Insulin-like growth factors (IGFs) are important mediators of bowel growth and maturation (1–3). We have previously implicated IGF-I as a mediator of dexamethasone effect in the newborn mouse ileum (4). This model was originally designed to mimic a clinical disorder observed in extremely low-birth-weight infants, where early postnatal dexamethasone treatment results in an increased incidence in focal small bowel perforation within the ileum (5–7). Both dexamethasone-treated newborn mice and dexamethasone-treated extremely low-birth-weight infants who acquire these ileal-specific perforations demonstrate robust mucosal growth concomitant with a thinning of the bowel wall. This altered growth pattern parallels the redistribution of IGF-I, suggesting it as an important intermediate of dexamethasone effect (Fig. 1) (8).
Although steroids are known to affect tissue growth and maturation in many neonatal organs, the mechanism(s) by which glucocorticoids affect the IGF system remain poorly characterized. Several mechanisms may be involved in dexamethasone-mediated IGF-I redistribution, including regional changes in 1) IGF-1 synthesis, 2) IGF-I uptake and degradation, or 3) IGF binding protein (IGFBP) abundance.
IGFBPs bind IGF-1 with variable affinity, and their individual effects can either inhibit or augment the availability of IGF-I to bind to its receptor (1). In addition, individual IGFBPs also have separate affinities for different extracellular matrix constituents, and these interactions may be important for region-specific localization of individual IGFBPs. We hypothesized that dexamethasone alters the distribution of IGF-I in the ileum by changing the mucosal composition of IGFBPs. To test this hypothesis, we assayed for changes in the abundance and distribution of all six IGFBPs and their mRNAs within the tissues of the newborn mouse ileum after dexamethasone treatment.
MATERIALS AND METHODS
Dexamethasone Administration and Tissue Collection
The animal protocols in this study were reviewed and approved by the institutional animal care and use committee at the University of North Carolina. Newborn C57 littermates from six litters were pooled and then equally distributed between the six mothers. Three litters received daily intraperitoneal injections of dexamethasone (1 μg/g), while three litters received saline beginning on the day of birth and repeated each 24 hours for three doses. This dose and length of therapy approximates the original daily dose used in a clinical trial where 17% of dexamethasone-treated infants in the first phase of the study developed focal small bowel perforation (this prompted the safety review committee to lower the dexamethasone dose, and there was a significant decrease in perforations over the remainder of the study) (6). We chose to use intraperitoneal injections rather than subcutaneous or intramuscular injections because this best approximated intravenous delivery, which was not feasible in this animal model.
Six litters of mice were used. Pups were killed after 72 hours of treatment (in the satiated state), and the ileum was immediately removed, fixed in 10% buffered paraformaldehe (Fisher Scientific, Pittsburgh, PA), and embedded in paraffin.
The cDNAs used in this study have been previously described and include rat IGFBP-l (a gift from Dr. L. Murphy, University of Mannitoba, Winnipeg, Manitoba, Canada), rat IGFBP-2, rat IGFBP-3, rat IGFBP-4, rat IGFBP-5, and rat IGFBP-6 (9).
Riboprobes for use in the in situ hybridization analyses were generated from rat IGFBP-1–6 cDNAs by in vitro transcription using fluorescein-isothiocyanate conjugated–UTP (Boerhinger Mannheim Corp., Indianapolis, IN) and T7 DNA-dependent RNA polymerase (antisense probes) or T3 DNA-dependent polymerase (sense strand probes) (Stratagene, La Jolla, CA) (9).
In Situ Hybridization
In situ hybridization was performed as previously described with some modification (9,10). The ileums from 3-day-old mice were dissected immediately on euthanasia and placed directly in 10% buffered paraformaldehyde (Fisher Scientific) at 4°C overnight before being paraffin embedded and sectioned for microscopy. Sections from dexamethasone and control mice were paired and processed on the same slide for the purpose of comparison. Sections were deparaffinized in xylene, then hydrated in decreasing ethanol concentrations to distilled water, digested in proteinase K (10 mg/ml in 100 mmol/L Tris-HC1, pH 8.0, Boehringer Mannheim) for 15 minutes at 37°C, then dehydrated in increasing ethanol concentrations and air dried before hybridization. Hybridization solution (50% deionized formamide, 0.25 mg/ml yeast transfer RNA in 4X SSC [20X SSC = 3 mol/L NaCl, 0.3 mol/L sodium citrate, pH 6.0], 5 ng fluorescein-isothiocyanate conjugated UTP-labeled probe/section) was applied to each section and covered with a glass cover slip. A sense strand RNA probe for each of the IGFBPs was prepared as previously described (18).
Sections were incubated in a humidified chamber at 60°C for 16 to 18 hours. Coverslips were then removed in IX SSC, and sections washed in IX SSC for 60 minutes at 68°C. Slides were washed 4 times at 60°C in IX SSC, then twice in buffer 1 (150 mmol/L NaCl, l00 mmol/L Tris-HC1, pH 7.5). Sections were then soaked in buffer 1 containing 1% bovine serum albumin (Boerhinger Mannheim Corp.) for 30 minutes at 25°C. Anti–fluorescein-isothiocyanate conjugated horse radish peroxidase (1.5 U/mL; Boerhinger Mannheim Corp.) was added in the same buffer and incubated for 2 hours. This was then removed by aspiration, and the slides were washed four times with buffer 1 at 25°C. Biotinyl tyramine (0.007 mmol/L) plus 0.003% H2O2 in buffer 1 was added, and the sections were incubated for a further 7 minutes followed by addition of avidin-biotin-peroxidase (ABC-HRP; Vector Labs, Burlingame, CA) and incubation for 30 minutes at 25°C. Sections were then washed and detection performed in substrate solution (0.05% 3´-3-diaminobenzidine, 0.003% H2O2 in 0.1 mol/L sodium acetate, pH 6.0), which was applied for 5 minutes to allow signal detection. After detection, sections were stained with hematoxylin, coverslipped, and digitized using a Spot Jr. digital camera (Diagnostics, Inc., Sterling Heights, MI).
Specificity controls for in situ hybridization were performed by two methods. First, in situ hybridization was performed in parallel on adjacent sections, but RNA probe was excluded to determine the level of background staining. Second, in situ hybridization was performed as described except that adjacent sections of tissue were pretreated with a cocktail of three RNases (RNase A 1 mg/mL, Roche, Indianapolis, IN; Tl l:50 Ambion's RPA III kit, Ambion, Austin, TX; RNase H 10 U/mL, Promega, Madison, WI) for 1 hour at room temperature to test for nonspecific hybridization.
To allow direct comparisons, dexamethasone-treated and untreated tissues were processed on the same slide as previously described (4). Sections (4 μm) were deparaffinized in xylene and hydrated in descending ethanol concentrations to distilled water. Antigen retrieval was performed by incubating slides in 0.1 mol/L citric acid, pH 8.0, for 10 minutes at 100°C, followed by cooling to 25°C and washing three times in distilled H2O. Endogenous peroxidase activity was quenched in 3% H2O2 in 60% methanol for 15 minutes, followed by three washes of distilled water. Blocking solution (1% wt/vol bovine serum albumin in TBS [0.05 mol/L Tris-HCL, 0.138 mol/L NaCl, 0.0027 mol/L KCl, pH 8.0]) was applied for 1 hour at 25°C, followed by incubation with one of six antibodies, rabbit anti-human IGFBP-l (a gift from Dr. David Clemmons, University of North Carolina at Chapel Hill) (11,12), goat anti-human IGFBP-2–5, (Santa Cruz Biotechnology Inc., Santa Cruz, CA), and rabbit anti-human IGFBP-6 (a gift from Dr. Nicolas Ling) (13) for 16 to 18 hours at 4°C in a humidified chamber. The primary antibody solution was aspirated, and the slides were washed three times at 25°C in TBS. Biotinylated anti-mouse antibody (Vector Labs) or biotinylated anti-rabbit antibody (Jackson Labs, West Grove, PA) was prepared per the manufacturer's instructions and applied for 60 minutes at 25°C, followed by three TBS washes as above. ABC-HRP (Vector Labs) was prepared according to the manufacturer's instructions, applied for 30 minutes, and then washed as above. The peroxidase substrate solution was then applied to each section for 5 minutes. The reaction was terminated by aspiration of the substrate solution followed by washing for 10 minutes in distilled water at 25°C. The sections were counterstained with hematoxylin, dehydrated, and coverslipped, and images were digitized as with in situ hybridization.
Specificity controls were performed by preabsorption of primary antibody with an excess of the antigen (1 g/mL in blocking solution, blocking peptides for IGFBP-1–6 were obtained from Santa Cruz Biotechnology Inc.). Preabsorption was performed overnight, and our standard immunochemistry protocol was then performed in parallel with slides being incubated either in primary antibody or primary antibody preabsorbed with excess antigen.
Ileal sections processed for either in situ hybridization or immunohistochemistry were immediately counterstained with hematoxylin, coverslipped, and imaged with an Olympus microscope using a 20× magnification objective. Digitized images were obtained using a Spot Jr. digital camera (Diagnostics, Inc.).
Paired sections for each IGFBP (for both in situ hybridization and immunohistochemistry) were blinded by an independent party (J.B.P.), such that neither the histochemical technique nor the individual reagents were known to the grader (P.V.G.). The mucosal staining abundance for each section was then assessed with a modified binomial grading system as follows: mucosal staining equivalent to background = 0, mucosal staining above background = 0.5, abundant mucosal staining = 1.0, and highly abundant mucosal staining (evident by color saturation of the horseradish peroxidase substrate) = 2.0. The data were then unblinded, and the median and interquartile ranges were determined for each condition. The abundances were tabulated according to the following predefined definitions: 0 to 0.49 = background staining (−), 0.5 to 0.99 = mucosal staining greater than background (+/−), 1.0 to 1.99 = abundant mucosal staining (+), 2.0 = highly abundant mucosa staining (++). This nonlinear grading paradigm was felt to most accurately reflect the semiquantitative nature of comparative histochemistry. Significant differences between treatment conditions (defined as P ≤ 0.05) were determined by the Mann–Whitney U test.
Dexamethasone Increases Insulin-like Growth Factor Binding Protein-4 and -5 mRNA Abundance
Daily intraperitoneal injections were comparably tolerated between dexamethasone-treated and untreated litters as previously reported (4). In situ hybridization of IGFBP transcripts revealed three distinct patterns of localization in control ileums at 3 days. IGFBP-l, -2, and -3 mRNA transcripts were widely distributed throughout the sections (Table 1 and Figs. 2A, C, E). IGFBP-4 and -6 mRNA transcripts were present above background and localized predominantly in the crypt epithelia (Table 1 and Figs. 2G, K). IGFBP-5 mRNA staining was faint and near background levels (Table 1 and Fig. 2I).
Results from in situ hybridization performed in ileum from dexamethasone-treated mice were similar to those of controls in most cases. No differences in abundance were found between treatment conditions for IGFBP-1–3 and -6 in any of the animals (Table 1 and Figs. 2A–F, K–L). In contrast, there were modest categoric increases in both IGFBP-4 and -5 mRNA transcript abundance, although only IGFBP-4 was significantly different by the Mann–Whitney U test (Table 1 and Figs. 2G–J). Control hybridization, where riboprobe was omitted or the sections were preincubated with a cocktail of RNases, showed minimal to no detection (data not shown).
Dexamethasone Increases the Abundance of Insulin-like Growth Factor Binding Proteins 2–5 in the Distal Villus
Immunohistochemical localization of IGFBP-1, -2, -3, and -5 showed minimal staining in control ileum (Table 1 and Figs. 3A, D, G, M). In contrast, IGFBP-4 was faintly but reliably detected in epithelial cells along the distal portion of the villus in control ileum (Fig. 3J), and IGFBP-6 staining was abundant in all epithelial cells (Fig. 3P). Dexamethasone-treated ileum showed significant increases in the staining for IGFBP-2, -3, -4, and -5 within the distal villi when compared with controls (Table 1 and Fig. 3). This change was most evident for IGFBP-5 and was consistent across all animals. Because of the abrupt pattern change noted with these four IGFBPs, we did preabsorption controls in parallel to further confirm the specificity of the individual antibodies. Both preabsorption of primary antibody with competing antigen and omission of primary antibody demonstrates minimal background staining (Figs. 3F, I, L, O, R).
IGFBP-l was not detected in either treatment condition (Table 1 and Figs. 3A, B), despite our demonstration of abundant mRNA with in situ hybridization (Table 1 and Figs. 2A, B). Immunolocalization of IGFBP-l in hepatocytes processed in parallel (Fig. 3C) suggests that the amount of IGFBP-l in the ileum is below the detection limit for the antiserum used.
In this study, we characterized the changing distribution and abundance of all six IGFBP mRNA transcripts and peptides within a model of dexamethasone-induced mucosal maturation. Our in situ hybridization data demonstrated that dexamethasone significantly increases IGFBP-4 mRNA transcript abundance. IGFBP-5 mRNA transcript abundance also showed a modest increase with dexamethasone treatment (we noted this to be particularly evident in the crypt epithelial cells), but this was not significantly different in our blinded assessment. In contrast, no change in abundance was evident with the other four IGFBP mRNA transcripts.
Our immunolocalization data revealed significant increases in four IGFBPs (-2, -3, -4, and -5) after dexamethasone treatment. Of these, IGFBP-5 was the most dramatic (Table 1 and Fig. 3M, N). This finding is in contrast to the inhibitory effects of glucocorticoids on IGFBP-5 transcription in many in vitro models (including cultured osteoblasts, pituitary cells, and lung explants), leading us to suspect that IGFBP-5 is not directly governed by dexamethasone (14–16). However, IGFBP-5 may be obligatorily up-regulated during intestinal epithelial cell maturation, since progressive increases in IGFBP-5 expression and abundance have been demonstrated during lung epithelial cell development (17,18).
While IGFBP-5 was found to have the greatest change in abundance and distribution, immunolocalization of IGFBP-2, -3, and -4 was also increased in the distal villi of dexamethasone-treated animals. Whether these four IGFBPs accumulate within epithelial cells through increased synthesis or uptake is not known, but their accumulation in dexamethasone-treated ileum parallels our previous observations of IGF-I redistribution from the mesenchyme to the mucosa (Fig. 1) (4,8). Taken together, our findings characterize the changing homeostasis of the IGF system during dexamethasone-induced maturation of the ileal mucosa and are consistent with at least three possible regulatory models: 1) accumulation of select IGFBPs within the mucosa draws IGF-I from the mesenchyme, 2) increasing IGF-I in the mucosa causes the accumulation of select IGFBPs (through synthesis, uptake and/or reduced proteolysis), or 3) the accumulation of IGF-I and one or more IGFBPs are coincident but are not causally related (e.g., IGFBP-5 transcription may be indirectly regulated by glucocorticoids, and IGFBP-5 peptide has been shown to have multiple effects that are independent of IGF-I (15,16)). These models are not exclusive and, given that four different IGFBPs were found to accumulate in the villus, each could prove to be true.
There are some limitations and caveats to this study. First, although we did not find differences in mRNA transcript abundance for IGFBP-1, -2, -3, and -6, this semiquantitative technique is not sufficient to rule out small transcriptional changes between the two treatment conditions, particularly when an mRNA transcript is highly abundant in both conditions. Likewise, it is not clear that our increased immunolocalization of IGFBP-2, -3, -4, and -5 represents up-regulated translation. Increases in IGFBP peptide abundance could be secondary to changes in IGFBP-specific protease activity, since several reports have demonstrated that the abundances of individual IGFBPs are tightly governed by their respective proteases (1,17,18). Alternatively, dexamethasone-related increases in IGFBP immunolocalization could be unique to the suckling state. Theoretically, there could be enhanced lumenal uptake of these IGFBPs, since all four have been found in breast milk and could be accumulating because of induced maturation in digestive function (19–21). Finally, with regard to the IGF/growth axis in the postnatal ileum, the combined effect of these four IGFBPs within the mucosa cannot be assessed by this study. However, we have previously demonstrated that dexamethasone enhances mucosal growth in the newborn mouse ileum, and those findings occur in parallel with the IGFBP changes characterized in this study, suggesting a net increase in IGF-I availability.
In summary, our findings are consistent with the hypothesis that dexamethasone redistributes IGF-I in the neonatal mouse ileum by changing the mucosal composition of IGFBPs. These findings further implicate the IGF system as a participant in mucosal maturation and, since precocious maturation is also seen in extremely premature infants who acquire ileal perforations, we suggest that IGFBPs may be mediators of the dexamethasone-related events that make these infants vulnerable to perforations.
The authors thank Brian Brighton and Ward Jarvis for their technical assistance, Sarabeth Thomas, MS, for reading the manuscript and helpful discussions, and Jessica B. Paxton for the blinding of data.
1. Jones J, Clemmons DR. Insulin-like growth factors and their binding proteins: Biologic actions. Endo Rev 1995; 16:3–34.
2. MacDonald RS. The role of insulin-like growth factors in small intestinal cell growth and development. Horm Metab Res 1999; 31:103–113.
3. Lund K. IGFs and the digestive tract. In: Rosenfield R, Roberts C, eds. Contemporary endocrinology: the IGF system. Totowa, NJ: Humana Press, 1998, pp 519–544.
4. Gordon PV, Price WA, Stiles AD. Dexamethasone
administration to newborn mice alters mucosal and muscular morphology in the ileum
and modulates IGF-I localization. Pediatr Res 2001; 49:93–100.
5. Gordon PV, Rutledge J, Sawin R, Thomas S, Woodrurn D. Early postnatal dexamethasone
increases the risk of focal small bowel perforation
in extremely low birth weight infants. J Perinatal 1999; 19:573–577.
6. Garland JS, Colleen AP, Pauly TH, et al. A three-day course of dexamethasone
therapy to prevent chronic lung disease in ventilated neonates: A randomized trial. Pediatrics 1999; 104:91–99.
7. Stark AR, Carlo WA, Tyson JE, et al. Adverse effects of early dexamethasone
treatment in extremely-low-birth-weight infants. N Engl J Med 2001; 344:95–101.
8. Gordon PV, Marshal DD, Stiles AD, Price WA. The clinical, morphologic, and molecular changes in the ileum
associated with early postnatal dexamethasone
administration: From the baby's bowel to the researcher's bench. Molec Genet Metab 2001; 72:91–103.
9. Retsch-Bogart GZ, Moats-Staats BM, Howard K, D'Ercole AJ, Stiles AD. Cellular localization of messenger RNAs for insulin-like growth factors (IGFs), their receptors and binding proteins during fetal rat lung development. Am J Respir Cell Mol Biol 1996; 14:61–69.
10. Moats-Staats BM, Price WA, Xu L, Jarvis W, Stiles AD. Regulation of the insulin-like growth factor system during normal rat lung development. Am J Respir Cell Mol Biol 1995; 12:56–64.
11. Busby WH, Snyder DK, Clemmons DR. Radioimmunoassay of a 26,000-dalton plasma insulin-like growthfactor-binding protein: Control by nutritional variables. J Clin Endocrinol Metab 1988; 67:1225–1230.
12. Sheikh MS, Shao ZM, Hussain A, et al. Regulation of insulin-like growth factor-binding-protein-1, 2,3,4,5, and 6: Synthesis, secretion, and gene expression in estrogen receptor-negative human breast carcinoma cells. J Cell Physiol 1993; 155:556–567.
13. Chevally T, Strond DD, Mohan S, Baylink D, Linkhart TA. Evidence for a role for insulin-like growth factor binding proteins in glucocorticoid inhibition of normal human osteoblast-like cell proliferation. Eur J Endocrinol 1996; 134:591–601.
14. Cheng SL, Zhang SF, Mohan S, et al. Regulation of insulin-like growth factors I and II and their binding proteins in human bone marrow stromal cells by dexamethasone
. J Cell Biochem 1998; 71:449–458.
15. Fielder PJ, Tauber JP, Wilson KF, Pham HM, Rosenfeld RG. Insulin-like growth factors (IGFs) stimulate and dexamethasone
inhibits IGF binding protein (EP)-5 expression in a mouse pituitary cell line. Growth Regul 1993; 3:226–234.
16. van de Wetering JK, Elfring RH, Oosterlaken-Dijksterhuis MA, Mol JA, Haagsman HP, Batenburg JJ. Perinatal expression of IGFEPs in rat lung and its hormonal regulation in fetal lung explants. Am J Physiol 1997; 273:Ll174–Ll181.
17. Retsch-Bogart GZ, Moats-Staats EM, Howard K, D'Ercole AJ, Stiles AD. Cellular localization of messenger RNAs for insulin-like growth factors (IGFs), their receptors and binding proteins during fetal rat lung development. Am J Resp Cell Mol Biol 1996; 14:61–69.
18. Wallen LD, Myint W, Nygard K, Shimasake S, Clemmons DR, Han VK. Cellular distribution of insulin-like growth factor binding protein
mRNAs and peptides during rat lung development. J Endocrinol 1997; 155:313–327.
19. Donovan SM, McNeil LK, Jimenez-Flores R, Odle J. Insulin-like growth factors and insu1in-like growth factor binding proteins in porcine serum and milk throughout lactation. Pediatr Res 1994; 36:159–168.
20. Gibson CA, Staley MD, Baumracker CR. Identification of IGF binding proteins in bovine milk and the demonstration of IGFEP-3 synthesis and release by bovine mammary epithelial cells. J Anim Sci 1999; 77:1547–1557.
21. Baumrucker CR, Erondu NE. Insulin-like growth factor (IGF) system in the bovine mammary gland and milk. J Mammary Gland Biol Neoplasia 2000; 5:53–64.