Translational inhibition of colonic epithelial heat shock proteins by IFN-gamma and TNF-alpha in intestinal inflammation

Abstract

Background & aims: Inducible heat shock proteins (iHsp), Hsp25/27 and Hsp70, play essential roles in protecting cells against stress and, in intestinal mucosal inflammation, potentially lessening the extent and severity of injury. We examined the expression and regulation of iHsp in human and experimental inflammatory bowel diseases (IBD) and in vitro.

Methods: iHsp expression and regulation were assessed in normal and IBD colonic biopsy specimens, IL-10(-/-) mice, and young adult mouse colonic epithelial cells by immunohistochemistry, Western blot, and real-time polymerase chain reaction (PCR). Phosphorylation of double-stranded RNA-dependent protein kinase (PKR) and eukaryotic initiation factor-2alpha (eIF-2alpha) was determined by Western blot.

Results: Hsp25/27 and Hsp70 levels were selectively reduced in areas of active mucosal inflammation associated with human IBD and IL-10(-/-) mice with colitis. Wild-type mice treated in vivo with interferon (IFN)-gamma + tumor necrosis factor (TNF)-alpha also demonstrated reduced colonic Hsp25/27 and Hsp70. In young adult mouse colonic epithelial cells, IFN-gamma+TNF-alpha inhibited heat induction of Hsp25/27 and Hsp70, an effect not associated with changes in iHsp messenger RNA or protein half-lives but caused by suppressed de novo iHsp synthesis. IFN-gamma+TNF-alpha cotreatment activated PKR, resulting in phosphorylation and inactivation of eIF-2alpha, an essential factor in protein translation. These effects were not due to induced apoptosis and could be negated by PKR-inhibitor and short interfering RNA to PKR. Increased phosphorylation of PKR and eIF-2alpha were also observed in active IBD tissues.

Conclusions: Mucosal inflammation is associated with iHsp down-regulation, an effect that appears mediated by translational down-regulation by proinflammatory cytokines. In the context of IBD, we propose that this mechanism contributes to the severity, extent, and persistence of inflammation-induced mucosal injury.

Figures

Figure 1. iHsp protein, but not mRNA, expression is significantly and selectively down-regulated in human IBD

A) Immunohistochemical staining for Hsp27 (human homolog to murine Hsp25) and Hsp70 in sections from colon of normal and active colitis. Image shown is representative of six individual experiments. B) Western blots of Hsp27, Hsp70, Hsc70, villin and β-actin in colonic biopsies from normal control, uninvolved (UN) and actively inflamed (IN) areas of CD and UC patients. C) Real-time PCR analysis of mRNA abundance of Hsp27 and Hsp70 in colonic biopsies from normal control, uninvolved (UN) and actively inflamed (IN) mucosa of CD and UC patients. Changes in iHsp mRNA relative to GAPDH were determined as fold change over control, as described in the methods. Results for panel B and C are means ± SEM, n=6. *p<0.05 compared with Control by analysis of variance using a Bonferroni correction.

Figure 2. Colonic Hsp25 and Hsp70 protein, but not mRNA, expression is decreased in colon of IL-10−/−-gene-deficient mice with inflammation

IL-10−/−-gene-deficient mice on the C57Bl/6 background were sacrificed at the age of 20 weeks. Wild-type mice were sacrificed as control. A) Western blot of Hsp25, Hsp70, Hsc70, villin and β-actin; and B) Real-time PCR analysis of Hsp25 and Hsp70 mRNA abundance in colonic mucosa from wild-type mice (WT), IL-10−/−-gene-deficient mice without colitis (UN), and IL-10−/−-gene-deficient mice with colitis (IN). Results are means ± SEM, n=4. *p<0.05 compared with WT by analysis of variance using a Bonferroni correction.

Figure 3. IFN-γ +TNF-α co-treatment of mice down-regulates colonic Hsp25 and Hsp70 protein expression in vivo, without changing mRNA expression

Mice were given an intraperitoneal dose of IFN-γ (2500U), TNF-α (1000ng) or a combination of IFN-γ+TNF-α at 48 hours before sacrifice. Mice without cytokine injection were sacrificed as control. A) Western blot of Hsp25, Hsp70, Hsc70, villin and β-actin; B) Real-time PCR analysis of Hsp25 and Hsp70 mRNA abundance in colonic mucosa. Results are means ± SEM, n=4. *p<0.05 compared with Control by analysis of variance using a Bonferroni correction.

Figure 4. IFN-γ+TNF-α co-treatment to YAMCs down-regulates Hsp25 and Hsp70 protein, but not mRNA, expression

YAMCs were treated with IFN-γ (200U/ml), TNF-α (100ng/ml), or both IFN-γ+TNF-α for 8 hours and then stimulated with heat shock (HS, 42°C for 23 minutes). A) Cells were harvested for total protein at 120 minutes after HS. Cells without cytokine treatment and HS were harvested as basal, and cells with HS only were harvested as HS-control. Hsp25, Hsp70, Hsc70, villin and β-actin were analyzed by Western blot. B) Cells were harvested for total RNA at 0, 15, 30, 60, 90, 120, 150 and 210 minutes after HS. Cells without HS were harvested as basal and cells without IFN-γ and TNF-α treatment were harvested as control. Hsp25 and Hsp70 mRNA abundance was analyzed using real-time PCR. C) PARP and caspase 3 in YAMCs were analyzed on Western blots under basal, HS control, IFN-γ and/or TNF-α treated conditions using the same samples as in Figure 4A. Cells treated with staurosporine (1μM) for 3 hours were analyzed as positive control (+) for apoptosis. Image shown representative of four individual experiments. Results are means ± SEM, n=4. For panel A, *p<0.05 compared with HS-control by analysis of variance using a Bonferroni correction. For panel B, statistical comparisons were made by paired Student’s t-test between Control and IFN-γ+TNF-α treated samples at each time point. No significant differences were observed.

Figure 5. IFN-γ+TNF-α inhibits de novo Hsp70 protein synthesis in YAMCs

YAMCs were treated with 200U/ml IFN-γ and 100ng/ml TNF-α (+) for 8 hours, then subjected to HS at 42°C for 23 minutes and pulsed with 35S-methionine or 3H-leucine for 60 minutes at varying times. Cells without HS were harvested as basal, and cells without cytokine treatment were harvested as control (−). A) Fluorograph of 35S-methionine-labeled Hsp70 protein for 60 minutes under non-stress condition (Basal) and HS for determination of protein synthesis. B) Fluorograph of 35S-methionine-labeled Hsp70 and β-actin determination of protein synthesized for 120 minutes with or without IFN-γ+TNF-α co-treatment. C) Quantification of fold changes in de novo Hsp70 synthesis over basal, with and without IFN-γ+TNF-α co-treatment. D) TCA precipitatable 35S-methionine-labeled protein. E) Incorporation of 3H-leucine in Hsp25 protein for 60 minutes under non-stress condition (Basal) and HS for determination of protein synthesis. F) TCA precipitatable 3H-leucine-labeled protein. Results are means ± SEM, n=4. For panel B, C, D, E and F, *p<0.05 comparing Control and IFN-γ+TNF-α treated samples by paired Student’s t-test at each time point. For panel B, #p<0.05 comparing Hsp70 and β-actin in IFN-γ+TNF-α treated samples by paired Student’s t-test.

Figure 6. IFN-γ+TNF-α co-treatment has no effect on protein degradation of Hsp70 in YAMCs

Hsp70 protein degradation was determined using pulse chase experiment. YAMCs were incubated with 35S-methionine for 1 hour to pulse label de novo proteins. IFN-γ (200U/ml) and TNF-α (100ng/ml) were added to cells after pulse label. Cells were harvested for protein analysis at 0, 2, 4 and 8 hours after cytokine treatment. Cells without cytokine treatment were harvested as control. A) Schematic of time course for cell treatment. B) Fluorograph of immuno-precipitated Hsp70 in cells with (+) and without (−) cytokine treatment. Image is representative of four individual experiments. C) Quantification of fold changes of Hsp70 over basal. Results are means ± SEM, n=4. For panel C, statistical comparisons were made by paired Student’s t-test between Control and IFN-γ+TNF-α treated samples at each time point. No significant differences were observed.

Figure 7. IFN-γ+TNF-α inhibit iHsp production via activation of PKR and phosphorylation of eIF-2α in YAMCs

A) YAMCs were treated with IFN-γ (200U/ml) and TNF-α (100ng/ml) for 8 hours, PKR-I (30nM) for 30 minutes and then subjected to HS. Hsp25, Hsp70, Hsc70, phospho-PKR (p-PKR), PKR, phospho- eIF-2α (p-eIF-2α) and eIF-2α were analyzed by Western blot under basal (−), HS (+), IFN-γ+TNF-α (+) and/or PKR-I (+) treated conditions. B) YAMCs were treated with silencing RNA to PKR (si-PKR) and control silencing RNA (si-C) for 16 hours prior to IFN-γ+TNF-α and HS treatment. Hsp25, Hsp70, Hsc70, p-PKR, PKR, p-eIF-2α and eIF-2α were analyzed by Western blot. Image shown is representative of four individual experiments. Results are means ± SEM, n=4. For panel B, *p<0.05 compared with HS control (HS) by analysis of variance using a Bonferroni correction.

Figure 8. Phosphorylation of PKR and eIF-2α is up-regulated in inflamed mucosa in vivo

A) Phospho-PKR (p-PKR) and phospho-eIF-2α (p-eIF-2α) were analyzed by Western blot from colonic pinch biopsies from control, uninvolved (UN) and actively inflamed (IN) areas of CD and UC patients using the same samples as in Figure 1B. B) PKR, p-PKR, eIF-2α and p-eIF-2α were analyzed on Western blot for colonic mucosa from wild-type mice (WT), IL-10−/− mice without colitis (UN) and IL-10−/− mice with colitis (IN) using the same samples as in Figure 2. Results are means ± SEM. For panel A, *p<0.05 compared with Control analyzed on same blot by analysis of variance using a Bonferroni correction, n=6. For panel B, *p<0.05 compared with WT by analysis of variance using a Bonferroni correction, n=4.