Therapeutic potential of follistatin for colonic inflammation in mice
2005; Elsevier BV; Volume: 128; Issue: 2 Linguagem: Inglês
10.1053/j.gastro.2004.11.063
ISSN1528-0012
AutoresTaeko Dohi, Chieko Ejima, Rie Kato, Yuki I. Kawamura, Rei Kawashima, Noriko Mizutani, Yoshiaki Tabuchi, Itaru Kojima,
Tópico(s)Genetic factors in colorectal cancer
ResumoBackground & Aims: Activins belong to the transforming growth factor-β superfamily. Recent studies have shown that activin and its natural antagonist, follistatin, are involved in tissue repair and inflammatory processes. The aim of this study was to determine whether neutralization of activins with follistatin would have an in vivo anti-inflammatory effect in several murine models of colitis. Methods: We assessed activin levels in the colitis induced by intracolonic administration of trinitrobenzene sulfonic acid (TNBS). We subsequently tested the effects of an intraperitoneal injection of follistatin before or after induction of TNBS colitis. We also examined the established colitis induced by oral dextran sulfate sodium (DSS) as well as the spontaneous colitis that develops in interleukin (IL)-10 gene-deficient (IL-10−/−) mice. Results: Levels of activin transcripts in the colon during the acute phase of TNBS colitis were up-regulated. Epithelial cells, infiltrating macrophages (Mϕ), and endothelial cells produced excess activin βA. Pretreatment with follistatin increased the survival rate of mice with TNBS colitis from 33% to 82% and decreased the plasma levels of IL-6 and amyloid A. Administration of follistatin also reduced the histologic score and tissue myeloperoxidase activity in established TNBS and DSS colitis and reduced the severity of the colitis in IL-10−/− mice. Based on results obtained from 3 mouse models and from in vitro experiments, follistatin promoted the proliferation of colonic epithelial cells. Conclusions: Neutralization of activins by follistatin promoted epithelial cell division and tissue repair, clearly suggesting a treatment modality for intestinal inflammation. Background & Aims: Activins belong to the transforming growth factor-β superfamily. Recent studies have shown that activin and its natural antagonist, follistatin, are involved in tissue repair and inflammatory processes. The aim of this study was to determine whether neutralization of activins with follistatin would have an in vivo anti-inflammatory effect in several murine models of colitis. Methods: We assessed activin levels in the colitis induced by intracolonic administration of trinitrobenzene sulfonic acid (TNBS). We subsequently tested the effects of an intraperitoneal injection of follistatin before or after induction of TNBS colitis. We also examined the established colitis induced by oral dextran sulfate sodium (DSS) as well as the spontaneous colitis that develops in interleukin (IL)-10 gene-deficient (IL-10−/−) mice. Results: Levels of activin transcripts in the colon during the acute phase of TNBS colitis were up-regulated. Epithelial cells, infiltrating macrophages (Mϕ), and endothelial cells produced excess activin βA. Pretreatment with follistatin increased the survival rate of mice with TNBS colitis from 33% to 82% and decreased the plasma levels of IL-6 and amyloid A. Administration of follistatin also reduced the histologic score and tissue myeloperoxidase activity in established TNBS and DSS colitis and reduced the severity of the colitis in IL-10−/− mice. Based on results obtained from 3 mouse models and from in vitro experiments, follistatin promoted the proliferation of colonic epithelial cells. Conclusions: Neutralization of activins by follistatin promoted epithelial cell division and tissue repair, clearly suggesting a treatment modality for intestinal inflammation. The maintenance of the architectural features of the gastrointestinal (GI) tract is not a static process but instead occurs through vigorous and rapid cell regeneration. Epithelial cells arise by a process of clonal growth from stem cells present in the deep crypts, and progeny cells move in vertical columns toward the villus apex. In the small intestine, cells produced by mitosis in the crypt live only 2–3 days before reaching the villus tip and are then extruded.1Potten C.S. Stem cells in gastrointestinal epithelium numbers, characteristics and death.Philos Trans R Soc Lond B Biol Sci. 1998; 353: 821-830Crossref PubMed Scopus (449) Google Scholar These progeny cells, which differentiate into either absorptive or goblet cells, form a potent tissue polarity. This continuous renewal of the intestinal epithelium is also important for the regeneration of tissues damaged by inflammation. In addition to growth factors, both inflammatory cytokines and chemokines also contribute to this normal tissue repair. Activins belong to the transforming growth factor-β (TGF-β) superfamily and have numerous biologic effects on the proliferation and differentiation of several types of cells. Different forms of activin are known: (1) Activin A is a homodimer of activin βA subunits; (2) activin B is a homodimer of activin βB subunits; and (3) activin AB is a heterodimer of these 2 subunits.2Ying S.Y. Inhibins, activins, and follistatins gonadal proteins modulating the secretion of follicle-stimulating hormone.Endocr Rev. 1988; 9: 267-293Crossref PubMed Scopus (802) Google Scholar Activin βC, βD, and βE subunits also occur in mammals. The actions of the activins are controlled by the synthesis and expression of type I and type II receptors, as well as by the natural antagonist follistatin, which binds activins A and B with high affinity and blocks their function.3Mather J.P. Moore A. Li R.H. Activins, inhibins, and follistatins further thoughts on a growing family of regulators.Proc Soc Exp Biol Med. 1997; 215: 209-222Crossref PubMed Scopus (248) Google Scholar, 4Ying S.Y. Zhang Z. Furst B. Batres Y. Huang G. Li G. Activins and activin receptors in cell growth.Proc Soc Exp Biol Med. 1997; 214: 114-122Crossref PubMed Scopus (78) Google Scholar Another activin-binding protein, follistatin-like protein 3, was recently discovered, and its differential binding and neutralization of activins were described.5Schneyer A. Schoen A. Quigg A. Sidis Y. Differential binding and neutralization of activins A and B by follistatin and follistatin like-3 (FSTL-3/FSRP/FLRG).Endocrinology. 2003; 144: 1671-1674Crossref PubMed Scopus (64) Google Scholar In general, activin A is considered to be an autocrine factor for regulating the growth and differentiation of epithelial cells, including renal tubular cells6Maeshima A. Zhang Y.-Q. Nojima Y. Naruse T. Kojima I. Involvement of the activin-follistatin system in tubular regeneration after renal ischemia in rats.J Am Soc Nephrol. 2001; 12: 1685-1698PubMed Google Scholar and hepatocytes.7Kogure K. Zhang Y.Q. Maeshima A. Suzuki K. Kuwano H. Kojima I. The role of activin and transforming growth factor-beta in the regulation of organ mass in the rat liver.Hepatology. 2000; 31: 916-921Crossref PubMed Scopus (68) Google Scholar There are several studies that have addressed the expression of activins in the GI tract, in which significant cell proliferation occurs. Activin synthesis was detected in a human embryonic intestinal cell line,8Kawamura N. Nobusawa R. Mashima H. Kanzaki M. Shibata H. Kojima I. Production of activin A in human intestinal epithelial cell line.Dig Dis Sci. 1995; 40: 2280-2285Crossref PubMed Scopus (10) Google Scholar in the human colonic cancer cell line Caco 2, and in the rat intestine.9Sonoyama K. Rutatip S. Kasai T. Gene expression of activin, activin receptors, and follistatin in intestinal epithelial cells.Am J Physiol Gastrointest Liver Physiol. 2000; 278: G89-G97PubMed Google Scholar Exogenous activin A suppressed the proliferation of IEC-6 cells in a dose-dependent manner.10Dignass A.U. Jung S. Harder-d'Heureuse J. Wiedenmann B. Functional relevance of activin A in the intestinal epithelium.Scand J Gastroenterol. 2002; 37: 936-943Crossref PubMed Scopus (26) Google Scholar Recent studies have clearly shown an important role for activins in systemic inflammatory responses and tissue repair.11de Kretser D.M. Hedger M.P. Phillips D.J. Activin A and follistatin their role in the acute phase reaction and inflammation.J Endocrinol. 1999; 161: 195-198Crossref PubMed Scopus (59) Google Scholar, 12Munz B. Hubner G. Tretter Y. Alzheimer C. Werner S. A novel role of activin in inflammation and repair.J Endocrinol. 1999; 161 (review, 34 refs): 187-193Crossref PubMed Scopus (88) Google Scholar, 13Phillips D.J. Jones K.L. Scheerlinck J.Y. Hedger M.P. de Kretser D.M. Evidence for activin A and follistatin involvement in the systemic inflammatory response.Mol Cell Endocrinol. 2001; 180: 155-162Crossref PubMed Scopus (95) Google Scholar A potential role for activins in inflammatory bowel disease (IBD) was suggested by the following findings: No activin βA messenger RNA (mRNA) was present in the normal human GI tract; however, the levels of mRNA for activin βA positively correlated with the degree of inflammation in IBD patients, as manifested by histopathology and by an increase in the level of interleukin (IL)-1β.14Hubner G. Brauchle M. Gregor M. Werner S. Activin A a novel player and inflammatory marker in inflammatory bowel disease?.Lab Invest. 1997; 77: 311-318PubMed Google Scholar Activin receptors I and II were present throughout the GI tract of patients with IBD and in control subjects. In contrast, no expression of the activin βA protein was noted in controls, whereas βA was expressed in the intestinal tissues of the patients.10Dignass A.U. Jung S. Harder-d'Heureuse J. Wiedenmann B. Functional relevance of activin A in the intestinal epithelium.Scand J Gastroenterol. 2002; 37: 936-943Crossref PubMed Scopus (26) Google Scholar In the present study, we investigated the expression of activins and follistatin in the inflamed murine intestinal tract in a colitis model prepared by colonic administration of trinitrobenzene sulfonic acid (TNBS). We were successful in treating colitis by neutralizing the action of activin through administration of follistatin in both the TNBS and oral dextran sulfate sodium (DSS) colitis models, as well as in IL-10 gene-deficient (IL-10−/−) mice, which develop colitis spontaneously. Female 8-week-old BALB/c mice obtained from CLEA Japan Inc. (Tokyo, Japan) were maintained under pathogen-free conditions in a facility of the Research Institute, International Medical Center of Japan (IMCJ; Tokyo, Japan). IL-10−/− mice (C57BL/6 background)15Kuhn R. Lohler J. Rennick D. Rajewsky K. Muller W. Interleukin-10-deficient mice develop chronic enterocolitis.Cell. 1993; 75: 263-274Abstract Full Text PDF PubMed Scopus (3760) Google Scholar were purchased from the Jackson Laboratory (Bar Harbor, ME). All experiments were performed according to the Institutional Guidelines for the Care and Use of Laboratory Animals in Research and the approval of the local ethics committee in the IMCJ. TNBS colitis was induced as described previously.16Dohi T. Fujihashi K. Kiyono H. Elson C.O. McGhee J.R. Mice deficient in Th1-type and Th2-type cytokines develop distinct forms of hapten-induced colitis.Gastroenterology. 2000; 119: 724-733Abstract Full Text Full Text PDF PubMed Scopus (86) Google Scholar, 17Dohi T. Fujihashi K. Rennert P.D. Iwatani K. Kiyono H. McGhee J.R. Hapten-induced colitis is associated with colonic patch hypertrophy and Th2-type responses.J Exp Med. 1999; 189: 1169-1179Crossref PubMed Scopus (158) Google Scholar A 2% solution of TNBS (Research Organics, Cleveland, OH) in PBS:ethanol (1:1 by volume) was administered intrarectally to mice lightly anesthetized with ketamine (Sankyo Co. Ltd., Tokyo, Japan). Different doses of TNBS were used according to the requirements of each experiment. For comparison of the survival rate and analysis of acute inflammatory parameters at day 1, a dose of 60 μg/g (3 μL solution/g) body weight of TNBS was given, which caused death in approximately 20% and 70% of BALB/c mice within 24 hours and 72 hours, respectively, following the TNBS enema. For histologic and time-course analyses, a TNBS dose of 48 μg/g of body weight was administered, and this resulted in the death of 10% or 40% of the mice at 24 hours and 72 hours, respectively (data not shown). Specimens for analysis were collected among the surviving mice at each time point after death with an excess intraperitoneal (i.p.) injection of a mixture of ketamine and xylazine. Human recombinant follistatin was kindly provided by Dr. Yuzuru Eto (Ajinomoto Co., Inc., Tokyo, Japan). An aliquot of follistatin (5 μg in 0.1 mL in PBS with 0.5% normal mouse serum) was injected i.p. 30 minutes prior to the TNBS administration. Control groups received an injection of PBS containing 0.5% normal mouse serum. To examine the effects of follistatin on established TNBS colitis, we administered a TNBS dose of 48 μg/g of body weight on day 0 and day 7, as described previously16Dohi T. Fujihashi K. Kiyono H. Elson C.O. McGhee J.R. Mice deficient in Th1-type and Th2-type cytokines develop distinct forms of hapten-induced colitis.Gastroenterology. 2000; 119: 724-733Abstract Full Text Full Text PDF PubMed Scopus (86) Google Scholar and then 5 μg follistatin on day 8. For induction of dextran sulfate sodium (DSS)-colitis, DSS (M. W. 5000; Sigma-Aldrich Co., Inc, Tokyo, Japan) was added to the drinking water at the concentration of 2%. Mice were provided this water ad libitum. After 7 days, weight loss was not yet obvious; however, most of the mice exhibited bloody diarrhea, and histologic analysis revealed colitis. Follistatin (5 μg) was then given IP to DSS-treated mice on days 7, 9, and 11. The drinking water was supplemented with DSS continuously until day 12 for all groups of mice. On day 12, the colon was resected for analysis. To test the effects of follistatin on the recovery phase of DSS colitis, we gave mice a 5% solution of DSS for 8 days and then administered 5 μg follistatin on days 7, 9, and 11. Histologic examination was done at day 14 in this group. IL-10−/− mice, maintained in the IMCJ specific pathogen-free facility, were free from signs of enteritis, and a set of female littermates born within a week of each other from 2 breeder female mice were placed in a conventional environment, in which no major pathogens were detected, at the age of 7 weeks. Four weeks after transferring these mice to this environment, all mice showed transient anal prolapse upon defecation with soft feces; however, their general condition was not affected, and no weight loss was seen. At the age of 18 weeks, their condition remained stable, and administration of follistatin was initiated. These littermates were separated into 2 groups based on their body weight, and 1 group was given follistatin (5 μg) IP on days 0, 3, 6, and 9. The other control group received an injection of PBS containing 0.5% normal mouse serum. On day 12, colonic tissue samples were taken for histologic analysis. This experiment was also performed on a separated set of male littermates, which were kept in the conventional environment for only 4 weeks. Digoxygenin (DIG)-labeled activin βA antisense and sense probes were prepared by in vector transcription by using a DIG RNA labeling Kit (Roche Diagnostics GmbH, Mannheim, Germany). In brief, a 1.6-kb XhoI fragment from the human activin βA coding sequence was cloned into pSPT18 and pSPT19. Sense cRNA was transcribed from pSPT18 with SP6 polymerase, and antisense cRNA was transcribed from pSPT19 with T7 polymerase. Mouse tissues were snap-frozen and embedded in O.C.T. compound (Sakura Finetechnical Co. Ltd. Tokyo. Japan). Eight-μm-thick sections were prepared, dried briefly, and fixed with 4% paraformaldehyde in PBS for 10 minutes. After digestion with 0.8 mg/mL Proteinase K (DakoCytomation Inc., Kyoto Japan) for 6 minutes and fixation with 4% PFA at 4°C for 20 minutes, the sections were treated with acetic anhydride and then incubated at 50°C overnight in hybridization buffer containing 1 μg/mL DIG-labeled cRNA, as described previously.18Maeshima A. Zhang Y.-Q. Furukawa M. Naruse T. Kojima I. Hepatocyte growth factor induces branching tubulogenesis in MDCK cells by modulating the activin-follistatin system.Kidney Int. 2000; 58: 1511-1522Crossref PubMed Scopus (50) Google Scholar After hybridization, the sections were washed and incubated with RNAse A (Sigma Chemical Co., St. Louis, MO). The hybridized probe was detected with a DIG Nucleic Acid Detection Kit (Roche Diagnostics, Indianapolis, IN), used according to the manufacturer's protocol. Samples of large intestine were taken, opened longitudinally, and rinsed in PBS. They were then treated with 2 mmol/L EDTA in PBS for 30 minutes to remove the epithelial cells. The residue was completely denuded from epithelial cells by vigorous washing and then digested with type V collagenase (Sigma) for 20 minutes to obtain lamina propria mononuclear cells (LPMCs). This step was repeated once more. In some experiments, the LPMCs were resuspended in RPMI 1640 supplemented with 5% FCS, penicillin, streptomycin, and amphotericin B (RPMI 1640 medium) in 48-well plates (3 × 106 cells/well) and incubated in a humidified 5% CO2 incubator for 45 minutes. The nonadherent cells, removed by washing with RPMI 1640 medium 3 times, and the remaining adherent cells were both used in further experiments. In some experiments, nonadherent cells were further purified with biotinylated anti-CD3 or anti-B 220 mAbs (BD Pharmingen, San Diego, CA) and a magnetic cell separator (AutoMACS, Miltenyi Biotec GmbH, Germany). To assess activin A release in cultures, we incubated cells in RPMI 1640 supplemented with 10% FCS, sodium pyruvate, L-glutamine, HEPES, penicillin, streptomycin, and amphotericin B (complete medium) at 37°C in an atmosphere of 5% CO2 in air. After 24 hours of culture, the supernatants were removed and subjected to an ELISA, as described below. Cells of the murine colonic epithelial cell line MCE30119Tabuchi Y. Ohta S. Arai Y. Kawahara M. Ishibashi K. Sugiyama N. Horiuchi T. Furusawa M. Obinata M. Fuse H. Takeguchi N. Asano S. Establishment and characterization of a colonic epithelial cell line MCE301 from transgenic mice harboring temperature-sensitive simian virus 40 large T-antigen gene.Cell Struct and Funct. 2000; 25: 297-307Crossref PubMed Scopus (52) Google Scholar were cultured in collagen-coated dishes or in BD Biocoat Cell Culture Inserts coated with collagen I (Becton Dickinson Co., Franklin Lakes, NJ) in DME/F12 medium supplemented with ITES, 10 ng/mL EGF, and 2% FCS at 33°C. After the cells had reached confluence, they were cultured at the nonpermissive temperature (39°C) for 3 days to induce differentiation and used for experiments. The human colonic cell line HT-29 (American Type Culture Collection, Rockville, MD) was cultured in Dulbecco's modified Eagle medium containing 10% fetal bovine serum. In some experiments, recombinant human activin A (Techne Co., Minneapolis, MN) or follistatin was added to the cultures. For the TNBS-specific T-cell proliferation assay, mononuclear cells were isolated from sacral lymph nodes and treated with 0.3 mg/mL TNBS in RPMI 1640 for 15 minutes at room temperature. They were then extensively washed and cultured in complete medium for 3 days. In each well of 96-well plates, 2 × 105 cells were placed, and 0.5 μCi/well of tritiated [3H] thymidine was added 18 hours before harvesting of the cells. The amount of [3H] thymidine was determined by scintillation counting. The proliferation index was determined as counts per minute (cpm) of wells with TNBS/cpm of wells without TNBS. The GI tract tissues were taken, fixed in 5% glacial acetic acid in ethanol, and paraffin-embedded; then 4-μm sections were prepared and stained with H&E. The colon was cut into upper, middle, and lower parts for histopathologic analysis. In some experiments, frozen sections were also used. To visualize proliferating cells, we injected 1 mg bromodeoxyuridine (BrdU, Sigma) i.p. 1 hour before death. Paraffin-embedded sections were treated with 4 N HCl, blocked with 1% bovine serum albumin in PBS, and stained with rat anti-BrdU Ab (Oxford Biotech, Ltd, Oxford, United Kingdom), followed by FITC-labeled anti-rat IgG Ab (Southern Biotechnology Associates, Inc. [SBA], Birmingham, AL). The numbers of BrdU+ cells per crypt were counted, and at least 20 crypts in 3 separate fields were examined in each section. Histologic scores for TNBS colitis were determined as previously described.17Dohi T. Fujihashi K. Rennert P.D. Iwatani K. Kiyono H. McGhee J.R. Hapten-induced colitis is associated with colonic patch hypertrophy and Th2-type responses.J Exp Med. 1999; 189: 1169-1179Crossref PubMed Scopus (158) Google Scholar Histologic scores for DSS colitis were determined by summing scores for changes in the epithelium (0, normal; 1, focal loss of goblet cells; 2, diffuse loss of goblet cells; 3, focal loss of crypts; 4, diffuse loss of crypts) and for cell infiltration (0, no increase; 1, around bases of the crypts; 2, diffuse infiltration along the layer of Muscularis mucosae; 3, complete infiltration of the mucosal layer; 4, infiltration of both mucosal and submucosal layers). Scores for stool consistency and rectal bleeding were assessed according to previously published procedures.20Kihara N. de la Fuente S.G. Fujino K. Takahashi T. Pappas T.N. Mantyh C.R. Vanilloid receptor-1 containing primary sensory neurones mediate dextran sulphate sodium induced colitis in rats.Gut. 2003; 52: 713-719Crossref PubMed Scopus (192) Google Scholar For detection of activin receptor IIB, a rabbit polyclonal antibody was kindly provided by Dr. Kohei Miyazono (The University of Tokyo). Frozen sections fixed with cold acetone for 10 minutes were treated with Blockace (Dainippon Pharmaceuticals, Osaka, Japan), incubated with the antiactivin receptor IIB antibody (1:100 dilution) for 60 minutes then reacted with TRITC-labeled anti-rabbit IgG secondary antibody (SBA). Activin A in culture supernatants and in plasma was measured by using a Quantikine M mouse activin A immunoassay kit (R&D Systems, Minneapolis, MN). Both IL-6 and tumor necrosis factor (TNF)-α were assessed with an ELISA development kit (Techne Co., Minneapolis, MN). Plasma levels of amyloid A were quantified by using a Cytoscreen immunoassay kit (BioSource International, Inc. Camarillo, CA). Murine MCP-1 and human IL-8 in culture medium were determined with an ELISA Kit purchased from Pierce Endogen (Rockford, IL). The myeloperoxidase (MPO) activity in colonic tissues was determined by a method reported previously.21Vaveka A.P. Agah A. Rollins S.A. Matis L.A. Li L. Stahl G.L. Myocardial infection and apoptosis after myocardial ischemia and reperfusion.Circulation. 1998; 97: 2259-2267Crossref PubMed Scopus (372) Google Scholar Briefly, colonic tissues were homogenized in 5 mmol/L phosphate buffer (pH 6.0) and centrifuged at 30,000g for 30 minutes at 4°C. The pellet was suspended in 50 mmol/L phosphate buffer containing 0.5% hexadecyltrimethyl ammonium bromide. After centrifugation at 20,000g, the supernatant was subjected to the MPO assay by use of a spectrophotometer equipped with a rate assay system (U3200; Hitachi, Tokyo, Japan). Data were presented as MPO activity units/g tissue. Total RNA was prepared from colonic tissue or from separated cells by using RNA-Bee RNA isolation reagent (Tel-Test, Inc., Friendswood, TX). Complementary DNA was synthesized from RNA by reverse transcription (RT). The PCR primers for murine GAPDH used were 5′-AGCCAAACGGGTCATCATCTC and 5′-TGCCTGCTTCACCACCTTCTT; those for activin βA, 5′-TGCTGCACTTGAAGAAGAGA-CCC and 5′-TGGTCCTGGTTCTGTTAGCCTTG; those for activin βB, 5′-ATGGACTTTCGGCTCAT-CGG and 5′-CACGATCATGTTGGGCACATC; and those for follistatin, 5′- CCTACTGTGTGACCTGTAA-TC and 5′-CTCCTCTTCCTCCGTTTCTTC. The step-cycle program was set for denaturing at 95°C for 45 seconds, annealing at 60°C for 45 seconds, and extension at 72°C for 45 seconds for a total of 40 cycles. Expression of mRNA was assessed by quantitative PCR using a SYBR Green PCR Master Mix (Applied Biosystems, Warrington, United Kingdom) and ABI PRISM 7700 Sequence Detector (Applied Biosystems). Quantification of mRNA for IFN-γ was performed by using ABI Taqman probes (Applied Biosystems). The PCR primers for IFN-γ were 5′-TGATCCTTTGGACCCTCTGA and 5′-GCAAAGCCAGATGCAGTGT, and the Taqman probe was 5′-CCTCCTGCGGCCTAGCTCTGAGAC. Threshold cycle numbers (Ct) were determined with Sequence Detector Software (version 1.7; Applied Biosystems) and transformed by using the ΔCt/ΔΔCt method as described by the manufacturer, with GAPDH used as the calibrator gene. The results were compared by the Mann-Whitney test using the Statview II statistical program (Abacus Concepts, Berkeley, CA) adapted for the Macintosh computer. Differences in survival rate were tested by the Kaplan-Meier test. Changes in body weight were compared by use of the Wilcoxon matched-pair signed-rank test. We first investigated the expression of activins and follistatin in the inflamed murine colon by using the TNBS colitis model. In this model, acute erosive lesions with infiltrating leukocytes were observed on days 1–3. Activin βA mRNA was detected in the colon of normal mice. Although the signal was not a strong one, epithelial cells and LPMC contained activin βA mRNA (Figure 1A and 1C). In contrast, one day after the induction of colitis, activin βA transcripts were remarkably up-regulated in epithelial cells in the lower crypts, in the infiltrating cells, and also in the endothelial cells (Figure 1B and D). When the time course for mRNA expression of activin βA, activin βB, and follistatin in the colon was assessed by quantitative RT-PCR, up-regulation of activin βA was evident at day 1 (Figure 2). Follistatin mRNA was also increased at day 1 but to a lesser degree than activin βA mRNA (Figure 2A). Plasma levels of activin A peaked at day 3 and decreased by day 7 (Figure 2B). These results suggest that excess activin A was produced during the acute phase of TNBS colitis. For further evaluation of the fraction of the infiltrating cells that was responsible for the inflammation-related synthesis of activins, we separated total colonic LPMCs into adherent cells, which were mostly CD11b+ cells (>90%), and nonadherent, floating cells, which were mainly lymphocytes. Messenger RNA expression for activins βA and βB induced by inflammation took place mainly in the adherent LPMC fraction (Figure 2C). Activin A was detected by ELISA in the supernatants of CD3+ T-cell and B220+ B-cell fractions; however, up-regulation of activin A release after induction of TNBS colitis was evident only in the adherent cell fraction (Figure 2D). It was also clear that follistatin mRNA levels did not change at this time point in either fraction of LPMCs (Figure 2C). This result suggests that the increased levels of activins in the colon may not have been neutralized by follistatin. We next tested the effects of follistatin for its ability to neutralize activins during the acute phase of TNBS colitis. A single administration of follistatin 30 minutes before the induction of colitis significantly improved the survival rate of the mice at day 3 (Figure 3A). Histologic scores were lower in the group treated with follistatin (Figure 3B). In TNBS colitis, focal ulcers were prominent at day 3; however, mice given follistatin showed a lower frequency of these erosive lesions than the control group, although mild edema and evidence of cell infiltration were observed (Figure 3C). Plasma amyloid A and IL-6 levels were also significantly lower in the mouse group treated with follistatin than in the control, nontreated group (Figure 3D and E). Plasma obtained from mice within 24 hours after intracolonic administration of TNBS showed the characteristic yellow color derived from the trinitrophenyl (TNP) residue; however, the plasma from TNBS-treated mice given follistatin was not colored by the TNBS (Figure 3F). When body weight was measured on day 3 after a medium dose of TNBS had been given, the follistatin-treated group showed less weight loss than the control group, although this difference did not reach statistical significance (Figure 3G). MPO activity in the colonic tissue, which reflects the accumulation of neutrophils, was much lower in the follistatin-treated than in the control group, and coincided with less ulcer formation in the follistatin-treated group (Figure 3H). Of note, enhancement of epithelial cell proliferation was seen in the mouse group treated with follistatin (Figure 3I and J). Rapid provision of new epithelial cells by accelerated division may be one of the mechanisms for the anti-inflammatory effects of follistatin. In the acute phase of TNBS colitis, damage to the colonic mucosa resulted in bacteremia, which is likely the major reason for the early death seen in these mice. We speculate that follistatin accelerates prompt epithelial cell regeneration in the colon and thus improves the barrier function of the colonic mucosa. In our next experiments, we tested whether follistatin would affect established TNBS colitis. To do this, we injected follistatin IP on day 8 and then examined the tissue on day 10. Histologic scores revealed that the frequency of ulcer formation was lower in the follistatin-treated than in the control mice (Figure 4C), and the MPO activity was also lower in the former (Figure 4D). However, this treatment did not reduce the mucosal mononuclear cell infiltration, which clearly indicates a chronic inflammatory response (Figure 4A and B). On day 10, we examined the TNP-specific T-cell proliferative responses of sacral lymph nodes, which are the draining lymph nodes of the mouse colon. Unexpectedly, in the follistatin-treated mice, these responses were comparable with those seen in the control mouse group (Figure 4F). This finding indicates that treatment with follistatin suppressed ulcer formation and accumulation of neutrophils but did not decrease TNP-specific T-cell responses. In this regimen, increased numbers of proliferating epithelial cells and well-preserved crypts were seen in the treated group (Figur
Referência(s)