help button home button Endocrine Society Endocrinology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Garrouste, F. L.
Right arrow Articles by Pommier, G. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Garrouste, F. L.
Right arrow Articles by Pommier, G. J.
Endocrinology Vol. 138, No. 5 2021-2032
Copyright © 1997 by The Endocrine Society

Articles

Up-Regulation of Insulin/Insulin-Like Growth Factor-I Hybrid Receptors during Differentiation of HT29-D4 Human Colonic Carcinoma Cells1

Françoise L. Garrouste, Maryse M. Remacle-Bonnet, Maxime M.-A. Lehmann, Jacques L. Marvaldi and Gilbert J. Pommier

Unité Interactions entre Systèmes Protéiques et Différenciation dans la Cellule Tumorale, CNRS URA 1924, Faculté de Médecine, Marseille, France

Address all correspondence and requests for reprints to: Dr. Gilbert J. Pommier, URA CNRS 1924, Faculté de Médecine, 27 boulevard Jean Moulin, 13385 Marseille Cedex 5, France.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
To assess the autocrine function of insulin-like growth factor II (IGF-II) in the balance of proliferation and differentiation in HT29-D4 human colonic cancer cells, we studied the expression of IGF-I receptors (IGF-IR) and insulin receptors (IR) in relation to the state of cell differentiation. IGF-IR and IR were expressed in both undifferentiated and enterocyte-like differentiated HT29-D4 cells. IGF-IR had two isoforms with a 97-kDa and a 102-kDa ß-subunit. In addition, HT29-D4 cells expressed hybrid receptors (HR) formed by the association of two {alpha}ß heterodimers from both IR and IGF-IR. HR were evidenced through 1) inhibition of IGF-I binding by the B6 anti-IR antibody and 2) immunoprecipitation with the {alpha}-IR3 anti-IGF-IR antibody, which revealed an additional 95-kDa IR ß-subunit that disappeared when the heterotetrameric receptor was dissociated by disulfide reduction into {alpha}ß heterodimers before immunoprecipitation. Like IGF-IR, HR had a high affinity for IGF-I (Kd, ~1.5 nM), but did not bind insulin significantly; the latter interacted with the native IR only (Kd, ~4 nM). In the differentiated HT29-D4 cell monolayer, all receptor species were strongly polarized (>97%) toward the basolateral membrane. Moreover, HT29-D4 cell differentiation was accompanied by an approximately 2-fold increase in the number of IR, whereas the number of IGF-I-binding sites was unaltered. However, in differentiated HT29-D4 cells, ~55% of the latter were involved in HR vs. ~20% in undifferentiated HT29-D4 cells. Thus, HT29-D4 cell differentiation is characterized by an up-regulation (~3-fold) of the level of HR coupled to a down-regulation (~40%) of the level of native tetrameric IGF-IR. Alterations were induced early during the cell differentiation process, i.e. 5 days postconfluence, and remained unchanged for at least 21 days. Taken together, these results suggest that the IGF-II autocrine loop in HT29-D4 cells may trigger distinct signaling pathways if it activates native IGF-IR, which predominate in undifferentiated cells, or if it activates HR, which are up-regulated in differentiated cells.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
INSULIN-LIKE growth factors (IGF-I and IGF-II) are multifunctional regulatory peptides that share structural homology with proinsulin, but mediate primarily proliferative and/or differentiative effects depending on the target cell and the presence of other hormones and growth factors (1). At the cellular level, the type I IGF receptor (IGF-IR) mediates most, if not all, of the biological effects of both IGF-I and IGF-II. IGF-IR is structurally closely related to the insulin receptor (IR), which explains why IGF-I and insulin can cross-react with the opposite receptor at high ligand concentrations. Both receptors contain two extracellular {alpha}-subunits bearing the ligand-binding site and two ß-subunits bearing a tyrosine kinase activity that provides the signaling activity. The {alpha}- and ß-subunits are linked by disulfide bonds in a heterotetrameric ß-{alpha}-{alpha}-ß configuration (2). Analysis of IGF-IR and IR is further complicated by the formation of hybrid receptors (HR) from the association of an {alpha}ß heterodimer from an IGF-IR with an {alpha}ß heterodimer from an IR (2, 3, 4). In addition, subtypes of IGF-IR that can be distinguished by apparent molecular mass of multiple ß-subunit moieties and/or by a particularly elevated affinity for IGF-II or insulin have been reported in various cell types and tissues (4, 5, 6, 7, 8, 9). Moreover, the actions of IGFs are regulated by interactions with IGF-binding proteins (IGFBPs), which have been shown to positively or negatively modulate the bioavailability of IGFs to cell surface IGF-IR (for a review, see Ref.1).

Alterations in IGF signaling appear to be involved in neoplastic transformation and progression in various tissues (10, 11, 12). Human colorectal tumors and cancer cell lines have been reported to contain higher amounts of IGF-II messenger RNA and peptide than histologically normal colonic mucosa (13, 14, 15). Human normal intestine epithelial cells and colonic cancer cells have also been shown to express IGF-IR and IR (15, 16, 17, 18, 19, 20, 21, 22). In addition, secretion of IGFBPs (mainly IGFBP-2 and IGFBP-4) by human colonic cancer cells has recently been reported (15), and IGFBP-2 and IGFBP-3 serum levels were found to be perturbed in patients with colorectal carcinoma (23).

The HT29-D4 human colonic carcinoma cell line constitutes a unique inducible enterocyte-like cell differentiation model (24). HT29-D4 cells cultured under standard culture conditions, i.e. a medium containing 25 mM glucose (HT29-D4-GLU cells), are totally undifferentiated and proliferate in a continuous manner, generating cell multilayers. However, when 5 mM galactose is used in place of glucose in the culture medium, HT29-D4 cells differentiate in an enterocyte-like phenotype (HT29-D4-GAL cells), thus mimicking the colonocyte maturation during the cell migration along the crypt-villus axis in intestine in vivo. HT29-D4-GAL cells form highly polarized monolayers with mature junctional complexes, well organized microvilli, and functionally specialized apical and basolateral membrane domains that generate a transepithelial resistance (25).

We previously presented evidence that a regulatory IGF-II autocrine loop is involved in the control of the differentiation of HT29-D4 cells (26, 27). The maintenance of these cells in an undifferentiated phenotype appears to at least in part result from the incapacity of cells to use the regulatory potential of the secreted endogenous IGF-II because it is completely sequestered in the extracellular medium by various molecular species of IGFBPs, identified as IGFBP-2, -4, and -6 (28). Indeed, restoration of an IGF-II autocrine loop by adding suramin at a concentration that releases IGF-II from IGFBPs or adding des-(1, 2, 3)-IGF-I, a truncated IGF analog that does not bind to IGFBPs, induces HT29-D4 cells in an enterocyte-like differentiation pathway (26, 27, 28). Moreover, HT29-D4-GAL cells secrete IGFBPs differentially toward the apical or basolateral aspect; this could, in turn, contribute to modulating IGF-II bioavailability (29). On the contrary, autocrine expression of IGF-II has recently been reported to sustain proliferation in CaCo-2 cells, a human colon carcinoma cell line that spontaneously differentiates after reaching confluence in culture, whereas differentiation appears to require an attenuation of the mitogenic effects of IGF-II (30, 31, 32). Such a dual effect of IGF-II on cells from a comparable tissue is surprising because cell proliferation and differentiation are believed to be mutually exclusive in a wide variety of cell types.

To obtain further insight into the autocrine mechanism of action of IGF-II in colonic epithelial cells, we examined the IGF-IR/IR status in undifferentiated HT29-D4-GLU cells vs. differentiated HT29-D4-GAL cells. HT29-D4 cells expressed both IGF-IR and IR, and a proportion of them was involved in HR. During the enterocyte-like cell differentiation, the number of HR was up-regulated (~3-fold). At the same time, the level of native IR was increased, whereas that of native IGF-IR was down-regulated. In addition, all receptor species were strongly polarized (>97%) toward the basolateral membrane in HT29-D4-GAL cells. These results suggest a mechanism by which endogenous IGF-II may initially be mitogenic via IGF-IR and then may promote differentiation via HR in HT29-D4 cells.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Materials
Tissue culture flasks, multiwell plates, and Cyclopore transparent microporous polyester filters (25-mm diameter; 1.0-µm pore size) mounted on cell culture inserts were purchased from Falcon (Lincoln Park, NJ). DMEM, RPMI medium containing 20 mM HEPES (RPMI-HEPES), Hanks’ Balanced Salt Solution, FCS, and other cell culture reagents were purchased from Eurobio (Les Ulis, France). Ca2+-free Eagle’s medium was purchased from Life Technologies (Gaithersburg, MD). Human recombinant IGF-I and IGF-II were purchased from Bachem (Bubendorf, Switzerland). Bovine insulin was obtained from Sigma (L’Isle d’Abeau, France). The B6 monoclonal antibody that recognizes the extracellular domain of IR, and fluorescein isothiocyanate (FITC)-labeled goat antimouse Ig were purchased from Immunotech (Marseille, France). The {alpha}-IR3 monoclonal antibody highly specific for the {alpha}-subunit of IGF-IR and protein G-agarose were purchased from Oncogene Science (Uniondale, NY). [125I]IGF-I and [125I]insulin (~2000 Ci/mmol) were purchased from Amersham (Aylesbury, UK). Disuccinimidyl suberate (DSS) and sulfosuccinimidyl-6-(biotinamido) hexanoate (NHS-biotin) were obtained from Pierce Chemical Co. (Rockford, IL). Electrophoresis reagents and mol wt standards were obtained from Bio-Rad Laboratories (Richmond, CA). Nitrocellulose sheets (Hybond-C extra), horseradish-peroxidase (HRP)-coupled streptavidin and enhanced chemiluminescence (ECL) Western blotting detection reagents were purchased from Amersham. All other reagents were of analytical grade.

HT29-D4 cell culture conditions
The HT29-D4 human colon adenocarcinoma cell line was routinely cultured in DMEM containing 25 mM glucose and 10% FCS (HT29-D4-GLU cells) as previously described (27, 29). To induce enterocyte-like cell differentiation, we cultured HT29-D4 cells in glucose-free DMEM supplemented with 5 mM galactose and 10% dialyzed FCS (HT29-D4-GAL cells) (29). The medium was changed daily, as reported previously (29). Cells were grown at 37 C under 5% CO2 and monitored daily by light microscopy. It was routinely checked that HT29-D4-GAL cells cultured on plastic support were able to function as an intestinal epithelial cell barrier, i.e. to form domes about 5–6 days postconfluence (24). HT29-D4-GAL cell monolayers were also cultured on permeable supports, as previously reported (29). For this purpose, HT29-D4-GAL cells were plated at a density of 4.0 x 105 cells/cm2 on the inside of 25-mm diameter/1.0-µm pore size microporous polyester filters mounted in cell culture inserts and placed in a six-multiwell culture plate. Both compartments on each side of the cell layer were filled with glucose-free DMEM, with daily change of medium. The transepithelial resistance was routinely determined with a voltohmeter (Millipore, Bedford, MA). Filters without cells were used as controls, and only cell-covered filters giving a resistance of more than 250 ohm · cm2 were used. In addition, HT29-D4-GAL cell monolayers were periodically analyzed by electron microscopy to check the presence of a polarized morphological phenotype, as reported previously (24).

[125I]IGF-I and [125I]insulin competitive binding
Competitive binding was carried out at 4 C essentially as previously described (33). However, the access to the basolateral membrane domain in HT29-D4-GAL cells cultured on a plastic support required disrupting the tight junctions by pretreating the cells with Ca2+-free Eagle’s medium for 1 h at 37 C (34). HT29-D4-GLU cells and HT29-D4-GAL cells were further incubated at 4 C in binding medium (RPMI-HEPES containing 0.1% BSA, pH 7.2) with 0.15 nM [125I]IGF-I or [125I]insulin in the presence or absence of various concentrations of cold competitors, {alpha}-IR3, or B6 antibodies. For HT29-D4-GAL cell monolayers cultured on permeable supports, the labeled ligand was applied with or without cold competitors to either the upper (apical) or the lower (basolateral) side of the monolayer; the opposite side contained binding medium alone. At the end of incubation, i.e. 2 or 16 h (equilibrium conditions), depending on whether cells were cultured on plastic or permeable support, respectively, cells were washed three times with cold PBS containing 0.1% BSA. Cells cultured on plastic were then lysed with 1 ml 0.1 M NaOH, whereas cell-covered filters were directly cut from the inserts, and cell-associated radioactivity was counted in a {gamma}-counter. Nonspecific binding, determined as the radioactivity bound to the cells in the presence of 2.0 µM unlabeled IGF-I or insulin, was subtracted from total binding to obtain specific binding. Nonspecific binding represented about 2.5% and 5% of the total binding of [125I]IGF-I and [125I]insulin, respectively. Experimental points were estimated in triplicate, and replicate wells were used in each experiment to determine cell number. Binding data were analyzed by the EBDA/Ligand computer program (35).

[125I]IGF-I and [125I]insulin affinity cross-linking
HT29-D4-GLU cells and HT29-D4-GAL cells were first pretreated with Ca2+-free medium as described above. Cell monolayers were then incubated for 2 h at 4 C in binding medium containing about 500,000 cpm/well [125I]IGF-I or [125I]insulin in the presence or absence of various unlabeled peptides, {alpha}-IR3, or B6 antibodies. Cells were washed three times in binding medium without BSA at 4 C, then the radioligand was covalently cross-linked to the receptor by the addition of 0.5 mM DSS for 20 min at 4 C, and the reaction was quenched with 0.1 M Tris-HCl, pH 7.4, containing 1.0 mM EDTA. The cells were finally solubilized in the electrophoresis sample buffer (62 mM Tris-HCl, pH 6.8; 2% SDS; 10% glycerol; and 5% 2-mercaptoethanol), and the samples were boiled for 10 min and submitted to SDS-PAGE on 5% polyacrylamide slab gel. The gels were dried and autoradiographed using RX Fuji x-ray film in the presence of intensifying screens at -80 C for a minimum of 20 days.

Biotinylation of cell surface and immunoprecipitation by {alpha}-IR3
HT29-D4-GLU cells and HT29-D4-GAL cells, first pretreated with Ca2+-free medium as described above, were labeled with NHS-biotin (0.5 mg/ml in PBS for 20 min at 4 C) as previously described (36). Labeled cells were then lysed in RIPA buffer (20 mM Tris-HCl, pH 8.0; 1% Triton X-100; 200 mM NaCl; 0.5% BSA; and 1 mM EDTA) and a cocktail of protease inhibitors (aprotinin, leupeptin, iodoacetamin, and pepstatin, 1.0 µg/ml each) and 1 mM phenylmethylsulfonylfluoride. Extracted proteins were then immunoprecipitated at 4 C overnight with 5 µg {alpha}-IR3 and protein G-Sepharose. For reduction of receptors to {alpha}ß dimers before immunoprecipitation, lysates were incubated in the same buffer as that described above, but in the presence of 1 mM dithiothreitol (DTT) for 30 min at room temperature, and then 3 mM N-ethylmaleimide was added (8). Immunoprecipitated material bound to protein G-Sepharose was submitted to SDS-PAGE in 5% gel under the reducing conditions described in the preceding paragraph and then transferred to nitrocellulose membrane for 1 h at 100 V. Membranes were then blocked with 30 mM Tris-HCl, pH 7.5; 10% (wt/vol) glycerol; 1 M glucose; 0.4% BSA; and 0.1% Tween-20 for 2 h at room temperature. Biotinylated proteins were labeled with HRP-streptavidin in the same buffer for 45 min at room temperature, and HRP-streptavidin was revealed with the Amersham ECL System using the technique recommended by the manufacturer.

IGF-IR and IR expression by flow cytometry
Both HT29-D4-GLU cells and HT29-D4-GAL cells were recovered after an overnight incubation at 37 C in serum-free and calcium-free Eagle’s medium supplemented with 1% BSA. The cells were then washed, counted, and resuspended in DMEM containing 1% BSA at 1 x 106 cells/ml. {alpha}-IR3, B6, or irrelevant antibody [antiurokinase-type plasminogen activator (anti-u-PA)] was added at a concentration of 10 µg/ml for 90 min at 4 C. Cells were then washed twice with the medium described above and incubated with FITC-labeled goat antimouse Ig at a dilution of 1:150 for 30 min at 4 C. Cells were washed and fixed at 4 C in 2% paraformaldehyde, and then intensively washed and resuspended in PBS. Flow cytometry was performed on a FacSort (Becton Dickinson, San Jose, CA) flow cytometer. The samples were excited with a 15-mW argon laser at 488 nm. Green fluorescence was monitored through a 530/20-nm bandpass and a 488-nm laser-blocking filter. Debris and dead cells were excluded by gating on the basis of forward and side scatters. The relative fluorescence intensity of cells was compared with the fluorescence intensity of the same cells stained with the anti-u-PA irrelevant monoclonal antibody. The data were collected and analyzed using Cell Quest software (Becton Dickinson). Results were presented as the number of cells (10,000/analysis) vs. the log of fluorescence intensity.

Statistical methods
Results are expressed as the mean ± SD of the number of determinations indicated in the figures and tables. Significance was determined through two-tailed Student’s t test for unpaired data; P < 0.05 was taken as the level of significance.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Polarity of [125I]IGF-I and [125I]insulin binding in HT29-D4 cells
The distribution of binding sites for [125I]IGF-I and [125I]insulin between apical and basolateral membrane domains in the differentiated HT29-D4-GAL cell monolayer at 5 days postconfluence was first determined in cells grown on permeable filters. Such a culture system allowed the radiolabeled tracers to directly and selectively access either the apical or basolateral surface of the polarized cell. As shown in Fig. 1Go, binding of [125I]IGF-I or [125I]insulin was barely detectable if the radioligand was added in the apical compartment only. In contrast, binding of both [125I]IGF-I and [125I]insulin was detected when the radioligand was added to the basolateral compartment of the cell-covered filter (Fig. 1Go). The basolateral to apical binding ratio was 29:1 for [125I]IGF-I and 35:1 for [125I]insulin, i.e. more than 97%. Moreover, {alpha}-IR3 anti-IGF-IR and B6 anti-IR monoclonal antibodies were found to block by 75–85% the binding of [125I]IGF-I and [125I]insulin, respectively. This confirms the presence of true IGF-IR and IR in HT29-D4-GAL cells. Note also that the basolateral location of IGF-IR and IR in HT29-D4-GAL cells was observed with a similar ratio of polarity up to 21 days postconfluence.



View larger version (10K):
[in this window]
[in a new window]
  		Figure 1. Polarity of [125I]IGF-I and [125I]insulin binding in HT29-D4-GAL cell monolayers grown on permeable filters. Cell monolayers at 5 days postconfluence were washed and then incubated for 16 h at 4 C with 0.15 nM [125I]IGF-I ({blacksquare}) or [125I]insulin ({square}) alone or in the presence of 10 µg/ml {alpha}-IR3 () or B6 () antibody. All reagents were added to either the apical (A) or the basolateral (BL) side of the cell monolayer. Specific binding for each radioligand was determined as described in Materials and Methods. Values are the mean ± SD from three independent experiments performed in triplicate.

 
However, for practical reasons, further quantitative analysis of [125I]IGF-I and [125I]insulin binding to HT29-D4 cells requires cells grown on a conventional plastic support. Confluent undifferentiated HT29-D4-GLU cells grown on plastic specifically bound [125I]IGF-I and [125I]insulin (Fig. 2Go). Moreover, the amounts of binding of these radioligands did not change when cells were pretreated by Ca2+-free medium, which is known to disrupt polarity in cultured epithelial cells (34). In contrast, no significant [125I]IGF-I or [125I]insulin binding was detected in confluent HT29-D4-GAL cells cultured on a plastic support. Under these conditions, however, only the apical side of the polarized HT29-D4-GAL cell monolayer was accessible to the radioligands, as tight junctions prevented their access to the basolateral membrane domain. To allow the radioligands access to both apical and basolateral membrane domains, we pretreated HT29-D4-GAL cells with Ca2+-free medium to disrupt the tight junctions (34). [125I]IGF-I and [125I]insulin effectively bound to HT29-D4-GAL cells (Fig. 2Go). A similar amount of [125I]IGF-I was bound to Ca2+-free-treated HT29-D4-GAL cells and HT29-D4-GLU cells, whereas the amount of [125I]insulin was about 2-fold higher in HT29-D4-GAL cells than in HT29-D4-GLU cells (Fig. 2Go). Again, [125I]IGF-I and [125I]insulin binding was potently inhibited by {alpha}-IR3 and B6 antibody, respectively.



View larger version (16K):
[in this window]
[in a new window]
  		Figure 2. Polarity of [125I]IGF-I and [125I]insulin binding in HT29-D4-GLU cells and HT29-D4-GAL cells grown on plastic. Confluent HT29-D4-GLU cells or 5 days postconfluence HT29-D4-GAL cells grown on a conventional plastic support were preincubated for 1 h at 37 C with a standard medium (+Ca2+) or a Ca2+-free medium (-Ca2+) to disrupt tight junctions. [125I]IGF-I ({blacksquare}) or [125I]insulin ({square}) at 0.15 nM was then added for 2 h at 4 C, either alone or in the presence of 10 µg/ml {alpha}-IR3 () or B6 () antibody. Specific binding for each radioligand was determined as described in Materials and Methods. Values are the mean ± SD from three independent experiments performed in triplicate.

 
These results indicate that in differentiated HT29-D4-GAL cells, IGF-IR and IR are primarily localized in the basolateral membrane domain, independent of the culture support used. In contrast, no significant polarization of the receptor distribution existed in undifferentiated HT29-D4-GLU cells.

Analysis of [125I]IGF-I binding in HT29-D4 cells
Figure 3AGo shows a representative experiment of competition of unlabeled IGF-I, IGF-II, and insulin for [125I]IGF-I binding to either HT29-D4-GLU cells or HT29-D4-GAL cells. The same binding features were observed in both types of cells. [125I]IGF-I binding was inhibited by IGF-I and IGF-II with IC50 values of about 2 and about 10 nM, respectively, whereas insulin was a much weaker competitor; at 400 nM, insulin yielded approximately 35% and 60% inhibition in HT29-D4-GLU cells and HT29-D4-GAL cells, respectively. {alpha}-IR3 (400 nM) also inhibited [125I]IGF-I binding to both types of cells by about 75%. Ligand analysis of competitive inhibition for [125I]IGF-I binding by unlabeled IGF-I gave a linear Scatchard plot consistent with a single class of binding sites in both HT29-D4-GLU cells and HT29-D4-GAL cells (Kd, ~1.5 nM; binding capacity, ~25,000 sites/cell; Fig. 3BGo and Table 1Go). In addition, Ligand analysis of competition of unlabeled heterologous ligands, i.e. IGF-II and insulin, for [125I]IGF-I binding yielded a Kd of 6.0–8.0 nM for IGF-II whatever the state of HT29-D4 cell differentiation. We obtained Kd values of about 210 and about 95 nM for insulin in HT29-D4-GLU cells and HT29-D4-GAL cells, respectively (Table 1Go).



View larger version (18K):
[in this window]
[in a new window]
  		Figure 3. Binding of [125I]IGF-I in HT29-D4 cells in relation to the state of differentiation. A, Representative experiment of the binding of [125I]IGF-I (0.15 nM) to confluent HT29-D4-GLU cells (solid line, closed symbols) or 5-day postconfluence HT29-D4-GAL cells (dashed line, open symbols) measured in the presence of various concentrations of unlabeled IGF-I (• and {circ}), IGF-II ({blacksquare} and {square}), insulin ({blacktriangleup} and {triangleup}), or {alpha}-IR3 antibody ({diamondsuit} and {diamond}). HT29-D4-GAL cells were pretreated with a Ca2+-free medium as described in Materials and Methods. B, Scatchard representation of the data for the [125I]IGF-I binding displacement by unlabeled IGF-I analyzed by the Ligand program. The solid lines are the computer-generated best fits for a one-site binding model in HT29-D4-GLU cells (closed symbols) and HT29-D4-GAL cells (open symbols).

 

View this table:
[in this window]
[in a new window]
  		Table 1. Binding parameters for [125I]IGF-I and [125I]insulin in HT29-D4-GLU cells and HT29-D4-GAL cells

 
As a complementary approach, we cross-linked [125I]IGF-I to both HT29-D4-GLU cells and HT29-D4-GAL cells. SDS-PAGE under reducing conditions and then autoradiography revealed a labeled band at 135 kDa in HT29-D4 cells whatever the state of differentiation (Fig. 4Go, lanes a). The labeling of this band was totally blocked by 150 nM IGF-I (lanes b) or IGF-II (lanes c) and 5.0 µg/ml {alpha}-IR3 (lanes e) and was partially inhibited by an excess of 1.0 µM insulin (lanes d). Cross-linking of [125I]IGF-II to HT29-D4-GLU cells and HT29-D4-GAL cells gave similar results, except that {alpha}-IR3 was not an inhibitor (not shown), in agreement with the widely reported fact that this antibody does not inhibit IGF-II binding to IGF-IR at 4 C (37). In addition, no radioactive band migrating at about 260 kDa was detected even in overexposed autoradiographs, thus confirming previous reports that HT29-D4 cells do not express type II IGF receptors at the cell surface (33). We were also unable to detect complexes of [125I]IGF-I or [125I]IGF-II cross-linked to proteins migrating with an apparent mol wt between 25–50 kDa, suggesting the absence of HT29-D4 cell surface-associated IGFBPs.



View larger version (64K):
[in this window]
[in a new window]
  		Figure 4. Affinity cross-linking of [125I]IGF-I to HT29-D4 cells in relation to the state of differentiation. Confluent HT29-D4-GLU cells or 5-day postconfluence HT29-D4-GAL cells (pretreated with a Ca2+-free medium) were incubated for 2 h at 4 C with [125I]IGF-I alone (lanes a) and in the presence of 150 nM IGF-I (lanes b), 150 nM IGF-II (lanes c), 1.0 µM insulin (lanes d), or 5.0 µg/ml {alpha}-IR3 (lanes e). Cross-linking was performed with 0.5 mM DSS. The samples were solubilized, reduced, and analyzed by SDS-PAGE (5% gel), and then autoradiographed as described in Materials and Methods. Mol wt markers are indicated on the left.

 
Analysis of [125I]insulin binding in HT29-D4 cells
We next studied the competition of unlabeled insulin, IGF-I, and IGF-II for [125I]insulin binding (Fig. 5AGo). [125I]Insulin binding to HT29-D4-GLU cells and HT29-D4-GAL cells was inhibited by insulin with similar IC50 values of approximately 4 nM. The corresponding values for IGF-I and IGF-II were approximately 150 and 50 nM, respectively. In addition, the B6 antibody (400 nM) inhibited [125I]insulin binding to both types of cells by about 75% (Fig. 5AGo). [125I]Insulin displacement curves generated by unlabeled insulin gave linear Scatchard plots, demonstrating the presence of a single type of binding site in HT29-D4-GLU and HT29-D4-GAL cells (Fig. 5BGo). The Kd of IR for insulin calculated with the Ligand program was not significantly different (~4 nM) with respect to the state of HT29-D4 cell differentiation. However, HT29-D4-GAL cells expressed about 2-fold more IR than did HT29-D4-GLU cells (~27,500 vs. ~14,000 sites/cell; P < 0.005; Fig. 5BGo and Table 1Go). Table 1Go also shows Ligand analysis of inhibition of [125I]insulin binding by heterologous unlabeled IGF-I and IGF-II ligands. Whatever the state of HT29-D4 cell differentiation, binding of [125I]insulin was inhibited by IGF-II more potently than by IGF-I (Kd, ~20 vs. ~150 nM).



View larger version (19K):
[in this window]
[in a new window]
  		Figure 5. Binding of [125I]insulin to HT29-D4 cells in relation to the state of differentiation. A, Representative experiment on the binding of [125I]insulin (0.15 nM) to confluent HT29-D4-GLU cells (solid line, closed symbols) or 5-day postconfluence HT29-D4-GAL cells (dashed line, open symbols) measured in the presence of various concentrations of unlabeled insulin ({blacktriangleup} and {triangleup}), IGF-I (• and {circ}), IGF-II ({blacksquare} and {square}), or B6 antibody ({diamondsuit} and {diamond}). HT29-D4-GAL cells were pretreated with a Ca2+-free medium as described in Materials and Methods. B, Scatchard representation of the data for the [125I]insulin binding displacement by unlabeled insulin analyzed by the Ligand program. The solid lines are the computer-generated best fits for a one-site binding model in HT29-D4-GLU cells (closed symbols) and HT29-D4-GAL cells (open symbols).

 
Affinity cross-linking with [125I]insulin showed a band migrating at about 130 kDa under reducing conditions for both HT29-D4-GLU cells and HT29-D4-GAL cells (Fig. 6Go, lanes a). However, the labeling in HT29-D4-GAL cells was much higher than that in HT29-D4-GLU cells, which is consistent with the ligand binding studies reported above. The cross-linking of [125I]insulin was totally blocked by 150 nM insulin (Fig. 6Go, lanes b) or 5.0 µg/ml B6 antibody (Fig. 6Go, lanes e) and blocked slightly less by 1.0 µM IGF-II (Fig. 6Go, lanes c), but was not blocked by 5.0 µg/ml {alpha}-IR3 (Fig. 6Go, lanes d).



View larger version (76K):
[in this window]
[in a new window]
  		Figure 6. Affinity cross-linking of [125I]insulin to HT29-D4 cells. Confluent HT29-D4-GLU cells or 5-day postconfluence HT29-D4-GAL cells (pretreated with a Ca2+-free medium) were incubated for 2 h at 4 C with [125I]insulin alone (lanes a) and in the presence of 150 nM insulin (lanes b), 1.0 µM IGF-II (lanes c), 5.0 µg/ml {alpha}-IR3 (lanes d), or 5.0 µg/ml B6 (lanes e). Cross-linking was performed with 0.5 mM DSS. The samples were solubilized, reduced, and analyzed by SDS-PAGE (5% gel), and then autoradiographed as described in Materials and Methods. Mol wt markers are indicated on the left.

 
Flow cytometric analysis of IR and IGF-IR in HT29-D4 cells
To structurally study the expression of IGF-IR and IR at the surface of HT29-D4 cells, we performed a flow cytometric analysis with {alpha}-IR3 and B6 antibodies (Fig. 7Go and Table 2Go). HT29-D4 cells expressed both immunoreactive IGF-IR and IR regardless of their state of differentiation (Fig. 7Go). The mean fluorescence intensity with {alpha}-IR3 was 7-fold that of the anti-u-PA-negative control; with B6 antibody, it was 1.6- to 2.5-fold (Table 2Go). However, HT29-D4-GAL cells expressed significantly more immunoreactive IR on their surfaces than did HT29-D4-GLU cells. The specific mean fluorescence intensity with B6 antibody in HT29-D4-GAL cells was about 3-fold that in HT29-D4-GLU cells (7.2 vs. 2.6; P < 0.001; n = 5; Table 2Go). In contrast, the specific {alpha}-IR3 staining of HT29-D4-GAL cells was not significantly different from that of HT29-D4-GLU cells (Fig. 7Go and Table 2Go).



View larger version (22K):
[in this window]
[in a new window]
  		Figure 7. Flow cytometric analysis of binding of anti-IGF-IR ({alpha}-IR3) and anti-IR (B6) antibodies to HT29-D4 cells with respect to the state of differentiation. Single cell suspensions (1 x 106 cells) of HT29-D4-GLU cells (A) or 5-day postconfluence HT29-D4-GAL cells (B) were incubated with 10 µg/ml {alpha}-IR3, B6, or anti-u-PA (irrelevant) monoclonal antibody for 90 min. The cells were washed and then incubated with FITC-conjugated goat antimouse Ig for 30 min. Samples were washed and analyzed through a FacSort fluorescent cell analyzer. Cells were excited at 488 nm, and the fluorescence was monitored at 525 nm. A minimum of 10,000 cells was used in each analysis. Dashed line (IGF-IR), Fluorescence profile with {alpha}-IR3 antibody; solid line (IR), fluorescence profile with B6 antibody. The stippled curve is the nonspecific fluorescence obtained with the anti-u-PA irrelevant antibody. The data are representative of five independent experiments.

 

View this table:
[in this window]
[in a new window]
  		Table 2. Flow cytometric analysis of binding of {alpha}-IR3 and B6 antireceptor antibodies in HT29-D4 cells with respect to the state of cell differentiation

 
Finally, note that both the functional (ligand binding) and the structural (immunoreactivity) analyses of IGF-IR and IR were highly similar regardless of the time of culture of HT29-D4-GAL cells after confluence, i.e. at 5 (the experiments reported here), 15, or 21 days postconfluence.

Identification of insulin/IGF-I HR in HT29-D4 cells
Cross-inhibition of IGF-I and insulin binding by {alpha}-IR3 and B6 antibodies. {alpha}-IR3 and B6 monoclonal antibodies specifically altered the binding of the cognate ligand to IGF-IR and IR, respectively. Moreover, neither of these antibodies cross-reacted with the heterologous receptor (38). Thus, any inhibition of the binding of the noncognate ligand by an antireceptor antibody, e.g. IGF-I binding by B6 or insulin binding by {alpha}-IR3, can be interpreted as reflecting the presence of HR (3, 4, 5). [125I]Insulin or [125I]IGF-I was, therefore, incubated at 4 C with HT29-D4-GLU cells or HT29-D4-GAL cells in the presence of {alpha}-IR3 or B6 antibody. As expected, [125I]insulin binding to HT29-D4-GLU cells and HT29-D4-GAL cells was potently inhibited by B6 (~75%), whereas {alpha}-IR3 was unable to alter it significantly (Fig. 8Go). In HT29-D4 cells, the functional insulin-binding sites, i.e. those analyzed in the binding experiments described above, were, therefore, not involved in HR regardless of the state of cell differentiation. As also expected, the level of [125I]IGF-I binding to HT29-D4-GLU cells or HT29-D4-GAL cells was markedly inhibited by {alpha}-IR3 (~75%). However, Fig. 8Go (arrows) shows that approximately 20% of [125I]IGF-I binding in HT29-D4-GLU cells and approximately 55% in HT29-D4-GAL cells were inhibited by the B6 anti-IR antibody. This result suggests that as a rough estimate, about 20% of IGF-I-binding sites in HT29-D4-GLU cells and about 55% in HT29-D4-GAL cells are, in fact, involved in HR. Note also that the fraction of HR in HT29-D4-GAL cells did not further change throughout the postconfluent cell differentiation process, i.e. from days 5–21 postconfluence.



View larger version (21K):
[in this window]
[in a new window]
  		Figure 8. Cross-inhibition of [125I]IGF-I and [125I]insulin binding by {alpha}-IR3 and B6 antireceptor antibodies in HT29-D4 cells in relation to the state of differentiation. Confluent HT29-D4-GLU cells or 5-day postconfluence HT29-D4-GAL cells (pretreated with a Ca2+-free medium) were incubated for 2 h at 4 C with 0.15 nM [125I]IGF-I ({blacksquare}) or [125I]insulin ({square}) alone and in the presence of 10 µg/ml B6 () or {alpha}-IR3 ( ) antibody. Specific binding for each radioligand was determined as described in Materials and Methods. Values are the mean ± SD from five independent experiments performed in triplicate. The arrows denote inhibition of binding of the noncognate [125I]IGF-I ligand by the B6 anti-IR antibody.

 
Immunoprecipitation of HT29-D4 cell surface-biotinylated receptors by {alpha}-IR3. To demonstrate HR by another method, we biotinylated surface proteins of HT29-D4-GLU cells and HT29-D4-GAL cells and then subjected them to immunoprecipitation with {alpha}-IR3. Figure 9AGo, lane a, shows that three distinct proteins migrating at 135, 102, and 97 kDa were specifically immunoprecipitated from the HT29-D4-GLU cell lysate by {alpha}-IR3. The molecular mass of the 135-kDa protein is similar to that of the labeled band revealed by [125I]IGF-I cross-linking (Fig. 4Go), suggesting that it represented the {alpha}-subunit of IGF-IR. It was not possible to distinguish IR {alpha}-subunits that might be involved in HR, because IR and IGF-IR {alpha}-subunits did not differ enough in size (compare Fig. 4Go with Fig. 6Go). The 97- and 102-kDa proteins specifically immunoprecipitated by {alpha}-IR3 (Fig. 9AGo) were expected to correspond to ß-subunits labeled on the extracellular region of the protein. They can arise either from two isoforms of IGF-IR, as reported in the literature (4, 5, 6, 7, 8, 9) or from HR. As these two bands were still present when the HT29-D4-GLU cell lysate was treated with DTT to reduce {alpha}2ß2 heterotetramers into {alpha}ß heterodimers before immunoprecipitation by {alpha}-IR3 (Fig. 9AGo, lane b), they were ß-subunits of IGF-IR subtypes. {alpha}-IR3 immunoprecipitation experiments were similarly performed with HT29-D4-GAL cell lysates (Fig. 9BGo). The 135-kDa band ({alpha}-subunit) and the 97- and 102-kDa bands (ß-subunits) that appear to be associated with two IGF-IR isoforms in HT29-D4-GLU cells were still recovered in HT29-D4-GAL cells (Fig. 9BGo, lanes c and d). However, an additional 95-kDa protein was specifically immunoprecipitated by {alpha}-IR3 (Fig. 9BGo, lanes c and d). Note that the migration of this 95-kDa protein could not be confused with that of a nonspecifically precipitated protein observed in HT29-D4-GLU cells (Fig. 9BGo, compare lanes c and d with lane e). When the samples were treated with DTT before immunoprecipitation with {alpha}-IR3, the immunoprecipitation was slightly poorer. However, the additional 95-kDa band disappeared almost totally (Fig. 9BGo, lanes a and b), whereas the 97- and 102-kDa ß-subunits were still detectable at that time in the postreduction precipitate. Thus, the {alpha}ß half-receptor species containing the 95-kDa protein, probably another ß-subunit, did not interact with {alpha}-IR3. To determine the origin of this 95-kDa protein, we removed IR from the HT29-D4-GAL cell lysate by immunodepletion with B6 antibody (Fig. 9CGo). B6 treatment of the cell lysate before immunoprecipitation by {alpha}-IR3 nearly completely removed this band (Fig. 9CGo, lane a), which strongly suggests that it represented an IR-ß subunit. These experiments, therefore, confirmed that HT29-D4-GAL cells contain HR with an {alpha}ß heterodimer from IR, containing a 95-kDa ß-subunit, linked to an {alpha}ß heterodimer from IGF-IR containing a 97- or a 102-kDa ß-subunit.



View larger version (21K):
[in this window]
[in a new window]
  		Figure 9. {alpha}-IR3 immunoprecipitation of cell surface biotin-labeled proteins in HT29-D4 cells in relation to the state of differentiation. Confluent HT29-D4-GLU cells (A) and 5-day postconfluence HT29-D4-GAL cells (B and C; pretreated with a Ca2+-free medium) were biotinylated and then lysed with Triton X-100. In A and B, the samples were divided into two parts and incubated in the presence (+) or absence (-) of 1 mM DTT for 30 min before immunoprecipitation by {alpha}-IR3 (5 µg) as described in Materials and Methods. In C, the samples of HT29-D4-GAL cell lysates were first submitted (lane a) or not (lane b) to immunoprecipitation by B6 and then to immunoprecipitation of the resulting supernatant by {alpha}-IR3. The immunoprecipitated biotinylated proteins were submitted to SDS-PAGE (7.5% gel), and then transferred to a nitrocellulose sheet and revealed with HRP-streptavidin by the ECL system. Arrows indicate proteins migrating at 135, 102, 97, and 95 kDa. NS, Nonspecific binding, i.e. bands also detected in control experiments in the absence of {alpha}-IR3. In B, lane e (*), {alpha}-IR3 immunoprecipitation of biotinylated proteins from HT29-D4-GLU cell lysate is shown for comparison. Mol wt markers are indicated on the left.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
We provide further evidence that IGF-IR and IR are expressed at the surface of the HT29-D4 human colonic cancer cells. These receptors were evidenced by competitive binding and chemical cross-linking of radiolabeled IGF-I and insulin, and flow cytometric analyses with monoclonal antibodies specific for IGF-IR ({alpha}-IR3) and IR (B6). Our main finding, however, is that a significant part of these receptors is involved in HR formed by the association of {alpha}ß halves of both IGF-IR and IR. Moreover, the level of HR is regulated in a differentiation-dependent manner; it is increased approximately 3-fold by induction of undifferentiated HT29-D4-GLU cells in functional HT29-D4-GAL enterocyte-like cells. This up-regulation of HR is parallel to an increase in immunoreactive IR and functional insulin binding sites from about 14,000–27,000/cell, whereas the numbers of immunoreactive IGF-IR and IGF-I binding sites are unchanged (~25,000/cell). However, as a substantially greater percentage of IGF-IR {alpha}ß halves is involved in HR in HT29-D4-GAL cells than in HT29-D4-GLU cells, the number of native IGF-IR is reduced by about 40% (Table 3Go). These results on cancerous colonic epithelial cells agree with those on rodent fibroblasts transfected with human IR, which showed that the percentage of HR expressed by a cell is a function of the ratio of IR and IGF-IR expression (39). Reports have also evidenced HR in untransfected cells and in human placental tissue (3, 4). However, ours is the first to report that the level of HR is regulated as a function of cell differentiation in untransfected cells, which, in addition, express a relatively low number of endogenous IGF-IR and IR. In contrast to the differentiation-associated alterations in receptor number, we observed no change in the affinity for binding radiolabeled IGF-I (Kd, ~1.5 nM) or insulin (Kd, ~4 nM) in HT29-D4-GLU cells and HT29-D4-GAL cells (Table 1Go). In addition, it should be pointed out that cell surface-associated IGFBPs and plasma membrane type II IGF receptors were not evidenced in HT29-D4 cells regardless of the state of differentiation. This observation is consistent with the results of studies using parental HT29 cells (33).


View this table:
[in this window]
[in a new window]
  		Table 3. Number of IGF-IR, IR, and HR in HT29-D4 cells with respect to the state of differentiation

 
Another important finding is that all the receptor species are strongly polarized toward the basolateral surface of the HT29-D4-GAL cell. The basolateral to apical polarity ratio is ~30:1 for both [125I]IGF-I and [125I]insulin binding. Cell surface distributions of IGF-IR and IR were similar when evaluated in HT29-D4-GAL cells cultured on a plastic support before and after opening the tight junctions with a Ca2+-free medium and when these cells were cultured on a permeable filter that allows free access to the apical or the basolateral side of the cell. This accurate sorting to the basolateral membrane domain of receptors belonging to the IGF-IR/IR family agrees with the general view that receptors for hormones, growth factors, and neurotransmitters are located in the basolateral membrane of epithelial cells (40). However, a lower basolateral polarity ratio of 2.5:1 has been reported for IGF-IR in differentiated CaCo-2 human colonic cancer cells (21). Such a discrepancy should be investigated because it could have important implications in understanding the actions of IGF in the gut in vivo. Note also that the binding parameters of receptors, the polarization of their distribution, and the proportion of HR are nearly identical in early differentiated HT29-D4-GAL cells (5 days postconfluence) and in late differentiated HT29-D4-GAL cells, i.e. 21 days postconfluence, when secretion of carcino-embryonic antigen, a recognized marker of intestine epithelial cell differentiation (41), is increased by ~5-fold (42). Thus, acute alterations in HR/IGF-IR/IR expression take place at an early step of the HT29-D4 cell differentiation process and then remain unchanged.

We determined the presence of HR in HT29-D4 cells by two approaches. The first consists in analyzing cell surface biotinylated proteins, treated or not for disulfide reduction, by immunoprecipitation by {alpha}-IR3. In HT29-D4-GLU cells, immunoprecipitation of {alpha}2ß2 tetrameric receptors by {alpha}-IR3 allowed us to identify an {alpha}-subunit (135-kDa) and two distinct ß-subunit isoforms (97 and 102 kDa). As the same pattern is still found when the receptors are dissociated into {alpha}ß dimers before immunoprecipitation, we conclude that these ß-subunits belong to two IGF-IR subtypes, and that HR are not present in HT29-D4-GLU cells. IGF-IR subtypes have been widely reported in the literature (2, 4, 5, 6, 7, 8, 9). These variant receptors cannot be accounted for by an alternative splicing of the IGF-IR transcript, and it is not known whether they arise from posttranslational alterations or from an unidentified gene (2). The inability of immunoprecipitation experiments to locate HR in HT29-D4-GLU cells argues for their low number, i.e. about 5000/cell, as evaluated through inhibition of ligand binding by antireceptor antibodies (discussed below). In contrast, {alpha}-IR3 immunoprecipitation in HT29-D4-GAL cells evidenced a third ß-subunit moiety migrating at 95 kDa in addition to the 97- and 102-kDa ß-subunits detected in HT29-D4-GLU cells. Removal of IR with the B6 antibody allows examination of only native heterotetrameric IGF-IR in HT29-D4-GAL cell lysate. With such an immunodepleted preparation, we show that {alpha}-IR3 no longer immunoprecipitates this 95-kDa ß-subunit, thus indicating that it is an IR ß-subunit presumably involved in HR. As expected for such a molecular architecture, the 95-kDa IR ß-subunit is no longer detectable when tetrameric receptors are dissociated by disulfide reduction into {alpha}ß dimers before immunoprecipitation by {alpha}-IR3. We, therefore, conclude that HT29-D4-GAL cells express heterotetrameric HR containing an IGF-IR {alpha}ß half (with either a 97-kDa or a 102-kDa ß-subunit) disulfide-linked to an IR {alpha}ß half with a 95-kDa ß-subunit. Using a truncated IR transfected in fibroblasts, Langlois et al., reported that only the high mol wt IGF-IR ß-subunit is able to form HR with IR (43). We, however, could not determine what proportion of 97- and 102-kDa IGF-IR ß-subunits is involved with the 95-kDa IR ß-subunit to form HR. Also, we could not establish whether the 97- and 102-kDa IGF-IR ß-subunits are involved in a same molecule of tetrameric IGF-IR. Such important questions require further work with specific immunoreagents.

The second approach we used to show HR in HT29-D4 cells consisted of studying the interference of receptor-specific monoclonal antibodies with the binding of the noncognate ligand, e.g. B6 with [125I]IGF-I and {alpha}-IR3 with [125I]insulin, respectively. We report that the B6 monoclonal antibody inhibits [125I]IGF-I binding by about 20% in HT29-D4-GLU cells and by about 55% in HT29-D4-GAL cells. In a rough estimation, we propose that this proportion of B6 antibody-reactive IGF-I-binding sites, i.e. about 5,000/HT29-D4-GLU cell and about 15,000/HT29-D4-GAL cell, is, in fact, involved in HR (Table 3Go). As HT29-D4-GLU cells and HT29-D4-GAL cells express only one class of IGF-I-binding sites with the same affinity, HR and IGF-IR express a similar affinity for this peptide. In contrast, the {alpha}-IR3 antibody is totally unable to prevent [125I]insulin binding in HT29-D4-GLU cells and HT29-D4-GAL cells. This result suggests that as {alpha}-IR3 can interact with HR (4), [125I]insulin, at least at tracer concentrations, binds primarily to native IR and does not bind significantly to HR in HT29-D4 cells regardless of their state of differentiation. All of these data agree closely with those for HR purified from placental membranes (44, 45) or generated in IR-transfected cells (39). These studies showed that HR function essentially as IGF-IR, i.e. they have high affinity for IGF-I and do not bind insulin (39) or bind it with low affinity (44, 45). This lack of binding agrees with the Ligand analysis of [125I]insulin binding to HT29-D4-GLU cells and HT29-D4-GAL cells showing only one class of binding sites from which [125I]insulin was displaced much more potently by unlabeled insulin (Kd, ~4 nM) than by IGF-I (Kd, ~150 nM) and IGF-II (Kd, ~20 nM). Yet, the affinity of IR for insulin is relatively low in HT29-D4 cells compared with the values usually reported for IR, including those for intestine epithelial cells (Kd, ~0.5 nM) (22). This discrepancy is unexplained.

The biological role of HR is still unclear, and further studies are needed to ascertain their physiological significance. Taking into account their high affinity binding for IGFs vs. insulin, HR have been proposed to reduce insulin binding to cells expressing both receptors (3, 4, 39, 43). Another attractive possibility is that the interaction of IGFs with HR may initiate a unique signaling pathway that would be distinct from that induced by the interaction of IGFs with native IGF-IR. It has been clearly demonstrated that trans-tyrosine phosphorylation occurs between the two heterologous ß-subunits in HR (43). This leads to a simultaneous triggering of IR ß-subunit- and IGF-IR ß-subunit-coupled signaling cascades in response to binding of IGFs to HR. IR and IGF-IR are thought to use similar intracellular signaling substrates (2). However, recent work with mutant receptors has suggested that particular subdomains and specific amino acid residues of the C-terminal region of IGF-IR and IR ß-subunits may mediate activation of the signal transduction cascade in a receptor-specific fashion (46, 47, 48). In the inducible HT29-D4 cell differentiation model, IGF-II has been reported to be potentially available to cells in an autocrine manner (49). Thus, we propose that HR up-regulation coupled to IGF-IR down-regulation during HT29-D4 enterocyte-like differentiation may be a critical event that alters the nature of the regulatory autocrine function of IGF-II in HT29-D4 cells. This hypothesis of a divergence in IGF-IR vs. HR signaling pathways might support the multiplicity of biological functions assigned to IGF-II in intestinal epithelial cells. Addition of des-(1, 2, 3)-IGF-I or des-(1, 2, 3, 4, 5, 6)-IGF-II, which are truncated IGFs analogs that do not bind to IGFBPs (50), induces HT29-D4 cells into the first steps of enterocyte-like differentiation, as indicated by the increase in carcino-embryonic antigen secretion and by the induction of intercellular cysts exhibiting well organized microvilli and tight junctions (26, 27). By contrast with this positive regulatory function of enterocytic differentiation, constitutive elevated expression of IGF-II increases cell proliferation and interferes with the differentiation pathway in CaCo-2 cells, a colon carcinoma cell line that undergoes spontaneous enterocytic differentiation after cell confluence (30). IGF-II also acts as a survival factor for colon cancer cells (51, 52) and might facilitate the transition from proliferating to differentiating cells. Such multiple biological effects of IGF-II have been reported for other cell types and are especially well documented for the regulation of myoblast proliferation, differentiation, and cell survival (53, 54, 55).

Finally, defining the mechanisms involved in IGF-II action in colonic cells is further complicated by the existence of IGFBPs whose local secretory profile predominantly controls IGFs bioavailability to cell surface receptors (1). In HT29-D4-GLU cells, the secreted IGFBPs, i.e. IGFBP-2, IGFBP-4, and IGFBP-6, totally sequester IGF-II in the extracellular medium, thus preventing its interaction with cell surface receptors and a subsequent differentiative effect (26, 27, 28). If HT29-D4 cells are induced in a differentiative enterocytic pathway by glucose deprivation, the concentration of all of the IGFBPs species increases significantly in a transient manner when cells reach confluence (our unpublished observation). In early differentiated HT29-D4-GAL cells (~5 days postconfluence), the IGFBPs molecular profile is not fundamentally altered. However, once differentiation progresses, the levels of secreted IGFBP-6 strongly decrease, whereas those of IGFBP-2 and IGFBP-4 do not change (42). It should be emphasized that among IGFBPs, IGFBP-6 has the highest affinity for IGF-II (1), and this species has been reported to inhibit IGF-II-mediated differentiation in myoblasts (56). Moreover, these quantitative alterations are coupled with a mechanism of apical vs. basolateral differential sorting of IGFBPs that profoundly alters the molecular profile of IGFBPs (especially for IGFBP-6 and IGFBP-2) secreted in the basolateral compartment, i.e. the one facing the IGF-responsive membrane domain (29, 42). A complex modulation of IGF-II autocrine action by the secreted IGFBPs has been also described for CaCo-2 cells, but the reported results are highly divergent between them (31, 32).

The IGF axis in colonic epithelial cells, therefore, appears very complex. In addition to the differential modulation of IGF-II bioavailability by IGFBPs at the different stages of enterocytic differentiation, we suggest that IGF-II may induce distinct autocrine signaling in HT29-D4 cells whether it activates IGF-IR or HR. As IGF-IR predominate in HT29-D4-GLU cells, this could induce an early autocrine stimulation of cell proliferation by IGF-II, whereas a subsequent interaction of IGF-II with HR, up-regulated in HT29-D4-GAL cells, may favor enterocyte-like cell differentiation. Because IGF-IR have been shown to be key receptors in cancerous cell transformation (12, 57), elucidation of the distinctive features of signal transduction pathways initiated by HR may have important clinical implications.


   Acknowledgments
 
We thank Roselyne Rance and Fabrice Parat for expert technical assistance. We gratefully acknowledge Charles Prevot for his help with the flow cytometry analyses, and Bernard Khalil for the artwork. The help of Garry Burckard in completed the manuscript is gratefully acknowledged.


   Footnotes
 
1 This work was supported in part by Association pour la Recherche contre le Cancer. Back

Received September 25, 1996.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Jones JI, Clemmons DR 1995 Insulin-like growth factors and their binding proteins: biological actions. Endocr Rev 16:3–34[CrossRef][Medline]
  2. LeRoith D, Werner H, Beitner-Johnson D, Roberts Jr CT 1995 Molecular and cellular aspects of the insulin-like growth factor I receptor. Endocr Rev 16:143–163[CrossRef][Medline]
  3. Soos MA, Siddle K 1989 Immunological relationships between receptors for insulin and insulin-like growth factor I. Biochem J 263:553–563[Medline]
  4. Siddle K, Soos MA, Field CE, Navé BT 1994 Hybrid and atypical insulin/insulin-like growth factor I receptors. Horm Res 41:56–65
  5. Alexandrides TK, Smith RJ 1989 A novel fetal insulin-like growth factor (IGF) I receptor. J Biol Chem 264:12922–12930[Abstract/Free Full Text]
  6. Garofalo RS, Barenton B 1992 Functional and immunological distinction between insulin-like growth factor I receptor subtypes in KB cells. J Biol Chem 267:11470–11475[Abstract/Free Full Text]
  7. Barenton B, Domeyne A, Garandel V, Garofalo RS 1993 A developmentally regulated form of insulin-like growth factor receptor ß-subunit in C2 myoblasts exhibiting altered requirements for differentiation. Endocrinology 133:651–660[Abstract]
  8. Moss AM, Livingston JN 1993 Distinct ß-subunits are present in hybrid insulin-like growth factor-1 receptors in the central nervous system. Biochem J 294:685–692
  9. Quiroga S, Garofalo RS, Pfenninger KH 1995 Insulin-like growth factor I receptors of fetal brain are enriched in nerve growth cones and contain a ß-subunit variant. Proc Natl Acad Sci USA 92:4309–4312[Abstract/Free Full Text]
  10. Macaulay VM 1992 Insulin-like growth factors and cancer. Br J Cancer 65:311–320[Medline]
  11. LeRoith D, Baserga R, Helman L, Roberts Jr CT 1995 Insulin-like growth factors and cancer. Ann Intern Med 122:54–59[Abstract/Free Full Text]
  12. Baserga R 1995 The insulin-like growth factor I receptor: a key role to tumor growth? Cancer Res 55:249–252[Abstract/Free Full Text]
  13. Tricoli JV, Rall LB, Karakousis CP, Herrera L, Petrelli NJ, Bell GI, Shows TB 1986 Enhanced levels of insulin-like growth factor messenger RNA in human colon carcinomas and liposarcomas. Cancer Res 46:6169–6173[Medline]
  14. Lambert S, Collette J, Gillis J, Franchimont P, Desaive C, Gol-Winkler R 1991 Tumor IGF-II content in a patient with a colon adenocarcinoma correlates with abnormal expression of the gene. Int J Cancer 48:826–830[Medline]
  15. Singh P, Rubin N 1993 Insulinlike growth factors and binding proteins in colon cancer. Gastroenterology 105:1218–1237[Medline]
  16. Pillion DJ, Ganapathy V, Leibach FH 1985 Identification of insulin receptors on the mucosal surface of colon epithelial cells. J Biol Chem 260:5244–5247[Abstract/Free Full Text]
  17. Park JHY, Vanderhoof JA, Blackwood D, Macdonald RG 1990 Secretion of insulin-like growth factor II (IGF-II) and IGF-binding protein-2 by intestinal epithelial (IEC-6) cells: implications for autocrine growth regulation. Endocrinology 126:2998–3005[Abstract]
  18. Rouyer-Fessard C, Gammeltoft S, Laburthe M 1990 Expression of two types of receptor for insulinlike growth factors in human colonic epithelium. Gastroenterology 98:703–707[Medline]
  19. Heinz-Erian P, Kessler U, Funk B, Gais P, Kiess W 1991 Identification and in situ localization of the insulin-like growth factor-II/mannose-6-phosphate (IGF-II/M6P) receptor in the rat gastrointestinal tract: comparison with the IGF-I receptor. Endocrinology 129:1769–1778[Abstract]
  20. Zarrilli R, Pignata S, Romano M, Gravina A, Casola S, Bruni CB, Acquaviva AM 1994 Expression of insulin-like growth factor (IGF)-II and IGF-I receptor during proliferation and differentiation of CaCo-2 human colon carcinoma cells. Cell Growth Differ 5:1085–1091[Abstract]
  21. Oguchi S, Walker WA, Sanderson IR 1995 Differentiation and polarity alter the binding of IGF-I to human intestinal epithelial (CaCo-2) cells. J Pediatr Gastroenterol Nutr 20:148–155[Medline]
  22. McRoberts JA, Riley NE 1994 Role of insulin and insulin-like growth factor receptors in regulation of T84 cell monolayer permeability. Am J Physiol 267:G883–G891
  23. El Atiq F, Garrouste F, Remacle-Bonnet M, Sastre B, Pommier G 1994 Alterations in serum levels of insulin-like growth factors and insulin-like growth factor binding proteins in patients with colorectal cancer. Int J Cancer 57:491–497[Medline]
  24. Fantini J, Abadie B, Tirard A, Rémy L, Ripert JP, El Batari A, Marvaldi J 1986 Spontaneous and induced dome formation by two clonal cell populations derived from a human adenocarcinoma cell line. J Cell Sci 83:235–249[Abstract]
  25. Fantini J, Rognoni JB, Verrier B, Lehmann M, Roccabianca M, Mauchamp J, Marvaldi J 1990 Suramin-treated HT29–D4 cells grown in the presence of glucose in permeable culture chambers form electrically active epithelial monolayers. A comparative study with HT29–D4 cells grown in the absence of glucose. Eur J Cell Biol 51:110–119[Medline]
  26. Pommier GJ, Garrouste FL, El Atiq F, Roccabianca M, Marvaldi JL, Remacle-Bonnet MM 1992 Potential autocrine role of insulin-like growth factor-II (IGF-II) during suramin-induced differentiation of HT29–D4 human colonic adenocarcinoma cell line. Cancer Res 52:3182–3188[Abstract/Free Full Text]
  27. Remacle-Bonnet M, Garrouste F, El Atiq F, Roccabianca M, Marvaldi J, Pommier G 1992 Des-(1–3)-IGF-I, an insulin-like growth factor analog used to mimic a potential IGF-II autocrine loop, promotes the differentiation of human colon-carcinoma cells. Int J Cancer 52:910–917[Medline]
  28. Pommier G, Garrouste F, El Atiq F, Marvaldi J, Remacle-Bonnet M 1993 Potential role of IGFBPs in the regulation of the differentiation state of human colonic carcinoma cells. Growth Regul 3:80–82[Medline]
  29. Remacle-Bonnet M, Garrouste F, El Atiq F, Marvaldi J, Pommier G 1995 Cell polarity of the insulin-like growth factor system in human intestinal epithelial cells. Unique apical sorting of insulin-like growth factor binding protein-6 in differentiated human colon cancer cells. J Clin Invest 96:192–200
  30. Zarrilli R, Romano M, Pignata S, Gravina A, Casola S, Bruni CB, Acquaviva AM 1996 Constitutive insulin-like growth factor-II expression interferes with the enterocyte-like differentiation of CaCo-2 cells. J Biol Chem 271:8108–8114[Abstract/Free Full Text]
  31. Park JHY, Corkins MR, Vanderhoof JA, Caruso NM, Hrbek MJ, Schaffer BS, Slentz DH, McCusker RH, MacDonald RG 1996 Expression of insulin-like growth factor-II and insulin-like growth factor binding proteins during CaCo-2 cell proliferation and differentiation. J Cell Physiol 166:396–406[CrossRef][Medline]
  32. Singh P, Dai B, Yallampalli U, Lu X, Schroy PC 1996 Proliferation and differentiation of a human colon cancer cell line (CaCo-2) is associated with significant changes in the expression and secretion of insulin-like growth factor (IGF) IGF-II and IGF binding protein-4: role of IGF-II. Endocrinology 137:1764–1774[Abstract]
  33. Remacle-Bonnet MM, Culouscou JM, Garrouste FL, Rabenandrasana C, Marvaldi JL, Pommier GJ 1992 Expression of type I, but not type II, insulin-like growth factor receptor on both undifferentiated and differentiated HT-29 human colonic carcinoma cell line. J Clin Endocrinol Metab 75:609–616[Abstract]
  34. Godefroy O, Huet C, Blair LAC, Sahuquillo-Merino C, Louvard D 1988 Differentiation of a clone from the HT29 cell line: polarized distribution of histocompatibility antigens (HLA) and of transferrin receptors. Biol Cell 63:41–55[CrossRef][Medline]
  35. Munson PJ, Rodbard D 1980 LIGAND: a versatile computerized approach for the characterization of ligand binding systems. Anal Biochem 107:220–239[CrossRef][Medline]
  36. Sargiacomo M, Lisanti MP, Graeve L, Le Bivic A, Rodriguez-Boulan E 1989 Integral and peripheral protein composition of the apical and the basolateral membrane domains in MDCK cells. J Membr Biol 107:277–286[CrossRef][Medline]
  37. Germain-Lee EL, Ja