Insulin-like Growth Factor-binding Protein-5 (IGFBP-5) Stimulates Growth and IGF-I Secretion in Human Intestinal Smooth Muscle by Ras-dependent Activation of p38 MAP Kinase and Erk1/2 Pathways*

Insulin-like growth factor-binding protein-5 (IGFBP-5) and insulin-like growth factor-I (IGF-I) are produced by human intestinal smooth muscle cells. Endogenous IGF-I stimulates growth and increases IGFBP-5 secretion. IGFBP-5 augments the effects of IGF-I by facilitating interaction of IGF-I with the IGF-I receptor tyrosine kinase. Andress (Andress, D. L. (1998) Am. J. Physiol. 274, E744–E750) and Berfield et al. (Berfield, A. K., Andress, D. L., and Abrass, C. K. (2000) Kidney Int. 57, 1991–2003) have shown that in osteoblasts and kidney mesangial cells, IGFBP-5 stimulates proliferation and filopodia formation independently of IGF-I, presumably by activating a distinct IGFBP-5 receptor serine kinase. The present study determined whether IGFBP-5 exerts direct effects on growth in human intestinal smooth muscle cells and identified the intracellular signaling pathways involved. IGFBP-5 caused a concentration-dependent increase in [3H]thymidine incorporation and an increase in IGF-I secretion that occurred independently of IGF-I and the IGF-I receptor tyrosine kinase. IGFBP-5-induced phosphorylation of p38 MAP kinase, which was abolished by SB203580, or expression of a dominant negative Ras mutant, Ras(S17N), and phosphorylation of Erk1/2, which was abolished by a Raf1 kinase inhibitor, U1026, or expression of Ras(S17N). IGFBP-5-stimulated [3H]thymidine incorporation and IGF-I secretion were partly inhibited by SB203580 or U1026 and abolished by the combination of the two inhibitors or by expression of Ras(S17N). These data show that IGFBP-5 stimulates growth and IGF-I secretion in human intestinal smooth muscle cells by activation of p38 MAP kinase-dependent and Erk1/2-dependent pathways that are independent of IGF-I. A positive feedback mechanism therefore links IGFBP-5 and IGF-I secretion that reinforces their individual effects on growth.


that in osteoblasts and kidney mesangial cells, IGFBP-5 stimulates proliferation and filopodia formation independently of IGF-I, presumably by activating a distinct IGFBP-5 receptor serine kinase. The present study determined whether IGFBP-5 exerts direct effects on growth in human intestinal smooth muscle cells and identified the intracellular signaling pathways involved. IGFBP-5 caused a concentration-dependent increase in [ 3 H]thymidine incorporation and an increase in IGF-I secretion that occurred independently of IGF-I and the
IGF-I receptor tyrosine kinase. IGFBP-5-induced phosphorylation of p38 MAP kinase, which was abolished by SB203580, or expression of a dominant negative Ras mutant, Ras(S17N), and phosphorylation of Erk1/2, which was abolished by a Raf1 kinase inhibitor, U1026, or expression of Ras(S17N). IGFBP-5-stimulated [ 3 H]thymidine incorporation and IGF-I secretion were partly inhibited by SB203580 or U1026 and abolished by the combination of the two inhibitors or by expression of Ras(S17N). These data show that IGFBP-5 stimulates growth and IGF-I secretion in human intestinal smooth muscle cells by activation of p38 MAP kinase-dependent and Erk1/2-dependent pathways that are independent of IGF-I. A positive feedback mechanism therefore links IGFBP-5 and IGF-I secretion that reinforces their individual effects on growth.
Insulin-like growth factor-I (IGF-I) 1 mediates three distinct regulatory effects on cell growth by activation of the IGF-I receptor: IGF-I stimulates proliferation of cells and is required for sustained growth of many cells (1); transformation and maintenance of the transformed state also require IGF-I receptor activation in some cells (2); and IGF-I protects cells from apoptosis (3). The central role of IGF-I in the regulation of smooth muscle cell growth in both the normal and pathologic states is manifested by the hyperplasia of intestinal and vascular smooth muscle in transgenic animals overexpressing a human IGF-I cDNA (4,5). The effects of IGF-I are modulated by IGF-binding proteins. Six IGF-binding proteins (IGFBP-1-6) have been identified that can either augment the effects of IGF-I by facilitating the interaction of IGF-I with its cognate receptor or inhibit the effects of IGF-I by diminishing the interaction of IGF-I with its receptor (6). The presence and effect of each IGF binding protein, however, is both tissue-and species-specific.
IGFBP-1, IGFBP-3, and IGFBP-5 indirectly influence cell growth by modulating the interaction of IGF-I with the IGF-I receptor and also directly influence cell growth by interacting with distinct cell surface receptors. IGFBP-1 interacts with the ␣ 5 ␤ 1 integrin receptor expressed by placental cells and Chinese hamster ovary cells (7). IGFBP-3 interacts with the Type V TGF-␤ receptor expressed in T47D breast cancer cells and mink lung epithelial cells (8). Recently, an IGFBP-5-specific receptor has been characterized in mouse osteoblasts and rat kidney mesangial cells (9). IGFBP-5 binds with high affinity to this ϳ420-kDa membrane-bound receptor protein and elicits autophosphorylation of serine residues (10). One intracellular signaling pathway coupled to this receptor is the small Gprotein, Cdc42, through which IGFBP-5-dependent mesangial cell filopodia formation is mediated (11).
Human intestinal smooth muscle cells produce IGF-I, and three IGF-binding proteins, IGFBP-3, IGFGBP-4 and IGFBP-5, each of which plays an autocrine role in the regulation of growth in human intestinal muscle cells (12,13). Binding of IGF-I to the IGF-I receptor tyrosine kinase activates distinct PI3-kinase-dependent and Erk1/2-dependent pathways that stimulate both proliferation and IGFBP-5 production (13,14). IGF-I-dependent stimulation of growth in these cells is inhibited by the indirect actions of IGFBP-3 and IGFBP-4 and is augmented by the indirect actions of IGFBP-5 (12,13). IGF-I and IGFBP-5 expression is increased within the intestinal muscle layer in regions of active inflammation and stricturing in Crohn's disease and in models of experimental enterocolitis (15,16). It is not known whether an IGFBP-5-specific receptor is expressed by human intestinal muscle cells or what role this receptor plays in the regulation of growth.
Isolation and Culture of Muscle Cells from Human Jejunum-Muscle cells were isolated from the circular muscle layer of human jejunum as described previously (12,13,17). Segments of normal jejunum were obtained from patients undergoing surgery according to a protocol approved by the Institutional Committee on the Conduct of Human Research. Briefly, muscle cells were isolated by enzymatic digestion overnight at 37°C in Dulbecco's modification of Eagle's medium containing 10% fetal bovine serum (DMEM-10), penicillin 200 units/ml, streptomycin 200 g/ml, gentamycin 100 g/ml, and amphotericin B 2 g/ml, with added 0.0375% collagenase (CLS type II), and 0.1% soybean trypsin inhibitor. The cells were plated at a concentration of 5 ϫ 10 5 cells/ml in DMEM-10 with antibiotics and incubated in a 10% CO 2 environment at 37°C. Medium was replaced every 3 days. Studies were performed in first passage after 14 days, at which time the cells are post-confluent and the production of endogenous IGF-I and IGFBP-5 are low (18).
[ 3 H]Thymidine Incorporation Assay-Proliferation of smooth muscle cells in culture was measured by the incorporation of [ 3 H]thymidine as described previously (14,18,19). Briefly, the cells were washed free of serum and incubated for 24 h in serum-free DMEM. The quiescent muscle cells were incubated for an additional 24 h with a maximally effective concentration of IGFBP-5 (50 nM) in the presence and absence of various test agents. During the final 4 h of this incubation period, 1 Ci/ml [ 3 H]thymidine was added to the medium. [ 3 H]Thymidine incorporation into the perchloric acid extractable pool was used as a measure of DNA synthesis.
Western Blot Analysis-The phosphorylation of Raf1, MKK1/2, Erk1/2, MKK3/6, and p38 MAP kinase was measured by Western blot analysis using standard methods (12)(13)(14). Briefly, post-confluent muscle cells were rendered quiescent by incubation for 24 h in serum-free medium. The cells were stimulated with recombinant human IGFBP-5 in the presence of the IGF-I receptor antagonist for periods of time from 0 to 60 min. The cells were rapidly washed with ice-cold phosphatebuffered saline and lysed in sample buffer. Lysates containing equal amounts of protein were boiled for 5 min, and the proteins were separated with SDS-PAGE under denaturing conditions. The proteins were electrotransferred to nitrocellulose membranes. Nitrocellulose membranes were incubated overnight with a 1:1000 -1:2000 dilution of antibodies specifically recognizing phosphorylated (activated) signaling intermediates in the p38 MAP kinase and Erk1/2 pathways: Raf1 (Ser 259 ), MKK1/2 (Ser 217 /Ser 221 ), Erk1/2 (Thr 202 /Tyr 204 ), MKK3/6 (Ser 189 /Ser 207 ), or p38 MAP kinase (Thr 180 /Tyr 182 ). Bands of interest corresponding to these phosphorylated signaling intermediates were visualized with chemiluminescence and quantitated with densitometry.
Measurement of Ras Activation-Activation of Ras by IGFBP-5 was measured by immunoprecipitation of GTP-bound (activated) Ras and subsequent Western blot of Ras according to the method of Taylor et al. (20). Confluent muscle cells growing in 100-mm plates were incubated in serum-free DMEM for 24 h. The cells were stimulated with IGFBP-5 (0.5 -5 nM) for 0 -5 min. The reaction was terminated by washing with ice-cold phosphate-buffered saline. The cells were lysed in immunopre-cipitation buffer consisting of 25 mM HEPES (pH 7.5), 150 nM NaCl, 1% Igepal CA-630, 10 mM MgCl 2 , 1 mM EDTA, and 10% glycerol (v/v) to which was added 10 g/ml aprotinin, 10 g/ml leupeptin, 25 mM NaF, and 1 nM sodium orthovanadate. Cell lysates containing equal amounts of protein were incubated for 30 min at 4°C with 5 g of a glutathione S-transferase fusion protein corresponding to the Ras-binding domain (residues 1-149) of Raf1 coupled to glutathione-agarose (20). The immunoprecipitated proteins were washed three times with lysis buffer and resuspended in 2ϫ Laemmli sample buffer. The proteins were boiled for 5 min and then separated on 15% agarose gels by SDS-PAGE. The separated proteins were electrotransferred to nitrocellulose membranes. The membranes were incubated overnight with 1 g/ml anti-Ras antibody (clone RAS10). The bands of interest corresponding to activated Ras were visualized with enhanced chemiluminescence and quantitated with densitometry.
Measurement of IGF-I Production by Radioimmunoassay-IGF-I production was measured as described previously (18). Confluent muscle cells growing in 100-mm plates were washed free of serum. Serumfree DMEM was conditioned by incubation for 24 h with the muscle cells. Samples of conditioned medium were subjected to acid-ethanol extraction to remove IGF-binding proteins from the secreted IGF-I as described previously (18) according to the method of Daughaday et al. (21). Briefly, aliquots of conditioned medium were added to an acidethanol mixture (87.5% ethanol:1.5% 2 N HCl (v/v)) at a ratio of 1:4. Samples were incubated at room temperature for 30 min. Samples were centrifuged, and the supernatant was neutralized with 0.855 M Tris base at a ratio of 5:2 and incubated at 4°C for an additional 2 h. After centrifugation the resultant IGFBP-free supernatants were assayed for immunoreactive IGF-I by radioimmunoassay using a polyclonal antibody raised in rabbits against human IGF-I. This antibody reacts fully with human IGF-I, has Ͻ0.02% cross-reactivity with IGF-II, and has no cross-reactivity with insulin. The limit of detection was 10 pg/tube, and the IC 50 was 187 pg/tube. IGF-I was measured in duplicate using 100-l aliquot samples. Production was expressed as pmol of IGF-I/mg of protein/24 h.
Transient Transfection of Human Intestinal Muscle Cells-Substitution of asparagine for serine at residue 17 (S17N) of Ras results in a 24 -40-fold decrease in its affinity for GTP without affecting its affinity for GDP (22). When expressed in cells, Ras(S17N) exerts a dominant negative-like effect by sequestering guanine-nucleotide exchange factors for Ras. cDNA for Ras(S17N) in the pUSEamp(ϩ) expression vector was purified, and human intestinal smooth muscle cells were transiently transfected with either pUSEamp(ϩ)-Ras(S17N) cDNA or with pUSEamp(ϩ) alone as control using LipofectAMINE PLUS ™ Reagent Kit (Invitrogen). Cells are incubated for 3 h at 37°C with the transfection reagent-DNA complexes. The DNA-containing medium was replaced with DMEM ϩ 10% FCS. After a 48-h incubation, the dominant negative effect of Ras(S17N) was confirmed by assaying IGFBP-5-stimulated Ras activity as described above.
Statistical Analysis-Values given represent the mean Ϯ S.E. of n experiments, where n represents 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.

IGFBP-5 Stimulates Growth and IGF-I Secretion Independently of IGF-I-Our
previous work has shown that IGFBP-5 augments, in a concentration-dependent fashion, the stimulatory effects of IGF-I on the growth of human intestinal smooth muscle cells in culture (13). In the present study, we hypothesized that, in addition to its IGF-I-dependent effects, IGFBP-5 might exert direct effects on the growth of human intestinal smooth muscle cells. Cells were examined during the postconfluent phase of culture, when the endogenous levels of both IGFBP-5 and IGF-I are lowest, and in the presence of an IGF-I receptor antagonist, IGF-I analog (1 M) (17,23). IGF-I analog, an IGF-I receptor antagonist that blocks the ability of IGF-I to initiate autophosphorylation of the IGF-I receptor tyrosine kinase, was used to eliminate the effects of IGFBP-5 mediated by facilitation of IGF-I binding to its cognate receptor. We have previously shown that in the presence of this antagonist, the ability of IGF-I to cause phosphorylation of its receptor is abolished. Incubation of quiescent muscle cells with IGFBP-5 (50 nM) for 2 min in the presence of the IGF-I antagonist did not elicit IGF-I receptor phosphorylation (1.03 Ϯ 0.10% of basal). The results of these initial studies implied that effects attributed to IGFBP-5 represented its IGF-I-independent effects, i.e. when the IGF-I-dependent effects of IGFBP-5 were abolished in the presence of the IGF-I receptor antagonist.
In normal human intestinal smooth muscle cells, incubation of quiescent muscle cells with IGFBP-5 for 24 h directly caused concentration-dependent (0.5-50 nM IGFBP-5) increase in [ 3 H]thymidine incorporation (50 nM, 145 Ϯ 9% above basal; basal, 84 Ϯ 3 cpm/mg protein) (Fig. 1). Incubation of human intestinal muscle cells for 24 h with 50 nM IGFBP-5, increased the secretion of IGF-I by 90 Ϯ 12% above basal levels (basal, 3.1 Ϯ 0.2; IGFBP-5, 5.7 Ϯ 0.6 pmol/mg protein/24 h, p Ͻ 0.05). The ability of IGFBP-5 to stimulate thymidine incorporation and IGF-I secretion in the presence of the IGF-I receptor antagonist implied that these effects were distinct from those mediated via the IGF-I receptor by augmentation of IGF-I binding.
IGFBP-5 Activates the p38 MAP Kinase Signaling Pathway-Activation of p38 MAP kinase was measured using a phospho-specific antibody recognizing the Thr 180 /Tyr 182 phosphorylated (activated) p38 MAP kinase isoform. IGFBP-5 caused rapid, time-dependent phosphorylation of p38 MAP kinase, which attained a maximum within 5 min and declined to lower levels by 60 min (Fig. 2A). The increase in p38 MAP kinase phosphorylation, measured at the 5 min maximum, was concentration-dependent (0.5-50 nM IGFBP-5) (Fig 2B).
IGFBP-5 Activates the Erk1/2 Signaling Pathway-Incubation of quiescent muscle cells with IGFBP-5 in the presence of the IGF-I receptor antagonist elicited a prompt, time-dependent phosphorylation of both Erk1/2 isoforms on Thr 202 /Tyr 204 .

IGFBP-5 Activates
Ras-Two methods were used to identify the role of Ras in the signaling pathways activated by IGFBP-5. The first method measured IGFBP-5-induced Ras activation as Ras-GTP using an immunoprecipitation-based assay as described under "Experimental Procedures" (20). The second method identified the requirement for Ras activation in the signaling pathways leading to p38 MAP kinase and Erk1/2 activation by expression of a dominant negative Ras(S17N) mutant in human intestinal muscle cells (22).
Incubation of quiescent human intestinal smooth muscle cells with IGFBP-5 caused time-dependent activation of Ras that was rapid, occurring within 30 s, sustained for up to 2 min, and then declined rapidly to lower levels by 5 min (Fig. 7A). Activation of Ras by IGFBP-5 (0.5-50 nM), measured at the 2 min maximum, was also concentration-dependent (Fig. 7B).
The requirement for Ras activation in IGFBP-5-induced p38 MAP kinase and Erk1/2 activation was examined in cells transiently transfected with a dominant negative Ras(S17N) mutant. In these cells, the ability of 50 nM IGFBP-5 to stimulate p38 MAP kinase phosphorylation was abolished (110 Ϯ 10% inhibition versus vector transfected control, p Ͻ 0.05) (Fig. 8A). In cells expressing the dominant negative Ras mutant, the ability of 50 nM IGFBP-5 to stimulate Erk1/2 phosphorylation was also abolished (98 Ϯ 8% inhibition versus vector-transfected control, p Ͻ 0.05) (Fig. 8B). The dominant negative effect of the Ras(S17N) mutant on Ras activation was confirmed in separate studies. Incubation of quiescent muscle cells transiently transfected with empty vector for 2 min with 50 nM IGFBP-5 elicited an increase in activated, GTP-bound Ras (270 Ϯ 40% above basal), whereas in cells expressing RAS(S17N), the ability of IGFBP-5 to activate Ras was abolished (9 Ϯ 2% above basal, p Ͻ 0.05 versus empty vector).  (Fig. 9). At the 1 M concentrations used in the present study, SB203580 selectively inhibits p38 MAP kinase activation without affecting other protein kinases (25,26).
In the presence of the combination of the p38 inhibitor, SB203580 (1 M), and the MKK1/2 inhibitor, U1026 (10 M), the ability of 50 nM IGFBP-5 to stimulate [ 3 H]thymidine incorporation in human intestinal muscle cells was nearly abolished at 89 Ϯ 2% inhibition (p Ͻ 0.01) (Fig. 9). The results suggest that activation of these two Ras-dependent pathways, p38 MAP kinase and Erk1/2, in human intestinal smooth muscle cells IGFBP-5 Stimulates IGF-I Secretion by Activation of p38 MAP Kinase and Erk1/2-We have previously shown that IGF-I stimulates IGFBP-5 production (13) by activation of the same signaling pathways, Erk1/2 and PI3-kinase, that mediate the growth stimulatory effects of IGF-I on human intestinal muscle cells. In the present study, the possibility that the increase in IGF-I secretion mediated by IGFBP-5 might be regulated by the same pathways, p38 MAP kinase and Erk1/2, was also examined. The increase in IGF-I secretion induced by IGFBP-5 was partially inhibited by the p38 MAP kinase inhibitor, SB203580 (1 M) (63 Ϯ 8% inhibition), or the MKK1/2 inhibitor, U1026 (10 M) (56 Ϯ 12% inhibition). In the presence of the combination of the two inhibitors, the increase induced by IGFBP-5 was abolished, and basal levels of IGF-I secretion were slightly inhibited (Fig. 10). DISCUSSION The present study provides the first evidence that IGFBP-5 has direct, IGF-I receptor-independent effects in normal human intestinal smooth muscle cells. The results show that in human intestinal smooth muscle cells, IGFBP-5 acting independently of the IGF-I receptor causes Ras-dependent activation of both p38 MAP kinase and Erk1/2, which jointly stimulate both muscle cell proliferation and IGF-I secretion. One explanation for the IGF-I-independent effects of IGFBP-5 on growth was provided by Andress and colleagues (9, 11) whose work suggested first in murine bone cells and later in rat glomerular mesangial cells the presence of an ϳ420-kDa IG-FBP-5-specific receptor. In cultured neonatal mouse osteoblasts, the receptor undergoes autophosphorylation on serine residues in response to . In rat glomerular mesangial cells, IGFBP-5 mediates Cdc42-dependent reorganization of the actin cytoskeleton and formation of filopodia (11). The evidence supporting the operation of IGFBP-5-dependent, IGF-I-independent effects in human intestinal smooth muscle cells can be summarized as follows. 1) Incubation of muscle cells with IGFBP-5 causes time-and concentration-dependent activation of both the p38 MAP kinase and the Erk1/2 signaling pathways. 2) Incubation of muscle cells with IGFBP-5 causes concentration-dependent stimulation of proliferation. 3) Incubation of muscle cells with IGFBP-5 does not elicit activation of PI3-kinase as would occur if IGF-I/IGF-I receptor activation were involved (14). 4) Incubation of muscle cells with IGFBP-5 causes a p38 MAP kinase-and Erk1/2-dependent increase in IGF-I secretion. 5) In the presence of the IGF-I receptor antagonist, IGFBP-5 does not elicit IGF-I receptor phosphorylation. IGFBP-5 has been shown to stimulate [ 3 H]thymidine incorporation in osteoblast and bone cells independently of IGF-I and the IGF-I receptor (27). The mechanisms mediating the direct, proliferative effects of IGFBP-5 were not delineated. On the basis of the current study and previous work, two potential mechanisms have been identified. The first involves a membrane-bound, IGFBP-5-specific receptor first characterized by Andress in osteoblast cells (9 -11), and the second involves nuclear transport of IGFBP-5 and direct nuclear effects mediated in the fashion of a ligand-dependent transcription factor. The latter mechanisms was identified in T47D breast cancer cells by Schedlich and colleagues (28,29). The carboxyl terminus of IGFBP-5 (and IGFBP-3, which also possesses IGF-Iindependent effects (30,31)), share a common nuclear localization sequence, IGFBP-3-(215-232) and IGFBP-5-(201-218). This sequence has been shown to be required for the importin-␤-dependent nuclear translocation of both IGFBP-5 and IGFBP-3 (8,29). Although this portion of IGFBP-5 also has been shown to participate in binding to the IGFBP-5 receptor, deletional studies have demonstrated that this sequence of the IGFBP-5 peptide is not required for activation of its cell surface receptor (27). Nuclear transport of IGFBP-5 was not specifically addressed in the current study; however, the ability of the p38 and MKK1/2 inhibitors in combination to abolish both proliferation and IGF-I secretion in response to IGFBP-5 suggests that if nuclear transport of IGFBP-5 does occur in these cells, it is not required for IGFBP-5-stimulated proliferation or IGF-I secretion.
The intracellular signaling cascades coupled to activation of the IGFBP-5 receptor are largely unknown but appear to begin following the autophosphorylation of serine residues of the receptor (10). In rat glomerular mesangial cells, activation of the IGFBP-5 receptor results in Cdc42-GTPase-activating protein aggregation, reorganization of the actin cytoskeleton, and filopodia formation (11). Whether this signaling pathway is involved in the proliferative responses to IGFBP-5 is unknown. Growth factors utilize distinct small GTPases to mediate specific cytoskeletal reorganizations such as lamellipodium formation (mediated via Rac), stress fiber formation (mediated via Rho), or filopodia formation (mediated via Cdc42). The small G-proteins Rho and Cdc42 are expressed by intestinal smooth muscle cells and mediate neurotransmitter-induced contraction (32). In human intestinal smooth muscle and other muscle cell types, the proliferative effects of IGFBP-5 are mediated by the activation of distinct Raf-Erk1/2 and p38 MAP kinase pathways. In COS-7 and HeLa cells, the small GTPases Rac, Rho, and Cdc42 are requisite cofactors in Ras-dependent Raf activation and subsequent activation of Erk1/2 (33)(34)(35). These monomeric G-proteins have also been shown to play a role in the activation of MKK3/6 leading to p38 MAP kinase activation (36,37). It remains to be determined whether IGFBP-5-induced growth mediated by the Raf-Erk1/2 or p38 MAP kinase pathways involves the participation of the GTPases Rho, Rac, and Cdc42.
Based on our previous work and the results of this current study we propose the following model whereby a positive feedback mechanism links IGF-I and IGFBP-5 (Fig 11). IGF-I, acting via its cognate receptor and facilitated by the presence of IGFBP-5, activates the PI3-kinase and Erk1/2 pathways that jointly mediate increased proliferation and secretion of IGFBP-5 (14,17). IGFBP-5, in turn, acting via its cognate receptor, activates the p38 MAP kinase and Erk1/2 pathways that jointly mediate increased proliferation and further secretion of IGF-I.
The concomitant presence of IGF-I and IGFBP-5 and the resultant interplay between their signaling pathways may be important factors in mediating their individual effects. We and others (13,38) have shown that IGFBP-5 protein levels are highly sensitive to protein synthesis inhibitors. Inhibition of RNA stabilizing factor translation by inhibitors such as cycloheximide leads to rapid IGFBP-5 mRNA degradation and results in IGFBP-5 levels falling rapidly to undetectable levels. The 3Ј-untranslated region of IGFBP-5 mRNA contains several adenosine-uridine-rich elements (38), which confer instability to IGFBP-5 mRNA but can be stabilized by binding cytoplasmic proteins. Hou and colleagues (38) have suggested that IGF-I might stimulate the production of AU-binding proteins that stabilize IGFBP-5 mRNA. In NIH 3T3 cells, the regulation of mRNA abundance is a dynamic process in which stabilizing and destabilizing AU-binding proteins compete (39). In these cells, PI3-kinase and p38 MAP kinase function by activating pathway specific stabilizing AU-binding proteins that regulate interleukin-3 mRNA levels (39). We have previously shown that in human intestinal muscle cells, IGF-I increases IGFBP-5 levels via Erk1/2-dependent and PI3-kinase-dependent mechanisms (17). We hypothesize that the ability of IGF-I to increase IGFBP-5 mRNA levels in human intestinal muscle cells might reflect in part the ability of IGF-I-activated signaling intermediates, such as PI3-kinase, to act as stabilizing kinases for IGFBP-5 mRNA.
The results of this study show that IGFBP-5 stimulates the secretion of IGF-I in intestinal muscle cells by activation of distinct p38 MAP kinase-dependent and Erk1/2-dependent pathways. The IGF-I gene has two alternative first exons that are spliced to a common block of exons that contain the mature peptide-containing sequence. In most tissues, IGF-I gene expression is regulated through activation of exon 1 (40). Several potential regulators of IGF-I gene transcription are known to be downstream targets for p38 MAP kinase including members of the CCAAT/enhancer-binding protein (C/EBP) and ATF/ CREB family; downstream targets of Erk1/2 also include members of the ATF/CREB family and the AP-1 family (40 -43). The specific transcriptional regulators of the IGF-I gene in human intestinal smooth muscle remain to be determined.
The positive feedback loop linking IGFBP-5 and IGF-I secretion may have clinical importance in the setting of intestinal inflammation, e.g. Crohn's disease. The presence of increased IGF-I and IGFBP-5 expression by intestinal smooth muscle cells in animal models of enterocolitis has been appreciated for a number of years (15), and this observation has recently been extended to include Crohn's disease in humans (16). The expression of both IGF-I and IGFBP-5 is increased in parallel in regions of active inflammation and stricture formation. In addition to muscle proliferation induced by IGFBP-5 and IGF-I, IGFBP-5 also stimulates collagen secretion (44). The mechanisms responsible for initiating this response have yet to be fully elucidated, but the resulting muscle hyperplasia and extracellular matrix production may be responsible, in part, for luminal narrowing and stricture formation in the intestine.
In summary, the present paper shows that IGFBP-5 exerts direct, IGF-I-receptor-independent effects in human intestinal smooth muscle cells. By jointly activating p38 MAP kinase and Erk1/2, IGFBP-5 stimulates both muscle cell proliferation and IGF-I secretion. A positive feedback mechanism linking IGFBP-5 and IGF-I secretion in these cells reinforces their individual ability to cause intestinal smooth muscle cell growth.