Role of alkaline phosphatase in colitis in man and rats

Transcription

1 Role of alkaline phosphatase in colitis in man and rats A Tuin, 1 K Poelstra, 1 A de Jager-Krikken, 1 L Bok, 2 W Raaben, 3 M P Velders, 3 G Dijkstra 2 c Competing interests: Declared (the declaration can be viewed on the Gut website at supplemental). 1 Department of Pharmacokinetics and Drug Delivery, University Centre for Pharmacy, University of Groningen, The Netherlands; 2 Department of Gastroenterology and Hepatology, University Medical Centre Groningen, The Netherlands; 3 AM-Pharma, Bunnik, The Netherlands Correspondence to: Dr A Tuin, Department of Pharmacokinetics and Drug Delivery, University Centre for Pharmacy, University of Groningen, Antonius Deusinglaan 1, 9713 AV Groningen, The Netherlands; a. tuin@rug.nl KP and GD contributed equally to this work. Revised 11 September 2008 Accepted 12 September 2008 Published Online First 13 October 2008 ABSTRACT Background and aims: Crohn s disease (CD) and ulcerative colitis (UC) are chronic multifactorial inflammatory bowel diseases (IBDs) with unknown aetiology, but a deregulated mucosal immune response to gutderived bacterial antigens is thought to be involved. Tolllike receptor ligands, especially lipopolysaccharide (LPS), contribute to the maintenance of the disease. It has previously been shown that the enzyme alkaline phosphatase (AP) is able to detoxify LPS, and the aim of this study was to examine a possible role in IBDs. Methods: Intestinal AP (iap) mrna expression and LPS dephosphorylation in intestinal biopsies of control subjects and patients with IBD were examined, and the effect of orally administered iap tablets on the progression of dextran sodium sulfate-induced colitis in rats was subsequently studied. Results: In healthy persons, iap mrna and enzyme activity was high in the ileum relative to the colon. In patients with UC and CD, iap mrna expression was found to be markedly reduced when inflamed tissue was compared with non-inflamed tissue. Oral administration of iap tablets to colitic rats resulted in a significant attenuation of colonic inflammation as reflected by reduced mrna levels for tumour necrosis factor a, interleukin 1b, interleukin 6 and inducible nitric oxide synthase NOS (inos), a reduced inos staining and inflammatory cell influx, and a significantly improved morphology of the intestinal wall. Conclusions: The present study shows that epithelial iap mrna expression is reduced in patients with UC and CD. The rat model demonstrates that oral administration of active iap enzymes in the intestinal tract results in a significant reduction of inflammation. This provides new insight on IBD pathology and a novel treatment approach to this severe inflammatory disease. Crohn s disease (CD) and ulcerative colitis (UC) are inflammatory bowel diseases (IBDs) of the digestive tract that are the result of an inappropriate and ongoing activation of the mucosal immune system driven by the presence of the normal luminal flora. 1 The exact causes of IBD are still unclear, but environmental factors, genetic predisposition and immunological disorders are suggested to be involved. Studies have shown that mutations in several genes, such as NOD2/CARD15 in CD 2 and Toll-like receptor 4 (TLR4) in CD and UC, 34 are associated with the occurrence of IBD. The intracellular protein encoded by the NOD2 gene interacts with bacterial products such as peptidoglycans, 56 and TLR4 is the signalling receptor for lipopolysaccharide (LPS). So, deficiencies in response mechanisms against bacterial products, in particular LPS, are an important factor in IBD. We demonstrated previously that alkaline phosphatase (AP) dephosphorylates LPS, 7 10 which was also confirmed by others, 11 resulting in the formation of a non-toxic lipid A group within the LPS molecule. In general, the lipid A group of LPS contains two phosphate groups that are responsible for the toxicity of LPS, and AP removes at least one of these phosphate groups. This enzyme is abundantly present along the microvilli in the small intestine of all species, 12 indicating a possible role in the protection of the host against LPS. As AP is able to detoxify LPS and response mechanisms against LPS are changed during IBD, we wondered whether the levels of AP are changed in the intestines of IBD patients. Therefore, iap mrna expression and LPS dephosphorylation in intestinal biopsies of control subjects and patients with IBD were determined. Furthermore, we studied the efficacy of orally administrated acidresistant iap tablets on dextran sodium sufate (DSS)-induced colitis in rats. In this study, we show that epithelial iap expression is decreased in patients with UC and CD, and that oral iap administration ameliorates inflammation in colitic rats. These observations provide novel insights and a rationale for new therapeutic strategies against IBD through augmentation of AP in the intestinal lumen. MATERIALS AND METHODS Patient characteristics/specimen collection Intestinal mucosal biopsy specimens were obtained during endoscopy following informed consent (approved by the Ethics Committee of the University Medical Centre Groningen) from patients with CD, UC and control subjects. Patient characteristics are described in table 1. Diagnosis of IBD was established by endoscopic and histopathological examination. The control subjects were referred to our endoscopy centre because of polyp surveillance or changed stool frequency. In control subjects, biopsies were obtained from four different intestinal areas (ileum, ascending colon, transverse colon and rectum). Biopsies from patients with CD were obtained from the edge of ulceration s or aphtoid lesions if present, and from macroscopic noninflamed areas using a standard biopsy forceps. Only patients with CD in whom the disease had affected the colon were included in our study that is, only colonic biopsies were used. In patients with UC, biopsies were not taken from the proximal upper level (transition zone) but from the inflamed and non-inflamed part of the colon. Intestinal specimens were immediately snap-frozen in liquid nitrogen for mrna and protein Gut 2009;58: doi: /gut

2 Table 1 Patient s characteristics Total number (males/females) Mean age (years; range) Medication (without/with) CD 10 (3/7) 38 (16 65) 5/5 Thiopurine (3), corticosteroids (2) methotrexate (1), infliximab (1) UC 14 (9/5) 35 (18 47) 6/8 Corticosteroids (5), mesalamine (6), thiopurine (4) Healthy controls 14(8/6) 41 (17 58) 12/2 Glimepiride/atorvastatine and ramipril (1), macrogol (1) CD, Crohn s disease; UC, ulcerative colitis. analysis or liquid nitrogen-cooled isopentane for immunohistochemical staining, and stored at 280uC until further processing. Enzyme histochemical detection of LPS dephosphorylation and AP activity LPS-dephosphorylating activity of human iap was examined in cryostat sections (5 mm) of biopsies of human small intestine (ileum) and colon (ascendens, descendens and rectum) with LPS (Escherichia coli serotype O55:B5, Sigma, St Louis, Missouri, USA) as a substrate according to a modified method of Wachstein and Meisel. 13 Briefly, cryostat sections were mildly fixed for 10 min in 4% formalin macrodex at 4uC and subsequently incubated in Tris HCl buffer ph 7.6 containing LPS (final concentration 3.2 mg/ml), MgSO 4 (final concentration 0.01 M) and Pb(NO 3 ) 2 (final concentation 0.06% w/v) at 37uC for 1 h. A lead phosphate precipitate will be formed at sites of enzyme activity, which is converted to a lead sulfate precipitate by incubating the sections with Na 2 S. The lead sulfate precipitate will appear as a dark brown staining. The specificity of this staining has been demonstrated previously by inhibition of AP activity using L-phenylalanine, a specific inhibitor of iap. 14 Control sections were incubated in the same medium without the substrate LPS. AP activity using the conventional substrate b-glycerophosphate was visualised according to the method of Gomorri. 15 RNA isolation and real-time PCR RNA was isolated from human intestinal biopsies and rat colon using the QIAGEN RNeasy Mini Kit (Qiagen, Hilden, Germany), and subsequently converted to cdna with the Promega Reverse Transcription System (Promega, Madison, Wisconsin, USA). The cdna was amplified with appropriate primers (tables 2 and 3) by quantitative real-time PCR using SYBR Green (Applied Biosystems, Warrington, UK)) and products were detected using the ABI PRISM 7900HT Detection System (Applied Biosystems). Relative quantification of the genes was calculated using the comparative threshold cycle (CT) method as described by Van de Bovenkamp, using glyceraldehyde phosphate dehydrogenase (GAPDH) as a housekeeping gene. 16 Table 2 Primers used for amplification of the listed genes in human cdna Gene Forward primer Reverse primer GAPDH CCATCACCATCTTCCAGGAG CCTGCTTCACCACCTTCTTG iap ACGCGGCAATGAGGTCATCT CCGCCAAGGATCACGTCAAT Villin TGACCCTGAGACCCCATC TCAGCAGTGATCTGGCTCCA GAPDH, glyceraldehyde phosphate dehydrogenase; iap, intestinal alkaline phosphatase. In vivo experiment To examine whether exogenous iap affects experimental colitis, a colonic inflammation was induced in male Sprague Dawley rats by DSS (catalogue no , MP Biomedicals, Solon, Ohio, USA). The rats were divided into four groups; (1) normal drinking water and placebo tablets (n = 5); (2) normal drinking water and iap tablets (n = 5); (3) 5% DSS in drinking water and placebo tablets (n = 10); and (4) 5% DSS in drinking water and iap tablets (n = 10). Both the iap and placebo tablets had a diameter of 5.3 mm and an enteric coating, consisting of eudragit L, triethylcitrate and talc, to prevent dissolution in the stomach, which would destroy the activity of acid-sensitive AP enzymes. The ph at which the tablets dissolved was determined to be 5.5. The iap tablets contained 1250 glycine units of iap (specific activity: 1035 U/mg protein), as determined by a standard enzyme activity assay. Oral treatment consisted of daily administration of a tablet, under isoflurane/o 2 /N 2 O anaesthesia (isoflurane from Abbott Laboratories, Queensborough, Kent, UK), from day 1 to day 7 after the start of DSS administration. Rats in Groups 3 and 4 received DSS in their drinking water throughout the whole experiment (day 1 to day 8). From day 1 to day 8, the rats were weighed daily, their consumption of drinking water was measured and their condition was scored using a standard scoring procedure. At day 8, the rats were anaesthetised with isoflurane/o 2 /N 2 O and sacrificed by heart puncture. Faeces was collected for measurement of AP activity. The colon was harvested and scored macroscopically by examining whether there was distension (score 1), partial distension (score 0.5) or no distension (score 0), and whether the serosa was thickened (score 1), partially thickened (score 0.5) or not thickened at all (score 0). The proximal, middle and distal parts of the colon were scored separately and all scores were summed. Thereafter, the colon was weighed, the length was measured and tissues samples of the distal part were stored for RNA isolation and real-time PCR analysis, as described above, for several genes (table 3). Table 3 Primers used for the amplification of the listed genes in rat cdna Gene Forward primer Reverse primer GAPDH CCATCACCATCTTCCAGGAG CCTGCTTCACCACCTTCTTG TNFa ATGTGGAACTGGCAGAGGAG GGCCATGGAACTGATGAGAG IL1b AGGCAGTGTCACTCATTGTG GGAGAGCTTTCAGCTCACAT IL6 CCGGAGAGGAGACTTCACAG ACAGTGCATCATCGCTGTTC inos CGTTCGATGTTCAAAGCAAA CCCTGGACTTCTCACTCTGC IL10R GCCCAGAGACTCTCGATGAC AAGACCCTTCCTTTCCCAGA Villin TGTGGAACTGGCAGGGAG GGGGTGGGTCTTGAGGTATT GAPDH, glyceraldehyde phosphate dehydrogenase; iap, intestinal alkaline phosphatase; IL, interleukin; inos, inducible nitric oxide synthase; TNFa, tumour necrosis factor a. 380 Gut 2009;58: doi: /gut

3 The rest of the colon was filled with Tissue-Tek, rolled up and frozen in isopentane for histochemical analysis. AP activity in faeces Homogenates of,1 g/ml rat faeces in water were centrifuged at 2000 rpm to spin down insoluble materials. The supernatant was removed and centrifuged at rpm to remove insoluble materials completely. Samples were diluted 0, 2, 4, 8, 16, 32, 64 and 128 times in a 96-well plate in 0.05 M ammediol buffer containing 2 mm MgCl 2. After addition of 10 ml of 10 mg/ml 4-nitrophenyl phosphate disodium salt, the plate was incubated for 30 min at 37uC. The reaction was stopped by adding 105 ml of 1 N NaOH. The optical densities were measured at 405 nm on a Thermomax microplate reader. Histochemistry The H&E staining was performed according to standard procedures. Myeloperoxidase (MPO) activity in activated neutrophils was visualised according to Poelstra et al. 17 This staining was inhibitable by catalase. AP activity in neutrophils was shown using the substrate b-glycerophosphate as described earlier in the Materials and Methods section. The staining for inos was done according to standard indirect immunoperoxidase techniques with a rabbit polyclonal antibody directed against inos, and GARPO (DAKO, Glostrup, Denmark) as the secondary antibody. The inos antibody was developed in the laboratory of Dr H Moshage (University of Groningen, The Netherlands) and has been described previously. 18 Peroxidase activity was visualised with 3-amino-9-ethylcarbazole. The staining for villin was performed like the inos-staining using a goat polyclonal antibody against villin (sc-7672, Santa Cruz Biotechnology Inc., Santa Cruz, California, USA) and RAGPO and GARPO as a secondary and third antibody, respectively (both from DAKO). Statistical analysis Statistical analysis of patient data was done by an unpaired two-tailed Student t test, assuming similar variances, and they are expressed as the mean (SD). Calculation of the correlation between iap and villin mrna levels in patients was done using the Pearson correlation test. The data of the animal experiment were subjected to a non-parametric one-sided Mann Whitney U test, because these data were not normally distributed. Differences were considered significant at p,0.05. RESULTS LPS dephosphorylation, AP activity and iap mrna levels in healthy human intestinal biopsies To investigate whether the human small intestine and colon have LPS-dephosphorylating activity, cryostat sections of biopsies of human terminal ileum, colon ascendens, colon transversum and rectum of healthy persons were examined. Results from the enzyme histochemical analysis showed that the epithelial cells in sections of human ileum have a high LPSdephosphorylating activity as demonstrated by a brown lead sulfate precipitate along the apical side of the microvilli of the enterocyte (fig 1A). In contrast, LPS dephosphorylation was absent in human colon sections; colon ascendens (fig 1C), colon transversum (fig 1E) and rectum (not shown). Occasional cells stained positive, which probably reflects AP activity in macrophages and endothelial cells of small blood vessels. When LPS was omitted from the incubation medium, no staining was detected (fig 1G). Figure 1 Lipopolysaccharide (LPS) dephosphorylation and alkaline phosphatase (AP) activity in intestinal cryostat sections of a normal healthy person (representative sections of n = 3 healthy persons). A brown staining in the case of LPS dephosphorylation and a black staining in the case of AP activity was clearly visible along the apical side of the microvilli of the enterocyte in the terminal ileum (A and B). In contrast, biopsies from the colon ascendens (C and D), colon transversum (E and F) and rectum (not shown) showed hardly any LPS-dephosphorylating activity and AP activity along the enterocytes. Occasional cells stained positive (arrows). Control sections (without LPS) were completely negative (G). Magnification The staining for LPS dephosphorylation at physiological ph levels was performed on the same tissue as the staining for AP activity using the conventional substrate b-glycerophosphate at ph 9. The staining for LPS dephosphorylation showed exactly the same pattern in ileum (fig 1B), colon ascendens (fig 1D), colon transversum (fig 1F) and rectum as the AP activity. In addition to LPS dephosphorylation, the biopsies were inspected for their iap mrna levels. As previous studies have shown that AP can dephosphorylate LPS, 7 8 AP expression levels may be indicative of the LPS-detoxifying capacity of the intestine. iap mrna levels in the human ileum were found to be,30 times higher than those in the human colon (fig 2). Gut 2009;58: doi: /gut

4 Figure 2 Intestinal alkaline phosphatase (iap) expression throughout the human colon and in the ileum. Biopsies were taken from six healthy persons and analysed for iap mrna levels by real-time reverse transcription-pcr. Expression levels of iap mrna were related to the average expression level in the colon transversum. Values are expressed as mean (SD), *p,0.005 vs ileum. LPS dephosphorylation and iap mrna levels in colon biopsies of patients with UC and CD After assessing LPS dephosphorylation at ph 7.4 and AP activity with b-glycerophosphate as a substrate at ph 9.0 in the small intestine and colon of healthy persons, we examined whether these activities had changed during IBD. Similar to normal colon biopsies, no LPS dephosphorylation could be detected in colon biopsies from patients with CD or UC (data not shown). We subsequently examined mrna levels of iap in colon biopsies. Real-time PCR revealed clear differences in mrna levels between inflamed and non-inflamed tissue within each patient also compared with control (healthy) persons. A significant decline in mrna levels was found in patients with UC as well as in those with CD; 2.8-fold and 2.4-fold decreases, respectively, in iap mrna levels were observed when inflamed and non-inflamed tissue of the same patient was compared (fig 3A,B). Compared with healthy controls, iap mrna levels were significantly decreased in non-inflamed tissue of patients with UC and CD. To examine whether this decline in mrna levels was due to the destruction of colonic epithelial cells by ulceration, we measured the amount of villin mrna, which is a marker for intestinal epithelial cells. Villin expression was significantly reduced in inflamed mucosa compared with noninflamed mucosa in patients with UC and CD (fig 3C,D). A significant correlation was found between the iap and villin mrna levels for patients wih UC (fig 3E; p,0.001, R 2 = 0.61) and those with CD (fig 3F; p,0.001, R 2 = 0.61), which links the iap expression to the intact intestinal epithelium. Figure 3 Intestinal alkaline phosphatase (iap) and villin mrna expression in control subjects (n = 11), non-inflamed and inflamed intestinal tissue of patients with ulcerative colitis (UC) (n = 14; A and C) and those with Crohn s disease (CD) (n = 10; B and D). Total RNA was isolated from biopsies of non-inflamed and inflamed intestinal tissue of patients with UC and CD, and of intestinal tissue of healthy subjects. mrna levels for iap and villin were related to glyceraldehyde phosphate dehydrogenase mrna levels. (E and F) The correlation between iap expression levels and villin levels in patients with UC and CD, respectively (open circles). iap and villin expression levels of healthy subjects (triangles) are also included in these graphs. 382 Gut 2009;58: doi: /gut

5 Figure 4 Effects of acid-resistant intestinal alkaline phosphatase (iap) or placebo tablets on the macroscopic score of the colons in male Sprague Dawley rats with (n = 10 per group) or without (n = 5 per group) dextran sodium sulfate (DSS)-induced colitis after 8 days of treatment. Horizontal bars represent the median of each group. The difference between untreated and AP-treated DSS rats was significant at p,0.05. Effect of oral administration of enteric-coated iap tablets on DSS-induced colitis in rats To examine whether oral administration of iap affects experimental colitis, inflammation was induced in the colon in male Sprague Dawley rats by oral intake of DSS. Oral administration of AP in tablets resulted in a higher AP activity within the intestine. AP activity in faeces at 24 h after treatment had risen from 0.13U/g of faeces in rats that received normal drinking water plus placebo tablets to 7.08 U/g of faeces in rats that received normal drinking water plus AP tablets. The total recovery of AP activity from acid-resistant AP tablets in the intestine is,30%, as determined previously. 21 The daily consumption of drinking water by the rats showed no significant differences, confirming that both treated and untreated rats received the same amount of DSS. In this study no difference in weight loss between the different groups that received DSS in their drinking water was found. The rats in the AP-treated and placebo-treated group lost 11.0% and 11.9% of their initial body weight, respectively. The control groups that is, the groups that received normal drinking water plus either placebo or AP tablets, showed an average increase of body weight of 4.5% and 4.6%, respectively. Macroscopic evaluation of the colons revealed that the colon morphology of AP-treated DSS rats was better than the colon morphology of placebo-treated DSS rats, as illustrated by a significantly lower macroscopical score (fig 4, p,0.05). Measurement of the colon length revealed a significant difference between DSS-treated rats receiving placebo tablets and DSS-treated rats receiving AP tablets; 8.1 (0.8) cm in the placebo-treated group vs 9.5 (1.3) cm in the AP-treated group (p,0.005). The shortening of the colon in DSS-treated rats was associated with a clearly visible thickening of the intestinal wall, as reflected by a significant increase in weight per length in DSS-treated rats. AP treatment reduced this thickening (p,0.01). In the distal part of the colon, relative mrna expression levels of the following genes were determined: tumour necrosis factor a (TNFa), interleukin 1b (IL1b), IL-6, inducible nitric oxide synthase (inos), IL10R and villin (fig 5). The mrna levels of the inflammatory genes, TNFa, IL1b, IL6, inos and IL10R were all strongly upregulated in DSS-treated rats. Expression levels of all these genes were significantly lower in DSS rats that received AP tablets compared with DSS rats that received placebo tablets (p,0.05). Only 2 out of 10 rats treated with AP still had elevated levels of inflammatory parameters. In contrast, mrna levels for the epithelial marker villin, which were profoundly reduced in placebo-treated DSS rats, were significantly increased in colitic rats receiving AP tablets (p,0.05). Interestingly, the two rats that displayed aberrant cytokine, inos and IL10R levels also had reduced villin levels. So again a strong correlation between villin levels and inflammatory response was found. These data indicate that the epithelial layer in the colon of rats receiving DSS plus placebo tablets is severely damaged whereas the epithelial layer in rats receiving DSS plus AP tablets is significantly less damaged. The colon was also histochemically evaluated by H&E, inos, MPO and villin staining. The colons of DSS rats receiving AP tablets were clearly less damaged than those of DSS rats that received placebo tablets (fig 6C vs B). The epithelial lining of the colon was more intact, confirming the PCR results for villin mrna levels, and, in line with the macroscopic data (fig 4), the serosa was clearly less thickened in AP-treated rats. In both groups of rats that received normal drinking water, no staining for the inflammation marker inos was observed (fig 6D). However, in placebo-treated DSS rats, all colon segments displayed a profound inos staining (fig 6E). In contrast, colons from AP-treated DSS rats showed no or only minor inos staining (fig 6F). The number of MPO-positive cells in the colon was greatly increased in DSS-treated rats receiving placebo tablets (fig 6H) compared with normal rats (fig 6G). Administration of AP tablets to DSS-treated rats markedly reduced this influx of MPO-positive cells (fig 6I). As MPOpositive cells also possess AP activity, the staining for AP activity with the conventional AP substrate b-glycerophosphate shows the same pattern as the MPO staining. Control rats show some AP-positive cells (fig 6L), whereas DSS-treated rats show a high influx of AP-positive cells (fig 6K), which is clearly reduced when the rats are treated with AP tablets (fig 6L). The epithelial marker villin was abundantly present along the epithelial lining of the mucosa in control rats (fig 6M), but completely absent in colons of DSS rats that received placebo tablets (fig 6N). In contrast, DSS rats that received AP tablets displayed a clear villin staining along the epithelial cells which was comparable with control rats (Fig. 6O), which is in line with the PCR data for villin mrna. Taken together, all data indicate that the inflammation of the colon and the structural damage after DSS exposure is significantly reduced upon AP treatment. DISCUSSION IBD is one of the major chronic inflammatory diseases, affecting millions of people world-wide. The disease is the result of a deregulated immune response in the intestinal wall and, although bacterial products may not be the initiators of disease, there are many studies showing that components of the normal luminal flora do play a role in its progression by stimulating the local inflammatory process During inflammation, a local influx of TLR4- and CD14-positive cells occurs, thereby introducing LPS-responsive cells at the site of intestinal damage. Several reports have shown that the enzyme AP is able to dephosphorylate and detoxify this LPS The resulting product, LPS with a monophosphoryl lipid A moiety, is nontoxic 26 and may even antagonise the effects of LPS. 27 Because of Gut 2009;58: doi: /gut

6 Figure 5 Effects of oral treatment with acid-resistant intestinal alkaline phosphatase (iap) or placebo (plac.) tablets on mrna expression of several genes in male Sprague Dawley rats with (n = 10 per group) or without (n = 5 per group) dextan sodium sulfate (DSS)-induced colitis after 8 days of treatment. Dot plots show the expression levels of tumour necrosis factor a (TNFa), interleukin 1b (IL1b), IL6, inducible nitric oxide synthase (inos), IL10R and villin mrna in the distal part of the rat colons. Horizontal bars represent the median of each group. Expression levels were normalised to glyceraldehyde phosphate dehydrogenase. For each parameter, the difference between untreated and AP-treated DSS rats was significant at p,0.05 (*). 384 Gut 2009;58: doi: /gut

7 Inflammatory bowel disease Figure 6 H&E staining (A C), histochemical staining for inducible nitric oxide synthase (inos) (D F), activated neutrophils (G I), alkaline phosphatase (AP) (J L) and villin (M O) in male Sprague Dawley rats with dextran sodium sulfate (DSS)-induced colitis after 8 days of (continued on next page) Gut 2009;58: doi: /gut

8 the perpetuating inflammatory response towards LPS during IBD, we hypothesised that administration of LPS-detoxifying molecules would attenuate the inflammatory response during periods of severe inflammation. In previous studies, we found high LPS-dephosphorylating activity along the intestinal wall of rats, 8 9 and in the present study this is confirmed in human intestinal biopsies. This enzyme activity was predominantly found along the villi of the small intestine. Previous studies have indicated that this LPSdephosphorylating activity represents AP activity, which is, with this particular substrate, expressed at physiological ph levels. LPS-dephosphorylating activity was absent along the luminal wall of the colon. Despite this, mrna for iap could be detected in biopsies of human colon (fig 2). These iap mrna levels were significantly decreased in biopsies of patients with UC and CD (fig 3), which demonstrates a relationship between this enzyme and IBD. In fact, mrna levels were already reduced in non-inflamed tissue compared with normal tissue, which might indicate that this reduction is an early marker for disease. Another explanation might be that reduction of food intake, which is common in IBD patients, induces this decline in iap mrna levels in non-inflamed biopsies of IBD patients. 28 Why no enzyme activity could be detected despite the presence of mrna for this enzyme remains to be established. Because the substrate concentration in our study is relatively low, at least a factor 10 lower than in conventional methods, 7 enzyme activity might be below detection levels but it may also be inactivated here. The enzymatic product (monophosphoryl lipid A) can reduce this enzyme activity, 9 and this product is most probably present within the colon. Attempts to detect the enzyme itself with specific antibodies failed. In any case, local LPS-dephosphorylating activity is very low within the colon, relative to the small intestine, whereas bacterial titres are very high at this site. If AP represents a true protective enzyme against LPS, high expression levels would be anticipated here. It is, however, also conceivable that at a site where bacterial growth is high and LPS is continuously shed from growing Gram-negative bacteria, continuous LPS dephosphorylation is not efficient. An intact structural barrier and an absence of LPSresponsive elements might be better safeguards. Many data including our own illustrate that when the epithelial layer within the colon is intact (reflected in our studies by the high mrna levels for the epithelial marker villin), the response on LPS is absent in rats (fig 5). Only when the integrity of the wall is affected might additional LPS-degrading enzymes be needed. So, based on this, we hypothesise that within the ileum, AP exerts a protective role by detoxification of LPS, whereas in the colon AP activity only plays a role after damage to the intestinal wall. We therefore explored whether administration of exogenous iap reduces intestinal inflammation when the natural intestinal barrier is affected. Bovine iap was for that reason orally administered to rats with DSS-induced colitis. Chronic DSS administration is a well known model for human UC and CD in rats 30 and associated with damage to the epithelial cell layer in the colon. In our studies this was evidenced by a significant reduction in mrna levels for villin in this area (fig 5). AP enzymes are very acidsensitive and hence will be completely inactivated within the stomach. Therefore, we employed tablets that offered full protection in the stomach and release of enzyme activity within the colon. 21 We assessed colonic integrity, disease parameters and inflammatory responses. Strong effects of exogenous iap were found on the integrity and morphology of the intestinal wall as well as on LPS-induced inflammatory responses (fig 6), but not on weight loss. Our treatment aims at detoxifying LPS, yet several studies show that LPS does not influence all disease activity markers, such as weight loss, in the DSS model: in DSStreated LPS-sensitive (C3H/He) and LPS-insensitive (C3H/HeJ) mice, the reported disease index (a score based on weight loss, diarrhoea and rectal bleeding) was the same in both groups. 31 In another study, the effect of DSS was studied in germ-free and conventional mice. 32 Again, bacterial products did not influence the weight of the animals after DSS. However, germ-free rats did display a reduced intestinal inflammation in the wall of the colon, compared with conventional mice. 32 Also in the study of Lange et al, 33 an LPS-hyporesponsive mouse strain treated with DSS displayed reduced TNFa levels, rectal bleeding and mortality compared with a normoresponsive mouse strain, but morphological signs in the initial phase of disease were the same. Apparently, in the initial phase, LPS does not play a role in DSS-induced clinical symptoms, but it does play a role in the progression of disease by enhancing the inflammatory response Some studies have found an increase in iap activity after DSS treatment. 34 This increase in enzyme activity was strongly associated with influx of inflammatory cells, 34 that contain high AP activity (see also fig 6J L). Inflammatory cells express the enzyme in the same granula where LPS accumulates, 36 which fits with the proposed function of the enzyme. In our study, treatment of rats with iap was associated with a significant protective effect on the epithelial cell layer and a strong reduction in all inflammatory parameters examined. In the DSS-treated group, mrna levels for villin were.90% reduced in all rats (10/10) compared with control animals (n = 10). This was associated with enhanced mrna levels for TNFa, IL1b, IL6, inos and IL10R in all rats (p,0.001). However, in DSS-rats that received a daily treatment with iap, only 3 out 10 rats had villin mrna levels around zero (p,0.001), and mrna levels for all inflammatory markers were significantly reduced compared with untreated DSS rats (p,0.05). Only two animals had a high inflammatory response, and these very same rats were also characterised as having low villin mrna levels. Why the treatment failed in 2 out of 10 animals could not be explained, but tablets might have been chewed or not swallowed properly by the rats. Yet, our results clearly show that DSS invariably and very strongly affected villin mrna levels, and AP tablet administration prevented this in 7 out of 10 cases. In addition, DSS invariably enhanced all inflammatory parameters, whereas iap prevented this significantly, with the exception of 2 rats. The effects of iap on epithelial integrity and inflammatory parameters were corroborated by the immunohistochemical treatment with placebo (middle column) or acid-resistant intestinal AP (iap) tablets (right-hand column). Normal rats are depicted in the left-hand column. Note the mucosal ulceration and thickening of the serosa in placebo-treated colitic rats (B) compared with oral iap-treated colitic rats (C). The inflammation marker inos was absent in non-colitic rats (D), enhanced in placebo-treated colitic rats (E) and strongly decreased in oral iap-treated colitic rats (F). The amount of activated neutrophils was low in non-colitic rats (G), high in placebo-treated colitic rats (H) and clearly reduced in iap-treated colitic rats (I). The AP activity corresponded to the amount of neutrophils; low AP activity in non-colitic rats (J), high AP activity due to neutrophil influx in DSS-treated rats (K) and a clearly lowered activity in iap-treated colitic rats (L). The red staining for the epithelial marker villin was visible along the epithelial lining in non-colitic rats (M), absent in placebo-treated colitic rats (N) and retained in iap-treated DSS rats (O). Magnification of all figures is Gut 2009;58: doi: /gut

9 studies. The staining for villin was absent in DSS-treated rats and present in animals treated with iap. In parallel with this, the inflammatory markers inos and neutrophil influx were prominently present in DSS-treated animals, and strongly reduced in diseased animals receiving iap. Also, general morphology as assessed by H&E staining was better in animals receiving iap compared with placebo-treated animals. AP tablet administration did not change endogenous AP activity along the epithelial cell lining, but strongly reduced the number of APpositive cells in the lamina propria, reflecting the reduced influx of inflammatory cells that express AP. Based on these data, we conclude that iap has protective effects within the colon. A role for LPS during colitis has been demonstrated by many others, and iap has been found to dephosphorylate and detoxify LPS, so it is therefore conceivable that AP exerts this protective effect via LPS detoxification. LPS-dephosphorylating activity is high in the human small intestine but absent in the intact wall of the colon. However, during disease the protective effects of exogenous iap become apparent in vivo in an experimental animal model of IBD. A high AP activity within the intestinal wall has been observed by many authors, but its function has been unknown. The fact that LPS dephosphorylation is found at physiological ph levels and both localisation and expression levels are associated with the presence of LPS (in activated neutrophils and macrophages, along vascular walls, in liver and intestine) supports our hypothesis that LPS detoxification is a physiological function of this enzyme. These data also support the notion that oral administration of iap in patients with UC and CD may be therapeutically effective. This prompted us to perform a phase II proof of concept study in patients with UC which will be reported later. Funding: This work was financially supported by the Dutch Organization for Scientific Research (NWO). Ethics approval: This study was approved by the Ethics Committee of the University Medical Centre Groningen. REFERENCES 1. Podolsky DK. Inflammatory bowel disease. N Engl J Med 2002;347: Hugot JP, Chamaillard M, Zouali H, et al. Association of NOD2 leucine-rich repeat variants with susceptibility to Crohn s disease. Nature 2001;411: 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: Franchimont D, Vermeire S, El HH, et al. Deficient host bacteria interactions in inflammatory bowel disease? The toll-like receptor (TLR)-4 Asp299gly polymorphism is associated with Crohn s disease and ulcerative colitis. Gut 2004;53: Inohara N, Ogura Y, Fontalba A, et al. Host recognition of bacterial muramyl dipeptide mediated through NOD2. Implications for Crohn s disease. J Biol Chem 2003;278: Girardin SE, Boneca IG, Viala J, et al. Nod2 is a general sensor of peptidoglycan through muramyl dipeptide (MDP) detection. J Biol Chem 2003;278: Poelstra K, Bakker WW, Klok PA, et al. Dephosphorylation of endotoxin by alkaline phosphatase in vivo. Am J Pathol 1997;151: Poelstra K, Bakker WW, Klok PA, et al. A physiologic function for alkaline phosphatase: endotoxin detoxification. Lab Invest 1997;76: Bentala H, Verweij WR, Huizinga-Van der Vlag A, et al. Removal of phosphate from lipid A as a strategy to detoxify lipopolysaccharide. Shock 2002;18: Beumer C, Wulferink M, Raaben W, et al. Calf intestinal alkaline phosphatase, a novel therapeutic drug for lipopolysaccharide (LPS)-mediated diseases, attenuates LPS toxicity in mice and piglets. J Pharmacol Exp Ther 2003;307: Bates JM, Akerlund J, Mittge E, et al. Intestinal alkaline phosphatase detoxifies lipopolysaccharide and prevents inflammation in zebrafish in response to the gut microbiota. Cell Host Microbe 2007;2: Domar U, Nilsson B, Baranov V, et al. Expression of intestinal alkaline phosphatase in human organs. Histochemistry 1992;98: Novikoff AB. Enzyme localizations with Wachstein Meisel procedures: real or artifact. J Histochem Cytochem 1967;15: Tillyer CR, Rakhorst S, Colley CM. Multicomponent analysis for alkaline phosphatase isoenzyme determination by multiple linear regression. Clin Chem 1994;40: Josza L, Lehto MU, Jarvinen M, et al. A comparative study of methods for demonstration and quantification of capillaries in skeletal muscle. Acta Histochem 1993;94: van de Bovenkamp M, Groothuis GM, Draaisma AL, et al. Precision-cut liver slices as a new model to study toxicity-induced hepatic stellate cell activation in a physiologic milieu. Toxicol Sci 2005;85: Poelstra K, Hardonk MJ, Koudstaal J, et al. Intraglomerular platelet aggregation and experimental glomerulonephritis. Kidney Int 1990;37: Vos TA, Gouw AS, Klok PA, et al. Differential effects of nitric oxide synthase inhibitors on endotoxin-induced liver damage in rats. Gastroenterology 1997;113: Aono K, Isobe K, Kiuchi K, et al. In vitro and in vivo expression of inducible nitric oxide synthase during experimental endotoxemia: involvement of other cytokines. J Cell Biochem 1997;65: Hofer D, Drenckhahn D. Cytoskeletal markers allowing discrimination between brush cells and other epithelial cells of the gut including enteroendocrine cells. Histochem Cell Biol 1996;105: Eriksson HJ, Verweij WR, Poelstra K, et al. Investigations into the stabilisation of drugs by sugar glasses: II. Delivery of an inulin-stabilised alkaline phosphatase in the intestinal lumen via the oral route. Int J Pharm 2003;257: Brand S, Staudinger T, Schnitzler F, et al. The role of Toll-like receptor 4 Asp299Gly and Thr399Ile polymorphisms and CARD15/NOD2 mutations in the susceptibility and phenotype of Crohn s disease. Inflamm Bowel Dis 2005;11: Gazouli M, Mantzaris G, Kotsinas A, et al. Association between polymorphisms in the Toll-like receptor 4, CD14, and CARD15/NOD2 and inflammatory bowel disease in the Greek population. World J Gastroenterol 2005;11: Ogura Y, Bonen DK, Inohara N, et al. A frameshift mutation in NOD2 associated with susceptibility to Crohn s disease. Nature 2001;411: Cuthbert AP, Fisher SA, Mirza MM, et al. The contribution of NOD2 gene mutations to the risk and site of disease in inflammatory bowel disease. Gastroenterology 2002;122: Schromm AB, Brandenburg K, Loppnow H, et al. The charge of endotoxin molecules influences their conformation and IL-6-inducing capacity. J Immunol 1998;161: Astiz ME, Rackow EC, Still JG, et al. Pretreatment of normal humans with monophosphoryl lipid A induces tolerance to endotoxin: a prospective, double-blind, randomized, controlled trial. Crit Care Med 1995;23: Goldberg RF, Austen WG Jr, Zhang X, et al. Intestinal alkaline phosphatase is a gut mucosal defense factor maintained by enteral nutrition. Proc Natl Acad Sci USA 2008;105: Su F, Brands R, Wang Z, et al. Beneficial effects of alkaline phosphatase in septic shock. Crit Care Med 2006;34: Elson CO, Sartor RB, Tennyson GS, et al. Experimental models of inflammatory bowel disease. Gastroenterology 1995;109: Stevceva L, Pavli P, Buffinton G, et al. Dextran sodium sulphate-induced colitis activity varies with mouse strain but develops in lipopolysaccharide-unresponsive mice. J Gastroenterol Hepatol 1999;14: Kitajima S, Morimoto M, Sagara E, et al. Dextran sodium sulfate-induced colitis in germ-free IQI/Jic mice. Exp Anim 2001;50: Lange S, Delbro DS, Jennische E, et al. The role of the Lps gene in experimental ulcerative colitis in mice. APMIS 1996;104: Camuesco D, Comalada M, Concha A, et al. Intestinal anti-inflammatory activity of combined quercitrin and dietary olive oil supplemented with fish oil, rich in EPA and DHA (n-3) polyunsaturated fatty acids, in rats with DSS-induced colitis. Clin Nutr 2006;25: Smith GP, Harris H, Peters TJ. Studies of the biochemical and immunological properties of human neutrophil alkaline phosphatase with comparison to the established alkaline phosphatase isoenzymes. Clin Chim Acta 1984;142: Detmers PA, Zhou D, Powell D, et al. Endotoxin receptors (CD14) are found with CD16 (Fc gamma RIII) in an intracellular compartment of neutrophils that contains alkaline phosphatase. J Immunol 1995;155: Gut 2009;58: doi: /gut