Previous studies have shown that glucocorticoids accelerate intestinal maturation and that this process is mediated by the mesenchymal cells. The possible involvement of laminin (LN), a basement membrane component, in this mesenchymal mediation has been analyzed. For this purpose, the influence of dexamethasone (DX) on the synthesis of LN, its chain composition and its cellular distribution has been examined biochemically and immunocytochemically in two different mesenchyme-derived cell populations, fetal intestinal mesenchymal cells and fetal skin fibroblasts, as well as in cocultures of intestinal endodermal cells seeded on top of confluent fetal skin fibroblasts. Neither the amount of metabolically labeled LN purified by affinity chromatography (expressed per mg cell proteins), nor the A versus B chain ratio monitored after separation on gel electrophoresis and immunoblotting, showed significant differences after 5 days of DX treatment. However, glucocorticoids induced a shift from secreted to cell-associated LN molecules paralleling a striking difference in the immunostaining pattern of intracellular and surface LN in the mesenchyme-derived cell monocul-tures; the granular intracytoplasmic LN staining in the control cultures was replaced by a fibrillar organization of LN molecules concomitantly with an increased accumulation at the cell surface. In 2-day DX-treated cocultures, there was an acceleration of LN deposition at the epithelial-fibroblastic interface, which accompanied the enhanced expression of epithelial cell differentiation markers (brush border digestive enzymes). These DX-induced changes can be blocked by the addition of antiLN antibodies in the culture medium.

These findings further support the concept that glucocorticoid action on intestinal epithelial cells involves alterations in the extracellular microenvironment, assessed here for LN molecules, occurring at the level of the mesenchymal cell compartment. These changes may contribute to an accelerated organization of LN at the epithelial-mesenchymal interface and subsequently to epithelial differentiation.

Glucocorticoids (GC) are known to accelerate intestinal maturation during the developmental period up to weaning. This phenomenon can be easily visualized, in vivo and in vitro, by the precocious induction of a normally late-appearing digestive enzyme in rodents, sucrase-isomaltase, and the stimulation of other enzymes, like maltase-glucoamylase and lactase, localized in the apical brush borders of absorptive cells (for reviews see Kedinger et al. 1986; Henning, 1987).

The intestinal tissue is composed of the endoderm-derived epithelium and of the mesenchyme-derived lamina propria and muscle coat. Several experimental models of inter-species or -topic embryonic tissue recombinations allowed us to show that the maturation effect of GC on epithelial cells was indirect, the primary target of the hormones being the mesenchymal compartment (for review see Kedinger et al. 1989a). Indeed, the response of chick embryonic intestinal enzymes to GC was different according to the origin of the mesenchymal support (Lacroix et al. 1985). Sucrase, which is never expressed in the colonic epithelium, can be induced by GC when colonic mesenchyme is replaced by small intestinal mesenchyme (Foltzer-Jourdainne et al. 1989). In vitro, GC act on epithelial cells only in the presence of a fibroblastic support (Kedinger et al. 1987). A similar conclusion could be drawn for the androgen-induced changes in vaginal or mammary epithelium with the very illustrative experiments of epithelial-mesenchymal hybrids between wild and receptor-deficient Tfm (testicular feminization mutation) mice (Cunha et al. 1983; Kratochwil, 1986).

Intestinal mucosa is an integrated system comprising epithelial and mesenchymal cells, and also the extracellular microenvironment. It is postulated that extracellular matrix components are involved in the dynamic tissue interactions that occur during development of the intestine as well as of other organs. Among these molecules are those that compose the basement membrane (BM) found at the epithelial-mesenchymal interface, like type IV collagen, laminin (LN) and heparan sulfate proteoglycan, (Timpl, 1989). In earlier experiments, we showed that the deposition of these components is progressive and that the elaboration of a complete BM precedes the expression of intestinal epithelial differentiation markers (Kedinger et al. 1989b). Furthermore, BM molecules originate from both tissue compartments, heparan sulfate proteoglycan being produced exclusively by the epithelial cells, type IV collagen mainly by the mesenchymal ones, and LN by both cell types (Simon-Assmann et al. 1988, 1989; Simo et al. 1992).

Laminin, a glycoprotein composed of three polypeptide chains, one A chain and two genetically different B chains (Bl and B2), is considered as a key molecule in the processes of morphogenesis and/or differentiation in various organs. During intestinal development, the important morphogenetic events, i.e. villus formation, crypt invagination and muscle coat development, are accompanied by noticeable quantitative and qualitative changes in the A/B1-B2 chain ratio and in A chain localization (Simo et al. 1991). Using chickμodent epithelial-mesenchymal hybrids on which A and B chains can be visualized with rodent-specific antibodies, it appeared that whilst B chains are produced by both tissue comportments, A chains are first deposited by epithelial cells only and later by mesenchymal cells as well (Simo et al. 1992). Furthermore, GC administered to postnatal rats have been shown to induce alterations in the intestinal biosynthetic pattern of extracellular matrix molecules, including increased laminin synthesis, in parallel with accelerated enzymatic maturation (Walsh et al. 1987).

On the basis of the data and observations described above, the question arises as to whether the effects of GC on intestinal tissue could be attributed to changes in the level or nature of LN molecules produced by the mesenchymal compartment.

In the present paper we approach this aspect in vitro by the study of neosynthesized laminin, of the relative proportion of the constituent laminin A and B chains and of the distribution of laminin molecules under the influence of dexamethasone (DX), a synthetic glucocorticoid, in mesenchymal-derived cell cultures, in cocultures of intestinal endodermal and mesenchymal cells, and in fetal intestinal expiants maintained in organ culture. The direct role of LN molecules on the epithelial enzymatic expression was analyzed by adding anti-LN antibodies to the cocultures. Both coculture and organ culture models have allowed us to demonstrate the inductive effects of GC on epithelial digestive enzymes (Simon-Assmann et al. 1982; Kedinger et al. 1987).

Reagents

Dulbecco’s modified Eagle’s medium (DMEM), CMRL 1066, Ham’s F10 and F12 media, trypsin and fetal calf serum were purchased from Gibco Laboratories and the SerXtend serum substitute from NEN products, collagenase (0.54 i.u. mg-1) from Boehringer-Mannheim. D-[6-3Hjglucosamine hydrochloride (specific activity 20–40 Ci mmol-1) was purchased from Amersham, dexamethasone from Sigma. HeparinSepharose 6B was a product of Pharmacia. Fluorescein-conjugated goat anti-rabbit γ-globulins and Affinipure goat anti-rat IgG were obtained from Nordic and from Jackson Immunoresearch Laboratories, respectively. Fluorescein-coupled sheep anti-mouse IgG antibodies were from the Institut Pasteur (Paris). Affinity-purified goat anti-rabbit IgG conjugated to alkaline phosphatase and the BCIP/NBT color development solution from Bio-Rad were used for revelation of the immunoblots.

Antibodies

Polyclonal anti-laminin antibodies were raised in rabbits immunized with laminin molecules extracted from the murine Englebreth-Holm-Swarm (EHS) tumour using NaCl-contain-ing buffers and purified by anion-exchange chromatography and by molecular-sieve chromatography as described by Timpl et al. (1979) and by Paulsson et al. (1987), respectively. The purification and determination of the specificity of the anti-laminin Ig were described earlier (Simo et al. 1991).

The rat monoclonal antibody, mAb 193, which reacts specifically with the A chain of laminin is a generous gift from Dr Sorokin (Sorokin et al. in preparation).

The mouse monoclonal anti-lactase (mAb YBB 2/61) has been kindly provided by Dr Quaroni (1983).

Animals

Fetuses from pregnant Wistar rats, whose gestation had been accurately timed, were used at 14, 15, 19 and 20 days of gestation; the day on which a vaginal plug was found was designated as day 0.

Organ cultures

Fetal rat (19-day) intestines were cut into expiants (2-4 mm) and cultured in organ culture dishes on stainless steel grids as described previously (Simon-Assmann et al. 1982) for 48 h in the presence or absence of 10−7 M dexamethasone (DX).

Cell cultures

Intestinal mesenchymal cells were derived from 14/15-day fetal rat intestines. The endoderms were separated mechanically from the mesenchymes after a 1 h incubation period at 37°C in CMRL 1066 synthetic medium containing 0.025% collagenase. Small mesenchymal microexplants were plated in culture dishes.

Skin fibroblastic cells were obtained by enzymatic dissociation of 20-day fetal rat skin dermis after incubation for 1 h at 37°C in Ca2+/Mg2+-free Ham’s F10 containing 0.01% trypsin, 0.01% collagenase. Isolated cells and small fragments were seeded in culture dishes.

The characteristics of both cell types were described in detail earlier (Kedinger et al. 1987). The culture medium used for the two types of cells was composed of a mixture (1:1) of DMEM and Ham’s F12 supplemented with 7.5% inactivated (60°C, 30 min) fetal calf serum (containing 2.5% SerXtend serum substitute) and gentamicin (200 ng ml-1). Subcultures were performed after trypsinization of confluent monolayers in Ca2+/Mg2+-free PBS containing 0.25% trypsin solution and 0.02% EDTA for 10–15 min at 37°C. Cells were seeded at a plating density of 3×104 cells cm-2. When necessary, 10−7 M DX was added to the medium at the beginning of confluency (3–4 days of culture). The cells (±DX) were harvested 5 days later.

Cocultures of intestinal endodermal cells with skin fibroblastic cells

Fetal rat (14/15 day) intestinal endoderms (separated from the mesenchymes as described above) cut into small fragments, were seeded on top of preformed confluent monolayers of skin fibroblastic cells (3–4 days). The use of fetal skin fibroblasts rather than intestinal mesenchymal cells in the cocultures is based on the fact that both cell types exhibit the same ability to support intestinal endodermal cell differentiation; intestinal mesenchymal cells, however, frequently retract the whole coculture (Kedinger et al. 1987). Where appropriate, 10−7 M DX was added to the culture medium at the beginning of the cocultures. Investigations were done 4–6 days later.

Purification of neosynthesized laminin molecules

Laminin molecules synthesized by the cells, cocultures or expiants were quantified after a 24 h period in the presence of [3H]glucosamine as previously described (Simo et al. 1991). Briefly, proteins were extracted from cells or organ pellets, 4 times consecutively in 4 M urea (60°C for 20 min) after a brief sonication. The lysates of each extraction were then clarified by centrifugation at 10,000 g for 15 min and pooled. The labeled medium from cell cultures and cocultures was dialysed against phosphate buffer. The neosynthesized laminin molecules were purified from the lysate or medium by affinity chromatography on a heparin-Sepharose 6B column. The level of labeled laminin was determined by counting the incorporated labeled precursor in the fractions eluted with 1 M NaCl.

Determination of the molecular form of the neosynthesized laminin

The material purified by affinity chromatography has been analyzed by immunoblotting. The constituent chains were separated by SDS-PAGE (on 4% to 15% gradient gels) after solubilisation of the proteins (100°C, 5 min in Laemmli sample buffer with 5% (v/v) /J-mercaptoethanol). The gels were calibrated with a standard of laminin-nidogen complex extracted from EHS tumor. After electrotransfer onto nitrocellulose, laminin was identified with the polyclonal-specific antibodies (8 μg ml-1 in 0.01 M Tris-HCl, pH 7.4, 0.155 M NaCl containing 0.3%, v/v, Tween 20) and goat antirabbit IgG conjugated to alkaline phosphatase as described earlier (Simo et al. 1991). The staining intensity of the bands was quantified by linear scanning with a densitometer (Shimadzu, Roucaire).

Immunocytochemistry

Sheets of cocultures detached mechanically from the culture dish were embedded in Tissue-Tek, frozen in Freon cooled in liquid nitrogen and stored at — 70°C. Cryostat sections (5–6 μm thick) were made and incubated in the presence of the specific polyclonal anti-laminin (1/1000), monoclonal anti-laminin A chains (1/100) or monoclonal anti-lactase (1/75) antibodies; the antigen-antibody complexes were identified with the appropriate fluorescein-conjugated second antibodies.

Detection of surface or intracellular LN was performed on endodermal or fibroblastic cells grown on glass coverslips in the presence or absence of DX. In this case, DX was added from the beginning of the culture for 3 or 4 days only, to avoid getting too densely packed cells. The cells used for the surface detection of laminin were postfixed in 2% paraformaldehyde for 30 min after the antigen-antibody reaction. For intracellular detection of LN, cells were prefixed with 1% paraformaldehyde for 10 min and permeabilized with 0.05% Triton ×100 for 10 min before incubation with the polyclonal or monoclonal antibody.

The cryosections or cells grown on coverslips were mounted in glycerol/PBS/phenylenediamine under coverslips and observed with an Axiophot Microscope (Zeiss).

Inhibition experiments

Anti-laminin polyclonal antibodies were added in endodermal-fibroblastic cocultures for 2–4 days at a concentration of 20 μg ml-1 culture medium. The sole incubation with the second fluorescein-conjugated goat anti-rabbit antibody on cryosections of the cocultures permitted the detection of LN molecules blocked by the antibodies.

Brush border enzyme activities

Sucrase, maltase-glucoamylase and lactase activities were measured in a fraction enriched in brush border membranes isolated from cultured intestinal fragments as described previously (Simon-Assmann et al. 1982). Enzymatic activities are expressed as milliunits per milligram of brush border proteins.

Laminin synthesis by mesenchyme-derived cells

To test the effect of GC on the synthesis of LN, intestinal mesenchymal cells or fetal skin fibroblastic cells cultured in the absence or presence of DX were incubated with [3H]glucosamine during 24 h prior to the cell harvesting. The radioactivity recovered after affinity chromatography of the material extracted from the cultured cells and of the corresponding culture medium indicated the level of neosynthesized LN.

Intestinal mesenchymal cells

The overall rate of LN synthesized by intestinal mesenchymal cells cultured in the absence or presence of DX was similar (Fig. 1A). However, DX induced an increase in the amount of neosynthesized cell-associated LN molecules (P<0.05; paired Student’s t-test; Fig. 1B). This effect of DX on the distribution of the newly synthesized molecules is illustrated in Fig. 2: although LN was found to be mainly secreted into the culture medium in both experimental conditions, DX increased the proportion of the cell-associated molecules from 12 to 22.5% (P < 0 001) with a reduction in the amount of LN molecules secreted. It must be noted that DX did not affect either the morphology of the cells (Fig. 3A,B) or cell proliferation as assessed by the number of cells recovered at the end of the experimental period (1.5×106versus 1.6×106 cells per dish without or with DX).

Fig. 1.

Laminin synthesis in monocultures of intestinal mesenchymal cells (IM), of fetal skin fibroblasts (SF) and in cocultures of intestinal endodermal cells and skin fibroblasts (E+SF) in the absence (—) and presence (+) of 10−7 M dexamethasone. The cells were incubated in medium containing [3H]glucosamine for 24 h, prior to purification of laminin by affinity chromatography. (A) Total amount of neosynthesized LN present in the cell pellet plus that released in the culture medium. (B)Neosynthesized LN associated with the cell monolayers or cocultures. Rates are expressed as disints min-1 (×10−3) [3H]glucosamine incorporated in laminin per dish (±S.E.M.). Numbers in parenthesis represent the number of experiments.

Fig. 1.

Laminin synthesis in monocultures of intestinal mesenchymal cells (IM), of fetal skin fibroblasts (SF) and in cocultures of intestinal endodermal cells and skin fibroblasts (E+SF) in the absence (—) and presence (+) of 10−7 M dexamethasone. The cells were incubated in medium containing [3H]glucosamine for 24 h, prior to purification of laminin by affinity chromatography. (A) Total amount of neosynthesized LN present in the cell pellet plus that released in the culture medium. (B)Neosynthesized LN associated with the cell monolayers or cocultures. Rates are expressed as disints min-1 (×10−3) [3H]glucosamine incorporated in laminin per dish (±S.E.M.). Numbers in parenthesis represent the number of experiments.

Fig. 2.

Neosynthesized LN recovered in the cell fraction (stippled bars) of the various cell populations and expressed as percentage of the total laminin synthesized; the open bars represent the percentage of LN molecules secreted in the culture medium. IM, intestinal mesenchymal cells; SF, fetal skin fibroblasts; E+SF, endodermal-fibroblastic cocultures in the absence (—) or presence (+) of 10−7 M dexamethasone.

Fig. 2.

Neosynthesized LN recovered in the cell fraction (stippled bars) of the various cell populations and expressed as percentage of the total laminin synthesized; the open bars represent the percentage of LN molecules secreted in the culture medium. IM, intestinal mesenchymal cells; SF, fetal skin fibroblasts; E+SF, endodermal-fibroblastic cocultures in the absence (—) or presence (+) of 10−7 M dexamethasone.

Fig. 3.

Phase-contrast micrographs of intestinal mesenchymal cells (A,B) and of fetal skin fibroblastic cells (C,D) cultured in the absence (-DX; A,C) or in the presence of 10−7 M dexamethasone for 5 days (+DX; B,D). ×80.

Fig. 3.

Phase-contrast micrographs of intestinal mesenchymal cells (A,B) and of fetal skin fibroblastic cells (C,D) cultured in the absence (-DX; A,C) or in the presence of 10−7 M dexamethasone for 5 days (+DX; B,D). ×80.

Fetal skin fibroblastic cells

The behavior of this cell type in response to DX was somewhat different from that of the intestinal mesenchymal cells. Indeed, the overall rate of neosynthesized LN in the presence of DX was decreased about 2fold as compared to the controls (P<0.01; Fig. 1A). Another difference from the intestinal mesenchymal cells was that the cellular morphology was strikingly altered by DX (Fig. 3C,D). The cells lost their typical elongated shape to acquire an epithelioid-like configuration, becoming round and flattened. In parallel the overall cell number decreased (2.02×106 to 1.55×106 cells per dish) as did the protein content per dish (from 0.63±0.04 mg to 0.39±0.09 mg). As a result, the specific LN synthetic activity of the skin fibroblastic cells cultured in the presence or absence of DX was not significantly different (396,555±72,803 versus 495,119±88,144 disints min-1 mg-1 protein, respectively). Yet the effect of DX on LN distribution was similar to that observed in the mesenchymal cell cultures. Indeed, in the control culture conditions, 6% of the newly synthesized molecules were associated with the cells, a percentage that increased to 25% under the influence of GC (P<0.01; Fig. 2), resulting in a significant increase in the cell-associated LN (P<0.02; Fig. 1B).

To analyze the molecular form of LN at the end of the culture period, the constituent chains of the affinity-purified molecules were separated on gels, transblotted and revealed with the polyclonal anti-LN antibody. The relative intensities of the bands were estimated with a scanning densitometer. In both mesenchyme-derived cell populations and regardless of the culture conditions (±DX), A and B1-B2 chains were found at, respectively, 400 and 210 kDa (Fig. 4), with an average of 5–10 A chains per 100 B chains. In both cell types, B chains were always visible in the cell pellet as well as in the culture medium. However, A chains were detected more sporadically on the immunoblots, mainly in the material purified from cell pellets. The sporadic presence of A chains in the cells may be due to the smaller amounts of LN molecules in the cell fraction as compared to the secreted one (Fig. 1), and to the threshold of detection of A chains, which are always found in very low amounts in intestine (Simo et al. 1991) as well as in other tissues (Kleinman et al. 1987; Olsen and Uitto, 1989; Timpl, 1989).

Fig. 4.

Immunoblot analysis of laminin isolated by affinity chromatography (A) from culture media or (B) from cell pellets and electrophoresed on SDS-polyacrylamide gels. Lanes a,b, intestinal mesenchymal monocultures; c,d, skin fibroblastic monocultures; and, e,f, endodermal-fibroblastic cocultures maintained in the absence (—) or presence (+) of 10−7 M dexamethasone. Arrowheads mark the positions of EHS laminin A (A: ∼400 kDa) and B (B: —210 kDa) chains and nidogen (N: —150 kDa). Note that A chains are always present at low levels; they are hardly detectable in most cell pellets.

Fig. 4.

Immunoblot analysis of laminin isolated by affinity chromatography (A) from culture media or (B) from cell pellets and electrophoresed on SDS-polyacrylamide gels. Lanes a,b, intestinal mesenchymal monocultures; c,d, skin fibroblastic monocultures; and, e,f, endodermal-fibroblastic cocultures maintained in the absence (—) or presence (+) of 10−7 M dexamethasone. Arrowheads mark the positions of EHS laminin A (A: ∼400 kDa) and B (B: —210 kDa) chains and nidogen (N: —150 kDa). Note that A chains are always present at low levels; they are hardly detectable in most cell pellets.

Laminin synthesis in cocultures of intestinal endodermal cells with skin fibroblastic cells Although it is impossible to attribute the respective part of the synthesis to the endodermal or the fibroblastic cell compartment, the overall effect of DX on LN synthesis in the cocultures was similar to that described in the monocultures of skin fibroblastic cells. Indeed, addition of DX resulted in a reduction of neosynthesized LN as compared to the control cocultures (Fig. 1A). However, the quantification of neosynthesized LN molecules in the cell pellets did not reveal differences between both control and DX-treated conditions (Fig. 1B), probably due to the participation of the endodermal cells. As in the monocultures of fibroblastic cells, DX induced a rise in the percentage of LN molecules associated with the cells (18 versus 12%, Fig. 2). Both A and B1-B2 chains were present in the cocultures in the presence or absence of DX (Fig. 4), with no noticeable changes in their relative ratio.

The efficiency of DX on brush border enzyme maturation was assessed immunocytochemically as described earlier (Kedinger et al. 1987, 1989b) by the stimulation of lactase and the induction of sucrase at the apical region of the epithelial cells (not illustrated).

Immunocytochemical detection of LN

LN deposits in the monocultures of intestinal mesenchymal cells or skin fibroblasts were analyzed on cells either after or without permeabilization, to distinguish intracellular from surface molecules (Fig. 5). In mesenchymal or fibroblastic cells cultured under control conditions, LN could be visualized within the whole cytoplasm of the cells as a granular fluorescent pattern (Fig. 5A). In the presence of DX, the staining pattern was strikingly different; LN was arranged as fibers within the cytoplasm (Fig. 5B). A dense network of LN fibers was obvious at the surface of DX-treated fibroblastic cells (Fig. 5D), whereas only minimal staining could be seen at the surface of the control cultures (Fig. 5C). Immunodetection of intracellular laminin A chains in permeabilized fibroblastic cells (Fig. 5E,F) revealed a shift from a faint granular staining pattern to a more filamentous network when DX was added. No surface labeling could be observed, regardless of the culture conditions, with anti-A-chains antibodies (not illustrated).

Fig. 5.

Immunocytochemical localization of laminin with a polyclonal anti-laminin antibody (A-D,G,H) and of laminin A chains (LNA) with a monoclonal anti A-chain antibody (E,F) on Triton X-100-permeabilized cells (A,B,E-H) or on the cell surface (C,D). (A-F) Monocultures of fetal skin fibroblastic cells, and (G,H) monocultures of intestinal endodermal cells cultured in the absence (-DX) or in the presence (+DX) of 10−7 M dexamethasone for 3–4 days. The greenish colour of nuclei in E,F is due to the use of an anti-rat secondary antibody to reveal the anti-LNA mAb raised in rat. ×400.

Fig. 5.

Immunocytochemical localization of laminin with a polyclonal anti-laminin antibody (A-D,G,H) and of laminin A chains (LNA) with a monoclonal anti A-chain antibody (E,F) on Triton X-100-permeabilized cells (A,B,E-H) or on the cell surface (C,D). (A-F) Monocultures of fetal skin fibroblastic cells, and (G,H) monocultures of intestinal endodermal cells cultured in the absence (-DX) or in the presence (+DX) of 10−7 M dexamethasone for 3–4 days. The greenish colour of nuclei in E,F is due to the use of an anti-rat secondary antibody to reveal the anti-LNA mAb raised in rat. ×400.

With intestinal endodermal cells, the presence of GC did not alter LN staining, which always exhibited an intracytoplasmic granular pattern (illustrated with the polyclonal anti-LN antibodies in Fig. 5G,H).

In the cocultures of intestinal endodermal cells and skin fibroblasts, the analysis of transverse sections of the associated cells at the end of the experimental period, i.e. 5-6 days±DX, revealed a continuous deposition of LN at the epithelial-fibroblastic interface, regardless of the presence of DX (Fig. 6A,C). Simultaneously, there was a noticeable decrease in LN molecules within the fibroblastic cell layer as compared to monocultures of fibroblastic cells (Fig. 6E). When the cells were cocultured for 2 days only, there was a striking difference between the control and DX-treated cocultures (Fig. 6B,D). In the former, there was no deposition of LN at the epithelial-fibroblastic interface, whereas the presence of DX induced an accelerated LN deposition visible at this stage as a continuous bright fluorescent line separating the two cell populations. Similarly, A chains could be visualized at the epithelial-mesenchymal interface only in the presence of DX after 2 days in cocultures (Fig. 6G,H). However, this hormonal effect was not obvious for all basement membrane molecules, since the deposition of heparan sulfate proteoglycan was not visible after 2 days in the presence of DX (not illustrated).

Fig. 6.

Immunocytochemical localization of laminin with a polyclonal anti-laminin antibody (A-E) and of laminin A chains (LNA) with a monoclonal anti-A chain antibody (G,H) on 5 gm cryosections of endodermal(e)-fibroblastic (f) cocultures (A-D,F-H) or on a section of fibroblastic cells cultured alone (E). The cocultures were examined after 5-6 (A,C) or 2 days (B,D,F-H) in the absence (-DX) or presence (+DX) of 10−7 M dexamethasone. (F) A coculture maintained for 2 days in the presence of DX and of 20 gg ml-7 polyclonal anti-laminin antibodies (+Ab); this section has been exposed to the secondary antibody (fluorescein-conjugated goat anti-rabbit IgG) alone. Arrows point to the epithelial-fibroblastic interface. ×320.

Fig. 6.

Immunocytochemical localization of laminin with a polyclonal anti-laminin antibody (A-E) and of laminin A chains (LNA) with a monoclonal anti-A chain antibody (G,H) on 5 gm cryosections of endodermal(e)-fibroblastic (f) cocultures (A-D,F-H) or on a section of fibroblastic cells cultured alone (E). The cocultures were examined after 5-6 (A,C) or 2 days (B,D,F-H) in the absence (-DX) or presence (+DX) of 10−7 M dexamethasone. (F) A coculture maintained for 2 days in the presence of DX and of 20 gg ml-7 polyclonal anti-laminin antibodies (+Ab); this section has been exposed to the secondary antibody (fluorescein-conjugated goat anti-rabbit IgG) alone. Arrows point to the epithelial-fibroblastic interface. ×320.

Inhibition experiments

To analyze the importance of LN deposition at the BM level on epithelial differentiation, 20 μg ml polyclonal anti-LN antibodies were added for 2-4 days to the cocultures in the presence of DX; these conditions were selected on the basis of data showing an accelerated deposition of LN and brush border enzyme expression under the influence of the hormone.

When the cocultured cells were incubated with antiLN antibodies, lactase, which was present in the DX-treated cocultures at the brush border level of the epithelial cells (Fig. 7A), could no longer be detected (Fig. 7B).

Fig. 7.

Immunocytochemical localization of lactase (an epithelial brush border digestive enzyme) with monoclonal antilactase antibodies on cryosections of DX-treated endodermal-fibroblastic cocultures examined after 2 days of culture in the presence of 10−7 M dexamethasone (+DX) without (A) or with (B) 20 μg ml-1 anti-laminin antibodies (+Ab). Arrows point to the apical pole of epithelial cells. A, ×200; B, ×125.

Fig. 7.

Immunocytochemical localization of lactase (an epithelial brush border digestive enzyme) with monoclonal antilactase antibodies on cryosections of DX-treated endodermal-fibroblastic cocultures examined after 2 days of culture in the presence of 10−7 M dexamethasone (+DX) without (A) or with (B) 20 μg ml-1 anti-laminin antibodies (+Ab). Arrows point to the apical pole of epithelial cells. A, ×200; B, ×125.

The binding of the antibody to LN in the cocultures was checked after 2 (Fig. 6F) and 4 days with a fluorescein-conjugated secondary antibody on cryosections. At 2 days, a clear specific binding of the antibodies at the epithelial-fibroblastic interface was obvious, being more intense at 4 days. The binding of laminin to the antibody did not interfere with the deposition of heparan sulfate proteoglycan, another BM molecule; indeed after 4 days in cocultures, despite the presence of anti-LN antibodies, its deposition at the epithelial-fibroblastic interface was obvious, as in nontreated cocultures (Kedinger et al. 1989b).

Influence of DX on LN synthesis by intestinal expiants in organ culture

Intestinal expiants taken from 19-day fetuses were cultured in the presence of [3H]glucosamine with or without DX for 48 h. This developmental stage was chosen as it corresponds: firstly, to a phase of low LN synthesis that follows the very active period of LN synthesis paralleling the villus upsurge at 17 days of gestation (Simo et al. 1991); and secondly, to a very sensitive phase of GC stimulation. Indeed, a significant stimulation or induction of brush border enzymes by DX treatment was obvious during the 48 h period (Simon-Assmann et al. 1982; and Table 1).

Table 1.

Effect of DX on brush border enzyme activities and laminin synthesis in 19-day fetal intestinal expiants maintained in organ culture for 48 h

Effect of DX on brush border enzyme activities and laminin synthesis in 19-day fetal intestinal expiants maintained in organ culture for 48 h
Effect of DX on brush border enzyme activities and laminin synthesis in 19-day fetal intestinal expiants maintained in organ culture for 48 h

In parallel, the level of neosynthesized LN was increased by 27% (P<0.05) under the influence of DX. It must be noted that the weights of the starting intestinal fragments were identical and that the tissue protein content at the end of the experimental period was exactly similar in both conditions.

The present data show that in a cell (co)-culture system glucocorticoids (1) induce qualitative changes in laminin molecules synthesized by mesenchyme-derived cells from either fetal intestinal or skin origin; and (2) lead to an accelerated organization of laminin at the epithelial-mesenchymal interface when epithelial cells are present. This process is accompanied by an accelerated maturation of the intestinal embryonic epithelial cell layer. The direct link between laminin deposition and epithelial differentiation is further demonstrated by the inhibition of lactase expression in the presence of anti-laminin antibodies in dexamethasone-treated cocultures. These studies provide data consistent with the hypothesis that the effect of GC on intestinal epithelial maturation involves an accelerated maturation of the extracellular environment.

The use of two types of mesenchyme-derived cell populations is based on earlier experiments demonstrating that fetal skin fibroblastic cells like intestinal mesenchymal cells had a permissive effect on the differentiation of intestinal epithelial cells (Haffen et al. 1989). We show here that the action of GC on laminin is similar in monocultures of either cell type as far as the specific synthetic activity, the proportion of secreted versus cell-associated LN molecules, and the cellular distribution are concerned. The difference between cell types concerning the GC-induced drop in total synthesis of LN in the skin fibroblasts, not obvious in the intestinal mesenchymal cells, seems to be attributable to an effect of GC on cell replication. This effect has been described as being dependent on the cell type and experimental conditions (for review see Durant et al. 1986).

Although dexamethasone does not induce significant quantitative changes in the level of LN specific synthetic activity or in the relative proportion of A/B chains in the fibroblastic cell populations, there is a dramatic alteration in the secretion and organization of these molecules. Indeed, there is a shift from secreted neosynthesized LN molecules to cell-associated insoluble molecules with the addition of DX. These changes are paralleled by an organization of LN molecules revealed immunocytochemically by a shift from a granular pattern to a filamentous network. It is noteworthy that a similar effect of GC on LN distribution has been described in melanoma cells (Lopes et al. 1991).

From our data, it is not possible to know if the effect of GC on LN molecule synthesis and/or assembly is direct or secondary to modifications of the overall matrix organization due to changes in the proportion of connective tissue molecules. Indeed, numerous papers devoted to the effect of GC on several cell types in vitro report various changes in the synthesis and/or assembly of ECM components. Although the effects described vary among cell types and, according to developmental stages and malignant states, the presence of GC led in general to a decrease in the expression of type I and III collagens and of hyaluronic acid (for reviews see Diegelmann, 1986; Durant et al. 1986; Oikarinen et al. 1986; Ohyama et al. 1990). Conversely, GC exhibit a stimulatory effect on heparan sulfate proteoglycan and mainly on fibronectin expression, maturation and polymerization, and possibly on the binding of the molecules to cell membrane (Furcht et al. 1979; Marceau et al. 1980; Oikarinen et al. 1986; Begemann et al. 1988; Dean et al. 1988; Kasinath et al. 1990). It is interesting to note that the alterations in extracellular matrix molecules described above in various cell culture systems in response to GC reproduce the changes observed during spontaneous or GC-induced intestinal maturation (Walsh et al. 1987; Bouziges et al. 1991).

The absence of a direct epithelial response to GC is suggested by two observations: (1) examination of LN and of the constituent chains in isolated intestinal endodermal cell cultures does not reveal any modification under the influence of DX, in contrast to the important changes observed in the fibroblastic cells; (2) the deposition in cocultures at the epithelial-fibroblastic interface of heparan sulfate proteoglycan molecules, known to be produced by the epithelial cell compartment (Simon-Assmann et al. 1989), is not accelerated by the addition of GC.

The present data confirm and support previous work that showed that cooperation of intestinal epithelial and mesenchyme-derived cells is required to bring about a state of competence for hormonal stimulation (for a review see Kedinger et al. 1989a). Moreover, in various epithelial-fibroblastic coculture systems, a clear correlation has been demonstrated between the development of structural differentiation characteristics, the deposition or rearrangement of ECM molecules and the competence of epithelial cells to respond to hormones (Clement et al. 1988a,b; Bouziges et al. 1989; Reich-mann et al. 1989; Kedinger et al. 1989a). There are several arguments suggesting that the assembly of BM molecules in these processes is controlled by mesenchymal cells (Bohnert et al. 1986; Delvoye et al. 1988). Our conclusions corroborate the evidence for a role for LN molecules on epithelial polarization and morphogenesis, already demonstrated in vitro in embryonic kidney and lung by using anti-LN antibodies (Klein et al. 1988; Schuger et al. 1990).

Finally, the 25% increase in LN synthesis observed in organ cultures of fetal intestine under the influence of DX and paralleling a dramatic enzymatic maturation, is in accordance with previous data reported by Walsh et al. (1987). By injecting DX to neonatal rats, these authors showed, among other changes in ECM molecules, an increased amount of LN molecules and of the corresponding mRNA. The apparent contradiction in the effect of GC on LN between the cell and organ systems raises the possibility that the modulation of the extracellular microenvironment by GC could involve an intermediary’ step through cytokines or other factors produced by cells that are absent in the cell culture systems.

In summary, our results demonstrate that glucocorticoids induce changes in the organization and cell association of laminin molecules in mesenchyme-derived cells, leading to their accelerated deposition at the basement membrane level when epithelial cells are present, paralleling the accelerated epithelial maturation. Taken together these data further strengthen the concept of mesenchymal involvement in the hormonally induced maturation of intestinal epithelium and argue in favour of a role for LN molecules in these processes.

P. Simo is the recipient of a fellowship from the Government of Cameroun.

We are indebted to Dr L. Sorokin (Max-Planck-Gesellschaft, Erlangen, Germany) and Dr A. Quaroni (Cornell University, Ithaca, NY, USA) for providing monoclonal antibodies against laminin A chains and brush border enzymes, respectively. We are grateful to E. Alexandre and C. Leberquier for excellent technical assistance, to C. Haffen for photographic processing and to L. Mathem for typing the manuscript. Financial support was given by the Institut National de la Santé et de la Recherche Médicale and by the Centre National de la Recherche Scientifique.

Begetnann
,
M.
,
Voss
,
B.
and
Paul
,
D.
(
1988
).
Hydrocortisone-induced accumulation of fibronectin mRNA and cell surface-associated fibronectin
.
J. Cancer Res. Clin. Oncol
114
,
477
481
.
Bohnert
,
A.
,
Hornung
,
J.
,
Mackenzie
,
I. C.
and
Fusenig
,
N. E.
(
1986
).
Epithelial-mesenchymal interactions control basement membrane production and differentiation in cultured and transplanted mouse kératinocytes
.
Cell Tissue Res
.
244
,
413
429
.
Bouziges
,
F.
,
Simon-Assmann
,
P.
,
Leberquier
,
C.
,
Haffen
,
K.
and
Kedinger
,
M.
(
1989
).
Glycosaminoglycan expression in intestinal epithelial/skin fibroblastic cell cocultures. Fibroblastic cell-mediated effects of glucocorticoids
.
J. Cell Sci
.
92
,
679
685
.
Bouziges
,
F.
,
Simon-Assmann
,
P.
,
Simo
.
P.
and
Kedinger
,
M.
(
1991
).
Changes in glycosaminoglycan expression in the rat developing intestine
.
Cell Biol. Int. Rep
.
15
,
97
106
.
Clement
,
B.
,
Guguen-Guillouzo
,
C.
,
Grimaud
,
J. A.
,
Rissel
,
M.
and
Guillouzo
,
A.
(
1988a
).
Effect of hydrocortisone on deposition of types I and FV collagen in primary culture of rat hepatocytes
.
Cell. Mol. Biol
.
34
,
449
460
.
Clement
,
B.
,
Rescan
,
P. Y.
,
Baffet
,
G.
,
Loreai
,
O.
,
Lehry
,
D.
,
Campion
,
J. P.
and
Guillouzo
,
A.
(
1988b
).
Hepatocytes may produce laminin in fibrotic liver and in primary culture
.
Hepatology
8
,
794
803
.
Cunha
,
G. R.
,
Chung
,
L. W. K.
,
Shannon
,
J. M.
,
Taguchi
,
O.
and
Fujii
,
H.
(
1983
).
Hormone-induced morphogenesis and growth: role of mesenchymal-epithelial interactions
.
Recent Progress in Hormone Research
39
,
559
598
.
Dean
,
D. C.
,
Newby
,
R. F.
and
Bourgeois
,
S.
(
1988
).
Regulation of fibronectin biosynthesis by dexamethasone, transforming growth factor ft, and cAMP in human cell lines
.
J. Cell Biol
.
106
,
21592170
.
Delvoye
,
P.
,
Pierard
,
D.
,
Noel
,
A.
,
Nusgens
,
B.
,
Foidart
,
J. M.
and
Lapiere
,
C. M.
(
1988
).
Fibroblasts induce the assembly of the macromolecules of the basement membrane
.
J. Invest. Dermatol
.
90
,
276
280
.
Diegelmann
,
R. F.
(
1986
).
Synthesis of extracellular matrix components by cultured hepatocytes
.
In Research in Isolated and Cultured Hepatocytes
(ed.
A.
Guillouzo
and
C.
Guguen-Guillouzo
), pp.
209
224
.
John Libbey Eurotext Ltdμnserm, Montrouge
,
France
.
Durant
,
S.
,
Duval
,
D.
and
Homo-Delarche
,
F.
(
1986
).
Factors involved in the control of fibroblast proliferation by glucocorticoids: a review
.
Endocr. Rev
.
7
,
254
269
.
Foltzer-Jourdalnne
,
C.
,
Kedinger
,
M.
and
Raul
,
F.
(
1989
).
Perinatal expression of brush border hydrolases in rat colon: hormonal and tissue regulations
.
Am. J. Physiol
.
257
,
G496
G503
.
Furcht
,
L. T.
,
Mosher
,
D. F.
,
Wendelschafer-Crabb
,
G.
and
Foidart
,
J. M.
(
1979
).
Reversal by glucocorticoid hormones of the loss of a fibronectin and procollagen matrix around transformed human cells
.
Cancer Res
.
39
,
2077
2083
.
Haffen
,
K.
,
Kedinger
,
M.
and
Simon-Assmann
,
P.
(
1989
).
Cell contact dependent regulation of enterocytic differentiation
.
In Human Gastrointestinal Development
(ed.
E.
Lebenthal
), pp.
1939
.
New York
:
Raven Press
.
Henning
,
S. J.
(
1987
).
Functional development of the gastrointestinal tract
.
In Physiology of the Gastrointestinal Tract, 2nd edn
(ed.
L.R.
Johnson
), pp.
285
300
.
New York
:
Raven Press
.
Kasinath
,
B. S.
,
Singh
,
A. K.
,
Kanwar
,
Y. S.
and
Lewis
,
E. J.
(
1990
).
Dexamethasone increases heparan sulfate proteoglycan core protein content of glomerular epithelial cells
.
J. Lab. Clin. Med
.
115
(
2
),
196
202
.
Kedinger
,
M.
,
Bouziges
,
F.
,
Simon-Assmann
,
P.
and
Haffen
,
K.
(
1989b
).
Influence of cell interactions on intestinal brush border enzyme expression
.
In Highlights of Modern Biochemistry
, vol.
2
, (ed.
A.
Kotyk
et al. ), pp.
1103
1112
. VSP International Science Publishers, Zeist.
Kedinger
,
M.
,
Haffen
,
K.
and
Simon-Assmann
,
P.
(
1986
).
Control mechanisms in the ontogenesis of villus cells
.
In Molecular and Cellular Basis of Digestion
(ed.
P.
Desnuelle
,
H.
Sjostrom
,
O.
Noren
), pp.
315
326
.
Amsterdam
:
Elsevier Science Publishers B.V
.
Kedinger
,
M.
,
Simon-Assmann
,
P.
,
Alexandre
,
E.
and
Haffen
,
K.
(
1987
).
Importance of a fibroblastic support for in vitro differentiation of intestinal endodermal cells and for their response to glucocorticoids
.
Cell Differ
.
20
,
171
182
.
Kedinger
,
M.
,
Simon-Assmann
,
P.
,
Bouziges
,
F.
,
Simo
,
P.
and
Haffen
,
K.
(
1989a
).
Mesenchyme-mediation of glucocorticoid effects on the expression of epithelial cell markers
.
Exp. Clin. Endocrinol
.
8
,
119
135
.
Klein
,
G.
,
Langegger
,
M.
,
Tlmpl
,
R.
and
Ekblom
,
P.
(
1988
).
Role of laminin a chain in the development of epithelial cell polarity
.
Cell
55
,
331
341
.
Kleinman
,
H. K.
,
Ebihara
,
I.
,
Killen
,
P. D.
,
Sasaki
,
M.
,
Cannon
,
F. B.
,
Yamada
,
Y.
and
Martin
,
G. R.
(
1987
).
Genes of basement membrane proteins are coordinately expressed in differentiating F9 cells but not in normal adult murine tissues
.
Develop. Biol
.
122
,
373
378
.
Kratochwil
,
K.
(
1986
).
Tissue combination and organ culture studies in the development of the embryonic mammary gland
.
In Developmental Biology
, vol.
4
(ed.
R.B.L.
Gwatkin
), pp.
315
333
.
Plenum Publishing Corporation
,
New York
.
Lacroix
,
B.
,
Kedinger
,
M.
,
Simon-Assmann
,
P.
and
Haffen
,
K.
(
1985
).
Enzymatic response to glucocorticoids of the chick intestinal endoderm associated with various mesenchymal cell types
.
Biol. Cell
54
,
235
240
.
Lopes
,
M. T. P.
,
Sonohara
,
S.
,
Chammas
,
R.
and
Brentanl
,
M. M.
(
1991
).
Effects of steroids on laminin-binding integrins in a human melanoma cell line
.
Int. J. Cancer
48
,
73
80
.
Marceau
,
N.
,
Goyette
,
R.
,
Valet
,
J. P.
and
Deschenes
,
J.
(
1980
).
The effect of dexamethasone on formation of a fibronectin extracellular matrix by rat hepatocytes in vitro
.
Exptl Cell Res
.
125
,
497
502
.
Ohyama
,
K.
,
Seyer
,
J. M.
,
Raghow
,
R.
and
Kang
,
A. H.
(
1990
).
Extracellular matrix phenotype of rat mesangial cells in culture: biosynthesis of collagen Type-I, Type-Ill, Type-FV, and Type-V and a low molecular weight collagenous component and their regulation by dexamethasone
.
J. Lab. Clin. Med
.
116
,
219
227
.
Oikarinen
,
A. L
,
Uitto
,
J.
and
Oikarinen
,
J.
(
1986
).
Glucocorticoid action on connective tissue: from molecular mechanisms to clinical practice
.
Med. Biol
.
64
,
221
230
.
Olsen
,
D. R.
and
Uitto
,
J.
(
1989
).
Differential expression of type FV procollagen and laminin genes by fetal vs adult skin fibroblasts in culture: determination of subunit mRNA steady-state levels
.
J. Invest. Dermatol
.
93
,
127
131
.
Paulsson
,
M.
,
Aumailley
,
M.
,
Deutzmann
,
R.
,
Tlmpl
,
R.
,
Beck
,
K.
and
Engel
,
J.
(
1987
).
Laminin-nidogen complex, extraction with chelating agents and structural characterization
.
Eur. J. Biochem
.
166
,
11
19
.
Quaroni
,
A.
(
1983
).
Use of monoclonal antibodies in the study of intestinal structure and function
.
In Brush Border Membranes, vol. 95
, pp.
113
131
.
London
:
Pitman Books Ltd, Ciba Foundation Symposium
.
Reichmann
,
E.
,
Ball
,
R.
,
Groner
,
B.
and
Friis
,
R. R.
(
1989
).
New mammary epithelial and fibroblastic cell clones in coculture form structures competent to differentiate functionally
.
J. Cell Biol
.
108
,
1127
1138
.
Schuger
,
L.
,
Oshea
,
K. S.
,
Nelson
,
B. B.
and
Varanl
,
J.
(
1990
).
Organotypic arrangement of mouse embryonic lung cells on a basement membrane extract. Involvement of laminin
.
Development
110
,
1091
1099
.
Simo
,
P.
,
Bouziges
,
F.
,
Lissitzky
,
J. C.
,
Sorokin
,
L.
,
Kedinger
,
M.
and
Simon-Assmann
,
P.
(
1992
).
Dual and asynchronous deposition of laminin chains at the epithelial-mesenchymal interface in the gut
.
Gastroenterology (in press)
.
Simo
,
P.
,
Simon-Assmann
,
P.
,
Bouziges
,
F.
,
Leberquier
,
C.
,
Kedinger
,
M.
,
Ekblom
,
P.
and
Sorokin
,
L.
(
1991
).
Changes in the expression of laminin during intestinal development
.
Development
112
,
477
487
.
Simon-Assmann
,
P.
,
Bouziges
,
F.
,
Arnold
,
C.
,
Haffen
,
K.
and
Kedinger
,
M.
(
1988
).
Epithelial-mesenchymal interactions in the Dexamethasone effect on laminin in cocultures 171 production of basement membrane components in the gut
.
Development
102
,
339
347
.
Simon-Assmann
,
P.
,
Bouziges
,
F.
,
Vigny
,
M.
and
Kedinger
,
M.
(
1989
).
Origin and deposition of basement membrane heparan sulfate proteoglycan in the developing intestine
.
J. Cell Biol
.
109
,
1837
1848
.
Simon-Assmann
,
P. M.
,
Kedinger
,
M.
,
Grenier
,
J. F.
and
Haffen
,
K.
(
1982
).
Control of brush border enzymes by dexamethasone in the fetal rat intestine cultured in vitro
.
J. Pediatr. Gastroenterol. Nutr
.
1
,
257
265
.
Timpl
,
R.
(
1989
).
Structure and biological activity of basement membrane proteins
.
Ear. J. Biochem
.
180
,
487
502
.
Timpl
,
R.
,
Rohde
,
H.
,
Gehron-Robey
,
P.
,
Rennard
,
S. I.
,
Foidart
,
J.-M.
and
Martin
,
G. R.
(
1979
).
Laminin: a glycoprotein from basement membranes
.
J. Biol. Chem
.
254
,
9933
9937
.
Walsh
,
M. J.
,
Leleiko
,
N. S.
and
Sterling
,
K. M.
(
1987
).
Regulation of type I, III and IV procollagen mRNA synthesis in glucocorticoid-mediated intestinal development
.
J. Biol. Chem
.
262
,
10
,81410,818.