Toll-like receptor expression in crypt epithelial cells, putative stem cells and intestinal myofibroblasts isolated from controls and patients with inflammatory bowel disease

The aim of our studies was to investigate the expression of Toll-like receptor (TLR)-2 and TLR-4 (and in some studies TLR-5) in myofibroblasts and small and large intestinal crypt epithelial cells from control patients and those affected by Crohn's disease and ulcerative colitis. Isolated and disaggregated crypt epithelial cells and monolayers of myofibroblasts were used for studies by reverse transcription–polymerase chain reaction (RT–PCR), real-time RT–PCR, flow cytometry, immunocytochemistry and Western blot analysis. Compared to control cells, crypt epithelial cells isolated from active ulcerative colitis and Crohn's disease colonic mucosal samples showed significantly higher expression of TLR-2 and TLR-4 transcripts and protein (on the cell surface). There was also enhanced expression of TLR-4 in crypt cells from ileal Crohn's disease. Expression of TLR-2 and TLR-4 transcripts in crypt epithelial cells isolated from inflamed mucosa of distal ulcerative colitis did not differ significantly from such cells obtained from the normal proximal colon. Crypt epithelial cells with side population characteristics (putative stem cells) also expressed transcripts and protein for TLR-2, TLR-4 and TLR-5. Colonic myofibroblast expression of these TLRs was much weaker than in crypt epithelial cells. In conclusion, enhanced TLR-2 and TLR-4 expression by crypt epithelial cells in active inflammatory bowel disease likely reflects greater ability to respond to microbial products. Results from our studies using mucosal samples from patients with distal ulcerative colitis suggest that the enhanced expression of these TLRs could be constitutive. TLR-2, TLR-4 and TLR-5 expression by stem cells imply ability to respond to distinct bacterial products.


Introduction
The inflammatory bowel diseases (IBD), ulcerative colitis and Crohn's disease, are a group of chronic conditions affecting the gastrointestinal tract characterized by a relapsing and remitting course. Although the pathogenesis of IBD remains to be fully understood, studies have implicated the epithelium, innate and adaptive immunity and resident (commensal) bacteria in disease pathogenesis [1,2].
The intestinal epithelium consists of a monolayer of subpopulations of cells of distinct phenotype and function, which are derived from stem cells located in crypts [3,4]. There is increasing recognition of the importance of interactions between intestinal epithelial cells and commensal bacteria (and their products) in the maintenance of normal mucosal homeostasis [5]. Changes in the nature of these interactions are also believed to be required for the development of chronic inflammatory disease of the intestine, as seen in IBD [1,2]. In addition to providing a physical barrier to penetration by resident bacteria and their products, epithelial cells may also shape immune responses mediated by cells in the lamina propria. This may occur via specific receptors which recognize and respond to bacterial products. Toll-like receptors (TLRs) are the best-known sensors of microbial components [6], and act by regulating gene expression.
Studies in mice have shown that TLR-2, TLR-4 and TLR-5 control intestinal epithelial homeostasis and provide protection from injury, such as that mediated by dextran sodium sulphate and radiation [7][8][9][10]. TLR-2 and TLR-4 sense Gram-positive and Gram-negative bacterial wall components (lipoteichoic acid and lipopolysaccharide, respectively) and TLR-5 binds monomeric flagellin of motile bacterial cell walls. Although agonists for these TLRs have been reported to provide protection against radiation injury [10,11], there are conflicting reports regarding their role in established murine models of IBD [12].
Human intestinal epithelial expression of TLR-2 and TLR-4 has been studied in tissue sections. In histologically normal colonic tissue the findings have been inconsistent, with epithelial expression reported to be confined to crypts [13,14], minimally detectable [15] or absent [16]. In tissue affected by IBD, epithelial expression of these TLRs was reported to be absent [16] or abundant for TLR-4 [15].
In patients with active IBD, changes in the epithelium are prominent and include loss of barrier function and loss of monolayer continuity (ulceration). During remission, epithelial repair and regeneration occurs via stem cell-derived progeny and processes such as restitution. Epithelial barrier function [17] and restitution [18] can be regulated by myofibroblasts that are located under the basement membrane. Myofibroblasts are also believed to be important regulators of intestinal stem cell function, via secretion of Wnt ligands [19] and bone morphogenetic protein (BMP) antagonists [20]. Moreover, myofibroblasts are capable of responding to luminal bacterial products via expression of TLRs [21,22], but relative expression compared to epithelial cells is unknown.
The aims of our studies were to investigate the expression of TLR-2, TLR-4 and TLR-5 in human intestinal crypt epithelial cells and putative stem cells isolated from control tissue and that affected by IBD. Myofibroblast expression of the TLRs was also studied.
Histological examination of mucosal samples from patients with inflammatory bowel disease showed mild to severe inflammation and they were on the following treatment at the time of intestinal resection: mesalazine (14), corticosteroids (eight), azathioprine/6-mercaptopurine (15), methotrexate (four), infliximab/adalimumab (eight), cyclosporin (one), metronidazole (one) (see Supporting information, Table S1). Specimens from patients who had received pre-operative chemotherapy or radiotherapy of any type or duration were excluded.
The above mucosal samples, which were surplus to clinical requirements, were used following informed consent from patients. This research was approved by the Nottingham Research Ethics Committee.

Isolation and disaggregation of crypt epithelial cells
Intestinal crypts were isolated and disaggregated as described previously [23,24]. In brief, after washing with calcium-and magnesium-free Hanks's balanced salt solution (HBSS), mucosal strips were incubated (for 30 min at 37°C, with shaking), on three occasions in 1 mM ethylenediaminete traacetic acid (EDTA) plus 0·05 mM dithiothreitol (DTT). Between the incubation steps, the mucosal strips were washed with HBSS. Released crypts were subsequently disaggregated using 0·25% pancreatin (Sigma, St Louis, MO, USA) and the cell suspensions were stored on ice prior to use in experiments.

Myofibroblast isolation and co-culture with crypt epithelial cells
Primary colonic myofibroblasts were isolated and cultured as described previously [25]. Briefly, mucosal samples denuded of epithelial cells (as described above), were cultured [at 37°C, in RPMI-1640 supplemented with 10% fetal calf serum (FCS)] to allow myofibroblasts to migrate out via basement membrane pores and to establish in culture. Established colonies of myofibroblasts were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% FCS, 1% non-essential amino acids (Gibco, Carlsbad, CA, USA), and 200 mM glutamine (Sigma). Following passage, the myofibroblasts were kept frozen prior to their use in experiments. Expression of α-smooth muscle actin and vimentin (by immunohistochemistry) confirmed the phenotype of the myofibroblasts.
For co-culture, myofibroblasts were grown to confluency on sterile glass coverslips before application of isolated and disaggregated crypt epithelial cells, as described previously [24]. Following co-culture for 30 min, non-adherent cells were removed by washing before the coverslips were fixed in cold acetone for 1 min and stored at −20°C until required.

Flow cytometry and cell sorting
Isolated and disaggregated crypt epithelial cells were incubated (in the dark at 4°C for 1 h) with the following fluorophore-conjugated monoclonal antibodies: BerEP4fluorescein isothiocyanate (FITC) (Dako), immunoglobulin (Ig)G2aκ isotype control-allophycocyanin (APC) (eBioscience), TLR-2-APC (eBioscience), TLR-4-APC (eBioscience) or CD45-AF488 (BioLegend, London, UK). Additional control included incubation in medium only (no primary antibody). Cells were subsequently washed, resuspended in 0·5% formaldehyde in phosphate-buffered saline (PBS) and stored at 4°C in the dark until analysis. A minimum of 20 000 total events per sample tube were collected for analysis on Moflo XDP (Beckman Coulter, High Wycombe, UK). Initial analysis was by forward-and sidescatter to exclude aggregates and non-viable cells. Side population cells were identified as reported previously [24,26].
Disaggregated crypt epithelial cells were used with and without prior incubation (at 37°C for 15 min) with either 50 μmol/l verapamil (Sigma-Aldrich, St Louis, MO, USA) or 10 μmol/l fumitremorgin C (Alexis Biochemicals, Exeter, UK). Hoechst 33342 (Sigma) was added to a final concentration of 2·5 μg/ml and the cells incubated (in the dark) for 30 min at 37°C, followed by 30 min at 4°C. Following centrifugation, the cells were resuspended in 0·5 ml of 2% fetal calf serum in HBSS with Ca/Mg and 10 mM HEPES, followed by incubation with normal mouse serum and fluorophore-conjugated monoclonal antibodies (above) for 1 h at 4°C. Following resuspension in medium at 4°C, the cells were analysed immediately on Moflo XDP (Beckman Coulter).
Viable and non-aggregated crypt cells were identified using forward-and side-scatter analysis and lack of cellular uptake of propidium iodide (PI), and analysed on a Beckman Coulter MoFlo cell sorter. Hoechst 33342 was excited at 405 nm and fluorescence emission measured using a 450/50 nm band-pass filter ('Hoechst blue') and a 620 nm long-pass filter ('Hoechst red').
Side population cells were demonstrated as those with low fluorescence in both the blue and red channels, which was ameliorated in cells pre-incubated with verapamil or fumitremorgin C, which block multi-drug resistance protein (mdr) or mdr-like mediated efflux of the Hoechst dye [26]. Following measurement of the fluorescence signal in the relevant gated region, 5 × 10 3 side population cells were sorted into Eppendorf tubes on ice and centrifuged, before total RNA was isolated for subsequent mRNA expression analysis, as described below. All flow cytometric data were analysed using Weasel version 3 software.

RT-PCR (conventional and real-time)
Total RNA was extracted using Qiagen RNeasy Plus Mini Kit (Qiagen, Venlo, the Netherlands), as per the manufacturer's instructions for eukaryotic cellular RNA. Synthesis of complementary DNA (cDNA) from mRNA was undertaken using the Qiagen QuantiTect RT kit (Qiagen), according to the manufacturer's instructions.

Western blot analysis
Disaggregated crypt epithelial cells (10 6 ) or colonic myofibroblasts were washed in PBS and incubated in CelLytic M reagent (Sigma) supplemented with phosphatase inhibitor cocktail 2 (Sigma) and protease inhibitor cocktail (Sigma), according to the manufacturer's instructions. The lysate was centrifuged at 10 000 g and the protein-containing supernatant was stored at −80°C until required.
Immunostaining was performed using a Vectastain ABC Universal kit (Vector Laboratories), according to the manufacturer's instructions.

Statistical analyses
Normally distributed data were analysed using paired or unpaired Student's t-test, as appropriate. Non-normally distributed data were analysed using non-parametric tests, Kruskal-Wallis test and either a Wilcoxon signed-rank test or Mann-Whitney U-test. Categorical data were analysed using Fisher's exact test. Statistical analyses were undertaken using spss (version 19) and Graphpad Prism (version 5) statistical software packages. All statistical tests were twotailed and those with P-values less than 0·05 (5%) were deemed statistically significant.

Studies in isolated intestinal myofibroblasts
Using conventional RT-PCR, myofibroblasts isolated from normal control and active IBD mucosal samples showed PCR products specific for TLR-2 and TLR-4 (not shown). TLR-2 and TLR-4 protein expression was confirmed by Western blot analysis, but the level of expression was much lower than that for isolated crypt epithelial cells (Fig. 8).

Discussion
To date, the role of TLRs in intestinal epithelial cells has been investigated predominantly in mice and human cell lines, with only limited studies in primary human mucosal epithelial cells. Heterogeneity in expression of TLR-4 has been reported in epithelial cell lines [28][29][30]. In tissue sections of human intestinal mucosal samples, reports of epithelial expression of TLR-2 and TLR-4 have been inconsistent [13][14][15][16].
In findings that we believe have not been reported previously, our studies using isolated and disaggregated colonic crypt epithelial cells consistently showed expression of not only transcripts, but also TLR-2 and TLR-4 protein on the cell surface. Compared to histologically normal controls, crypt epithelial cells isolated from colonic mucosal samples affected by ulcerative colitis and Crohn's disease demonstrated enhanced expression of TLR-2 and TLR-4 transcripts and cell surface protein. These studies suggest greater capacity for colonic crypt epithelial cells in inflammatory bowel disease to respond to luminal microbial products that bind these receptors.
The enhanced epithelial expression of TLR-2 and TLR-4 is likely to have occurred in response to proinflammatory cytokines [28,31]. However, we report for the first time that  Limitations of our studies include the use of relatively small numbers of samples, which were obtained from operation resection specimens. Although the control histologically normal mucosal samples were obtained distant from the cancer in the resection specimen, it is conceivable that the presence of the neoplasm may affect TLR expression in the adjacent tissue. We believe this is unlikely, but future studies using samples from patients without cancer can address this issue. Studies using epithelial cells isolated from endoscopic biopsies from IBD patients while not on any treatment will also be of interest.
Isolated and disaggregated crypt epithelial cells used in our studies also contain stem cells, which give rise to the progeny that differentiate as they migrate to the surface of the mucosa.
Stem cells with so-called side population characteristics (based on the ability to efflux the DNA-binding dye Hoechst 33342) have been characterized in the bone marrow [26] and murine intestine [32,33]. We have shown previously that isolated and purified (using cell sorter) putative human colonic epithelial stem cells with side population characteristics adhere to monolayers of primary human colonic myofibroblasts [24]. In novel studies, we now report that these putative human colonic stem cells express TLR-2, TLR-4 and TLR-5. Wnt signalling is important in regulating stem cell function and a recent study has reported the ability of TLR-4 to activate the canonical Wnt pathway in colonic epithelial cell lines [34]. Expression of TLR-4 in Lgr5-positive murine small intestinal stem cells has also been reported [35]. Moreover, loss of TLR-4 in murine intestinal epithelial cells has been shown to lead to goblet cell differentiation, probably via suppression of Notch signalling in stem cells [36].
Studies suggest that, in contrast to the epithelium, TLR signalling in lamina propria cells leads to proinflammatory responses [37]. Beneficial effects mediated by TLR-2 and TLR-4 receptors in the intestinal epithelium have been observed predominantly in models of radiation injury [11] and colitis induced by dextran sulphate sodium [9,38,39] and Citrobacter rodentium [40]. Stem cells are sensitive to radiation [3], and our studies suggest that the protective effects of TLR-4 [11] and TLR-5 [10] ligands could be mediated directly via receptors expressed on the surface of these cells.
Human intestinal myofibroblasts, which demonstrate characteristics of fibroblasts and smooth muscle cells [25], are located immediately subjacent to the epithelium. In the crypt, they represent an important component of the stem cell niche [19,20], and have also been implicated in adenoma initiation and growth [41]. The demonstration of intestinal myofibroblast expression of TLRs [21,22] represents an increasing appreciation of their role in mediating mucosal immunological responses [42,43]. Our studies have confirmed the expression of TLR-2 and TLR-4 in myofiboblasts isolated from normal colonic mucosal samples, and report for the first time that levels of the proteins were lower than in isolated crypt epithelial cells. Future studies can investigate the significance of the differences between crypt epithelial cells and myofibroblasts in expression of these TLRs.

Supporting information
Additional Supporting information may be found in the online version of this article at the publisher's web-site:  .  Table S1. Details of patients studied. TNFα = tumour necrosis factor-α. *P < 0·05; **P < 0·01 versus healthy controls. Table S2. Relative quantitative expression of Toll-like receptor (TLR)-2 and TLR-4 mRNA transcripts in isolated and disaggregated colonic and small intestinal crypt epithelial cells obtained from histologically normal control mucosal samples and those affected by active ulcerative colitis (UC), Crohn's colitis and ileal Crohn's disease. Extracted RNA was used for real-time reverse transcription-polymerase chain reaction (RT-PCR) and data for UC and Crohn's disease are presented as 'fold change' in expression of transcripts compared to mean expression in the control group in which the crypt epithelial cells were obtained from histologically normal colonic and small intestinal mucosal samples. IQR = interquartile range. Table S3. Quantitative surface Toll-like receptor (TLR)-2 and TLR-4 protein expression by colonic crypt epithelial cells. Isolated and disaggregated crypt epithelial cells were obtained from mucosal samples affected by active Crohn's colitis, active ulcerative colitis or from histologically normal control colonic tissue. The cells were labelled with anti-BerEP4-fluorescein isothiocyanate (FITC) antibody and either anti-TLR-2 allophycocyanin (APC), anti-TLR-4-APC or isotype control monoclonal antibodies and analysed by flow cytometry. Surface TLR-2 and TLR-4 proteinassociated median fluorescence intensity was determined in BerEP4-positive (gated) epithelial cells. IQR = interquartile range.