Tumor necrosis factor-receptor 2 is up-regulated on lamina propria T cells in Crohn\'s disease and promotes experimental colitis in vivo

3142

M. H. Holtmann et al.

Eur. J. Immunol. 2002. 32: 3142–3151

Tumor necrosis factor-receptor 2 is up-regulated on lamina propria T cells in Crohn’s disease and promotes experimental colitis in vivo Martin H. Holtmann1, Eleni Douni2, Michael Schütz1, Geraldine Zeller1, Jonas Mudter1, Hans-Anton Lehr3, Jeanette Gerspach4, Peter Scheurich4, Peter R. Galle1, George Kollias2 and Markus F. Neurath1 1

Laboratory of Immunology, I. Medical Clinic, University of Mainz, Mainz, Germany Institute of Immunology, Biomedical Sciences Research Center Al. Fleming, Vari, Greece 3 Institute of Pathology, University of Mainz, Mainz, Germany 4 Department of Cell Biology and Immunology, University of Stuttgart, Stuttgart, Germany 2

Tumor necrosis factor (TNF) plays a pivotal role in the pathogenesis of Crohn’s disease (CD). However, little is known about the role of TNF receptors (TNF-R) in this disease. Here, we found that TNF-R2 (in contrast to TNF-R1) was significantly up-regulated on lamina propria and peripheral blood T cells in CD compared to control patients. To directly test the functional role of TNF-R2 in Th1-mediated experimental colitis in vivo, we took advantage of transgenic animals overexpressing TNF-R2 in T cells. Reconstitution of SCID mice with CD4+ CD62L+ T cells from TNF-R2 transgenic mice led to an earlier wasting syndrome, a more severe colitis and augmented Th1 cytokine production than reconstitution with cells from wild-type littermates. In addition, TUNEL staining revealed a significantly decreased apoptosis rate of lamina propria mononuclear cells in mice reconstituted with TNF-R2 transgenic T cells compared to mice reconstituted with wild-type T cells. In summary, our data suggest a critical regulatory role of TNF-R2 signaling for disease exacerbation in Th1mediated chronic colitis. Taken together with the increased expression of TNF-R2 in CD, selective targeting of TNF-R2 signaling thus emerges as a potentially novel approach to the treatment of CD. Key words: TNF-R2 / Cytokine / IBD / Apoptosis / T lymphocyte

1 Introduction

Received Revised Accepted

6/2/02 15/7/02 27/8/02

The last two authors share senior authorship of this manuscript.

induction of cell proliferation in some experimental systems [10] but was thought to contribute to TNF signaling mainly indirectly by passing over sTNF to TNF-R1 [5]. Later studies have demonstrated that TNF-R2 can be strongly activated by mTNF rather than sTNF [11]. TNFR2 could thus be a crucial mediator of pro-inflammatory TNF effects at the local level in a paracrine or autocrine fashion [12]. However, most of these data on TNF-R signaling have been generated in in vitro cell systems. In vivo models have shown that both TNF-R are involved in mediating important functions of TNF, such as TNFmediated cytotoxicity, cell death and tissue necrosis [13, 14]. Furthermore, overexpression of TNF-R2 in transgenic mice was associated with a general inflammatory syndrome [15]. However, in that model TNF-R2 was overexpressed ubiquitously, and only little is known about the function of TNF-R2 in the mucosal immune system.

Abbreviations: CD: Crohn’s disease LP: Lamina propria RE: Relative expression LPMC: LP mononuclear cells mTNF: Membrane-bound TNF sTNF: Soluble TNF HPF: High-power field

Based on previous observations that active CD is associated with increased production of TNF by intestinal mononuclear cells [16, 17] and direct evidence from vari-

TNF is a proinflammatory 17-kDa protein that exists in membrane-bound (mTNF) and soluble (sTNF) forms [1]. The biological effects of TNF are mediated via two distinct cell surface receptors (TNF-R1 and TNF-R2) also designated p60 and p80 based on their molecular weights [2–6]. The importance of TNF signaling via TNFR1, rather than TNF-R2, was explained by the higher binding affinity of TNF-R1 to soluble TNF. TNF-R2 on the other hand had been mainly associated with the activation and nuclear translocation of NF- ‹ B [7–9] and the

[I 22847]

0014-2980/02/1111-3142$17.50 + .50/0

© 2002 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Eur. J. Immunol. 2002. 32: 3142–3151

TNF-R2 involvement in IBD

3143

ous murine models of chronic intestinal inflammation [18, 19], TNF is assumed to be a key player in the pathogenesis of CD. This has led to the application of various antiTNF strategies in the treatment of CD [20]. However, to date all strategies to block the TNF/TNF-R system target both the soluble and membrane-bound forms of the ligand, and thus block all biological effects of TNF. This may account, at least in part, for various adverse side effects associated with anti-TNF therapies, in particular the increased susceptibility to infections. To devise more specific ways to block TNF action, we need to learn more about the signaling pathways of TNF in CD. In an initial approach towards this goal, we investigated the expression of TNF-R on lamina propria (LP) T cells in patients with active CD, and studied the role of TNF-R2 in a Th1 T cell-mediated murine model of experimental colitis induced by reconstitution of SCID mice with CD62L+ CD4+ T cells [18, 21].

2 Results 2.1 Up-regulation of TNF-R2 but not TNF-R1 on LP T lymphocytes in CD To determine whether active Crohn’s disease is associated with changes in the expression of TNF-R, we analyzed TNF-R expression by flow cytometry. Based on initial studies showing a linear relation between relative expression (RE = mean fluorescence intensity of TNF-R antibody minus intensity of an isotype control antibody) values of TNF-R determined by FACS and absolute numbers of TNF-R on the cell surface determined by competitive binding studies with 131I-radiolabeled TNF (not shown), we analyzed TNF-R levels on peripheral blood (PB) and LP T lymphocytes by FACS. TNF-R1 expression on PB CD3+ T cells was virtually identical in patients with CD (n=15) and controls (n=10) with low RE values (Fig. 1, Table 1), consistent with previous reports showing that TNF-R1 is constitutively expressed in normal resting and activated T cells [4]. Subsequently, we analyzed TNF-R1 expression on LP T cells from CD patients and controls. Since isolation of T cells from the LP involved enzymatic digestion of the tissue with DNase, hyaloronidase and collagenase for 5 min, we performed control experiments to rule out the possibility that such treatment had an impact on receptor expression on the cell surface (data not shown). TNF-R1 expression on LP T cells was hardly above the lower detection level and no differences were noted between patients with active CD or controls. We observed that TNF-R2 expression on PB CD3+ T cells was much higher than expression levels of TNF-R1. Furthermore, there was an almost twofold-increased RE of

Fig. 1. Assessment of TNF-R1 and TNF-R2 expression in Crohn’s disease. TNF-R1 and TNF-R2 expression on PBMC and LPMC from patients with CD and healthy controls was determined as RE. Cell populations were identified by double staining with PE-conjugated antibodies against CD3, CD4, CD8, and CD14 plus a TNF-R specific antibody. Shown are the individual data points from patients with CD and controls. Statistically significant changes in receptor expression between CD and control patients are indicated by * (p X 0.05). There was a significant up-regulation of TNFR2 but not TNF-R1 expression on lamina propria CD4+ T cells in CD compared to controls.

TNF-R2 on CD3+ T cells in patients with active CD compared to controls (p X 0.05) (Fig. 1). Subpopulation analysis revealed that this up-regulation on CD3+ T cells was predominantly due to up-regulation on CD8+ T cells but not on CD4+ T cells (Fig. 1). With regard to LP mononuclear cells (LPMC), we found that TNF-R2 expression was higher than TNF-R1 expression (Fig. 1). On LP CD4+

3144

M. H. Holtmann et al.

Eur. J. Immunol. 2002. 32: 3142–3151

Table 1. Side by side comparison of the RE of TNF-R1 and TNF-R2 exemplified for the PBMC populations of control patients CD3 Control

CD4

CD8

CD14

TNF-R1

TNF-R2

TNF-R1

TNF-R2

TNF-R1

TNF-R2

TNF-R1

TNF-R2

1

0.21

1.05

0.24

0.84

0.09

0.84

3.22

3.37

2

0.15

0.66

0.19

0.91

0.14

0.65

1.29

1.29

3

0.14

1.46

0.24

1.21

0.13

0.59

0.98

1.21

4

0.15

1.05

0.22

0.98

0.15

0.92

0.60

0.71

5

0.12

0.78

0.15

0.84

0.12

0.54

1.06

1.21

6

0.16

0.91

0.16

0.98

0.17

1.94

2.52

2.28

7

0.14

0.98

0.19

1.37

0.09

1.46

2.79

2.52

8

0.15

0.98

0.12

1.13

0.11

0.84

0.71

0.65

9

0.19

1.13

0.16

1.05

0.14

1.21

2.17

2.16

10

0.15

1.05

0.20

1.06

0.13

1.29

1.95

1.84

Mean

0.16

1.01

0.18

1.04

0.13

1.03

1.73

1.72

SE

0.01

0.14

0.01

0.05

0.01

0.14

0.29

0.27

T cells relative expression of TNF-R2 was significantly (p X 0.05) increased in CD compared to control patients, suggesting a potential local pathophysiological role of TNF-R2 rather than TNF-R1 on LP T cells in patients with CD.

2.2 Overexpression of TNF-R2 on T lymphocytes augments and accelerates the development of chronic intestinal inflammation in vivo To analyze whether the increased expression of TNF-R2 in CD could be pathogenetically relevant, we examined the role of TNF-R2 in a Th1-mediated adoptive transfer model of chronic intestinal inflammation. Accordingly, T cell-deficient SCID mice on a BALB/c background were reconstituted with splenic CD4+ CD62L+ T cells from transgenic mice overexpressing TNF-R2 (CD4+ CD62L+ TNF-R2-TG T cells) and from wild-type (WT) control BALB/c mice (CD4+ CD62L+ WT T cells). It was found that SCID mice reconstituted with WT T cells showed diarrhea and wasting disease 8 weeks after the T cell transfer and displayed maximal weight loss (about 13%) after 10 weeks (Fig. 2). In contrast, SCID-mice reconstituted with CD4+ CD62L+ TNF-R2-TG T cells started to develop chronic diarrhea, rectum prolapse and wasting disease much earlier, at 2–4 weeks after T cell transfer. Furthermore, disease development was associated with a much more dramatic weight loss of 20% after 6 weeks and almost 30% after 10 weeks (Fig. 2, 3A). Develop-

ment of colitis was due to transfer of donor cells, since unreconstituted SCID mice treated with PBS intraperitoneally showed no signs of wasting disease or colitis (Fig. 2). Macroscopic analysis showed that colons of SCID mice reconstituted with CD4+ CD62L+ TNF-R2 TG T cells were enlarged and showed a more thickened, hyperemic colon wall compared to SCID mice reconstituted with WT T cells (Fig. 3C). Histopathological analysis of colon cross-sections from SCID mice reconstituted with WT T cells showed inflammation with low to moderate levels of mononuclear cell infiltrates in the lamina propria (Fig. 4). The epithelial surface, however, was intact and no crypt abscesses could be found. An average colitis score of 0.9±0.9 was found in a blinded quantitation of colitis severity. In contrast, cross-sections from SCID mice reconstituted with CD4+ CD62L+ TNF-R2 TG T cells exhibited a much more pronounced lymphohistiocytic infiltrate, epithelial lesions, crypt abscesses and enlarged microvessels as signs of active inflammation (Fig. 4). The average colitis score in this group was 2.88±0.25, significantly increased compared to the WT reconstituted control group (p X 0.01). In contrast, colon sections from SCID mice treated with PBS showed no signs of inflammation (Fig. 4).

Eur. J. Immunol. 2002. 32: 3142–3151

Fig. 2. Overexpression of TNF-R2 promotes experimental T cell-dependent colitis in vivo. Weight changes (% of the initial body weight) ± SEM of C.B-17 SCID mice reconstituted with splenic CD4+ CD62L+ T cells from WT donor mice or TNF-R2 transgenic donor mice (TNF-R2 TG). The donor cells were prepared as described [24]. Donor cells (106) were injected i.p. into recipient SCID mice, which were 10–12 weeks old by the time of reconstitution with an average body weight of 21±2.5 g. To ensure, that development of colitis was due to reconstitution with donor cells, unreconstituted control mice injected with PBS were included. Mice reconstituted with WT T cells developed a wasting syndrome with weight loss. Mice reconstituted with TNF-R2 TG T cells started to show weight loss 2 weeks after reconstitution. The latter mice had, on average, a weight loss of 27% after 10 weeks, while mice reconstituted with WT T cells exhibited a less-pronounced wasting disease. Mice injected with PBS did not develop a wasting syndrome. Data represent one out of two representative experiments with n=7 for the groups given CD4+ CD62L+ T cells and n=4 for the group treated with PBS.

2.3 Reconstitution of SCID mice with TNF-R2 transgenic T cells results in augmented IFN- q but not IL-4 production Next, we quantitated cytokine production of splenic mononuclear cells from reconstituted SCID mice. Splenic mononuclear cells from SCID mice reconstituted with CD4+ CD62L+ TNF-R2 TG T cells produced large amounts of IFN- + with levels that were beyond the range of quantitation of 10,000 pg/ml in the supernatant (Fig. 5). Mononuclear cells from control mice reconstituted with WT T cells produced significantly less IFN- + with a mean of 7,370±2,795 pg/ml (p X 0.05). In contrast, IL-4 production by mononuclear cells did not differ between the two groups. Cells from WT reconstituted mice had IL-4 levels of 366±407 ng/ml in the superna-

TNF-R2 involvement in IBD

3145

Fig. 3. Overexpression of TNF-R2 promotes experimental colitis: Macroscopic findings. (A) Photographs of a representative C.B.-17 SCID mouse given TNF-R2 TG T cells 10 weeks after reconstitution (left) compared to a control mouse given WT T cells (right). C.B.-17 SCID-mice reconstituted with TNF-R2 TG T cells developed more severe chronic colitis with a more pronounced wasting syndrome than control mice given WT T cells. (B) Mice given TNF-R2 TG T cells lost on average a quarter of their body weight by 10 weeks after the T cell transfer and showed a rectal prolapse more frequently (left) than mice given WT T cells (right). (C) Representative photographs of the colon of a SCID mouse given TNF-R2 TG T cells (bottom) and of a control SCID mice given T cells from WT mice (top). The colon of the SCID mouse given TNF-R2 transgenic T cells is strikingly enlarged, swollen and hypervascularized.

tant, and cells from CD4+ CD62L+ TNF-R2 TG T cell reconstituted mice showed levels of 354±76 ng/ml.

3146

M. H. Holtmann et al.

Eur. J. Immunol. 2002. 32: 3142–3151

Fig. 4. Overexpression of TNF-R2 promotes experimental colitis: histological findings. (A) Photomicrographs of hematoxylin/ eosin-stained cross-sections of the colon from SCID mice reconstituted with TNF-R2 TG T cells (left) and WT T cells (right), respectively. SCID mice reconstituted with TNF-R2 TG T cells showed a severely thickened and edematous colon wall compared to WT T cell reconstituted mice. (B) Both SCID mice reconstituted with WT T cells (center) and TNF-R2 TG T cells (left) showed lymphohistiocytic infiltration that was not seen in unreconstituted SCID-mice treated with PBS (right). However, lymphohistiocytic infiltration in TNF-R2 TG T cell reconstituted SCID mice was much more severe and transmural extending to the lamina muscularis propria.

2.4 Suppression of mononuclear cell apoptosis in the LP of mice reconstituted with T cells from TNF-R2 transgenic mice An increased LP infiltrate in experimental colitis as observed in mice reconstituted with TNF-R2 transgenic T cells can reflect either increased cell migration, proliferation and expansion or else decreased apoptosis of inflammatory cells. In fact, T cell resistance against apoptosis has been demonstrated in patients with active CD and is thought to play an important role in disease perpetuation [21–24]. To determine whether regulation of mononuclear cell apoptosis plays a role in the aggravation of experimental colitis observed by overexpression of TNF-R2, we performed in situ detection of apoptotic cells in colon section from mice with experimental colitis (Fig. 6). Of note, the number of fluorescence-positive apoptotic cells detected by TUNEL assays remained strikingly low in the LP of mice reconstituted with TNFR2 transgenic T cells in spite of the dramatically increased cell infiltrate in the colon (Fig. 6A). This finding suggests a decreased rate of cell death in the inflamed colon. In fact, quantitative analysis of the percentage of apoptotic mononuclear cells showed a 7.1% apoptosis rate in the colon of the control group (Fig. 6C). In con-

trast, the apoptosis rate in the colon of mice reconstituted with TNF-R2 transgenic T cells was significantly lower (3.3%; p X 0.001), suggesting a significant inhibition of cell death through increased signaling via TNR-R2 on T cells.

3 Discussion Previous data suggested that the cytokine TNF plays a key role in the immunopathogenesis of Crohn’s disease and Th1-mediated animal models of chronic intestinal inflammation [17–20, 25]. However, little information has been available so far on the role of TNF-R and TNFdependent signal transduction in chronic intestinal inflammation. In the present study, we provide clinical and experimental evidence that TNF exerts proinflammatory effects through increased signaling via TNF-R2. We found that TNF-R2 was up-regulated on CD4+ T cells in the LP from patients with Crohn’s disease compared to controls. The pathophysiological relevance of TNF-R2 overexpression was further underlined by in vivo studies in an animal model of CD induced by reconstitution of SCID mice with CD62L+ CD4+ T cells. In this model, transgenic overexpression of TNF-R2 caused an

Eur. J. Immunol. 2002. 32: 3142–3151

TNF-R2 involvement in IBD

3147

transgenic T cells led to a much faster development of severe colitis than reconstitution of mice with CD4+ CD62L+ T cells from WT mice. Production of the Th1 cytokine IFN- + was significantly increased in mononuclear cells isolated from the spleen of TNF-R2 transgenic T cell-reconstituted mice, while IL-4 production showed no difference to cells isolated from SCID mice reconstituted with T cells from WT mice. This finding suggests that augmented TNF-R2 signaling in T cells favors and augments the development of a Th1-like cytokine profile, as it is found in CD. Thus, the above-mentioned increased expression of TNF-R2 in CD LP T cells could augment the Th1 T cell response that is characteristically observed in this disease.

Fig. 5. Overexpression of TNF-R2 augments Th1 cytokine production. Shown are cytokine levels in the supernatant of splenic mononuclear cells from reconstituted SCID mice. Cells were isolated 10 weeks after the T cell transfer and cultured in RPMI with 10% FCS for 72 h with and without activation of the accessory pathway using anti-CD3/anti-CD28 antibodies. Cytokine levels were determined by ELISA. Cells from the TNF-R2 TG T cell-reconstituted mice produced significantly more IFN- + than cells from control mice upon stimulation with anti-CD3 plus anti-CD28 antibodies.

aggravation of Th1-dependent chronic intestinal inflammation, presumably due to the induction of apoptosis resistance and augmented Th1 cytokine production. The observed increase in TNF-R2 expression on LP T cells in CD is most likely associated with augmented TNF-R2 signaling in this disease, since CD is associated with enhanced production of TNF by LP cells [16, 17, 26]. Since TNF-R2 signaling results in NF- ‹ B p65 activation and since NF- ‹ B has been identified as principal effector molecule in TNF-R2 signaling [7, 9], enhanced TNF-R2 signaling might also contribute to the previously described activation of the transcription factor NF- ‹ B in CD [27–29] that is known to regulate caspase activation and apoptosis [30]. Hence, our data are consistent with a model in which augmented TNF signaling via TNF-R2 leads to activation of NF- ‹ B, thereby preventing apoptosis of LP mononuclear cells in chronic intestinal inflammation. To investigate the role of increased TNF-R2 signaling in vivo, we studied the role of TNF-R2 in the CD4+ CD62L+ T cell transfer model of experimental colitis [21] taking advantage of transgenic mice that overexpress TNF-R2. Reconstitution of SCID mice with CD4+ CD62L+ TNF-R2

In addition, we were able to show that the apoptosis rate of LPMC is significantly decreased in mice reconstituted with CD4+ CD62L+ TNF-R2 transgenic T cells compared to controls. The long-term effects of a decreased apoptosis rate by 50% during several months of chronic colitis are probably quite remarkable and could well contribute to inappropriate lymphocyte accumulation and cell expansion in experimental colitis in vivo. These data from our experimental mouse model system are thus in agreement with the observation that CD is associated with a resistance of mucosal cells against apoptosis that is most likely driven by proinflammatory cytokines such as IL-6 and TNF [22–24]. In fact, it has been recently demonstrated that antibodies to TNF induce apoptosis of LPMC in CD [31], suggesting that TNF signaling is a key factor in the regulation of programmed cell death in the inflamed mucosa. It has recently been discovered that TNF-R2 preferentially binds mTNF rather than sTNF with high affinity and that TNF-R2 can be activated in an autocrine and paracrine fashion [11, 12]. Taken together with the observed increased expression of TNF-R2 in CD, these data suggest that signaling via mTNF and TNF-R2 on LP macrophages and T cells in CD could be an important mechanism for activation of these cells during local cell-cell interactions. Unfortunately, little is known about the role of mTNF in chronic intestinal inflammation. However, Corazza et al. [32] recently observed in an adoptive transfer model of colitis that TNF production by non-T cells in the recipient mice rather than by T cells from the donor mice is critical for colitis development. Taken together with our observations, this finding is consistent with a pathogenesis model of colitis in which mTNF on non-T cells is important for activation of donor T cells that carry mTNF-R. Interestingly, a recent clinical trial in patients with active CD showed that treatment with a human sTNF-R:Fc fusion protein that inactivates sTNF but not mTNF had no clinical effect [33, 34]. This further supports the emerging notion that TNF signaling by

3148

M. H. Holtmann et al.

Eur. J. Immunol. 2002. 32: 3142–3151

Fig. 6. Overexpression of TNF-R2 suppresses apoptosis of LPMC in experimental colitis. (A) Shown are representative colon sections from control SCID mice reconstituted with WT cells (left) and mice reconstituted with TNF-R2 TG T cells (right) with TUNEL staining. Apoptotic cell bodies appear green by fluorescence analysis. In spite of a significantly increased lymphohistiocytotic infiltrate in the LP of the TNF-R2 TG T cell-reconstituted SCID mice compared to controls, the absolute number of apoptotic cells was not increased. (B) For quantification of apoptotic LPMC, the sections were surveyed carefully for epithelial crypts and villi, which were highlighted by help of a graphics program (Power Point from Microsoft). This way the dramatically increased cellular infiltrate (green) in SCID mice reconstituted with TNF-R2 TG T cells (right) compared to SCID mice reconstituted with WT cells (left) became apparent. This method allowed the counting of absolute numbers of mononuclear cells and apoptotic cells per HPF (40 ? m2). (C) Numbers of LPMC (left), number of apoptotic cells (center) and percentage of apoptotic LPMC (right) in reconstituted SCID mice. The absolute number of LPMC (± SEM) was increased in sections from TNF-R2 TG T cell reconstituted mice (n=6; black bars) compared to control SCID mice reconstituted with WT T cells (n=4; white bars), while the absolute number of apoptotic cells was not significantly changed. The apoptotic rate of LPMC was thus significantly decreased in the TNF-R2 TG T cell reconstituted mice (3.3% versus 7.1%) (p X 0.001).

mTNF via TNF-R2 plays a critical role in chronic inflammation. However, further studies are necessary to prove this hypothesis. In summary, our data on TNF-R2 expression in patients with CD and on the role of TNF-R2 in experimental colitis indicate a critical pathogenic role for TNF-R2-mediated signaling in chronic intestinal inflammation. The findings indicate that overexpression of TNF-R2 could be an

important mechanism for T cell resistance to apoptosis and the perpetuation of intestinal inflammation. Selective targeting of TNF-R2 signaling thus emerges as a potentially novel approach for the treatment of CD and presumably other chronic inflammatory disorders associated with increased TNF production.

Eur. J. Immunol. 2002. 32: 3142–3151

4 Materials and methods 4.1 Isolation of PBMC from patients with CD and controls PB was obtained by venipuncture of patients with active CD from our outpatient clinic or from healthy volunteers as controls. The CD group (n=20) comprised 7 men and 13 women with age ranging from 19 to 60 years. At the time of sample collection, 8 patients of the CD group were treated steroids, 10 with an oral mesalazine preparation and 7 with immunosuppressive drugs (azathioprine). The control group consisted of 10 healthy volunteers between 22 and 35 years of age, none of them receiving any anti-inflammatory or other medication. PBMC were isolated using Ficoll-Hypaque gradients.

4.2 Isolation of LPMC from patients with CD and controls Colonic specimens (biopsy samples and surgical resections) were obtained from 8 patients with active CD and from 15 controls. None of the control patients had evidence of mucosal inflammation. Patients with CD were between 26 and 68 years of age. At the time of sample collection 4 patients were treated with steroids, 4 with an oral mesalazine preparation and 2 with immunosuppressive drugs (azathioprine). The collection of surgical and biopsy samples was approved by the ethical committee and the institutional review board of the University of Mainz. Briefly, samples were washed in calcium- and magnesiumfree Hanks’ balanced salt solution (HBSS-CMF), cut into 0.5-cm pieces, and incubated twice in HBSS containing EDTA (0.37 mg/ml) and dithiothreitol (DTT; 0.145 mg/ml) at 37°C for 15 min followed by subsequent digestion of 5 min with collagenase V 400 (1 mg/ml; Sigma, Deisenhofen, Germany), deoxyribonuclease (0.1 mg/ml DNase; Roche Mannheim, Germany) and hyaloronidase (0.1 mg/ml; Sigma) in a shaking incubator at 37°C. The digested tissue was washed twice in RPMI medium containing 10% FCS and centrifuged. The tissue was finally mechanically homogenized (Dako homogenizer) for 4 min and cells were filtered through a sterile 40- ? m cell strainer (Falcon for Becton Dickinson, Heidelberg, Germany). The cells were then stained for flow cytometry as specified below.

TNF-R2 involvement in IBD

3149

washing with PBS, the FITC-conjugated goat anti-mouse Fab fragment (Dianova, Hamburg, Germany) was added as secondary antibody (final concentration: 15 ? g/ml) and incubated for 20 min. Isotype control antibodies (IgG1 for 80M2 and IgG2a for H398) were included to exclude unspecific staining. Then cells were stained with PE-conjugated antiCD3, anti-CD4, anti-CD8, anti-CD14 (PharMingen, Heidelberg, Germany) and anti-CD33 (Coulter/Immunotech, Krefeld, Germany) antibodies. All steps of the staining procedure were performed at 4°C. Labeled cells were analyzed on a FACScan (Becton Dickinson, Heidelberg, Germany).

4.4 Cell isolation and purification of spleen mononuclear cells BALB/c WT and TgE1335 BALB/c transgenic mice were maintained under specific pathogen-free conditions at the animal facility of the I. Medical Clinic. To exclude the effects of different background genes, transgenic TgE1335 CBA/ C57BL/6 mice overexpressing TNF-R2 under the control of its own promoter [15] were backcrossed seven generations to the BALB/c background before further analysis. Spleen cells were isolated from 2–4-month-old BALB/c WT and BALB/c transgenic mice, as previously described [21]. In brief, spleens were removed from mice killed by cervical decapitation and put through a 40- ? m cell strainer to obtain a single cell suspension. Erythrocytes were eliminated by hypotonic lysis in ammonium chloride and potassium chloride (ACK) buffer. CD4+ T cells were then isolated using FITC-conjugated monoclonal mouse anti-CD4 antibodies (PharMingen, San Diego, CA), anti-FITC immunomagnetic beads and magnetic cell sorting (Miltenyi, Bergisch Gladbach, Germany) resulting in 99% purity. Cells were then purified further by immunomagnetic beads for CD62L. This procedure led to separation of CD4+CD62L+ T cells from CD4+CD62– T cells with a purity of more than 93 %.

4.5 Induction of colitis by transfer of CD4+ CD62L+ T cells Colitis was induced by adoptive transfer of splenic CD4+ CD62L+ cells, as previously described [21]. Briefly, 1×106 CD4+ CD62L+ T cells were injected intraperitoneally into 8–10-week-old syngenic C.B.-17 SCID mice (Charles River Laboratories, Wilmington, MA) bred in our animal facility. Reconstituted SCID mice were maintained in isolated ventilated cages in the S2 animal facility of our Department.

4.3 Double staining for TNF-R and FACS analysis To determine TNF-R expression, 500,000 mononuclear cells were utilized per analysis. Double staining was performed in two steps using monoclonal mouse anti-human antibodies against TNF-R1 (clone H398), 6 ? g/ml and TNF-R2 (clone 80M2) at a final concentration of 3 ? g/ml as first antibodies. Cells were washed 2× in PBS and incubated with the receptor specific antibodies for 30 min at room temperature. After

4.6 Histological analysis of colon sections Reconstituted SCID mice were killed 10–12 weeks after the T cell transfer. Paraffin sections were made and stained with hematoxylin and eosin. The extend of inflammation on microscopic cross-sections of the colon was evaluated as follows: 0, no leukocyte infiltration; 1, low level of leukocyte

3150

M. H. Holtmann et al.

infiltration; 2, moderate to high level of leukocyte infiltration; 3, transmural leukocyte infiltration, high vascular density, intraepithelial abscesses and epithelial lesions. Grading was performed in a blinded fashion by the same pathologist (H.A.L.) as previously described in detail [21].

4.7 Cytokine assays Splenic cells (1×106) from reconstituted SCID mice were cultured in complete RPMI 1640 medium supplemented with penicillin/streptomycin, L-glutamine and 10% heatinactivated fetal calf serum in a humidified atmosphere containing 5% CO2 at 37°C. Cells were stimulated with antimouse CD3 4 plus anti-mouse CD28 antibodies (PharMingen) for 72 h. Culture supernatants were removed and cytokine concentrations were determined by specific enzymelinked immunosorbent assay (ELISA) using commercially available recombinant cytokines and antibodies (PharMingen, San Diego, CA; Genzyme, Cambridge, MA; R&D Systems, Minneapolis, MN). Monoclonal rat anti-mouse cytokine antibodies were diluted at 2 ? g/ml in 2 mM carbonate buffer, pH 9.6. Antibody solution (50 ? l/well) was then aliquoted into Immulon-4 96-well microtiter plates (Dynatech Laboratories Inc., Chantilly, VA). Plates were incubated at 4°C overnight and washed. Supernatant or standards of recombinant mouse cytokines (Genzyme, Boston, MA) (100 ? l) were added and incubated for 2 h. After repeated washing, 100 ? l biotin-labeled rat anti-mouse cytokine antibody (2 ? l/ml in PBS + 3% BSA) was added to each well and incubated for 45 min at room temperature. Plates were then washed six times followed by addition of 100 ? l of 1:1,000diluted solution of horseradish peroxidase-labeled streptavidin (Zymed, San Francisco, CA) in PBS with 3% BSA. After incubation for 30 min at room temperature and repeated washing, 100 ? l o-phenylemediamine dihydrochloride (15 mg tablet/15 ml phosphate-citrate buffer with 0.01% H2O2 from Sigma) was added. Absorbance was measured on a Dynatech MR 5000 ELISA reader at a wavelength of 490 nm.

4.8 TUNEL staining For detection of apoptotic cells terminal deoxynucleotidyl transferase-mediated dUTP-X fluorescein nick end labeling (TUNEL) was performed utilizing the In Situ Cell Death Detection Kit (Roche) following the manufacturer’s protocol. In brief, paraffin sections were deparaffinated with xylol and graded concentrations of ethanol, treated with 3% H2O2 for 30 min and pepsin HCl for 20 min. Next, the TUNEL reaction mixture was applied on each slide and incubated in a humid chamber for 60 min followed by three washes in PBS. Slides were mounted with mounting medium for fluorescence (Vector Laboratories, Burlingame, CA) and analyzed with an Olympus Fluorescence Microscope. Sections were analyzed within 1 h to avoid loss of fluorescence.

Eur. J. Immunol. 2002. 32: 3142–3151 4.9 Quantitative assessment of apoptotic cells in the inflamed colon For quantification of apoptotic LPMC, epithelial crypts and villi in colon sections were highlighted using a device in a graphics program (Powerpoint) using a Macintosh computer (G4, San Jose, CA). To determine the apoptotic rate of LPMC, cells in the LP were counted in six representative high-power fields (HPF) displaying 40 ? m2 of TUNEL-stained sections from mice reconstituted with CD4+CD62L+ TNF-R2 transgenic T cells and four representative HPF from control SCID mice reconstituted with CD4+ CD62L+ T cells from WT mice. Apoptosis rates of LPMC were expressed in percent of all counted cells.

4.10 Statistical analysis Statistical differences were assessed using the Student’s ttest. p values of X 0.05 and 0.0 are indicated in the figures as one or two asterisks, respectively.

Acknowledgements: This work was supported by Deutsche Morbus Crohn/Colitis ulcerosa Vereinigung (DCCV) for M.H.H.

References 1 Tracey, K. J. and Cerami, A., Tumor necrosis factor: a pleiotropic cytokine and therapeutic target. Annu. Rev. Med. 1994. 45: 491–503. 2 Brockhaus, M., Schoenfeld, H. J., Schlaeger, E. J., Hunziker, W., Lesslauer, W. and Loetscher, H., Identification of two types of tumor necrosis factor receptors on human cell lines by monoclonal antibodies. Proc. Natl. Acad. Sci. USA 1990. 87: 3127–3131. 3 Hsu, H., Xiong, J. and Goeddel, D. V., The TNF receptor 1associated protein TRADD signals cell death and NF- kappa B activation. Cell 1995. 81: 495–504. 4 Scheurich, P., Grell, M., Meager, A. and Pfizenmaier, K., Agonistic and antagonistic antibodies as a tool to study the functional role of human tumor necrosis factor receptors. In Fiers, W. and Buurman, W. A. (Eds.) Tumor necrosis factor: molecular and cellular biology and clinical relevance. Karger, Basel 1993, pp 52–57. 5 Tartaglia, L. A., Pennica, D. and Goeddel, D. V., Ligand passing: the 75-kDa tumor necrosis factor (TNF) receptor recruits TNF for signaling by the 55-kDa TNF receptor. J. Biol. Chem. 1993. 268: 18542–18548. 6 Schutze, S., Berkovic, D., Tomsing, O., Unger, C. and Kronke, M., Tumor necrosis factor induces rapid production of 1,2,diacylglycerol by a phosphatidylcholine-specific phospholipase C. J. Exp. Med. 1991. 174: 975–988. 7 Rothe, M., Sarma, V., Dixit, V. M. and Goeddel, D. V., TRAF2mediated activation of NF-kappa B by TNF receptor 2 and CD40. Science 1995. 269: 1424–1427. 8 Rothe, J., Lesslauer, W., Lotscher, H., Lang, Y., Koebel, P., Kontgen, F., Althage, A., Zinkernagel, R., Steinmetz, M. and Bluethmann, H., Mice lacking the tumour necrosis factor recep-

Eur. J. Immunol. 2002. 32: 3142–3151 tor 1 are resistant to TNF-mediated toxicity but highly susceptible to infection by Listeria monocytogenes. Nature 1993. 364: 798–802. 9 Takeuchi, M., Rothe, M. and Goeddel, D. V., Anatomy of TRAF2. Distinct domains for nuclear factor-kappaB activation and association with tumor necrosis factor signaling proteins. J. Biol. Chem. 1996. 271: 19935–19942. 10 Grell, M., Becke, F. M., Wajant, H., Mannel, D. N. and Scheurich, P., TNF receptor type 2 mediates thymocyte proliferation independently of TNF receptor type 1. Eur. J. Immunol. 1998. 28: 257–263. 11 Grell, M., Douni, E., Wajant, H., Lohden, M., Clauss, M., Maxeiner, B., Georgopoulos, S., Lesslauer, W., Kollias, G. and Pfizenmaier, K., The transmembrane form of tumor necrosis factor is the prime activating ligand of the 80 kDa tumor necrosis factor receptor. Cell 1995. 83: 793–802. 12 Haas, E., Grell, M., Wajant, H. and Scheurich, P., Continuous autotropic signaling by membrane-expressed tumor necrosis factor. J. Biol. Chem. 1999. 274: 18107–18112. 13 Erickson, S. L., de Sauvage, F. J., Kikly, K., Carver-Moore, K., Pitts-Meek, S., Gillett, N., Sheehan, K. C., Schreiber, R. D., Goeddel, D. V. and Moore, M. W., Decreased sensitivity to tumour-necrosis factor but normal T cell development in TNF receptor-2-deficient mice. Nature 1994. 372: 560–563. 14 Loetscher, H., Gentz, R., Zulauf, M., Lustig, A., Tabuchi, H., Schlaeger, E. J., Brockhaus, M., Gallati, H., Manneberg, M. and Lesslauer, W., Recombinant 55-kDa tumor necrosis factor (TNF) receptor. Stoichiometry of binding to TNF alpha and TNF beta and inhibition of TNF activity. J. Biol. Chem. 1991. 266: 18324–18329. 15 Douni, E. and Kollias, G., A critical role of the p75 tumor necrosis factor receptor (p75TNF-R) in organ inflammation independent of TNF, lymphotoxin alpha, or the p55TNF-R. J. Exp. Med. 1998. 188: 1343–1352. 16 Breese, E. and MacDonald, T. T., TNF-alpha secreting cells in normal and diseased human intestine. Adv. Exp. Med. Biol. 1996. 371B: 821–824. 17 Plevy, S. E., Landers, C. J., Prehn, J., Carramanzana, N. M., Deem, R. L., Shealy, D. and Targan, S. R., A role for TNF-alpha and mucosal T helper-1 cytokines in the pathogenesis of Crohn’s disease. J. Immunol. 1997. 159: 6276–6282. 18 Powrie, F., Leach, M. W., Mauze, S., Menon, S., Caddle, L. B. and Coffman, R. L., Inhibition of Th1 responses prevents inflammatory bowel disease in scid mice reconstituted with CD45RBhi CD4+ T cells. Immunity 1994. 2: 553–562. 19 Neurath, M. F., Fuss, I., Pasparakis, M., Alexopoulou, L., Haralambous, S., Meyer zum Büschenfelde, K.-H., Strober, W. and Kollias, G., Predominant role of tumor necrosis factoralpha in experimental colitis in mice. Eur. J. Immunol. 1997. 27: 1743–1750. 20 Targan, S. R., Hanauer, S. B., Deventer, S. J. V., Mayer, L., Present, D. H., Braakman, T., Woody, K. L. D., Schaible, T. F. and Rutgeerts, P. J., A short-term study of chimeric monoclonal antibody cA2 to tumor necrosis factor alpha for Crohn’s disease. N. Engl. J. Med. 1997. 337: 1029–1035. 21 Atreya, R., Mudter, J., Finotto, S., Müllberg, J., Jostock, T., Wirtz, S., Schütz, M., Bartsch, B., Holtmann, M., Becker, C., Strand, D., Czaja, J., Schlaak, J. F., Lehr, H. A., Autschbach, F., Schürmann, G., Nishimoto, N., Yoshizaki, K., Ito, H., Kishimoto, T., Galle, P. R., Rose-John, S. and Neurath, M. F., Blockade of IL-6 trans-signaling suppresses T cell resistance against apoptosis in chronic intestinal inflammation: evidence in Crohn’s disease and experimental colitis in vivo. Nat. Med. 2000. 6: 583–588.

TNF-R2 involvement in IBD

3151

22 Boirivant, M., Marini, M., Di-Felice, G., Pronio, A. M., Montesani, C., Tersigni, R. and Strober, W., Lamina propria T cells in Crohn’s disease and other gastrointestinal inflammation show defective CD2 pathway-induced apoptosis. Gastroenterology 1999. 116: 557–565. 23 Ina, K., Itoh, J., Fukushima, K., Kusugami, K., Yamaguchi, T., Kyokane, K., Imada, A., Binion, D. G., Musso, A., West, G. A., Dobrea, G. M., McCormick, T. S., Lapetina, E. G., Levine, A. D., Ottaway, C. A. and Fiocchi, C., Resistance of Crohn’s disease T cells to multiple apoptotic signals is associated with a Bcl2/Bax mucosal imbalance. J. Immunol. 1999. 163: 1081–1090. 24 Neurath, M. F., Finotto, S., Fuss, I., Boirivant, M., Galle, P. R. and Strober, W., Regulation of T cell apoptosis in inflammatory bowel disease: to die or not to die, that is the mucosal question. Trends Immunol. 2001. 22: 21–26. 25 Duchmann, R. and Zeitz, M., Crohn’s disease. In Ogra, P. and Strober, W. (Eds.) Handbook of mucosal immunology. Academic Press, New York 1998 26 Ogura, Y., Bonen, D. K., Inohara, N., Nicolae, D. L., Chen, F. F., Ramos, R., Britton, H., Moran, T., Karaliuskas, R., Duerr, R. H., Achkar, J. P., Brant, S. R., Bayless, T. M., Kirschner, B. S., Hanauer, S. B., Nunez, G. and Cho, J. H., A frameshift mutation in NOD2 associated with susceptibility to Crohn’s disease. Nature 2001. 411: 603–606. 27 Rogler, G., Brand, K., Vogl, D., Page, S., Hofmeister, R., Andus, T., Knuechel, R., Baeuerle, P. A., Scholmerich, J. and Gross, V., Nuclear factor kappaB is activated in macrophages and epithelial cells of inflamed intestinal mucosa. Gastroenterology 1998. 115: 357–369. 28 Jobin, C. and Sartor, R. B., NF-kappaB proteins as therapeutic targets for inflammatory bowel diseases. Inflamm. Bowel. Dis. 2000. 6: 206–213. 29 Neurath, M. F., Becker, C. and Barbulescu, K., 1998. Role of NF-kappaB in immune and inflammatory responses in the gut. Gut 43: 856–860. 30 Wang, C. Y., Mayo, M. W., Korneluk, R. G., Goeddel, D. V. and Baldwin, A. S. Jr., NF-kappaB anti-apoptosis: induction of TRAF1 and TRAF2 and c-IAP1 and c-IAP2 to suppress caspase8 activation. Science 1998. 281: 1680–1683. 31 Lugering, A., Schmidt, M., Lugering, N., Pauels, H. G., Domschke, W. and Kucharzik, T., Infliximab Induces apoptosis in monocytes from patients with chronic active Crohn’s disease by using a caspase-dependent pathway. Gastroenterology 2001. 121, in press. 32 Corazza, N., Eichenberger, S., Eugster, H. P. and Mueller, C., Nonlymphocyte-derived tumor necrosis factor is required for induction of colitis in recombination activating gene (RAG)2(–/–) mice upon transfer of CD4(+)CD45RB(hi) T cells. J. Exp. Med. 1999. 190: 1479–1992. 33 Sandborn, W. J., Hanauer, S. B., Katz, S., Safdi, M., Wolf, D. G., Baerg, R. D., Tremaine, W. J., Johnson, T., Diehl, N. N. and Zinsmeister, A. R., Etanercept for active Crohn’s disease: a randomized, double-blind, placebo-controlled trial. Gastroenterology 2001. 121: 1088–1094. 34 Van Deventer, S. J. H., Transmembrane TNF-alpha, induction of apoptosis, and the efficacy of TNF-targeting therapies in Crohn’s disease. Gastroenterology 2001. 11: 1242–1246.

Correspondence: Markus F. Neurath, Laboratory of Immunology, I. Department of Medicine, University of Mainz, Langenbeckstrasse 1, D-55101 Mainz, Germany Fax: +49-6131-175508 e-mail: neurath — 1-med.klinik.uni-mainz.de