalling pathways in a way not elucidated so far in endometrial cells. This ability may be linked to the degree of polar organisation observed in Ishikawa cells indicating an essential role of cell phenotype modification in apical adhesiveness of uterine epithelium for trophoblast in vivo.

Figure 1

Ultrathin sections of Ishikawa cells. A: Ishikawa cells before treatment with toxin A. Cells grow as monolayers and show apico-basal polarity with nuclei located at the base of the cells and organelles found predominantly in the supranuclear region of the cells. Insert (light microscopy, cross section) shows overview of Ishikawa cells growing as monolayers. B: Lateral cell membranes show tight junctions, adherens junctions and desmosomes in varying combinations, whereas regular junctional complexes consisting of tight junction, adherens junction and desmosomes in apico-basal sequence were rarely seen. C: Ishikawa monolayers after treatment with toxin A. Standard electron microscopy showed no substantial differences between untreated (A) and toxin A-treated cells (C). Me: cell culture medium; N: nucleus; oo: coverslip; *: desmosome; arrow: adherens junction.

Figure 2

Localisation of E-cadherin and integrin subunit α_v. A-D: E-cadherin; E-H: integrin subunit α_v, both detected by confocal laser scanning microscopy. Note that both adhesion molecules are located in all plasma membrane domains, including the apical one. Typical patterns are presented for apical (apical) and basal (basal) cell poles as well as for the middle part (mid) of cells. D,H: negative controls.

Figure 3

Localisation of F-actin and RhoA in Ishikawa monolayers before binding of JAR spheroids. A, D: F-actin (red); B, E: RhoA (green); C, F: Merger of F-actin and RhoA signal. xy-sections representing second slices from the apical (apical) and basal (basal) cell poles, respectively. Typical patterns are presented. G, H: negative controls.

Figure 4

Quantification of F-actin and RhoA. A: Quantification of F-actin in the apical and basal regions of monolayer cultured Ishikawa cells before (mono) and after (contact) binding of JAR spheroids as well as after treatment with toxin A (mono + toxin A). For semi-quantitative evaluation of fluorescence, grey scale values (gsv) of each colour channel were determined within double-labelled cells. Stacks of 6 xy-sections at 0.5 μm intervals were collected with the first marked slice at the apical cell surface and basal cell pole, respectively. Numbers of Ishikawa cells tested: n = 14 (mono), n = 12 (contact), n = 15 (mono + toxin A). Values differ significantly (p < 0.05) between these experimental groups but not between apical and basal cell poles within Ishikawa cells being in contact with a JAR spheroid (n.s.). Data are presented as medians (first – third quartile). B: Quantification of RhoA in the apical and basal regions of monolayer cultured Ishikawa cells before (mono) and after (contact) binding of JAR spheroids as well as after treatment with toxin A (mono + toxin A). Semi-quantitative evaluation of fluorescence was done as described under A. Values differ significantly (p < 0.05) between these experimental groups but not between apical and basal cell poles within Ishikawa cells being in contact with a JAR spheroid (n.s.). Data are presented as medians (first – third quartile). C: Diameter of RhoA-positive granules in Ishikawa cells before (mono) and after contact (contact) with JAR spheroids, number of cells tested: n = 14 (mono), n = 12 (contact). Values differ significantly (p < 0.05) between apical and basal cell poles in contact situation as well as between basal cell poles in contact vs. non-contact situation (*). n.d.: not detectable. Data are presented as medians (first – third quartile).

Figure 5

Immunoblot analysis of endogenously expressed RhoA. Ishikawa monolayers remained untreated (- toxin A) or were treated with 100 ng/ml toxin A for 24 h (+ toxin A). Preparation of cell lysates (lys), membranes (mem), cytoskeleton fraction (TX) and cytosol (cyt) of cells was performed as described. Proteins (lysates: 10 μg/lane; membranes, cytoskeleton fraction, cytosol: each 100 μg/lane) were separated by SDS-PAGE and subsequently immunoblotted. Data shown are typical for three independent experiments.

Figure 6

Apical adhesiveness of Ishikawa cells. Apical adhesiveness for JAR cell spheroids of monolayer-cultured Ishikawa cells was determined in the centrifugal force-based adhesion assay before (mono; n = 349) and after (mono + toxin A; n = 358) treatment with Clostridium difficile toxin A. For comparison, adhesiveness of poly-D-lysine-coated glass coverslips (glass) for spheroids is shown (n = 335). Values differ significantly (p < 0.05) between experimental groups except for Ishikawa cells treated with and without toxin A (n.s.). Data are presented as means ± SEM.

Figure 7

Localisation of F-actin and RhoA in Ishikawa monolayers after binding of JAR spheroids. A: xz-section through contact site of Ishikawa cells and JAR spheroid after tracking of JAR cells with the vital dye CMFDA (blue). red: F-actin cytoskeleton. B-I: Localisation of F-actin (red) (B, E) and RhoA (green) (C, F) in the apical (apical) and basal (basal) regions of Ishikawa monolayers after binding of JAR spheroids tracked with CMFDA (blue). xy-sections represent second slice from the apical and basal cell poles, respectively. D, G: Merger of F-actin and RhoA signal. Typical patterns are presented. H, I: negative controls.