Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Infiltration of the mucosa by leukocytes is a key feature in the pathogenesis of human inflammatory bowel disease. At inflammatory sites, proinflammatory cytokines such as tumor necrosis factor- (TNF-) stimulate endothelial cells to express adhesion molecules on their cell surface (Marui et al., 1993; Spiecker et al., 1997; Nakada et al., 1998). These molecules interact with receptors on leukocytes, which subsequently infiltrate from the blood vessels into the mucosa. TNF- also activates invading T cells and NK cells to produce interferon- (IFN-). In vitro, IFN- increases the sensitivity of colonic epithelial cells to several apoptotic stimuli via up-regulation of caspase-1 (O'Connell et al., 2000). Moreover, IFN- exerts indirect cytotoxicity by increasing the release of reactive oxygen species by macrophages.

Successful treatment of patients with steroid refractory or fistulizing Crohn's disease with anti-TNF- antibody (van Dullemen et al., 1998; Rutgeerts et al., 1999; Sandborn and Hanauer, 1999) demonstrates the anti-inflammatory potency of this specific cytokine blockade (Sandborn and Hanauer, 1999). Repeated administration of anti-TNF- antibodies, however, may lead to production of anti-chimeric (Sandborn and Hanauer, 1999) or anti-double-strand DNA antibodies (Elliott et al., 1994) or a possible flare up of tuberculosis due to the long-term alteration of the immune defense (Keane et al., 2001). Among the agents known to inhibit the synthesis of TNF-, attention has focused on cAMP-elevating phosphodiesterase (PDE) inhibitors (Eigler et al., 1997). The predominant PDE isoenzyme in macrophages, the main cellular source of TNF-, is the type-4 PDE (Gantner et al., 1997). Compared with the nonspecific PDE inhibitor pentoxifylline, the specific type-4 PDE inhibitor rolipram was 500-fold more potent in suppressing TNF- synthesis in human mononuclear cells (Semmler et al., 1993). In vitro, rolipram not only suppresses the synthesis of proinflammatory cytokines such as TNF- and IFN- but also induces anti-inflammatory cytokines such as interleukin-10 (Eigler et al., 1998). Nevertheless, several adverse effects of rolipram may limit its use in a clinical setting. The most frequent adverse effect reported in clinical trials was nausea (Zeller et al., 1984; Hebenstreit et al., 1989). Moreover, rolipram is a racemic compound with differing activities and kinetics of the two enantiomers. Thus, it may not meet the state-of-art demands on a new compound. Mesopram is a novel potent and selective enantiomeric PDE4 inhibitor. It has been shown to be effective in the treatment of experimental autoimmune encephalitis in rodents (Dinter et al., 2000).

In our model, colitis was induced by oral administration of dextran sulfate sodium (DSS). DSS-induced colitis is characterized histopathologically by mucosal infiltration of inflammatory cells, focal crypt damage, epithelial injury, and ulceration (Okayasu et al., 1990; Cooper et al., 1993; Dieleman et al., 1997). Like in Crohn's disease, pathology of DSS-induced colitis exhibits a diffuse alteration of the colon with transmural inflammation and aphthous erosions, although no fissures or granuloma are seen. The pathomechanism of DSS-induced colitis includes both toxic effects on the epithelium and production of proinflammatory cytokines by macrophages, which are stimulated after phagocytosis of DSS. In the present study, we induced colitis in BALB/c mice and investigated the efficacy of the specific type-4 PDE inhibitor mesopram both in the prevention and therapy of colitis. Endpoints were the clinical score, the colon length, the histologic score of the colon, and the local IFN- expression.

Additionally a dose finding study of mesopram p.o. was performed using the preventive model with clinical score and colon length as end points. As a reference compound in this trial, olsalazin was used since it was previously shown to be effective in DSS-induced colitis (Zijlstra et al., 1992; Axelsson et al., 1998).

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Mice. Female, 6- to 8-week old BALB/c mice weighing 20 to 22 g and weighing 16 to 19 g in the dose finding study were used as indicated. The animals were housed in temperature-controlled rooms with a 12-h light/dark cycle. They were fed standard mouse chow pellets, had access to bottled tap water ad libitum, and were acclimatized to the conditions at least 7 days before they were used in experiments. Mice were killed by cervical dislocation under isoflurane anesthesia (Forene; Abbott GmbH, Wiesbaden, Germany). All experiments were approved by the regional animal study commitee and are in agreement with the guidelines for the proper use of animals in biomedical research. Both animal handling and clinical and histologic scoring of colitis were performed in a blinded experimental design.

Reagents. Brefeldin A, phorbol myristyl acetate, and ionomycin were purchased from Sigma-Aldrich (Munich, Germany), RPMI 1640 medium from Biochrom (Berlin, Germany), and fetal calf serum from Invitrogen (Karlsruhe, Germany). Mesopram was kindly supplied by Schering AG (Berlin, Germany). It was stored and dissolved as indicated in the data sheet. Olsalazine (Dipentum) was purchased from Pharmacia and Upjohn (Erlangen, Germany).

Induction and Treatment of Colitis. Mice were fed 3.5% DSS (lot no. 4470 C; molecular mass 30-40 kDa; ICN, Eschwege, Germany) dissolved in sterile, distilled water ad libitum at days 1 to 11 in the preventive model or days 1 to 7 in the therapeutic model. Mesopram (10 mg/kg b.wt. q.d.) was administered p.o. or i.p. at days 1 to 10 in the preventive model or at days 8 to 14 in the therapeutic model. In the dose finding study, mesopram was administered orally in doses of 2, 10, or 50 mg/kg b.wt. q.d.. When applied orally, mesopram was suspended in a solution of 0.5% hydroxyethylcellulose (HEC; lot no. S29089015; Schuchard, Germany) as vehicle. When administered intraperitoneally, as done in the therapeutic model, mesopram was dissolved in a solution of 10% Cremophor EL (BASF, Ludwigshafen, Germany) as vehicle and filtered through syringe filters (0.2 µm; Gelman Sciences, Ann Arbor, MI). Placebo-treated animals received oral vehicle without mesopram. Control mice were given bottled tap water without DSS ad libitum and received mesopram (10 and 50 mg/kg b.wt. p.o. q.d.) or vehicle p.o. q.d.. The reference compound olsalazine, used in the dose finding study, was administered orally in a dose of 40 mg/kg b.wt. q.d. dissolved in 0.5% HEC. The volume of application was 200 µl both orally and intraperitoneally.

Evaluation of Clinical Score, Colon Length, and Histologic Score. Body weight, fecal blood, and stool consistency were determined daily (Cooper et al., 1993; Hartmann et al., 2000; Siegmund et al., 2001a). Two investigators blinded to the treatment groups assessed the clinical score as indicated in Table 1. The average of these three scores (body weight, stool consistency, and fecal blood) gave an overall clinical score ranging from 0 (healthy) to 4 (maximal activity of colitis).

Colon Cytokine Synthesis. To study colon homogenate at the end of the experimental course, colons were removed, and strips of the colon (about 1 cm) were mechanically crushed, vigorously vortexed for 1 min in 200 µl of tissue protein extraction reagent (Pierce Chemical, Rockford, IL) and shock frozen in liquid nitrogen, as previously described (Hartmann et al., 2000; Siegmund et al., 2001a). The homogenate was centrifuged at 10,000g at 4°C for 15 min. The amount of total extracted protein was determined by Bradford analysis using BioRad Protein Assay (BioRad Laboratories, Munich, Germany) as the dye reagent.

Statistical Analysis. Data are expressed as means ± S.E.M. Statistical significance was determined by factorial analysis of variance. Differences were considered statistically significant for p < 0.05. Statistical analyses were performed using StatView 4.51 software (Abacus Concepts, Calabasas, CA).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Preventive Model and Dose Finding Study

Clinical Score. Mice fed with 3.5% DSS developed signs of colitis at day 5, defined by a clinical score >0.5. Oral treatment with 10 mg/kg q.d. mesopram delayed the onset of colitis for 1 day and reduced the progression of the disease until the end of the study (Fig. 1A). At day 10, placebo-treated animals suffered from a severe colitis, with a clinical score of 3.2 ± 0.2 (n = 8), whereas colitis was significantly blunted in the 10 mg/kg q.d. mesopram-treated group (clinical score 1.8 ± 0.3; n = 7; p = 0.004). Each individual parameter of the clinical score was significantly improved in the mesopram-treated group. The score of weight loss was 3.3 ± 0.3 in the placebo group versus 2.3 ± 0.3 in the mesopram group (p = 0.041), the stool consistency score was 2.8 ± 0.4 versus 1.0 ± 0.5 (p = 0.017), and the fecal blood score was 3.5 ± 0.3 versus 2.0 ± 0.4 (p = 0.010). None of the control animals, which were not exposed to DSS and received either mesopram (10 mg/kg b.wt. q.d.; n = 4) or vehicle HEC (n = 4), showed signs of colitis throughout the experimental course.

View larger version (17K): [in this window] [in a new window]   Fig. 1.   A, effect of mesopram (10 mg/kg p.o. q.d.) in the preventive setting of experimental colitis. BALB/c mice were exposed to 3.5% DSS for 10 days. The clinical score (consisting of weight loss, stool consistency, and rectal bleeding; range from 0 = healthy to 4 = maximally ill) was assessed daily. Scores are depicted as means ± S.E.M. B, clinical score of the dose finding study of mesopram (2, 10, and 50 mg/kg p.o. q.d.) in the preventive setting of experimental colitis. BALB/c mice were exposed to 3.5% DSS for 10 days. The clinical score (consisting of weight loss, stool consistency, and rectal bleeding; range from 0 = healthy to 4 = maximally ill) was assessed every day. Scores are depicted as means.

Colon Length. DSS led to a reduction of mean colon length in the placebo-treated group (8.6 ± 0.4 cm, n = 8) compared with the nonexposed control groups (Fig. 2A) (14.3 ± 0.2 cm, n = 4, in the HEC-treated control group; 14.5 ± 0.3 cm; n = 4, in the 10 mg/kg q.d. mesopram-treated control group). Oral treatment with 10 mg/kg mesopram significantly (p < 0.001) reduced the extent of DSS-induced colon shortening to 11.5 ± 0.2 cm (n = 7).

View larger version (16K): [in this window] [in a new window]   Fig. 2.   A, reduction of DSS-induced colon length shortening by mesopram in the prevention model of DSS-induced colitis. BALB/c mice were exposed to 3.5% DSS in their drinking water for 10 days. Oral application of 10 mg/kg mesopram q.d. partially prevented colon shortening, a marker of inflammation. Values are depicted as means ± S.E.M length of the colon in each group. B, reduction of DSS-induced colon length shortening by mesopram in the dose finding study. BALB/c mice were exposed to 3.5% DSS in their drinking water for 10 days. Oral application of 2, 10, or 50 mg/kg mesopram q.d. partially prevented colon shortening, a marker of inflammation. Olsalazin, used as reference compound, did not affect colon length shortening. Values are depicted as means ± S.E.M. length of the colon in each group.

Histologic Score. The mucosal damage caused by DSS was quantified by a pathologist in a blinded manner. Colons of DSS-exposed mice revealed mucosal erosions and crypt destruction as well as an inflammatory infiltrate composed of macrophages, lymphocytes, eosinophils, and a few neutrophils. In DSS-exposed groups, oral treatment with mesopram reduced the mean histologic score (0 for no changes, 6 for maximal tissue damage and cell infiltration) from 4.2 ± 0.3 in the HEC-treated group (n = 7) to 2.9 ± 0.3 in the mesopram-treated group (n = 7). This effect was statistically significant (p = 0.003) (Fig. 3). Neither control group showed any histologic sign of colonic inflammation evidenced by a score of 0.3 ± 0.1 in the HEC-treated group (n = 4) and 0.4 ± 0.2 in the mesopram-treated group (n = 4).

View larger version (10K): [in this window] [in a new window]   Fig. 3.   Reduction of histologic score by treatment with mesopram in the prevention model. At day 11, the histologic score (0 for no inflammation, 6 for maximal tissue damage and cell infiltration) was determined in a blinded manner. Exposure to DSS led to typical histologic signs of colonic inflammation. Treatment with mesopram (10 mg/kg q.d.) significantly reduced the histologic score of DSS-exposed mice compared with placebo. Scores are depicted as means + S.E.M.

Colonic Content of Proinflammatory Cytokines. For measurement of the colonic content of the proinflammatory cytokines TNF- and IFN-, colons were removed and homogenized at day 11. TNF- and IFN- were quantified in the homogenate by ELISA. Non-DSS-exposed groups showed the lowest level of colonic TNF- (17 ± 3 pg/mg of protein, n = 4 in the HEC-treated control group; and 17 ± 1 pg/mg of protein, n = 4 in the mesopram-treated control group). DSS treatment led to an increased content of TNF- in the colon (41 ± 10 pg/mg of protein, n = 7) (Fig. 4A). Although not significantly different, an oral treatment condition with 10 mg/kg mesopram markedly reduced TNF- (25 ± 6 pg/mg of protein, n = 7) compared with the placebo-treated group.

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Content of TNF- (A) and IFN- (B) in colonic tissue quantified by ELISA. BALB/c mice were fed 3.5% DSS for 10 days. At day 11, colon strips of about 1 cm were homogenized and investigated. DSS-exposed mice showed a higher colonic content of both TNF- and IFN- than control animals. Administration of 10 mg/kg mesopram p.o. q.d. reduced the content of both proinflammatory cytokines in colon tissue of DSS-exposed mice. Values are depicted as means ± S.E.M.

Treatment of Established Colitis in BALB/c Mice

Clinical Score. The efficacy of mesopram was also investigated in the treatment of pre-existing colitis. Colitis was induced in BALB/c mice by feeding them 3.5% DSS dissolved in the drinking water ad libitum over a period of 7 days. After replacing DSS by normal drinking water at day 8, DSS-exposed mice were stratified into three groups according to their clinical score, each group containing eight equally ill mice. Two control groups containing four mice did never receive DSS. Starting at day 8, these groups were treated with 10 mg/kg q.d. mesopram i.p. (dissolved in the vector 10% Cremophor EL and ultra-filtered) or p.o. (dissolved in 0.5% HEC solution) or placebo (0.5% HEC solution p.o.). The two control groups received from day 8 mesopram (10 mg/kg b.wt. orally q.d.) or the vehicle (0.5% HEC). At day 7, the clinical score of all DSS-exposed groups was significantly (p < 0.05) higher than in control mice, which showed no signs of colitis (Fig. 5). Although DSS was discontinued at day 8, all DSS-exposed groups initially showed a progression of the clinical signs of colitis, which reached its maximum at day 9 in all groups. The extent of colitis was constant in the placebo-treated group at days 9 through 12. At day 15, the clinical score of this group (n = 8) had improved from a maximum of 2.1 ± 0.1 to 1.2 ± 0.2 reflecting the spontaneous course of the disease. Mice that received mesopram orally (n = 8) had the most severe colitis at day 9 compared with the other groups (3.0 ± 0.3 versus 2.1 ± 0.1 in the placebo group and 2.0 ± 0.2 in the intraperitoneally treated group). At day 15, the intraperitoneally mesopram-treated group (n = 8) had recovered from colitis with a clinical score of 0.4 ± 0.2. The clinical score was significantly lower compared with the placebo-treated group (1.2 ± 0.2, p = 0.020).

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Effect of mesopram in the treatment of pre-existing colitis. Mice were exposed to 3.5% DSS for 7 days. At day 8, DSS was discontinued, and colitis was treated with mesopram (10 mg/kg i.p. or p.o. q.d.) or placebo. The extent of colitis was quantified by the clinical score from day 9 to 15. i.p. mesopram treatment was effective in healing the DSS-induced colitis at the end of the treatment protocol (clinical score <0.5). Scores depicted as means ± S.E.M.

Colon Length and Histology. Treatment with DSS leads to a shortening of the colon as a post mortem marker of the extent of colitis (Fig. 6). DSS-exposed, placebo-treated mice (n = 8) showed significantly reduced colon lengths compared with mice that were not exposed to DSS (10.3 ± 0.3 versus 14.8 ± 0.7 cm, n = 4 in mesopram-treated control mice; and 14.5 ± 0.5 cm, n = 4 in placebo-treated control mice, p < 0.001). Mesopram treatment of established colitis over a period of 7 days reversed the DSS-induced colon shortening compared with placebo-treated, DSS-exposed mice (12.0 ± 0.2 cm, p < 0.001, n = 8, in the i.p. group; and 12.2 ± 0.2 cm, p < 0.001, n = 8 in the p.o. group). Histologic analysis of the colons confirmed the absence of inflammation in both non-DSS-exposed control groups (Fig. 7). In contrast, all DSS-exposed groups showed a histologic score that reflected colitis, which was induced in these animals during the 1st week of the experimental course. Intraperitoneally mesopram-treated mice showed a trend toward an improved histologic score (2.8 ± 0.2, n = 8), whereas orally mesopram-treated mice (3.3 ± 0.4, n = 8) did not differ from placebo-treated mice (3.6 ± 0.3, n = 8) in their histologic score.

View larger version (12K): [in this window] [in a new window]   Fig. 6.   Mesopram partially reverses DSS-induced colon shortening in established colitis. Mice were fed 3.5% DSS for 7 days. At day 8, DSS was discontinued, and mice received either mesopram (10 mg/kg i.p. or p.o.) or placebo. At day 15, colon length was measured. DSS induced a significant reduction of colon length compared with control animals. Both, intraperitoneal and oral treatment with mesopram reversed significantly the DSS-induced shortening of the colon. Bars represent means ± S.E.M.

View larger version (11K): [in this window] [in a new window]   Fig. 7.   Histologic analysis of colons with established DSS-induced colitis. Mice were fed 3.5% DSS for 7 days. At day 8, DSS was discontinued and mice received either mesopram (10 mg/kg i.p. or p.o.) or placebo. At day 15, histologic score (0 for no inflammation, 6 for maximal tissue damage, and cell infiltration) was determined in a blinded manner. Exposure to DSS led to an increase of the histologic score compared with control mice. There was a trend toward an improvement of the histologic score in the i.p. mesopram-treated group. Values are depicted as means ± S.E.M.

IFN- Synthesis by Colonic Tissue. To evaluate the ex vivo production of IFN- by colonic tissue, colons were removed at day 15. Strips of the colon of about 1 cm in length were prepared and incubated overnight, as indicated in the method section. After an incubation period of 18 h, the concentration of IFN- in the combined lysate plus conditioned medium was quantified by ELISA and adapted to the weight of the tissue (Fig. 8). The highest amounts of IFN- were detected in the combined lysate plus conditioned medium of colons obtained from DSS-exposed, placebo-treated mice (121 ng of IFN-/100 mg of colon, n = 8). Both, intraperitoneal (44 ± 4 ng of IFN-/100 mg of colon, n = 8) and oral (39 ± 7 ng of IFN-/100 mg of colon, n = 8) treatment with 10 mg/kg mesopram significantly reduced the ex vivo production of IFN- (p = 0.003).

View larger version (12K): [in this window] [in a new window]   Fig. 8.   Mesopram reduced the ex vivo production of IFN- in colonic tissue in the treatment model of DDS-induced colitis. At day 15, strips of colons were incubated without stimulus overnight. The content of IFN- in these conditioned medium plus cell lysate was determined by ELISA. Both p.o. and i.p. treatment of mice with mesopram in vivo significantly reduced the ex vivo production of IFN- by cultured colonic tissue compared with placebo. Values are depicted as means ± S.E.M.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

In the present study, we demonstrate the in vivo efficacy of the specific type-4 phosphodiesterase inhibitor mesopram in the treatment of established DSS-induced murine colitis and its efficacy in the prevention of DSS-induced colitis with two different routes of administration. In an additional dose finding study of orally administered mesopram (2, 10, and 50 mg/kg b.wt. q.d.), mesopram dose dependently ameliorated DSS-induced colitis with significant improvement in the clinical score and colon length.

Oral treatment with mesopram effectively prevented DSS-induced colitis in BALB/c mice. In concordance with its clinical efficacy, oral mesopram reduced colon shortening as a marker of colitis. Furthermore, oral administration of mesopram partially prevented histologic signs of colitis, and colonic specimens obtained from mesopram-treated animals showed a trend toward a reduced content of the proinflammatory cytokines TNF- and IFN-. In the therapeutic model, both intraperitoneal and oral mesopram led to an increased recovery from established colitis. Animals treated intraperitoneally with mesopram even recovered completely from colitis within the 7-day treatment period. The in vivo efficacy was paralleled by a partially reversed colon shortening and by a reduction of colonic IFN- synthesis to the level of non-DSS-exposed control mice length in both mesopram groups. In contrast, colons obtained from placebo-treated mice produced a significantly higher amount of IFN-.

The DSS model of murine colitis has shown to be useful for preclinical testing of new compounds for therapy of human inflammatory bowel disease (Cooper et al., 1993; Elson et al., 1995). Therapeutic agents that are in clinical use or are evaluated in clinical trials, such as olsalazine (Zijlstra et al., 1992; Axelsson et al., 1998), anti-TNF- antibodies (Murthy et al., 1999), or interleukin-10 (Tomoyose et al., 1998) have been successfully tested in this model. The advantages of this model include its simplicity and the high degree of uniformity and reproducibility of the colonic lesions (Elson et al., 1995). The principle endpoint of this study was the clinical disease activity, which was scored with a system that has been described to correlate with the pathologic changes (Cooper et al., 1993; Hartmann et al., 2000; Siegmund et al., 2001a). As a reference compound olsalazine (40 mg/kg b.wt.) was applied. Surprisingly, and in contrast to the study by Axelsson et al. (1998) orally administered olsalazine in our study did not show a significant improvement in either clinical score or colon length. This underlines the potency of phosphodiesterase type-4 inhibitors in this experimental model.

Post mortem, two further endpoints were evaluated: 1) the shortening of the colon as a morphometric surrogate parameter for the degree of inflammation, which correlates with the pathologic changes and has proved to be a consistent marker of colitis (Okayasu et al., 1990; Hartmann et al., 2000; Siegmund et al., 2001a) and 2) histologic assessment of the tissue damage and extent of infiltration by inflammatory cells.

TNF- and IFN- are both key cytokines in human Crohn's disease. The local production of these proinflammatory cytokines in murine colonic tissue was quantified ex vivo by ELISA. In the preventive model, all DSS-exposed groups produced sufficient amounts of proinflammatory cytokines to achieve levels in the colonic homogenate high above the detection limit of the assay. In contrast, animals in the therapeutic model whose colitis was substantially ameliorated in each group at the end of the study did not differ in the native colonic TNF- and IFN- content compared with non-DSS-exposed control mice (data not shown). Therefore, in this model, colonic strips had to be cultured over a period of 20 h to obtain enough IFN- in the combined conditioned medium and cell lysates to meet the detection limit of the ELISA. Suppression of TNF- and IFN- synthesis may be one of several mechanisms involved in the reduction of colonic inflammation.

Unwanted effects of PDE4 inhibition are described in clinical phase I trials. These studies revealed central nervous problems to be dose-limiting side effects (Zeller et al., 1984). Also the type-4 phosphodiesterase inhibitor mesopram, currently in clinical development as an anti-inflammatory compound for the treatment of multiple sclerosis, has been tested in normal healthy volunteers. It has been shown that it was well tolerated at lower doses, but the highest doses induced some nausea and vomiting in a dose-dependent fashion. With the highest doses, nausea and vomiting were accompanied by cardiac arrythmias, which were felt to be due to vomiting. Emesis, a side effect described by Robichaud et al. (2001) after the application of PDE4 inhibitors such as rolipram in ferrets, was not observed in our animals. Previous experiments showed that mesopram-treated mice do not alter their drinking behavior, however.

In our study, the following adverse effects of mesopram were observed: after administration of the drug, mice became dizzy or fell asleep for a few minutes. These symptoms never exceeded 30 min. After 5 days of administration of mesopram, we observed a change in the behavior of the mice in each of the study protocols performed. Despite the progression of colitis in the preventive studies, mesopram-treated mice became more vivid and were sometimes aggressive.

This psychomotor activation may be a side effect also seen in antidepressants, which facilitate the noradrenergic transmission in the central nervous system. Inhibition of type-4 phosphodiesterase is considered to have antidepressant effects in the treatment of patients with mood disorders (Norman et al., 1992). Several type-4 phosphodiesterase inhibitors, e.g., rolipram, were initially developed as antidepressant compounds (Zeller et al., 1984; Hebenstreit et al., 1989), and thus, agitation may be a class-specific side effect. Yet, in clinical trials conducted with other specific type-4 phosphodiesterase inhibitors, no central nervous side effects were reported (Doherty, 2000; Hay, 2000).

To our knowledge, this is the first study showing a beneficial dose-dependent effect of oral treatment with a specific type-4 phosphodiesterase inhibitor in DSS-induced murine colitis. This is a key finding since oral administration is of clinical relevance. Elevation of the intracellular cAMP levels with consecutive down-regulation of proinflammatory cytokines such as TNF- is supposed to be a main mechanism of action of type-4 phosphodiesterase inhibitors concerning their anti-inflammatory properties. TNF- is considered to be a critical cytokine that orchestrates the inflammatory response in inflammatory bowel diseases (van Dullemen et al., 1995; Eigler et al., 1997; van Dullemen et al., 1998; Rutgeerts et al., 1999; Sandborn and Hanauer, 1999). Moreover, TNF- induces expression of IFN- in lymphocytes, which directly and indirectly leads to increased mucosal damage (O'Connell et al., 2000). This cytokine was found to be highly expressed in colonic specimens of patients with Crohn's disease (Camoglio et al., 1998; Monteleone and MacDonald, 2000). Oral administration of mesopram partially prevented colitis clinically and histologically and, although statistically not significant, led to a reduced colonic content of TNF- and IFN-, supporting the hypothesis that the anti-inflammatory effect of mesopram results from inhibition of TNF- expression. This may consecutively lead to a reduced synthesis of other proinflammatory cytokines such as IFN- and a reduced infiltration of activated lymphocytes.

Additonally, we demonstrated in the therapeutic model that oral mesopram also improves clinical signs of an established colitis more efficiently than placebo. Yet, intraperitoneal administration of mesopram reduced pre-existing colitis more efficiently than oral administration. This finding may partially be explained by the fact that, compared with oral feeding, intraperitoneal injection is a technically more reliable route of administration. Another reason for the higher efficacy of intraperitoneal mesopram may be a lower bioavailability of mesopram when administered by the oral route. Since no data are available concerning the pharmacokinetics of mesopram in mice, this issue remains unresolved. The efficacy of mesopram in ameliorating an established inflammation is also reflected by the statistically significant suppression of IFN- synthesis by colonic tissue compared with controls. Given the fact that histologic improvement lags behind clinical improvement, even the nonsignficant trend toward an improved histologic score in the intraperitoneally treated group highlights the efficacy of mesopram in the treatment of established colitis.

In additional experiments (10 mg/kg mesopram i.p. q.d.) with the TH1-biased mouse strain C57BL/6J we could furthermore demonstrate a trend toward a suppressed synthesis of TNF- and IFN- in splenocytes by fluorescence-activated cell sorting analysis (data not shown). This hints at a systemic effect of mesopram on the inflammatory response. These results are consistent with the suppression of IFN- synthesis in peripheral blood mononuclear cells of patients with multiple sclerosis or atopic dermatitis by the specific type-4 phosphodiesterase inhibitor rolipram (Ostlere et al., 1995; Navikas et al., 1998).

Preclinical research and first clinical trials provide evidence that specific inhibitors of type-4 phosphodiesterase may represent a new class of anti-inflammatory drugs to treat different kinds of chronic inflammatory diseases. Since cAMP additionally mediates relaxation of bronchial smooth muscle cells, airway diseases such as asthma and chronic obstructive pulmonary disease have been selected as the first target indications (Doherty, 2000). Results from clinical trials with the specific type-4 phosphodiesterase inhibitor airflow in patients with chronic obstructive pulmonary disease are positive (Hay, 2000; Compton et al., 2001). A recent study reports on the successful inhibition of experimental autoimmune encephalitis in rodents by mesopram (Dinter et al., 2000). Future indications for these agents as anti-inflammatory drugs could comprise rheumatoid arthritis (Nyman et al., 1997; Francischi et al., 2000; Hogan et al., 2001) and, as demonstrated here, chronic inflammatory bowel diseases.

We conclude, that specific inhibition of type-4 phosphodiesterase by mesopram may form a novel attractive therapeutic strategy in the treatment of chronic inflammatory diseases and warrants to be evaluated in a clinical trial.