H

Fig. 9. Confocal laser scanning microscopy of F-actin localized by FITC-phalloidin binding in Caco-2 cells either untreated (A, C) or treated from the AP side with 50 ||M CuCl2 (B, D), 200 ||M CuCl2 (E, G) or 200 ||M Cu(His)2 (F, H). Arrowheads indicate where horizontal (A, B, E, F) and vertical (C, D, G, H) scans were taken. Bars = 20 |M.

Fig. 9. Confocal laser scanning microscopy of F-actin localized by FITC-phalloidin binding in Caco-2 cells either untreated (A, C) or treated from the AP side with 50 ||M CuCl2 (B, D), 200 ||M CuCl2 (E, G) or 200 ||M Cu(His)2 (F, H). Arrowheads indicate where horizontal (A, B, E, F) and vertical (C, D, G, H) scans were taken. Bars = 20 |M.

Double-labeling with an antibody for the adherens junction protein E-cadherin and with FITC-phalloidin showed that following treatment with 300 |M CuCl2 for 3 h, in those areas where the F-actin signal was strongly reduced (Fig. 8D), the E-cadherin signal was correctly localized at the cell periphery (Fig. 8C) and was of the same intensity as in control cells (Fig. 8A). This observation is in accord with recent data indicating that tight junctions in differentiated Caco-2 cells are uncoupled from adherens junctions (62). However, we had previously reported that also the tight-junction proteins ZO1 and occludin did not undergo gross changes in expression and localization following copper treatment (40). Therefore, the mechanisms of copper-induced alteration of tight-junction permeability do not appear to affect the localization of junctional proteins.

Confocal laser scanning microscopy allows a more accurate localization of intracellular signals in polarized cells such as the Caco-2 cells, as it permits acquisition of thin optical sections of relatively thick cell layers, allows exact microphotometry, gathers optical sections for vertical sectioning or three-dimensional reconstruction, and eliminates out-of-focus fluorescence in the plane of focus of the image (63).

As shown in Fig. 9A in a horizontal scan of control cells taken just below the AP surface, the F-actin signal was concentrated in a bundle running around the cell periphery. In a vertical scan of the same cells, the F-actin signal was prominent near the AP surface and along the brush border (Fig. 9C). In cells treated with 50 |M CuCl2 the horizontal scan showed areas of the cell monolayer com-

Fig. 10. Transmission electron microscopy of Caco-2 cells treated with copper. Control Caco-2 cells (A) and cells treated with 50 pM (B) or 200 pM CuCl2 (C) in the AP compartment for 3 h at pH 6.0. Ultrastructural changes to microvilli are evident after copper treatment with major disorganization (B), swelling and detachment occurring at increasing concentration of copper (C). Bars: A = 450 nm, B = 530 nm, C = 590 nm.

Fig. 10. Transmission electron microscopy of Caco-2 cells treated with copper. Control Caco-2 cells (A) and cells treated with 50 pM (B) or 200 pM CuCl2 (C) in the AP compartment for 3 h at pH 6.0. Ultrastructural changes to microvilli are evident after copper treatment with major disorganization (B), swelling and detachment occurring at increasing concentration of copper (C). Bars: A = 450 nm, B = 530 nm, C = 590 nm.

posed of a few cells in which the F-actin signal appeared to be absent (Fig. 9B). In the vertical scan taken across these areas, the loss of F-actin staining near the AP membrane and brush border was evident (Fig. 9D). Following treatment with 200 pM CuCl2, the areas showing reduced and altered F-actin staining increased in size and frequency. Figures 9E (horizontal scan below the AP surface) and 9G (vertical scan) showed a large area of the cell monolayer exhibiting strongly decreased staining for F-actin, and the cells all around this patch exhibited reduced and disorganized F-actin signal. Conversely, F-actin staining in cells treated with 50-200 pM Cu(HiS)2 was similar to that in untreated cells [Fig. 9F,H showing horizontal and vertical scans, respectively, of cells treated with 200 pM Cu(HiS)2 for 3 h].

A depolymerizing effect of copper on F-actin has previously been observed in Mytilus galloprovincialis emocytes (64). Although copper is able to bind with high affinity to a COOH-terminal cysteine residue of actin, this binding does not appear to affect its state of polymerization (65). However, the effect of copper on F-actin may be indirect and may be mediated by cytosolic factors such as actin-associated proteins, as recently demonstrated for cadmium in an in vitro actin polymerization assay employing cytosol preparations of mesangial cells (66).

4.2. Ultrastructural Studies by Transmission Electron Microscopy

4.2.1. Materials and Methods

To evaluate the effects of copper on the ultrastructure of intestinal Caco-2 cells, copper-treated and control cells were prepared for transmission electron microscopy (TEM).

One-half of the filters used for F-actin-localization experiments were fixed for 1 h in 2.5% glut-araldehyde in 0.1 M phosphate buffer (pH 7.4) and postfixed in 1% OsO4 for 30 min. The samples were dehydrated and embedded in Agar 100 resin (Agar Scientific Ltd., Stansted, Essex, UK). Ultra-thin sections were cut (Ultracut E, Reichert Jung Optische Werke, Vienna, Austria), stained, and observed in the electron microscope (CM 10; Philips, Eindhoven, The Netherlands). A rapid and convenient method for processing of filter-grown cells for TEM has been published (67) and may also be applied to Caco-2 cells.

4.2.2. Results and Discussion

Figure 10A shows the ultrastructure of microvilli in control-differentiated Caco-2 cells: The microvilli are well aligned and regular in shape and height. Following treatment with 50 | M CuCl2 for 3 h, the ultrastructure of microvilli appeared distorted, with some branched microvilli (Fig. 10B). At higher concentrations, copper induced gross distortions and dilation of microvilli, often detaching from the cell surface (Fig. 10C). Similar changes in microvilli morphology have been reported in kidney proximal tubule cells (68) and in isolated rat hepatocytes (69) in which F-actin depolymeriza-tion had been induced by ATP depletion or by changes in intracellular calcium. Hepatocytes of copper-loaded rats also exhibited ultrastructural alterations of the microvilli (70). The ultrastructural changes to microvilli are probably the result of the copper-induced depolymerization of the major scaffold of microvilli formed by F-actin bundles, as also shown by CLSM data demonstrating a complete loss of F-actin signal in the area of the brush border (Fig. 9).

0 0

Post a comment