Development of Fetal Monolayer Culture Rat Intestine in Organ and

Maturation and differentiation of intestinal epithelial cells was demonstrated in segments of fetal rat small intestine, maintained for more than a month in suspension organ culture, by ultrastructural, biochemical, and immunological criteria. Over a 5-7 d period, fragments of fetal intestine evolved into globular structures covered with a single columnar epithelium ultrastructurally similar to suckling villus cells. Loose mesenchymal cells, cellular debris, and collagen were present inside the structures. After 6 d in culture, goblet cells, not present in the fetal intestine at day 18, were numerous and well developed. Intestinal endocrine cells were also observed. Immunofluorescence studies employing monoclonal antibodies specific for villus and crypt cells in vivo, and various enzyme assays, have demonstrated a level of differentiation and maturation of the cultured epithelial cells similar but not identical to that of suckling intestinal mucosa in vivo. Crypts and crypt cell markers were not observed in the the cultures. Addition of glucocorticoids to the culture medium resulted in the induction of sucrase-isomaltase but failed to promote most of the functional changes characteristic of the intestinal epithelium at weaning in vivo. Epithelial cells were identified in explants derived from the organ cultures by their specific expression of intestinal cytokeratin. Differentiation-specific markers, present in the epithelial cells in primary cultures, were lost upon selection and subculturing of pure epithelial cell populations. These results suggest a requirement for mesenchymal and/or extracellular matrix components in the maintenance of the differentiated state of the epithelial cells. The fetal intestinal organ cultures described here present significant advantages over traditional organ and monolayer culture techniques for the study of the cellular and molecular interactions involved in the development and differentiation of the intestinal epithelium. Much effort in recent years has been devoted to the development of in vitro culture systems which would allow a detailed investigation of the properties of the intestinal epithelial cells under relatively simple and well-defined experimental conditions, Various techniques have been described for the organ culture of adult intestinal fragments or explants (13, 34), but are limited in use by the rapid necrosis and degeneration of the epithelium. Better tissue preservation and longer survival times have been reported for rat and human colon (1) and fetal or embryonic rat (5) and chick (2-4) small intestine. A different approach, pursued successfully in a number of laboratories, is represented by monolayer cultures of epithelioid cells derived from rat small (19, 24-27) and large (37) intestine, and from human benign tumors (10), but these cells have not been shown to perform typical differentiated functions of villus cells in vivo. Another approach to the problem of culturing differentiated intestinal epithelial cells is represented by the long-term suspension culture of fetal rat intestine (29, 35). Fetal intestinal organ cultures (FIOC) ~ differ from traditional organ culture systems in that the original architecture of the intestinal mucosa is lost within the first few days in culture, but they allow maintenance of morphologically and functionally stable differentiated intestinal cells in chemically defined medium for long periods of time. Normal development and differentiation in many epithelial systems appear to depend upon the interaction between epithelium and mesenchyme. In the case 1 Abbreviations used in thispaper: FCS, fetal calf serum; FIOC, fetal intestinal organ cultures; TBS-BSA, I mM CaCb, 5 mM KCI, 0.5 mM MgCI~, 0.136 M NaCI, 0.7 mM No2 HPO4, 25 mM Tris, 0.05% (wt/vol) NAN3, 0.2% (wt/vol) BSA, pH 7.4. THE JOURNAL OF CELL BfOLOGY • VOLUMe 100 MAY 1985 1611 1622 1611 © The Rockefeller University Press 0021-9525185/05[1611/12 $1.00 on A uust 7, 2017 jcb.rress.org D ow nladed fom of the intestine, experiments combining isolated epithelioid and fibroblastic cell populations, followed by transplantation into syngeneic hosts, have demonstrated the importance of epithelial-mesenchymal interactions for intestinal cell differentiation (1 l, 14, 20, 22). These interactions appear to be sufficiently preserved in the suspension cultures of fetal rat intestine, since normal differentiated functions develop and are expressed with time in culture. In this paper, an ultrastructural, immunological, and biochemical evaluation of the differentiated state of the intestinal epithelial cells in the FIOC cultures is reported. Starting from a relatively immature, stratified epithelium of fetal rat intestine at day 18 of gestation, the epithelial cells in culture achieve a level of differentiation comparable but not identical to that of villus cells in suckling intestinal mucosa. In contrast, epithelial monolayer cultures derived from FIOC explants have been found unable to express differentiated functions after subculturing and establishment of pure epithelial populations. These results are in accordance with previous studies suggesting a fundamental role for the intestinal mesenchyme, or mesenchyme-derived products, in the induction of intestinal cell differentiation. MATERIALS AND METHODS Materials: Sprague-Dawley rats (CD strain, of either sex, weighing 170175 g) and timed pregnant Sprague-Dawley rats (CD strain) were obtained from Charles River Breeding Laboratories, Inc. (Wilmington, MA). Dulbecco's modified Eagle's medium with 4.5 g/l glucose (DME), Roswell Park Memorial Institute 1640 medium, irradiated fetal calf serum (FCS), penicillin-streptomycin mixture, and trypsin (2.5% in Hanks' balanced salt solution without calcium and magnesium) were obtained from M.A. Bioproducts (Walkersville, MD). Tris(hydroxymethyl)aminomethane (Tris), HEPES, phenylmethylsulfonyl fluoride, aprotinin, leupeptin, and antipain were obtained from Sigma Chemical Co. (St. Louis, MO). Membrane Purification: The luminal (brush border) membrane of intestinal epithelial ceils was purified from adult rat jejunum and from FIOC by the method of Kessler et al. (15). FIOC were washed three times with phosphate-buffered saline (PBS) before homogenization. A mixture of proteasc inhibitors (1 mM phenylmethylsulfonyl fluoride, 50 #g/ml leupeptin, 50 t~g/ ml antipain, 0.1 mg/ml aprotinin) was added to all buffers and solutions used for homogenization and membrane purification. Immunoprecipitation and Characterization of Membrane Antigens: Membrane proteins of FIOC were labeled metabolically with [3H]lysine, and brush border membranes purified from adult rat jejunum were radiolabeled by reductive akylation with [~4C]folmaldehyde and sodium cyanoborohydride as described in the previous paper (30). Solubilization of membrane proteins, binding to monoclonal antibodiesSepharose 4B beads, SDS slab gel electrophoresis, two-dimensional slab gel electrophoresis, and visualization of radioactive bands by fluorography were performed as described in the previous paper (30). Immunofluorescence Staining: Immunofluorescence staining of frozen sections of rat small intestine and FIOC was done using the double antibody fluorescence technique (30). Immunofluorescence staining of monolayer cultures and FIOC explants was by the biotin-avidin method. Cells cultured on glass coverslips were washed three times with PBS, incubated for l h at 4*C with 1% formaldehyde in 100 mM sodium phosphate buffer (pH 7.4), washed three times with PBS, incubated for 1 h at 4"C with 100 mM glycine in PBS (pH 7.4), washed two times with PBS + 0.2% (wt/vol) bovine serum albumin (BSA). All subsequent incubations were performed at room temp. The cells were incubated for 30 min with mouse serum (controls) or monoclonal antibodies (ascites fluid) diluted 1:100 in PBS + 0.2% BSA, washed three times with PBS, incubated 30 rain with biotinylated anti-mouse IgG (H+L) immunoglobulin (affinity purified, obtained from Vector Laboratories, Inc., Burlingame, CA), 15 zg/ml in PBS + 0.2% BSA, washed three times with PBS, incubated 30 min with fluorescein isothiocyanate-labeled Avidin DCS (cell sorter grade, Vector Laboratories, Burlingame, CA), 10 ~g/ml in PBS + 0.2% BSA, washed three times with PBS, counterstained for 30 s with 0.01% (wt/ vol) Evans blue in PBS, washed two times with PBS, mounted.in glyceroI-PBS 9:1, and viewed in a Nikon Optiphot microscope equipped with epifluorescence 1612 THE IOURNAL OF CI-tL BIOLOGY VOLUME 100, 1985 attachment. Cells to be stained with the monoclonal antibody BBC 3/48/2, specific for intestinal cytokeratin, after fxation with 1% formaldehyde and incubation with I00 mM glycine in PBS (pH 7.4) were washed twice with PBS and treated for 5 rain at -20"C with methanol, then for 3 rain at -20"C with acetone, washed with distilled water, and then with PBS. Following procedures were as described above. Fluorescence was excited with the output of an Osram HBO 100 lamp filtered with a Nikon B2 filter set (excitation filter 460-485 nm; dichroic mirror DM 510, eyepiece-side absorption filter 520-560 nm). Pictures were recorded with Kodak Ektachrome 400 film (Eastman Kodak Co., Rochester, NY). Black and white negatives of the color slides were obtained using Kodak Technical Pan Film 2415 or Panatomic-x film (Eastman Kodak Co.). The red fluorescence of the counterstain (Evans blue) was filtered out with a green filter (Tiffen #58). Enzyme Assays: Sucrase, maltase, lactase, trehalase, glucoamylase, alkaline phosphatase, aminopeptidase, and "y-glutamyltransfemse activities in brush border membranes, purified from fetal intestinal mucosa and FIOC by the method of Kessler et al. (15) were determined as previously described ( 12, 25, 28). Results were expressed as mU of enzyme per milligram of protein; enzyme units were defined in all cases as the amount of enzyme transforming 1 umol of substrate per minute. Electron Microscopy: Samples of FIOC for transmission electron microscopy were rinsed twice with PBS, fixed at room temperature for 1 h in 8 mM CaCI2, 2% paraformaldehyde-2.5% glutaraldehyde buffered with 0.1 M sodium cacodylate at pH 7.4, rinsed with the same buffer and postfixed in 1.3% osmium tetroxide buffered with 0.1 M collidine at pH 6.8. The samples were stained en bloc in a 1% solution of uranyl acetate in the same buffer for 1 h at 4"C, washed with collidine buffer, dehydrated in a graded series of alcohols, and embedded in a mixture of Polybed 812 and Araldite 6005. Sections with gray interference colors were cut on an LKB Ultrome III ultramicrotome (LKB Instruments, Inc., Gaithersburg, MD), stained with uranyl acetate and lead citrate, and examined in a Philips 300 transmission electron microscope. Samples for scanning electron microscopy were rinsed with PBS, fixed with 2% paraformaldehyde-2.5% g]utaraldehyde, postfixed in 1.3% osmium tetroxide, and dehydrated in a graded series of alcohols as described above. The samples were dried in a Polaron E 3000 Critical Point Drying Apparatus (Polaron Instruments Inc., Hatfield, PA) using CO2 as the transition fluid, attached to an aluminum stub using conductive paint, coated with gold using a Polaron E 5000 sputter coater, and examined in an AMRAY 1000 transmission electron microscope at 20 kV potential. GeneraI CelI Culture Conditions: Monolayerculturesweregrown in plastic petri dishes (Lux, purchased from Miles Scientific Div., Naperville, IL) at 37"C in an atmosphere of 94% air, 6% CO2. Before being used for cell culture, glass coverslips were boiled in 50% nitric acid, rinsed extensively with distilled water, and autoclaved. Culture of the IEC-6, IEC17, and IEC18 cells, selection and subculturing of epithelioid cells with cloning cylinders (0.8-cm diameter), and other general cell culture procedures were as previously described (25). Growth rates (population doubling times), saturation densities, plating efficiencies, and ability of the cells to establish colonies in soft agar were determined as in our previous work (25). Monoclonal Antibodies: The mouse monoclonal antibodies to intestinal membrane proteins and intestinal cytokeratin used in this work have been prepared and characterized as described elsewhere (see Table II1 for list of antigen specificity of the antibodies): YBB 3/10, YBB 1/27, and CC 4/80 (30); BB 3/34 and BB 5/8 (12); BB 4/33, BB 4/35, BB 5/16, and BB 8/2 (28, 29, and Quaroni, A., submitted for publication); IEC 1/48, CC 4/11, CC 4/39, CC 4/91, BBC 1/35, BBC 3/88, BBC 3/90, BBC 3/91, YBB 1/57, YBB 2/61, YBB 2/54/3, and BBC 3/48/2 (Quaroni, A., submitted for publication). Binding of Monoclonal Antibodies to Intestinal Cells in Monolayer Cultures: Cells were cultured in 35-mm diameter tissue culture dishes, with the exception of secondary cultures of cells from FIOC explants which were seeded and tested in the wells of 24-well Costar plates (Data Packaging Co., Cambridge, MA): all volumes referred to below were halved for assays done in Costar plates. When the cultures were confluent, the medium was aspirated and the cells were washed three times with I mM CaCI2, 5 mM KC1, 0.5 mM MgCI2, 0.136 M NaCI, 0.7 mM Na2 HPO4, 25 mM Tris, 0.05% NaN3 (pH 7.4) containing 0.2% (wt/vol) BSA (TBS-BSA), incubated for 90 min at 4"C with l ml mouse serum (controls) or monoclonal antibodies (ascites fluid) diluted 1:500 in TBS-BSA, washed three times with 4 ml TBSBSA (5 min each time at 4"C), incubated for 90 min at 4"C with 1 ml TBS + 1% BSA containing l-l.5 x l0 s cpm of ~251-1abeled sheep anti-mouse lgG (H+L) F(ab')2 fragment (affinity purified, 7-9 uCi/~g, New England Nuclear, Boston, MA), washed three times with 4 ml TBS-BSA (5 min each time at 4"C), and then once with 4 ml of PBS. The cells were dissolved in 1 ml of 0. l N NaOH, neutralized with 1 M acetic acid, and counted with l0 ml Acquasol II (New England Nuclear, Boston, MA) in a Beckman LS 3800 liquid scintillation counter (Beckman Instruments, Inc., Pain Alto, CA). Parallel cultures on A uust 7, 2017 jcb.rress.org D ow nladed fom were used for cell counting, in triplicate. Data were expressed as cpm of ~2Sl bound per 106 cells. The background (cpm/106 cells bound to cultures incubated with mouse serum instead of monoclonal antibodies, diluted 1:500 in TBSBSA) was in no case greater than 2,000 cpm/106 cells, and was subtracted from all the data included in Table I11. Preparation of FIOC and FIOC Explants: FIOC were prepared and maintained in culture as previously described (29, 35) with minor modifications outlined here. Two different culture conditions were used. CULTURE IN SERUM-FREE MEDIUM: The culture medium was composed ofa l:l mixture of Dulbecco's modified Eagle medium (DME) with 4.5 g/l glucose, and Roswell Park Memorial Institute 1640 medium, supplemented with 10 mM HEPES, penicillin (50 U/ml), streptomycin (50 ug/ml), 2 mM glutamine, and 0.2% BSA (2x crystallized, Schwarz-Mann, Cambridge, MA). FIOC were cultured suspended in 15 ml complete culture medium/100-mm diameter tissue culture dishes. Half volume of medium was aspirated at 1to 2-d intervals, and replaced with fresh medium. CULTURE IN SERUM-CONTAINING MEDIUM: The culture medium was composed of a 1:1 mixture of DME and Roswell Park Memorial Institute 1640 medium, supplemented with 10 mM HEPES, 5% FCS, penicillin (50 U/ ml), streptomycin (50 t~g/ml), and 2 mM glutamine. FIOC were cultured suspended in 15 ml complete culture medium per 100-mm diameter tissue culture dishes. The surface of the dishes was covered with a Millipore LSWP090-25 filter (5.0 ~m pore size, Millipore/Continental Water Systems, Bedford, MA), sterilized by autoclaving. Half volume medium changes were performed at 1to 2-d intervals. The presence of 5% FCS in the medium was beneficial in promoting long-term preservation of the epithelial cells and all the results included in this paper were obtained with FIOC cultured in the presence of serum. However, maturation and differentiation of the epithelial cells were also observed in FIOC cultured for 2-3 wk in serum-free medium, as judged on the basis of ultrastructural studies, immunofluorescence staining with monoclonal antibodies, and enzyme assays. Hydrocortisone-supplemented media contained 250 ng/ml hydrocortisone2 l-succinate. Hydrocortisone (0.5 uM) added to the culture medium resulted in the induction of sucrase-isomaltase (Table 1) as well as a generalized increase in the specific activity (U/rag protein) of most other brush border enzymes tested, which may be an indication of a better functional preservation of the epithelial cells. The presence of FCS in the culture medium resulted in a marked tendency of the FIOC to attach to the surface of plastic tissue culture dishes, followed by cellular outgrowth (as described below for FIOC explants) and rapid degeneration of the fragments. It was therefore necessary to cover the surface of the dishes with a Millipore filter, to which the FIOC did not significantly attach. FIOC explants were prepared as follows: FIOC, in culture for 7 or more d, were suspended in serum-containing medium and added to 60or 100-ram diameter tissue culture dishes, or 35-mm diameter dishes containing acidwashed sterile circular glass coverslip. Most of the medium was aspirated, leaving just enough to cover the surface of the dishes and keep the FIOC fragments moist. The dishes were placed overnight in the incubator. The following day, the proper volume of serum-containing complete medium (2, 4, and 10 ml, for 35-, 60-, and 100-mm diameter dishes, respectively) was added to the dishes. Thereafter, FIOC explants were cultured using standard cell culture techniques. Medium changes were performed at 3-d intervals.