We have shown that human intestinal smooth muscle cells produce IGF-I and IGF binding protein-3 (IGFBP-3). Endogenous IGF-I acts in autocrine fashion to stimulate growth of these cells. IGFBP-3 inhibits the binding of IGF-I to its receptor and thereby inhibits IGF-I-stimulated growth. In several carcinoma cell lines and some normal cells, IGFBP-3 regulates growth independently of IGF-I. Two mechanisms for this effect have been identified: IGFBP-3 can directly activate transforming growth factor-β (TGF-β) receptors or it can undergo direct nuclear translocation. The aim of the present study was to determine whether IGFBP-3 acts independently of IGF-I and to characterize the mechanisms mediating this effect in human intestinal smooth muscle cells. The direct effects of IGFBP-3 were determined in the presence of an IGF-I receptor antagonist to eliminate its IGF-I-dependent effects. Affinity labeling of TGF-β receptors (TGF-βRI, TGF-βRII, and TGF-βRV) with 125I-labeled TGF-β1 showed that IGFBP-3 displaced binding to TGF-βRII and TGF-βRV in a concentration-dependent fashion. IGFBP-3 stimulated TGF-βRII-dependent serine phosphorylation (activation) of both TGF-βRI and of its primary substrate, Smad2(Ser465/467). IGFBP-3 also caused IGF-I-independent inhibition of basal [3H]thymidine incorporation. The effects of IGFBP-3 on Smad2 phosphorylation and on smooth muscle cell proliferation were independent of TGF-β1 and were abolished by transfection of Smad2 siRNA. Immunoneutralization of IGFBP-3 increased basal [3H]thymidine incorporation, implying that endogenous IGFBP-3 inhibits proliferation. We conclude that endogenous IGFBP-3 directly inhibits proliferation of human intestinal smooth muscle cells by activation of TGF-βRI and Smad2, an effect that is independent of its effect on IGF-I-stimulated growth.
- insulin-like growth factor-I
- short interfering RNA
igf binding protein-3 (IGFBP-3) is one of six IGFBPs that regulate the binding of IGF-I with the cognate IGF-I receptor tyrosine kinase (13). By modulating the binding of IGF-I to its receptor, individual IGFBPs can either inhibit or augment IGF-I-stimulated growth. IGFBP-3 is a 40- to 45-kDa glycoprotein produced largely by the liver and is the major IGFBP present in serum (4, 32). It is produced by other cells, but its expression and effects are both tissue and species specific. Visceral and vascular smooth muscle, including human intestinal smooth muscle, secrete both IGF-I and IGFBPs that regulate smooth muscle growth and development in autocrine fashion (2, 9, 14, 15). Studies in genetically altered mice have shown that these autocrine effects in smooth muscle are mediated independently of the IGF-I and IGFBPs produced in the liver and released into the circulation (33, 42). Locally produced IGF-I and IGFBPs, such as IGFBP-5, play a key role in the growth and development of smooth muscle throughout the body, including the intestine (33, 41).
IGFBP-3, like IGFBP-1 and IGFBP-5, is capable of regulating cell growth independently of its effects on IGF-I-stimulated growth (13, 26). Two distinct mechanisms mediating the IGF-I-independent effects of IGFBP-3 have been identified. The first mechanism involves interaction of IGFBP-3 with transforming growth factor-β (TGF-β) cell surface receptors. In mink lung epithelial cells and other cells, IGFBP-3 binds to TGF-β receptor (TGF-βR) type V (TGF-βRV) (23, 24). A direct inhibitory effect on growth mediated by this receptor has been proposed, but a mechanism of action has not been fully elucidated. Recent evidence (10) implicates the low-density lipoprotein receptor-related protein-1 in this pathway. IGFBP-3 also binds to and activates intracellular signaling via the TGF-βRII and TGF-βRI heteromeric complex (7, 8, 26). In the T47D breast cancer cell line, this mechanism requires the presence of TGF-β1 and both TGF-βRI and TGF-βRII and results in Smad2 [Homo sapiens mothers against decapentaplegic homolog 2 (Drosophila) (MADH2)] activation and inhibition of growth (8). A distinct, nonreceptor-based mechanism mediating IGF-I-independent inhibition of growth by IGFBP-3 has also been elucidated. IGFBP-3 possesses a consensus nuclear translocation sequence in its COOH terminus (29, 30). After nonreceptor-mediated nuclear translocation of IGFBP-3 via the β-importin pathway, IGFBP-3 binds to the nuclear retinoid X receptor-α (RXRα) and directly inhibits growth of opossum kidney cells, A549 lung cancer cells, T47D breast cancer cells, and various other cells (25, 29, 30).
We (2, 15, 19) have previously shown that human intestinal smooth muscle cells secrete IGF-I and IGFBP-3, -4, and -5. IGFBP-3 acts to inhibit IGF-I-stimulated proliferation (2). Intestinal smooth muscle cells also secrete TGF-β1, which inhibits growth directly and increases IGFBP-3 expression (14). Little is known, however, regarding the IGF-I-independent effects of IGFBP-3 on growth of smooth muscle, whether TGF-β1 is required for IGFBP-3 to have these effects as it does in breast cancer cells, or what mechanisms might mediate these effects. These mechanisms may play an important role in the altered growth regulation of intestinal smooth muscle during the intestinal inflammation of Crohn’s disease in which levels of IGFBP-3 and TGF-β1 are altered and may contribute to muscle hyperplasia and stricture formation.
The present study shows that IGFBP-3 directly inhibits human intestinal smooth muscle growth. IGFBP-3 binds to the TGF-β receptor types expressed by intestinal muscle cells. Binding of IGFBP-3 to TGF-βRII is followed by serine phosphorylation of TGF-βRI and Smad2, the primary substrate of activated TGF-βRI receptors. Activation of Smad2 mediates IGFBP-3-dependent inhibition of proliferation. Immunoneutralization of secreted IGFBP-3 increased basal proliferation. The ability of IGFBP-3 to inhibit proliferation does not require the presence of TGF-β1. The effects of IGFBP-3 were abolished in cells after Smad2 gene knockdown with short interfering RNA (siRNA). The results provide evidence that, in addition to the ability of IGFBP-3 to inhibit IGF-I-stimulated growth, endogenous IGFBP-3 also causes direct inhibition of human intestinal smooth muscle cell proliferation by activating TGF-β receptors and Smad2.
MATERIALS AND METHODS
Recombinant human (rh) IGF-I and rhTGF-β1 were obtained from Austral Biologicals (San Ramon, CA); collagenase and soybean trypsin inhibitor were obtained from Worthington Biochemical, Freehold, NJ); HEPES was obtained from Research Organics (Cleveland, OH); DMEM and HBSS were obtained from Mediatech (Herndon, VA); fetal bovine serum was obtained from Summit Biotechnologies (Fort Collins, CO); 125I-labeled TGF-β1 (specific activity: 3,000 Ci/mmol) and [3H]thymidine (specific activity: 6 Ci/mmol) were obtained from New England Nuclear (Boston, MA); rhIGFBP-3, rabbit polyclonal antibodies to Smad2(Ser465/467), total Smad2, and protein A/G agarose beads were obtained from Upstate Biotechnology (Lake Placid, NY); mouse monoclonal antibody to phosphoserine (22a) was obtained from BD Transduction Laboratories (San Jose, CA); rabbit polyclonal antibody to TGF-βRI was obtained from Santa Cruz Biotechnology (Santa Cruz, CA); neutralizing antibody to TGF-β1 was obtained from R&D Systems (Minneapolis, MN); anti-rabbit horseradish peroxidase conjugates were obtained from Cell Signaling Technology (Beverly, MA); disuccinimidyl suburate was obtained from Pierce (Rockford, IL); Western blotting materials and DC protein assay kit were obtained from Bio-Rad Laboratories (Hercules, CA); and plastic cultureware was obtained from Corning (Corning, NY). All other chemicals were obtained from Sigma (St Louis, MO).
Culture of smooth muscle cells isolated from normal human intestine.
Muscle cells were isolated and cultured from the circular muscle layer of human intestine as described previously (2, 14, 15, 18–20). Briefly, 4- to 5-cm segments of normal jejunum were obtained from patients undergoing surgery for morbid obesity according to a protocol approved by the Virginia Commonwealth University Institutional Review Board. The segments were opened along the mesenteric border, the mucosa was dissected away, and the remaining muscle layer was cut into 2 × 2-cm strips. Slices were obtained from the circular layer using a Stadie-Riggs tissue slicer. The slices were incubated overnight at 37°C in 20 ml DMEM plus 10% fetal bovine serum (DMEM-10) containing 200 U/ml penicillin, 200 μg/ml streptomycin, 100 μg/ml gentamycin, and 2 μg/ml amphotericin B (DMEM-10) to which was added 0.0375% collagenase (type II), and 0.1% soybean trypsin inhibitor. Muscle cells dispersed from the circular layer were harvested by filtration through 500-μm Nitex mesh and centrifugation at 150 g for 5 min. Cells were resuspended and washed twice by centrifugation at 150 g for 5 min. After resuspension in DMEM-10 containing the same antibiotics, the cells were plated at a concentration of 5 × 105 cells/ml as determined by counting in a hemocytometer. Cultures were incubated in a 10% CO2 environment at 37°C. DMEM-10 was replaced every 3 days until the cells reached confluence. Primary cultures of muscle cells were passaged on reaching confluence. All subsequent studies were performed in first-passage cultured cells after 7 days, at which time the cells are confluent. We (14) have previously shown that these cells express a phenotype characteristic of intestinal smooth muscle as determined by immunostaining for intestinal smooth muscle markers and expression of γ-enteric actin. Epithelial cells, endothelial cells, neurons, and interstitial cells of Cajal are not detected in these cultures (37, 38).
[3H]thymidine incorporation assay.
Proliferation of smooth muscle cells in culture was measured by the incorporation of [3H]thymidine as described previously (2, 14, 18–20). Briefly, the cells were washed free of serum and incubated for 24 h in serum-free DMEM in the presence or absence of various test agents. During the final 4 h of this incubation period, 1 μCi/ml [3H]thymidine was added to the medium. [3H]thymidine incorporation into the perchloric acid-extractable pool was used as a measure of DNA synthesis.
Measurement of Smad2 phosphorylation.
The levels of phospho-Smad2(Ser465/467) and total Smad2 were measured by immunoblot analysis using standard methods (2, 15, 18, 19, 21). Briefly, confluent muscle cells were washed free of serum and stimulated with various test agents. The reaction was terminated by two rapid washes in ice-cold PBS after which lysates were prepared from the cells. Lysates were separated by SDS-PAGE under denaturing conditions. After the proteins were electrotransferred to nitrocellulose, the membranes were incubated overnight with a 1:1,000 dilution of antibodies recognizing Ser465/467-phosphorylated Smad2 or total Smad2. Bands of interest were visualized with enhanced chemiluminescence on a FluorChem 8800 (Alpha Innotech, San Leandro, CA), and the resulting digital images were analyzed by using AlphaEaseFC version 3.1.2 software.
Measurement of TGF-βRI phosphorylation.
Serine phosphorylation (activation) of TGF-βRI was measured by immunoblot analysis after immunoprecipitation of TGF-βRI from whole cell lysates by modification of previously described methods (20, 21). Briefly, confluent muscle cells were washed free of serum and stimulated with various test agents. The reaction was terminated by two rapid washes in ice-cold PBS after which lysates were prepared from the cells. Cell lysates were prepared in a buffer consisting of (in mM): 50 Tris·HCl (pH 7.5), 150 NaCl, 50 NaF, 1 Na orthovanadate, 1 dithiothreitol, 1 PMSF, and 0.5% Nonidet P-40 to which was added 1 μg/ml leupeptin, 1 μg/ml pepstatin A, and 1 μg/ml aprotinin. The resulting lysates were clarified by centrifugation at 14,000 g for 10 min at 4°C. The lysates were precleared by incubation with protein A agarose beads for 1 h at 4°C. Samples containing equal amounts of protein (0.5 mg) were incubated for 2 h at 4°C with 2 μg of rabbit anti-TGF-βRI after which 10 μl of protein A/G agarose beads were added, and the incubation was continued overnight. The immune complex-agarose beads were washed three times with ice-cold lysis buffer. After the final wash, the beads were resuspended in 25 μl of sample buffer. The samples were boiled for 5 min after which 20 μl of each sample was loaded onto a 15% polyacrylamide gel, and the proteins were separated by SDS-PAGE under denaturing conditions. After the proteins were electrotransferred to nitrocellulose, the membranes were incubated overnight with a 1:2,500 dilution of a monoclonal antibody (22a) recognizing serine-phosphorylated proteins. Bands of interest were visualized with enhanced chemiluminescence on a FluorChem 8800, and the resulting digital images were analyzed by using AlphaEaseFC version 3.1.2 software. Values for serine phosphorylation were normalized to total TGF-βRI levels after the membranes were stripped and reblotted for TGF-βRI.
Ligand binding analysis of TGF-β receptors.
Binding of IGFBP-3 to TGF-β receptors was measured by 125I-labeled TGF-β1 affinity cross-linking and ligand blot analysis according to the methods of Cheifetz, et al. (3). Briefly, muscle cells growing in 60-mm dishes were washed twice in binding buffer consisting of (in mM): 50 HEPES (pH 7.5), 128 NaCl, 5 KCl, 1.2 CaCl2, and 1.2 MgSO4 with added 5 mg/ml (wt/vol) BSA. The cells were incubated for an additional 1 h at 37°C in binding buffer and equilibrated for 10 min at 4°C. The cells were then incubated with 100 pM 125I-labeled TGF-β1 at 4°C for 2 h. Nonspecific binding was determined in the presence of unlabeled 100 nM TGF-β1. Unbound 125I-labeled TGF-β1 was removed by washing the cells three times with ice-cold binding buffer. The cells were incubated for an additional 15 min in binding buffer without BSA and containing the bifunctional cross-linking reagent disuccinimidyl suberate (DSS, 25 mM). The cross-linking reaction was terminated by washing the cells in buffer consisting of (in mM): 10 Tris (pH 7.0), 250 sucrose, 1 EDTA, and 0.1 PMSF, with added pepstatin (1 μg/ml) and leupeptin (1 μg/ml). The cells were solubilized in sample buffer, and proteins were separated on 4–15% gradient polyacrylamide gels under reducing conditions. Gels were dried and then visualized by using autoradiography. The resulting autoradiograms were imaged on a FluorChem 8800. The resulting digital images were then analyzed by using AlphaEaseFC version 3.1.2 software. Bands corresponding to specific TGF-β receptor types were identified by their apparent molecular weight compared with known molecular weight standards.
Transfection of Smad2 siRNA.
Protein levels of Smad2 in human intestinal muscle cells were inhibited by transfection of a Smad2 siRNA (6, 36). siRNA sequences were determined based on the sequence of human Smad2. Briefly, muscle cells growing in six-well plates were transfected with 2 μg of Smad2 siRNA or 2 μg on control RNA (Upstate Biotechnology) using TransMessenger transfection reagent according to the manufacturer’s directions (Qiagen, Valencia, CA). Cells were incubated for 3 h at 37°C with the transfection reagent-RNA complexes. The cells were washed free of transfection reagent-RNA complexes with PBS, and 2 ml of fresh DMEM-10 were added to each well. After 24-h incubation, the ability of Smad2 siRNA, but not control RNA, to inhibit Smad2 levels was determined by Western blot analysis as described above. Initial experiments were performed by using a range of RNA amounts (0.5–4 μg) and transfection times (24–48 h) to determine the optimal conditions. Optimal inhibition of Smad2 was obtained by using 2 μg of RNA after a 24-h incubation period (data not shown).
Measurement of protein content.
The protein content of cell lysates was measured by using the Bio-Rad DC protein assay kit according to the manufacturer’s directions. Samples were adjusted to provide aliquots of equal protein content before SDS-PAGE.
Values given represent the means ± SE of the number of experiments on cells derived from separate primary cultures. Statistical significance was tested by Student’s t-test for either paired or unpaired data as appropriate. Analysis of relative densitometric values for Western blots was performed by using AlphaEaseFC version 3.1.2 software. Densitometric values for protein bands were reported in arbitrary units above basal values. Values for Smad2(Ser465/467) were normalized to total Smad protein levels after membranes were stripped and reblotted.
IGFBP-3 binds to TGF-β receptors.
The notion that IGFBP-3 might bind to the TGF-β receptor was investigated by using 125I-labeled TGF-β1 affinity labeling of TGF-β receptors (3). TGF-β receptors on human intestinal smooth muscle were identified in whole cell lysates prepared from cells in which 125I-labeled TGF-β1 was cross-linked to TGF-β receptors using DSS. Autoradiograms of 125I-labeled TGF-β1-labeled proteins showed that intestinal smooth muscle cells expressed three types of TGF-β receptors: TGF-βRI (∼53 kDa), TGF-βRII (∼70 kDa), and the high molecular mass (∼400 kDa) TGF-βRV (Fig. 1A) (3, 29). Binding of 125I-labeled TGF-β1 to TGF-βRV was inhibited in a concentration-dependent fashion by unlabeled IGFBP-3 with 89 ± 5% inhibition of binding in the presence of 500 ng/ml IGFBP-3 (Fig. 1, A and B). IGFBP-3 (500 ng/ml) also inhibited by 72 ± 7% the binding of 125I-labeled TGF-β1 to lower molecular weight type II receptors that heterodimerizes with TGF-βRI receptors and initiates TGF-β1-dependent intracellular signaling via the receptor-coupled Smad (R-Smad) pathway.
IGFBP-3 activates type I TGF-β receptors.
In response to the binding of TGF-β1 to TGF-β receptors, association of TGF-βRII receptor serine kinase with TGF-βRI elicits serine phosphorylation of TGF-βRI and initiates Smad signaling (5, 33). The ability of IGFBP-3 to not only bind to but activate TGF-β receptors was measured in phosphoserine immunoblots of TGF-βRI immunoprecipitated from quiescent muscle cells and in cells treated with IGFBP-3.
Incubation of muscle cells for 15 min with 500 ng/ml IGFBP-3 increased serine phosphorylation of TGF-βRI levels by 110 ± 30% above basal levels (Fig. 2). In control experiments, incubation of muscle cells for 15 min with 1 nM TGF-β1 also increased TGF-βRI serine phosphorylation by 130 ± 20% above basal levels (Fig. 2).
IGFBP-3 elicits Smad2 phosphorylation.
The effects of TGF-β1 are mediated by activation of R-Smads that link the activated TGF-β receptors to downstream signaling events (5, 7). The R-Smad homologs Smad2 and Smad3 are the primary substrates of the activated TGF-βRI. Smad2 is activated by serial serine phosphorylation on residues 465 and 467. The possibility that IGFBP-3 also activates Smad2 was therefore investigated by measurement of Smad2(Ser465/467) phosphorylation.
Smooth muscle cells were incubated for 15 min with increasing concentrations of IGFBP-3 (5–500 ng/ml) or with 1 nM TGF-β1 as a positive control. IGFBP-3 elicited concentration-dependent Smad2(Ser465/467) phosphorylation with 500 ng/ml, causing a 91 ± 5% increase above basal levels within 15 min (Fig. 3). In T47D breast cancer cells, the effects of IGFBP-3 on TGF-β receptor signaling were dependent on the presence of TGF-β1 (8). This possibility was examined by repeating the experiment in the presence of a neutralizing antibody to TGF-β1. We have previously used this antibody (50 ng/ml), shown it fully neutralizes endogenous TGF-β1 production in these cells, and shown that endogenous TGF-β1 exerts growth-inhibitory effects in human intestinal smooth muscle cells (2, 14). In the presence of the immunoneutralizing antibody to TGF-β1 (50 ng/ml), IGFBP-3 retained the ability to cause concentration-dependent Smad2(Ser465/467) phosphorylation (500 ng/ml: 73 ± 10% above basal) (Fig. 3). The results implied that IGFBP-3 does not require the presence of TGF-β1 to activate the TGF-β receptors and initiate intracellular signaling via Smad2 in human intestinal smooth muscle.
IGFBP-3 directly inhibits proliferation.
We have previously shown that IGFBP-3 inhibits IGF-I-stimulated proliferation of human intestinal muscle cells (2). The possibility that IGFBP-3 could affect muscle proliferation independently was investigated by measurement of [3H]thymidine incorporation in the presence of an IGF-I receptor antagonist to eliminate the IGF-I-dependent effects of IGFBP-3 on growth. IGFBP-3 (5–500 ng/ml) elicited a concentration-dependent inhibition of basal [3H]thymidine incorporation (basal: 132 ± 10 counts· min−1·mg protein−1) with 500 ng/ml IGFBP-3 inhibiting basal growth by 38 ± 7% (Fig. 4A).
In T47D breast cancer cells, direct growth inhibition by IGFBP-3 required the presence of TGF-β1. The requirement for TGF-β1 was also examined for IGFBP-3-induced proliferation after immunoneutralization of endogenous TGF-β1 produced by these cells as had been done when examining Smad2 activation. Cells were incubated with 50 ng/ml of a TGF-β1- neutralizing antibody. We (2, 14) have previously shown that this antibody effectively eliminates TGF-β1-induced inhibition of proliferation. As expected from our previous work, neutralization of endogenous TGF-β1 increased basal [3H]thymidine incorporation, confirming that endogenous TGF-β1 inhibits growth in these cells. Consistent with the ability of IGFBP-3 to activate Smad2 independently of TGF-β1, IGFBP-3 also retained its ability to inhibit [3H]thymidine incorporation. After immunoneutralization of endogenous TGF-β1, IGFBP-3 (5–500 ng/ml) still caused identical relative concentration-dependent inhibition of proliferation (500 mg/ml: 41 ± 8% inhibition of [3H]thymidine incorporation) (Fig. 4A).
The role of endogenous IGFBP-3 in regulating smooth muscle cell growth was also examined in the muscle cells. Our previous work (2, 14) has shown that endogenous IGF-I stimulates growth and that IGF-I-stimulated growth is inhibited by IGFBP-3. The direct effects of IGFBP-3 on growth were examined by immunoneutralization of endogenous IGFBP-3 and in the presence of the IGF-I receptor antagonist so that the modulatory role of IGFBP-3 on IGF-I-stimulated growth was eliminated. Muscle cells were incubated for 24 h in the presence of a neutralizing antibody to IGFBP-3 (1–25 μg/ml). Basal [3H]thymidine incorporation was increased in a concentration-dependent manner in the presence of increasing concentrations of IGFBP-3-neutralizing antibody (Fig. 4B). The results implied that endogenous IGFBP-3 directly inhibited muscle cell proliferation.
IGFBP-3-induced proliferation is Smad2-dependent.
The role of Smad2 in IGFBP-3-dependent direct regulation of smooth muscle cell growth was examined in cells in which Smad2 was eliminated by using a siRNA approach (6, 36). Initial experiments were performed to confirm that transfection of Smad2 siRNA resulted in a decrease in Smad2 protein levels. Transfection of muscle cells with 2 μg of control RNA did not alter the levels of Smad2 protein from that present in naive, untransfected cells. In cells transfected with 2 μg of Smad2 siRNA, Smad2 protein was decreased by 69 ± 4% (Fig. 5).
In muscle cells transfected with Smad2 siRNA, the ability of 500 ng/ml IGFBP-3 to inhibit [3H]thymidine incorporation (control: 27 ± 6% inhibition) was abolished. Similarly, the ability of TGF-β1 to inhibit proliferation (control: 41 ± 7% inhibition) was also abolished in cells transfected with Smad2 siRNA (Fig. 5). Transfection of muscle cells with control RNA did not affect the levels of basal [3H]thymidine incorporation (naive cells: 122 ± 14 counts·min−1·mg protein−1; control RNA: 126 ± 20 counts·min−1·mg protein−1). As expected, transfection of muscle cells with Smad2 siRNA increased basal [3H]thymidine incorporation to 147 ± 25 counts·min−1·mg protein−1. This is attributable to removal of the inhibitory effects of endogenous IGFBP-3 and TGF-β1 on cell growth.
Cellular growth represents the balance of cell division and cell death. IGF-I regulates the growth of human intestinal smooth muscle by stimulating proliferation and inhibiting apoptosis (14, 16). These IGF-I-dependent effects are modulated by the three IGFBPs expressed by human intestinal smooth muscle and by circulating IGFBPs. IGFBP-3 and -4 inhibit the binding of IGF-I to its receptor and inhibit IGF-I-stimulated growth (2, 15). IGFBP-5 has the opposite effect; it facilitates the binding of IGF-I to its receptor and augments IGF-I-stimulated growth (20). IGFBPs can also exert direct effects on cell growth independently of their ability to modulate the binding of IGF-I to its receptor. In this regard, we (20) have previously shown that IGFBP-5 acts directly on human intestinal muscle cells via Erk1/2- and p38 MAP kinase-dependent pathways to stimulate proliferation and IGF-I secretion. IGFBP-1 also has IGF-I-independent effects, but it is not expressed by intestinal smooth muscle cells. An IGF-I-independent effect of IGFBP-3 on growth has been identified in a number of mammalian cell types, including mink lung epithelial cells, rat chrondrocytes, and several cell lines derived from breast, lung, prostate, and colon cancers (8, 11, 26, 28, 35, 40). The ability of endogenous IGFBP-3 to directly regulate intestinal smooth muscle cell growth and the mechanisms involved has not previously been examined.
This paper shows that endogenous IGFBP-3 directly inhibits normal human intestinal smooth muscle cell proliferation. The effects of IGFBP-3 on proliferation are mediated by binding to TGF-β receptors, activation of TGF-βRI receptors, and initiation of intracellular signaling via Smad2 phosphorylation. The evidence that IGFBP-3 directly inhibits growth via TGF-βRI-dependent, Smad2-dependent pathways can be summarized as follows: 1) IGFBP-3 binds to several TGF-β receptor types in a concentration-dependent fashion, 2) IGFBP-3 stimulates serine phosphorylation (activation) of TGF-βRI receptors, 3) IGFBP-3 stimulates concentration-dependent phosphorylation (activation) of Smad2 on Ser465/467, 4) IGFBP-3 causes concentration-dependent inhibition of [3H]thymidine incorporation, 5) the effect of IGFBP-3 on proliferation is abolished when Smad2 expression is diminished, and 6) immunoneutralization of endogenous IGFBP-3 increases basal [3H]thymidine incorporation implying that endogenous IGFBP-3 inhibits proliferation.
Two distinct mechanisms by which IGFBP-3 directly regulates cellular growth have been identified (1). The first involves the interaction of IGFBP-3 with TGF-β receptors and TGF-β-dependent signaling mechanisms; the second involves the interaction of IGFBP-3 with nuclear RXRα (8, 23, 29). Via the first mechanism, described in mink lung epithelial cells, IGFBP-3 binds to the ∼400 kDa TGF-βRV (23, 24). This serine/threonine kinase is widely expressed in normal cells and is also expressed in some transformed cells. TGF-β1 is unable to inhibit growth in the human colorectal carcinoma cell lines HCT-116 and RII-37, which do not express TGF-βRV receptors but do express the lower molecular weight type I TGF-β (TGF-βRI) and type II TGF-β (TGF-βRII) receptors (24). Although IGFBP-3-dependent inhibition of growth coupled to the TGF-βRV receptor has been shown (24), a signaling mechanism activated by this receptor has yet to be identified. Recently, the bovine TGF-βRV receptor was shown to be identical to the human low-density lipoprotein receptor-related protein-1 involved in endocytosis (10) and may provide some insight into its mechanisms of action.
In the Hs578T, MCF-7, and T47D breast cancer cells lines, IGFBP-3 binds directly to TGF-βRII receptors. In a fashion similar to TGF-β1, binding of IGFBP-3 to the constitutively active TGF-βRII activates the TGF-βRI receptor serine/threonine kinase (7, 8, 26, 27). Studies performed in T47D cells have been particularly useful in elucidating the distinct TGF-β receptor-dependent mechanisms activated by IGFBP-3. In naive T47D cells, which lack TGF-βRII receptors, IGFBP-3 has no direct effect on growth because, at least in these cells, expression of TGF-βRII and the presence of TGF-β1 peptide are required in order for IGFBP-3 to inhibit growth (8). Under the experimental conditions of TGFβ-RII transfection and exogenous TGF-β1 addition, the ability of IGFBP-3 to activate TGF-βRI and Smad2 and inhibit growth of T47D cells is restored (8). Human intestinal smooth muscle cells, in contrast to T47D cells, express both type I and II TGF-β receptors and secrete TGF-β1. Immunoneutralization of endogenous TGF-β1, however, does not diminish IGFBP-3-dependent Smad2 activation or IGFBP-3-dependent inhibition of [3H]thymidine incorporation, implying that IGFBP-3 does not require TGF-β1 to directly inhibit proliferation in human intestinal smooth muscle cells.
One mechanism by which TGF-β1 inhibits growth is through the R-Smad signaling pathway (5). Once activated by TGF-β1-dependent association with TGF-βRII, TGF-βRI phosphorylates its primary R-Smad substrates, Smad2 and/or Smad3 (36). Initial phosphorylation of Smad2 on Ser467 in the COOH terminus is followed by phosphorylation of Ser465 (34). Dual serine phosphorylation of Smad2 is required in order for it to associate with the next signaling protein in this pathway, Smad4. The Smad2/3-Smad4 complex translocates to the nucleus in which it acts as a transcriptional regulator of TGF-β- responsive elements. One such target of IGFBP-3-activated Smad signaling is the promoter region of the TGF-β-responsive gene plasminogen activator inhibitor-1 (PAI-1). PAI-1 is involved in maintaining IGFBP-3 homeostasis by blocking the activation of plasmin (7). Plasmin degrades IGFBP-3; this pathway in effect delays IGFBP-3 degradation and augments IGFBP-3-dependent responses. In human intestinal smooth muscle cells, blockade of the Smad2 pathway using Smad2 siRNA abolished the ability of IGFBP-3 to directly inhibit growth. The results implied that although IGFBP-3 binds TGF-βRV receptors, the type V receptor is not coupled to inhibition of proliferation in these cells. This is not to say that this pathway might not affect cell growth, because IGFBP-3 regulates apoptosis in a variety of cells (25, 29, 30).
A distinct mechanism that mediates IGF-I-independent inhibition of growth by IGFBP-3, which does not involve TGF-β or other cell surface receptors, has also been elucidated. The COOH terminus of IGFBP-3 possesses a consensus nuclear translocation sequence. This region is distinct from the regions that activate TGF-β receptors and Smad (30). Direct nuclear translocation of IGFBP-3 via the β-importin pathway allows IGFBP-3 to bind to the nuclear RXRα, regulate transcriptional signaling, and directly inhibit growth in opossum kidney cells, A549 lung cancer cells, T47D breast cancer cells, and a number of other cells (7, 25, 29, 30). This mechanism has been shown to regulate not only cellular proliferation but the complementary aspect of growth as well as apoptosis. The ability of IGFBP-3 to regulate apoptosis and the mechanisms involved were not investigated directly in the present study. However, the ability of Smad2 siRNA to fully abolish the effects of IGFBP-3 on [3H]thymidine incorporation suggests that TGF-βRI receptor activation by IGFBP-3 accounts fully for its effects on cellular proliferation. The possibility that IGFBP-3 affects apoptosis via β-importin-dependent nuclear translocation and activation of RXRα was not examined in the present study.
The IGF-I-independent, direct effects of IGFBP-3 have potential clinical relevance in the setting of the intestinal inflammation of Crohn’s disease, muscle hyperplasia, and stricture formation. Circulating levels of IGF-I and IGFBP-3 are decreased in the setting of intestinal inflammation in Crohn’s disease (11). The muscularis propria, however, is known to be regulated by IGF-I and IGFBPs produced endogenously. Both IGF-I and TGF-β1 expression within the intestinal smooth muscle layer, for example, are increased in regions of muscle inflammation and stricture formation (22, 42). Although evidence suggests that IGF-I overexpression in intestinal muscle does not alter IGFBP-3 expression (39), it is not known what effect intestinal inflammation or TGF-β1 might have on local smooth muscle IGFBP-3 levels. Preliminary studies suggest that IGFBP-3 mRNA levels, measured by real-time PCR, are not altered within intestinal smooth muscle in patients with Crohn’s disease (J. F. Kuemmerle and J. G. Bowers, unpublished results). We (2) have previously shown, however, that IGFBP-3 protein levels in human intestinal smooth muscle are regulated largely by posttranslational mechanisms rather than by gene transcription. Whereas IGFBP-3 levels are increased by TGF-β1 (2), it is not known whether IGFBP-3 influences TGF-β1 levels in return. The effect of altered IGFBP-3 levels in the inflamed intestinal muscle may be to alter the net growth-regulatory signals, both IGF-I-dependent growth and IGF-I independent growth, as well as both TGF-β1-dependent and possibly IGFBP-3-dependent (via the TGF-β receptor) effects.
In summary, this study shows that endogenous IGFBP-3 directly inhibits the proliferation of human intestinal smooth muscle cells. The mechanism mediating this effect is distinct from the ability of IGFBP-3 to inhibit IGF-I-stimulated growth. After binding of IGFBP-3 to TGF-βRII receptors, type I TGF-β receptors are activated and intracellular signaling via Smad2 is initiated. This pathway is coupled to inhibition of proliferation. Although IGFBP-3-dependent inhibition of growth is mediated via TGF-β receptors, these effects are independent of endogenous TGF-β1.
This study was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grant DK-49691 (to J. F. Kuemmerle).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2004 the American Physiological Society