Growth Hormone Inhibits Signal Transducer and Activator of Transcription 3 Activation and Reduces Disease Activity in Murine Colitis  Xiaonan Han, Danuta Sosnowska, Erin L. Bonkowski, Lee A. Denson  Gastroenterology  Volume 129, Issue 1, Pages 185-203 (July 2005) DOI: 10.1053/j.gastro.2005.05.018 Copyright © 2005 American Gastroenterological Association Terms and Conditions

Figure 1 GH administration improves colon disease activity. (A) GHR-expressing cells were identified in PBS-treated WT and IL-10-null mice by immunohistochemistry with the AL-47 GHR antibody. A, WT; B, IL-10 null; C, negative control. Representative positive CECs and LPMCs are identified with closed arrows. (B) WT and IL-10-null mice received human GH or PBS for 2 weeks, and colon histology was scored. Representative H&E-stained colon sections are shown. A, WT/PBS; B, WT/GH; C, IL-10 null/PBS; D, IL-10 null/GH. (C) Scores for PBS- and GH-treated IL-10-null mice and the associated mean values (——) are shown; n = 6–8; P = .04. Gastroenterology 2005 129, 185-203DOI: (10.1053/j.gastro.2005.05.018) Copyright © 2005 American Gastroenterological Association Terms and Conditions

Figure 2 GH regulates colon CEC and LPMC apoptosis and proliferation. Apoptotic cells were identified by using (A) TUNEL (brown) or (B) an anti–activated caspase 3 antibody. (C) Proliferating cells were identified by using PCNA (green) or Ki-67 (red). Typical positive cells are identified with closed arrows. Note the inset showing representative PCNA-positive CECs in colitis. CON, negative control; KO, IL-10 null (original magnifications, respectively, 40× for the full-size image and 100× for the inset in C; bar = 50 μm). Gastroenterology 2005 129, 185-203DOI: (10.1053/j.gastro.2005.05.018) Copyright © 2005 American Gastroenterological Association Terms and Conditions

Figure 3 GH regulates colon fas/fasL and bcl-2/bax expression. (A) Colon fas/fasL and bcl-2/bax protein abundance were determined by immunoblot with cytosolic extracts from whole colon. Signal intensity was determined by densitometry and is shown as the mean ± SD (n = 6–8; *P < .05 vs untreated controls of same genotype; ^P < .05 vs untreated WT control). (B) CECs and LPMCs expressing bcl-2 or bax were identified by immunofluorescence; representative positive cells are identified with closed arrows. CON, negative control; KO, IL-10 null (original magnifications, 200× and 400×, respectively; bar = 50 μm). Gastroenterology 2005 129, 185-203DOI: (10.1053/j.gastro.2005.05.018) Copyright © 2005 American Gastroenterological Association Terms and Conditions

Figure 4 GH does not increase colon IGF-I expression in colitis. Colon IGF-I RNA expression was determined by real-time PCR and primers specific for total IGF-I, the exon 1-containing transcript, and the exon 2-containing transcript. (A) Total colon IGF-I RNA expression is shown as the mean ± SD (*P < .05 vs untreated control of same genotype; ^P < .05 vs untreated WT control; n = 6–8). (B) Colon IGF-I exon 1 and exon 2 transcript expression is shown as the mean ± SD (^P < .05 vs untreated WT control; n = 6–8). KO, knockout. Gastroenterology 2005 129, 185-203DOI: (10.1053/j.gastro.2005.05.018) Copyright © 2005 American Gastroenterological Association Terms and Conditions

Figure 5 GH reduces colon STAT3 activation and DNA binding. (A and B) WT and IL-10-null mice received rat GH (1 μg/g intraperitoneally; single dose) or PBS 30 minutes before death or received human GH (3 μg/g subcutaneously twice daily; chronic dose) or PBS for 2 weeks, and colon STAT3 tyrosine phosphorylation and DNA binding were determined. (A) Immunoblots for nuclear STAT3 (NE STAT3) and nuclear tyrosine-phosphorylated STAT3 (NE PSTAT3) are shown. SH-PTP1 was used to control for protein loading. KO, knockout. (B) Nuclear proteins were prepared from mouse colon and from colon biopsy samples obtained from children with normal tissue (Con) or newly diagnosed Crohn’s colitis (Colitis). STAT3 DNA binding was determined by EMSA; supershift was performed by using a STAT3 antibody and competition with an excess of the unlabeled STAT3 oligonucleotide. (C) Whole colon from WT and IL-10-null (KO) mice was maintained in short-term tissue culture, and the effect of human GH (500 ng/mL for 60 minutes) on nuclear pSTAT3 and STAT3 abundance was determined. Membranes were reprobed for SH-PTP1 to control for nuclear protein loading. Signal intensity was determined by densitometry and is shown as the mean ± SD (*P < .05 vs untreated control of the same genotype; ^P < .05 vs untreated WT control; n = 6–8). Gastroenterology 2005 129, 185-203DOI: (10.1053/j.gastro.2005.05.018) Copyright © 2005 American Gastroenterological Association Terms and Conditions

Figure 6 GH-dependent reduction in CEC and lamina propria T-cell STAT3 activation is associated with improved colon histopathology. (A) Cells expressing tyrosine-phosphorylated STAT3 were identified by immunohistochemistry. KO, IL-10 null; CON, negative control. (B) The correlation between colon pSTAT3 nuclear abundance determined by immunoblot and colon histology score in GH-treated IL-10-null mice was determined and is shown. (C) Cells expressing both tyrosine-phosphorylated STAT3 (pSTAT3; red) and CD3 (green) were identified by immunofluorescence and colocalization (orange). Representative positive cells are identified with closed arrows. Note that the upper row in (C) shows representative results for CECs, whereas the middle row shows representative results for lamina propria T cells in untreated colitic mice. The percentage of CD3 cells that were positive for pSTAT3 was determined in untreated and GH-treated IL-10-null mice and is shown in the bar graph as the mean ± SD (n = 6–8; *P < .05; original magnification, 400× bar = 50 μm). Gastroenterology 2005 129, 185-203DOI: (10.1053/j.gastro.2005.05.018) Copyright © 2005 American Gastroenterological Association Terms and Conditions

Figure 6 GH-dependent reduction in CEC and lamina propria T-cell STAT3 activation is associated with improved colon histopathology. (A) Cells expressing tyrosine-phosphorylated STAT3 were identified by immunohistochemistry. KO, IL-10 null; CON, negative control. (B) The correlation between colon pSTAT3 nuclear abundance determined by immunoblot and colon histology score in GH-treated IL-10-null mice was determined and is shown. (C) Cells expressing both tyrosine-phosphorylated STAT3 (pSTAT3; red) and CD3 (green) were identified by immunofluorescence and colocalization (orange). Representative positive cells are identified with closed arrows. Note that the upper row in (C) shows representative results for CECs, whereas the middle row shows representative results for lamina propria T cells in untreated colitic mice. The percentage of CD3 cells that were positive for pSTAT3 was determined in untreated and GH-treated IL-10-null mice and is shown in the bar graph as the mean ± SD (n = 6–8; *P < .05; original magnification, 400× bar = 50 μm). Gastroenterology 2005 129, 185-203DOI: (10.1053/j.gastro.2005.05.018) Copyright © 2005 American Gastroenterological Association Terms and Conditions

Figure 7 GH administration reduces colon SOCS-3 expression in colitis. (A) Colon SOCS-3 RNA expression was determined by using real-time PCR, (B) protein abundance was determined by using whole colon cytosolic proteins and immunoblot, and (C) cellular localization was determined by using immunohistochemistry. Immunoblots were reprobed for actin to control for protein loading. In (B), the effect of coincubation with a SOCS-3 blocking peptide is shown. In (C), representative positive cells are identified with closed arrows (original magnification, 400× bar = 50 μm). Bar graphs depict the mean ± SD (n = 6–8; *P < .05 vs untreated control of the same genotype; ^P < .05 vs untreated WT control). CON, negative control; WT, wild type; KO, IL-10 null. Gastroenterology 2005 129, 185-203DOI: (10.1053/j.gastro.2005.05.018) Copyright © 2005 American Gastroenterological Association Terms and Conditions

Figure 8 GH increases colon SH-PTP2/gp130 binding. Colon cytosolic (A) or membrane (B) protein SH-PTP2/gp130 binding was determined by using coimmunoprecipitation and immunoblot. Immunoprecipitation (IP) was performed with SH-PTP2 with immunoblot for gp130, followed by stripping of membranes and reprobing for SH-PTP2. Signal intensity for the SHP-2/gp130 complex and for SHP-2 was determined by densitometry and is shown as the mean ± SD (n = 6–8; *P < .05 vs untreated control of the same genotype). WT, wild type; KO, IL-10 null. Gastroenterology 2005 129, 185-203DOI: (10.1053/j.gastro.2005.05.018) Copyright © 2005 American Gastroenterological Association Terms and Conditions

Figure 9 GH induces SH-PTP2/gp130 association and inhibits IL-6 activation of STAT3 in human colon epithelial cells. T84 cells were treated with IL-6 (100 ng/mL) or human GH (500 ng/mL) for the times shown, and (A) nuclear tyrosine-phosphorylated STAT3 and STAT5b abundance (pSTAT3 and pSTAT5) and (B) cytosolic SH-PTP2/gp130 association and SH-PTP2 tyrosine phosphorylation (PY20 blot) and abundance (SH-PTP2 blot) were determined by immunoblot. (C) Cytosolic SOCS-3 expression was determined with Western blot. In (A), signal intensity was determined by densitometry for nuclear pSTAT3 abundance and is shown in the bar graph as the mean ± SD (n = 3; *P < .05 vs IL-6–treated control). In (B), protein was immunoprecipitated by using the SH-PTP2 antibody, and then membranes were probed for gp130, SH-PTP2, and PY20. Immunoblots shown in (A) and (B) are representative of 3 independent experiments. IL6-30m/GH, cells were cotreated with IL-6 and GH for 30 minutes before nuclear protein preparation. CE, cytosolic extracts; NE, nuclear extracts; m, minutes; Ig, immunoglobulin. Gastroenterology 2005 129, 185-203DOI: (10.1053/j.gastro.2005.05.018) Copyright © 2005 American Gastroenterological Association Terms and Conditions

Figure 10 GH induces SH-PTP2/gp130 association and inhibits IL-6 activation of STAT3 in Jurkat human T cells. Jurkat cells were treated with IL-6/sIL-6R (100 ng/mL), human GH (500 ng/mL), or both, and (A) nuclear tyrosine-phosphorylated STAT3 and STAT5, (B) cytosolic SH-PTP2/gp130 association, and (C) cytosolic SOCS-3/gp130 association were determined by immunoblot. In (A), signal intensity was determined by densitometry for nuclear pSTAT3 abundance and is shown in the bar graph as the mean ± SD (*P < .05). In (D), Jurkat T cells and IM-9 B cells were treated with human GH or IL-2 for the times shown, and cytosolic GHR and nuclear P-STAT5 abundance were detected by immunoblot. Immunoblots shown are representative of 3 independent experiments. IL6-30m/GH, cells were cotreated with IL-6 and GH for 30 minutes before nuclear protein preparation. NE, nuclear extracts; m, minutes. Gastroenterology 2005 129, 185-203DOI: (10.1053/j.gastro.2005.05.018) Copyright © 2005 American Gastroenterological Association Terms and Conditions