Search

The epithelium of normal human colon mucosa “in vivo” exhibits a gradual pattern of differentiation as undifferentiated stem cells from the base of the crypt of “lieberkuhn” rapidly divide, differentiate and migrate toward the free surface. The major differentiated cell type of the intestine observed are: absorptive cells displaying brush border, goblet cells containing mucous granules, Paneth and endocrine cells containing dense secretory granules. These different cell types are also found in the intestine of the 13-14 week old embryo.We present here morphological evidence showing that HT29, an adenocarcinoma of the human colon cell line, can differentiate into various cell types by changing the growth and culture conditions and mimic morphological changes found during development of the intestine in the human embryo.HT29 cells grown in tissue-culture dishes in DMEM and 10% FCS form at late confluence a multilayer of morphologically undifferentiated cell culture covered with irregular microvilli, and devoid of tight junctions (Figs 1-3).

Recent studies indicate that adhesion between cells is mediated by at least two different surface sites, one Ca2+-dependent and one Ca2+-independent. The first is highly sensitive to trypsin in the absence of Ca2+ but resistant in the presence of Ca2+, whereas inactivation of the second requires higher trypsin concentrations and Ca2+ has no effect on proteolysis.Tight junctions (TJ) in vertebrate epithelia provide a transepithelial permeability barrier, help to maintain cell polarity, and also are strongly adhesive. Trypsin and other proteases are often used in cell dissociation protocols but are known in several epithelia to induce TJ proliferation. Our previous study of LA7, a rat mammary epithelial cell line, showed as demonstrated by others that withdrawal of Ca2+ leads to disruption of TJ. The effect of trypsin on LA7 cells in the absence or presence of Ca2+ was studied to determine whether TJ are related to the Ca2+-dependent mechanism of cell adhesion in epithelia.

Human colon adenocarcinoma cell line HT29 was found to be a favorable system for the study of junctions assembly. Using scanning electron microscopy (SEM), thin sectioning (TEM) and freeze fracturing, we were able to demonstrate that HT29 cells grown to confluency have desmosomes but clearly lack tight or gap junctions. When treated with trypsin, cells were rounded up, made contacts, and within 15 min formed tight and gap junctions. The desmosomes, however, were split to hemi-desmosomes. These findings agree with results obtained for other experimental systems.In this work we have used cytochalasin B (CB), colchicine (C) and CB + C, drugs known to interfere with microfilaments and microtubules, in order to determine to what extent these structures are involved with the assembly of tight junctions in HT29 cells treated with trypsin. The drugs were added to confluent monolayered cells (CB:20 μg/ml - 1 h; C: 10-4 M - 3 h; CB:20 μg/ml + C: 10-4 M/ml - 3 h), then trypsin was added to a final concentration of 0. 25% for 15 min.

Tight junctions (TJ) at 3-cell intersections (3Cl) appear in freeze-fracture replicas as channels limited by 2 vertical strands in one cell's membrane, each representing the line of fusion with one of the other cells. Recent observations show in mammary cells exposed to a calcium chelator, retraction of microvilli from cell borders, starting at the 3Cl. Ca2+ depletion is also known to reduce transepithelial resistance. These observations led us to inquire whether 3Cl might be the sites most vulnerable to agents that increase junction permeability. We used the rat mammary adenocarcinoma cell line LA7 (from R. Dulbecco) to ask this question.LA7 monolayers were treated with reagents known to affect transepithelial permeability: Ca2− and Mg2+-free phosphate-buffered saline (PBS-CMF), the Ca2+ chelator EGTA (0.025%), colchicine (0.1mM), cytochalasin B (10 μg/ml), procaine (1 - 10 μg/ml), kinetin (0.4 μM), phalloidin (10-50 μM). Confluent LA7 on impermeable substrates form many domes (Fig I) as fluid transported basally is trapped between the monolayer and the plastic. Their presence is proof of TJ integrity and was the criterion used in this study. Functional effects of the treatments were measured by maintenance or change in dome numbers and morphological effects were followed by freezefracturing.