Interferon gamma bound to extracellular matrix changes the hyporesponsiveness to LPS in crypt but not villous intestinal epithelial cells

Immunology Letters 99 (2005) 109–112

Interferon gamma bound to extracellular matrix changes the hyporesponsiveness to LPS in crypt but not villous intestinal epithelial cells Jansi Alvarado a , Peter Taylor b , Jes´us R. del Castillo a , Luz. E. Thomas a, ∗ b

a Laboratorio de Fisiolog´ıa Gastrointestinal, Instituto Venezolano de Investigaciones Cient´ıficas (IVIC), Caracas, Venezuela Laboratorio de Patolog´ıa Celular y Molecular, Instituto Venezolano de Investigaciones Cient´ıficas (IVIC), Caracas, Venezuela

Received 15 November 2004; received in revised form 2 February 2005; accepted 3 February 2005 Available online 23 February 2005

Abstract Intestinal epithelial cells (IEC) are hyporesponsive to LPS. Responsiveness to luminal bacteria has been implicated in the pathogenesis of inflammatory bowel diseases (IBD). In support of this, previous studies have demonstrated that some intestinal epithelial cell lines are induced by IFN-␥ to respond to LPS. However, both the responsiveness to LPS and the effect of IFN-␥ in intestinal cell lines are heterogeneous. In addition, IFN-␥ may be sequestered in the extracellular matrix (ECM) compartment. The ECM-bound form is more effective than soluble IFN-␥ in producing its biological effects in several experimental models. We investigated the effect of ECM-bound and soluble IFN-␥ treatment on interleukin-8 (IL-8) secretion in response to LPS in freshly isolated villous and crypt cells. We demonstrate that ECM-bound, but not soluble IFN-␥, induced an increase in IL-8 secretion in response to LPS in undifferentiated crypt cells. This effect was associated with an increase in TLR4 expression. In contrast, mature villous cells did not modify their response to LPS when treated with IFN-␥ (ECM-bound or soluble). These results suggest that selective changes in immature crypt cells induced by IFN-␥ bound to extracellular matrix could contribute to inappropriate responsiveness to commensal bacteria in IBD. © 2005 Elsevier B.V. All rights reserved. Keywords: Intestinal epithelial cell; Lipopolysaccharide; Toll-like receptor

Inflammatory bowel diseases (IBD), including human ulcerative colitis and Crohn’s disease, are chronic immunemediated diseases of the intestinal tract with unknown etiologies [1]. Various pathogenic mechanisms have been proposed, including inflammatory responses to persistent luminal pathogens, abnormal luminal constituents, autoimmunity, or an overly aggressive immune response to normal luminal constituents such as commensal enteric bacteria. It has been shown that the presence of enteric bacteria is essential in the development of gut inflammation in most animal models of these diseases, and that modulation of the quantity ∗ Corresponding author. Present address: Centro de Biof´ısica y Bioqu´ımica, IVIC, Apartado 21827, Caracas 1020-A, Venezuela. Tel.: +58 212 504 1752; fax: +58 212 504 1093. E-mail address: [email protected] (Luz. E. Thomas).

0165-2478/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.imlet.2005.02.005

or quality of intestinal flora may be beneficial in patients with IBD [2]. Lipopolysaccharide (LPS), a glycolipid present in the outer membrane of Gram-negative bacteria, is a strong stimulator of the immune system, capable of activating neutrophils, lymphocytes, and macrophages. This effect is mediated by Toll-like receptor (TLR) 4 in conjunction with secreted MD-2 and soluble or membrane-bound CD14 [3,4]. Cultured intestinal cells are, in general, poorly stimulated by LPS [4]. However some intestinal cell lines can secrete IL-8 in response to this bacterial glycolipid [5,6]. This heterogeneity may be due to different degrees of maturation between cell lines [5]. In this sense, it have been demonstrated that differentiation of intestinal cell lines impairs the responsiveness to LPS [7]. In the villous–crypt axis of the intact intestinal epithelium, where mature epithelial cells are located at the villi and undifferentiated cells are in the crypts, cell differentiation

110

J. Alvarado et al. / Immunology Letters 99 (2005) 109–112

might be related to a different response to pro-inflammatory stimuli. Normal intestinal cells express low levels of TLR4 [8]. In contrast, colonic biopsies from patients with IBD show high TLR4 expression [9]. The cellular mechanism involved in the increase of TLR4 in IBD is not clear. Increased production of IFN-␥ at sites of inflammatory bowel disease has been reported, as well as its participation in the change of responsiveness of intestinal epithelial cells and cell lines to microbial products [5,10]. It has been proposed that this effect is mediated by an increase in TLR4 expression [5,6,9,10]. Several cytokines and growth factors, including IFN-␥ bind to glycosaminoglycan (GAG) [11]. The relevance of this interaction for the biological activity of IFN-␥ on human arterial smooth muscle cells (HASMC) has been well studied [11,12]. The binding of IFN-␥ to ECM may increase the residence time, localizes it to restricted areas within tissues [12] or may protect it from inactivation [13], thus inducing a higher response than to the unbound cytokine. In this paper, we evaluated the effect of ECM-bound or soluble IFN-␥ on interleukin-8 (IL-8) secretion in response to LPS in intestinal villous and crypt cells freshly isolated from swine. Villous and crypt cells were obtained by sequential palpation of the pig intestinal segments as described by del Castillo et al. [14]. Briefly, a segment of small intestine was excised and then rinsed, with saline solution (NaCl 0.9%) supplemented with penicillin (100 U/ml), five times to remove mucus. The segment was then filled, and incubated for 10 min at 37 ◦ C with solution 1 (7 mM K2 SO4 , 44 mM K2 HPO4 , 9 mM NaHCO3 , 10 mM HEPES-Tris, 180 mM glucose, pH 7.4 and 340 mosmol/l). The luminal content was discarded, and the intestinal segment was refilled and incubated for 3 min at 37 ◦ C with solution 2 (solution 1 with 0.25 mM EDTA). The intestine segment was then gently palpated, and the luminal solution, containing isolated cells, was collected in DMEM (100 ml) at 4 ◦ C, filtered through a nylon mesh (60 ␮m pore diameter), and centrifuged at 100 × g for 5 min. These steps were repeated six times to obtain six different fractions. The preparation of separate populations of differentiated and undifferentiated enterocytes was assessed using the capacity of proliferating cells, located principally in the intestinal crypts, to incorporate [3H]-thymidine and bromodeoxyuridine (BrdU) into their DNA. Cells isolated in the first two palpations (fractions 1 and 2) were regarded as villous cells and those isolated from the fifth and sixth palpation (fractions 5 and 6) as crypt cells. The differential expression of alkaline phosphatase activity between the villous and crypts fractions also indicated the successful separation of these two epithelial populations [villous: 72.1 ± 7.75 versus crypts: 9.3 ± 2.41 nmol Pi/mg/min, n = 6]. The isolated cells were suspended in DMEM modified to contain 116 mM NaCl, 5 mM KCl, 1 mM MgSO4 , 1 mM NaH2 PO4 , 7.5 mM K2 SO4 , 10 mM HEPES-Tris, pH 7.4 and 320 mosmol/l. All solutions were oxygenated with 100% O2 for 15 min before use. After isolation, cells were washed three times by

centrifugation at 100 × g with DMEM supplemented with N-methyl-l-cysteine as a mucolytic agent and suspended in the same solution. The cell viability was evaluated by Trypan blue exclusion. Cells were used only when viability was greater than 95%. We have noted that the centrifugation steps at 100 × g are extremely effective in eliminating contamination with lighter cells. Possible contamination of the cell preparations with non-epithelial LPS-responsive cells was evaluated by three methods. (a) CD45 expression indicating leukocyte contamination. No CD45 positive cells were detected by inmunofluorescence or western blot, using a specific polyclonal antibody (Santa Cruz Biotechnology cat. no. sc-1121), (b) both cell preparations were evaluated for cytokeratin expression, which is only found on epithelial cells, using two anti-cytokeratin antibodies (Santa Cruz Biotechnology cat. nos. sc-8020 and sc-8421), followed by a FITC-conjugated secondary antibody. Very few negative cells (<1%) were observed under the microscope. (c) LPS-induced TNF-␣ secretion. Neither cell preparation showed a TNF response to LPS, even when primed with IFN-␥. Any contamination with monocyte/macrophages could be expected to show up using the very sensitive WEHI bioassay for TNF [15]. For the stimulation of primary intestinal epithelial cells, plates coated with ECM were prepared from CCD-18Co, a human intestinal fibroblasts line (ATCC, Rockville, MD, USA). Cells were cultured with DMEM supplemented with 10% heat-inactivated fetal bovine serum (FBS). Confluent fibroblasts were fixed on plates with a solution of 75% cold methanol; air dried and stored at 4 ◦ C until use. In all cases 5 × 106 pig enterocytes were seeded into 12-wells ECMcoated plates in 2 ml DMEM supplemented with 1% nonessential amino acids and 2 mM glutamine. In each experiment, ECM-coated control plates were incubated in parallel with medium alone. After a 6 h incubation, the medium was recovered, centrifuged at 800 × g for 5 min and the supernatant stored at −20 ◦ C until assayed in the porcine IL-8 ELISA (Biosource International, CA, USA). TLR4 expression was assessed by Western blot from cell lysates using a specific polyclonal antiserum (H80, Santa Cruz Biotechnology). The blots were treated with stripping buffer (1 mM glycine, pH 3.5) for 1 h and then re-probed using anti-Na/K ATPase immune serum (AB1203, Chemicon) to monitor protein loading. IFN-␥ was bound to ECM by incubating ECM plates with 20 U of porcine recombinant IFN-␥ in DMEM, without FBS, overnight. Then, 5 × 106 isolated IEC were added to each well, brought up to 2 ml final volume without removing the IFN-␥, to maintain the same IFN-␥ concentration between ECM-bound and soluble IFN-␥ conditions. LPS (100 ng/ml) was added as indicated. We observed a basal secretion of IL-8 by both unprimed villous and crypt intestinal epithelial cells. Treatment with LPS alone did not alter IL-8 secretion in either cell type. In villous cells, neither ECM-bound nor soluble IFN-␥ changed the IL-8-response to LPS. In contrast, priming crypt cells with

J. Alvarado et al. / Immunology Letters 99 (2005) 109–112

111

Fig. 1. Effect of IFN-␥ bound to ECM or soluble IFN-␥ on LPS-induced IL-8 secretion in isolated villous and crypts intestinal epithelial cells. Cells were incubated for 6 h with LPS (100 ng/ml), soluble IFN-␥ (10 U/ml), ECM-bound IFN-␥ (20 U) or medium as indicated. Supernatants were harvested and IL-8 was determined by ELISA. Each assay was carried out in triplicate, and results are reported as mean ± S.D., * p < 0.05 in comparison to respective unstimulated cells.

ECM-bound IFN-␥ but not with soluble IFN-␥ resulted in a significant increase (p < 0.05) in the LPS response (Fig. 1). This result suggests that the proinflammatory cytokine IFN-␥ bound to extracellular matrix may act differentially on ma-

ture and undifferentiated cells in the intact organ. We tested whether the increase in IL-8 secretion in crypt cells could be associated with increased TLR4 expression. A significant increase in TLR4 expression was only observed under co-

Fig. 2. TLR4 protein expression in crypt intestinal epithelial cells. Effect of LPS and ECM-bound IFN-␥. Isolated cells were treated with LPS (100 ng/ml) and ECM-bound IFN-␥ (20 U) separately for 6 h as indicated. Cell lysates were analyzed by western blotting, using anti-human TLR4 antibody. A representative inmunoblot is shown. TLR4 protein is only detectable after treatment with LPS plus ECM-bound IFN-␥. Lysates of human peripheral mononuclear cells were used as the positive control for TLR4 expression. Equal loading of cell proteins to each lane was confirmed reprobed the nitrocellulose with polyclonal serum against Na/K ATPase.

112

J. Alvarado et al. / Immunology Letters 99 (2005) 109–112

stimulation with LPS plus ECM-bound IFN-␥ (Fig. 2), suggesting that TLR4 mediates the LPS response observed in immature crypt cells. We did not detect TNF-␣ production, by the crypt cell preparations, using a TNF-␣ bioassay (data not shown) under any of the evaluated experimental conditions, discarding therefore, the possibility of contaminating macrophages in our preparations. These findings support the idea that intestinal epithelial cells, rather than macrophages or other lamina propia cell populations, are the predominant cells expressing TLRs in IBD [9]. Luminal LPS is usually well tolerated in large quantities within the healthy intestine [2,8]. This tolerance could result from TLR4 downregulation minimizing LPS recognition. Normal tissue expresses very little or no TLR4 [10]. However, in IBD, host tolerance towards luminal bacterial toxins is altered [1,8], which could reflect an increase in LPS recognition as a result of TLR4 upregulation. Epithelial TLR expression appears to be key in the host defense to bacterial challenges, linking the innate and adaptive immune responses. In this study, we demonstrate that ECM-bound, but not soluble IFN-␥, induced an increase in IL-8 secretion in response to LPS in undifferentiated crypt cells. This effect was associated with an increase in TLR4 expression. Furthermore, IFN-␥ linked to ECM could contribute to perpetuate the intestinal inflammation, selectively altering the epithelial cell response to LPS. A local increase in IFN-␥ in the extracellular matrix could contribute to the inappropriate responsiveness to commensal bacteria and the pathogenesis of idiophatic IBD.

Acknowledgements We thank Dr. Max Ciarlet (Biologics and Clinical Research, Merck & Co., Inc.) for critical reading the manuscript. We also thank M. Barrios (CEA-IVIC) for technical assistance. This work was supported by a grant from FONACIT (S1-20000000554) and by the Instituto Venezolano de Investigaciones Cient´ıficas (IVIC). This study was part of the Masters thesis of Jansi Alvarado at the Centro de Estudios Avanzados (CEA-IVIC).

References [1] Fiocchi C. Inflammatory bowel disease: etiology and pathogenesis. Gastroenterology 1998;115:182–205.

[2] Danese S, Sans M, Fiocchi C. Inflammatory bowel disease: the role of environmental factors. Autoimmun Rev 2004;3:394–400. [3] Funda DP, Tuckova L, Farre MA, Iwase T, Moro I, TlaskalovaHogenova H. CD14 is expressed and released as soluble CD14 by human intestinal epithelial cells in vitro: lipopolysaccharide activation of epithelial cells revisited. Infect Immun 2001;69:3772– 81. [4] Abreu MT, Vora P, Faure E, Thomas LS, Arnold ET, Arditi M. Decreased expression of toll-like receptor-4 and MD-2 correlates with intestinal epithelial cell protection against dysregulated proinflammatory gene expression in response to bacterial lipopolysaccharide. J Immunol 2001;167:1609–16. [5] Suzuki M, Hisamatsu T, Podolsky DK. Gamma interferon augments the intracellular pathway for lipopolysaccharide (LPS) recognition in human intestinal epithelial cells through coordinated up-regulation of LPS uptake and expression of the intracellular toll-like receptor 4MD-2 complex. Infect Immun 2003;71:3503–11. [6] Bocker U, Yezerskyy O, Feick P, Manigold T, Panja A, Kalina U, et al. Responsiveness of intestinal epithelial cell lines to lipopolysaccharide is correlated with Toll-like receptor 4 but not Toll-like receptor 2 or CD14 expression. Int J Colorectal Dis 2003;18:25– 32. [7] Bocker U, Schottelius A, Watson JM, Holt L, Licato LL, Brenner DA, et al. Cellular differentiation causes a selective downregulation of interleukin (IL)-1beta-mediated NF-kappaB activation and IL-8 gene expression in intestinal epithelial cells. J Biol Chem 2000;275:12207–13. [8] Jump RL, Levine AD. Mechanisms of natural tolerance in the intestine: implications for inflammatory bowel disease. Inflamm Bowel Dis 2004;10:462–78. [9] Cario E, Podolsky DK. Differential alteration in intestinal epithelial cell expression of Toll-like receptor 3 (TLR3) and TLR4 in inflammatory bowel disease. Infect Immun 2000;68:7010–7. [10] Abreu MT, Arnold ET, Thomas LS, Gonsky R, Zhou Y, Hu B, et al. TLR4 and MD-2 expression is regulated by immunemediated signals in human intestinal epithelial cells. J Biol Chem 2002;277:20431–7. [11] Camejo EH, Rosengren B, Camejo G, Sartipy P, Fager G, Bondjers G. Interferon gamma binds to extracellular matrix chondroitinsulfate proteoglycans, thus enhancing its cellular response. Arterioscler Thromb Vasc Biol 1995;15:1456. [12] Hurt-Camejo E, Rosengren B, Sartipy P, Elfsberg K, Camejo G, Svensson L. CD44, a cell surface chondroitin sulfate proteoglycan, mediates binding of interferon-gamma and some of its biological effects on human vascular smooth muscle cells. J Biol Chem 1999;274:18957–64. [13] Lortat-Jacob H, Baltzer F, Grimaud JA. Heparin decreases the blood clearance of interferon-gamma and increases its activity by limiting the processing of its carboxyl-terminal sequence. J Biol Chem 1996;271:16139–43. [14] del Castillo JR, Sulbaran-Carrasco MC, Burguillos L. Glutamine transport in isolated epithelial intestinal cells. Identification of a Na+-dependent transport mechanism, highly specific for glutamine. Pflugers Arch 2002;445:413–22. [15] Espevik T, Nissen-Meyer J. A highly sensitive cell line, WEHI 164 clone 13, for measuring cytotoxic factor/tumor necrosis factor from human monocytes. J Immunol Meth 1986;95:99–105.