Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Current treatments for patients with inflammatory bowel disease (IBD) are based on recent advances in elucidating the pathophysiology of the disease. Several studies indicate that T helper (Th)1 immune responses have important roles in the development of IBD (Fuss et al., 1996; Sartor, 1997; Simpson et al., 1997). Furthermore, dysregulation of cytokine networks is involved in Th1-dominant immune responses in IBD (Fuss et al., 1996; Sartor, 1997; Simpson et al., 1997). Several murine colitis models with abnormalities of various cytokines or their receptors, which influence Th1-type immune responses, develop colitis (Sandlack et al., 1993; Neurath et al., 1995; Strober et al., 1998).

IL-10, which is a cytokine produced by activated macrophages and Th2-type T cells, has a pivotal inhibitory effect on the Th1-type immune response as well as on the antigen-presenting function of monocytes and macrophages (Fiorentino et al., 1991a,b). In addition, IL-10 induces an antigen-specific anergic state in human CD4-positive T cells (Fiorentino et al., 1991b). Kuhn et al. (1993) reported development of colitis in the IL-10 gene knockout mouse. Thus, IL-10 has an important role in maintaining the normal immune state in the intestine.

In patients with IBD, IL-10 levels in the intestine are abnormal and its kinetics differs from that in the whole body. Serum levels of IL-10 are elevated in patients with active IBD, whereas intestinal tissue concentrations of IL-10 are low or within the normal range in patients with IBD (Kucharzik et al., 1995; Schreiber et al., 1995). Moreover, in situ hybridization and immunohistochemistry studies indicate that local production of IL-10 by mucosal mononuclear cells in IBD is insufficient to down-regulate proinflammatory cytokines such as IL-1 in the lamina propria compartment (Autschbach et al., 1998). Together, these data suggest that impairment of IL-10 function is involved in the pathogenesis of IBD, and therefore, IL-10 is one of the most promising candidates for the treatment of IBD. The clinical efficacy of systemically injected IL-10 for patients with mild to moderately active Crohn's disease, however, has not been satisfactory (Fedorak et al., 2000; Sands, 2000; Schreiber et al., 2000). Moreover, several adverse side effects such as headache, high fever, and back pain are inevitable, although they are reversible (Fedorak et al., 2000; Sands, 2000; Schreiber et al., 2000). Administration of IL-10 using a more efficient drug delivery system is required to enhance its effect on the inflamed colon and to decrease side effects.

A new drug delivery system using polymer microspheres was recently developed to obtain sustained release of various drugs and successful targeting of specific organs (Tabata and Ikada, 1990). We previously reported a new oral drug delivery system using poly(DL-lactic acid) microspheres containing dexamethasone. This drug delivery system has more prominent anti-inflammatory effects on colonic inflammation than dexamethasone alone by specifically targeting the M cells in the inflamed colon (Nakase et al., 2000b, 2001). Poly(DL-lactic acid) microspheres cannot be used for the delivery of protein-based drugs such as the various cytokines, because the biologic activity of the proteins might be lost due to protein-polymer formation. To solve this problem, we recently developed gelatin microspheres (GM) that can release proteins without affecting their biologic activity (Tabata and Ikada, 1999). In the present study, we successfully incorporated IL-10 into gelatin microspheres, and evaluated its therapeutic effects on colitis in IL-10-deficient (IL-10^/) mice.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals. IL-10^/ mice on a C57BL/6 background were purchased from Jackson Laboratory (Bar Harbor, ME). C57BL/6 mice were used as controls at 4 weeks of age. All mice were housed in specific pathogen-free conditions in the animal facility of Kyoto University. The studies were approved by the Animal Protection Committee.

Preparation of Gelatin Microspheres. Gelatin microspheres were prepared as reported previously (Tabata et al., 1999). Briefly, 10 ml of an aqueous solution of acidic gelatin (10 wt% preheated to 40°C) was added dropwise into 375 ml of olive oil while stirring at 420 rpm at 40°C for 10 min to yield a water-in-oil emulsion. The emulsion temperature was decreased to 15°C, followed by further stirring for 30 min to allow for natural gelation of the gelatin aqueous solution. Acetone (100 ml) was added to the emulsion and stirring was continued for 1 h. The resulting microspheres were washed three times with acetone, recovered by centrifugation (5000 rpm, 4°C, 5 min), passed through sieves with different apertures for size fractionation, and air-dried. The average size of the microspheres was adjusted to be less than 12 µm. Gelatin aqueous solution (0.2 ml, 10 wt%) and olive oil (5 ml) were agitated with a vortex mixer for 1 min and then sonicated at 3.0 W/cm² for variable time periods. The prepared emulsion was cooled down, washed with acetone by centrifugation, and air-dried. The noncross-linked and dried gelatin microspheres (25 mg) were placed in 5 ml of 0.1 wt% Tween 80 aqueous solution containing glutaraldehyde (40 µg/ml) and stirred at 4°C for 15 h to facilitate cross-linking. Following collection by centrifugation (5000 rpm, 4°C, 5 min), the microspheres were agitated in 5 ml of 10 mM aqueous glycine solution at 37°C for 1 h to block the residual aldehyde groups on unreacted glutaraldehyde. The resulting microspheres were washed three times with double-distilled water by centrifugation and freeze-dried.

Radiolabeling of IL-10. Recombinant mouse IL-10 (rmIL-10) was purchased from Genzyme Corporation (Cambridge, MA). Recombinant mIL-10 was labeled with ¹²⁵I by radioiodination using chloramine-T essentially as described (Vaisman et al., 1990; Cohen et al., 1995). Briefly, 2.0 µg of rmIL-10 were labeled with about 100 mCi sodium (Amersham International UK, Ltd., Little Chalfont, Buckinghamshire, UK) and 10 µg of chloramine-T in 20 mM sodium phosphate buffer (pH 7.4) in a final volume of 120 µl. After 1 min, the iodination was quenched by the addition of 50 µl of sodium metabisulphite (2 mg/ml). ¹²⁵I-Labeled rmIL-10 was separated from free iodine using a heparin-Sepharose column. The specific activity of ¹²⁵I-labeled rmIL-10 was 0.5 to 1.5 × 10⁵ cpm/ng.

Incorporation of Protein into GM. Incorporation of rmIL-10 into GM was performed by allowing the freeze-dried microspheres to swell in the aqueous solution of the protein. Briefly, 10 µl of IL-10 solution (5 µg/ml) was dropped onto 2.5 mg of freeze-dried glutaraldehyde cross-linked microspheres, and allowed to sit at 37°C for 1 h. ¹²⁵I-Labeled and unlabeled GM were prepared in the same way.

Release Test and Blood Distribution of IL-10. The in vitro release test of rmIL-10 from the microspheres was conducted at 37°C. GM containing ¹²⁵I-labeled rmIL-10 were immersed under shaking in collagenase solution (1 ml, 0.8 mg/ml) prepared from phosphate-buffered saline (PBS). After centrifugation, the radioactivity of ¹²⁵I-labeled rmIL-10 in supernatants was determined using a gamma counter (ARC-300; Aloka Co., Tokyo, Japan).

Treatments. Female IL-10^/ mice at 4 weeks of age were divided into five groups (five mice each; groups A-E) and treated as follows: group A, no treatment; group B, GM (2.5 mg/body) alone; group C, rmIL-10 (0.05 µg) alone; group D, GM containing rmIL-10 (GM-IL-10; 2.5 mg/body that contained 0.05 µg of rmIL-10). C57BL/6 mice were used as controls (group E). GM, rmIL-10, and GM-IL-10 were suspended in 200 µl of PBS. Each mouse received rectal administrations three times per week as mentioned above and were sacrificed after 1 month. The spleen and colonic tissues were removed from each mouse and examined in further studies as described below.

Assessment of the Severity of Colitis. The colon was opened by longitudinal incision, washed in PBS, and subsequently excised for microscopic observation of damage and isolation of Mac-1-positive cells. Microscopic damage was assessed after fixation in 10% formalin, followed by hematoxylin and eosin staining. Histologic analysis was performed in a blind manner as reported previously (Fuss et al., 1999).

Isolation of Mac-1-Positive Cells from Lamina Propria and Spleen. Lamina propria and spleen mononuclear cells were isolated as described previously (Neurath et al., 1995), and the cell suspension was plated on a monocyte-seperating plate (Nihon-Koutai Laboratory, Gunma, Japan) for 60 min at 37°C. After incubation, nonadherent cells were washed three times with PBS and removed. Adherent cells containing Mac-1-positive cells were incubated with 0.2% EDTA in PBS for 30 min at 4°C. After washing twice with PBS, adherent cells were analyzed using flow cytometry, and more than 90% of the cells were determined to be Mac-1-positive using a CD11b antibody (Serotec, Oxford, UK).

mRNA Expression of IL-12 in Mac-1-Positive Cells in the Lamina Propria of the Colon. Total RNA from 10⁶ Mac-1-positive cells in the lamina propria of the colon was isolated using the guanidium isothiocyanate method as described previously (Khan and Collins, 1994). The concentration of RNA was determined by absorbance at 260 nm relative to that at 280 nm. The RNA was stored at 70°C until use. RNA (1 µl) was transcribed to cDNA using a Superscript preamplification system (Invitrogen, Carlsbad, CA). The reverse transcription product (1 µl) was added to 1 mM of each primer and a solution of 1 U of Taq DNA polymerase (Takara, Biochemicals, Ohtsu, Japan) in a final volume of 20 µl. Polymerase chain reaction (PCR) amplification was performed for 35 cycles (1 min at 94°C, 1 min at 52°C, and 20 s at 20°C). cDNA-free solution served as a negative control for each reaction. The sequences of primers for each cytokine are as follows: IL-12p35 forward 5'-GAGGACTTGAAGATGTACCAG-3'; IL-12p35 reverse 5'-TTCTATCTGTGTGAGGAGGGC-3'; IL-12p40 forward 5'-GACCCTGCCCATTGAACTGGC-3'; IL-12p40 reverse 5'-CAACGTTGCATCCTAGGATCG-3'; -actin forward 5'-TTGTAACCAACTGGGACGATATGG-3'; and -actin reverse 5'-GATCTTGATCTTCATGGTGCTAGG-3'.

Flow Cytometric Cell Sorting (FACS) Analysis. Mac-1-positive cells of lamina propria and spleen from IL-10^/ mice were resuspended in FACS buffer (1× PBS, 0.2% bovine serum albumin fraction V; Sigma-Aldrich, St. Louis, MO) to a final concentration of 10⁶ cells/ml. Cells (1 × 10⁶) were preincubated with mouse serum for 20 min on ice and stained with fluorescein isothiocyanate-conjugated antibodies against CD40 (BD PharMingen, San Diego, CA). Cells were washed with FACS buffer and analyzed using a FACS flow cytometer (EPICS XL; Beckman Coulter, Inc., Fullerton, CA). Results were analyzed using SYSTEM II software (Beckman Coulter, Inc.).

Statistical Analysis. The generalized Wilcoxon t test and the Mann-Whitney test were used where appropriate for statistical analysis. The data were presented as means ± S.E. A two-tailed P value of less than 0.05 was considered to be statistically significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

In Vitro Release of rmIL-10 from GM-IL-10. Recombinant mIL-10 was released from GM-IL-10 in a time-dependent manner: 38 ± 5% (2 h), 66 ± 4% (6 h), 83 ± 3% (12 h), and 90 ± 5% (48 h) (Fig. 1).

View larger version (14K): [in this window] [in a new window]   Fig. 1.   In vitro release of rmIL-10 from gelatin microspheres. The 2.5 mg of gelatin microspheres containing 125I-labeled rmIL-10 were immersed under shaking in 1 ml of PBS with collagenase (0.8 mg/ml). Each point represents the mean ± S.E. of the percentage of the released radioactivity (n = 3).

Retention Ratio of IL-10 in the Colon and Blood Distribution of IL-10. Retention ratios of rmIL-10 remaining in the colon were measured at 15, 30, and 45 min, and 1, 6, 12, 24, and 48 h after rectal administration of GM containing ¹²⁵I-labeled IL-10 or ¹²⁵I-labeled IL-10 alone (Fig. 2). The retention ratios of rmIL-10 in the colon of GM-IL-10-treated normal mice at 1, 6, and 12 h after administration were significantly higher than the respective values of rmIL-10-treated normal mice. Furthermore, the retention ratios of rmIL-10 in the colon of GM-IL-10-treated IL-10^/ mice at 6, 12, and 24 h were significantly higher than the respective values of the GM-IL-10-treated normal mice.

View larger version (19K): [in this window] [in a new window]   Fig. 2.   In vivo retention ratio of IL-10 in the colon after rectal administration of gelatin microspheres containing 125I-labeled IL-10 or 125I-labeled IL-10 alone. Each point represents the mean ± S.E. of the percentage of the radioactivity remaining in the colonic tissue (n = 3). #, P < 0.05 compared with IL-10-treated normal mice. , P < 0.01 compared with GM-IL-10-treated normal mice.

/

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Blood distribution of IL-10 after rectal administration of gelatin microspheres containing 125I-labeled IL-10 or 125I-labeled IL-10 alone. Each point represents the mean ± S.E. of the percentage of the radioactivity remaining in the blood (n = 3). #, P < 0.05 compared with IL-10-treated normal mice. , P < 0.05 compared with GM-IL-10-treated normal mice.

Animal Profile. IL-10^/ mice generally appeared healthy until 4 weeks of age, when they gradually developed diarrhea and spontaneous colitis. Histological studies in the colon of IL-10^/ mice at birth, 2, and 3 weeks of age were normal. However, at 4 weeks of age, the IL-10^/ mice displayed various degrees of colonic inflammation characterized by mucosal ulceration and mild to moderate epithelial hyperplasia. IL-10^/ mice had a normal small intestine macroscopically and histologically from birth to 4 or 8 weeks of age (data not shown).

Macroscopic Evaluation. Macroscopic examination of the colon from groups A (nontreated), B (GM alone), and C (rmIL-10 alone) revealed marked thickening of the colonic wall. In contrast, macroscopic examination of group D (GM-IL-10) revealed no thickening of the colonic wall. There was no difference in the GM-IL-10 treatment effect with respect to the site within the colon (data not shown).

Histologic Evaluation. Histologic findings revealed epithelial hyperplasia, mucosal ulceration, and remarkable infiltration of mononuclear cells in the colons of groups A (nontreated), B (GM alone), and C (rmIL-10 alone). In contrast, histologic findings in group D (GM-IL-10) were almost normal, except for low levels of infiltrating monocytes (Fig. 4a). The histologic score for group D was significantly lower than those for groups A, B, and C. There were no significant differences in the histologic scores among groups A, B, and C (Fig. 4b). The jejunum and ileum were histologically normal in all groups.

View larger version (89K): [in this window] [in a new window]   Fig. 4.   Effects of GM-IL-10 and IL-10 on the colonic mucosa of IL-10/ mouse. a, histologic findings of the colon in IL-10/ mouse with various treatments. (A), group A (nontreated IL-10/ mouse); (B), group B (GM-treated IL-10/ mouse); (C), group C (IL-10-treated IL-10/ mouse); (D), group D (GM-IL-10-treated IL-10/ mouse); (E), normal mice. b, histologic scores of the colon. A score of 4 represents maximal injury and a score of 0 no injury. Data are means ± S.E. (n = 5 for each group). , group A (nontreated IL-10/ mouse); , group B (GM-treated IL-10/ mouse); , group C (IL-10-treated IL-10/ mouse); , group D (GM-IL-10-treated IL-10/ mouse); , controls. *, P < 0.01 versus groups A, B, and C.

mRNA Expression in Mac-1-Positive Cells from the Lamina Propria of the Colon. The reverse transcription- PCR results demonstrated that mRNA expressions of IL-12 p40 and IL-12 p35 were up-regulated in Mac-1-positive cells in the colon of groups A (nontreated), B (GM alone), and C (rmIL-10 alone). Transcript levels of these cytokines in group D (GM-IL-10), however, were significantly lower than those in groups A, B, and C (Fig. 5).

View larger version (29K): [in this window] [in a new window]   Fig. 5.   Effect of several treatments on the transcript levels of IL-12 p40 and IL-12 p35 in Mac-1-positive cells. Amounts of cytokine transcripts are expressed as relative concentration of -actin (-actin = 1). Results are means ± S.E. #, P < 0.01 versus groups A, B, and C. , group A (nontreated IL-10/ mouse); , group B (GM-treated IL-10/ mouse); , group C (IL-10-treated IL-10/ mouse); , group D (GM-IL-10-treated IL-10/ mouse); , controls.

Expression of CD40 on Mac-1-Positive Cells from the Lamina Propria of the Colon and Spleen. Compared with untreated IL-10^/mice, CD40 expression on Mac-1-positive cells in the lamina propria and spleen were markedly decreased in group D (GM-IL-10) but not in group C (rmIL-10 alone) (Fig. 6).

View larger version (16K): [in this window] [in a new window]   Fig. 6.   Flow cytometric analysis of CD40 expression on Mac-1-positive cells in the lamina propria and spleen of IL-10/ mice after treatment with GM-IL-10 and IL-10. CD40 expression on Mac-1-positive cells was up-regulated in IL-10/ mice in the lamina propria and spleen (black line). CD40 expression was down-regulated in the GM-IL-10-treated group (red line), but not in the IL-10-treated group (blue line) in the lamina propria and spleen.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The present study clearly demonstrated that rectal administration of GM-IL-10 inhibits colonic mucosal inflammation in IL-10^/ mice more efficiently than treatment with IL-10 alone. Analysis of the retention ratio of ¹²⁵I-labeled IL-10 in the colon indicated that IL-10 remained in the colon of GM-IL-10 treated mice longer than in IL-10-treated mice. Thus, IL-10 appears to be released from GM-IL-10 gradually and continuously in the colonic mucosa, resulting in prolonged availability of IL-10 to the colon. Moreover, the retention ratio of IL-10 after administration of GM-IL-10 in IL-10^/mice was greater at each time point than the respective values in normal mice. The reason that IL-10 remains in the inflamed colon in greater concentrations as well as for a longer period of time than in normal colon is unclear. We previously reported a new oral drug delivery system using poly(DL-lactic acid) microspheres, and demonstrated that poly(DL-lactic acid) microspheres specifically target the inflamed colonic mucosa not only by preferentially adhering to the inflamed mucosa but also by being absorbed by M cells (Nakase et al., 2000b). Whether GM are more adhesive to the tissues than poly(DL-lactic acid) microspheres is unclear in the present study. Similar mechanisms might be considered for prominent targeting of GM for inflamed colon compared with normal colon. Additionally, the number of M cells is increased in inflamed mucosa (Kucharzik et al., 2000), which might also contribute to the targeting of GM to the inflamed colon.

As for blood distribution of IL-10, there was a longer period of distribution in the GM-IL-10-treated group than in the IL-10-treated group. This result is considered to be due to a difference in the colonic retention ratio between IL-10 and GM-IL-10, and reflects the local sustained release in the colon by GM-IL-10. The ratio of blood distribution is extremely low (less than 0.07% of administered ¹²⁵I-labeled rmIL-10) compared with that of intestinal distribution, and therefore, it is unlikely to induce systemic side effects. In this study, instead of directly measuring the concentration of IL-10 in the blood and colonic tissue, we measured radioactivity of blood and colonic tissue during the experiment. Although the results of radioactivity may partially include a ¹²⁵I-fragment from ¹²⁵I-labeled rmIL-10, it is considered to neglect this possibility because of very low levels in the preliminary study (data not shown). In any event, GM-IL-10 appears to exert a more favorable effect on the inflamed colon with few systemic side effects. Thus, rectal administration of GM-IL-10 might be an ideal cytokine delivery system for treating IBD.

Spontaneous colonic inflammation in IL-10^/ mice under specific pathogen-free conditions occurs at approximately 4 weeks of age as reported previously (Madsen et al., 2000). Therefore, in the present study, the treatments were started at 4 weeks, and the therapeutic effects on the colonic inflammation were observed at 8 weeks of age. The results of both macroscopic and histologic studies revealed that GM-IL-10 has more potent inhibitory effects on colitis than IL-10 alone. Furthermore, we previously reported that rectal administration of GM-IL-10 ameliorated established trinitrobenzene sulfonic acid-induced colitis in rats with strong inhibition of nuclear factor B (Nakase et al., 2000a). These results suggested that GM-IL-10 not only prevents but also improves colonic inflammation.

IL-12 is a cytokine involved in Th1 T-cell differentiation, which promotes the production of interferon- (Wenner et al., 1996). Systemic administration of monoclonal antibodies against IL-12 leads to the improvement of colitis in mice by elimination of the Th1 T cells through induction of Fas-mediated apoptosis (Fuss et al., 1999). These data suggest that IL-12 has a key role in the development of Th1 dominant colitis. In the present study, gene expression of IL-12 p40 as well as p35 was enhanced in Mac-1-positive mononuclear cells in the lamina propria of the colonic mucosa of IL-10^/ mice, and topical administration of GM-IL-10 to the colon reduced the expression of IL-12 p40 and p35 to levels similar to those of normal mice, whereas IL-10 alone had very little effect. Therefore, improvement of colitis in IL-10^/ mice following rectal administration of GM-IL-10 might be due, at least in part, to decreased mRNA expression of IL-12 p40 and p35. The data also support the idea that administration of GM-IL-10 with its sustained release of IL-10 is more advantageous for treatment of colitis than administration of IL-10 alone.

To further examine the mechanism of therapeutic action of GM-IL-10 on colitis, we investigated CD40 expression on Mac-1-positive mononuclear cells in the lamina propria of the colon and also in the spleen. CD40, a 45-kDa cell surface glycoprotein on B cells, dendritic cells, and activated macrophages, is a member of the tumor necrosis factor receptor superfamily, and CD154, a 39-kDa surface glycoprotein on activated T cells, has been identified as a ligand of CD40 (Freudenthal and Steinman, 1990; Hollenbaugh et al., 1992; Alderson et al., 1993). In monocytes/macrophages, the interaction of CD40 with CD154 results in the production of inflammatory cytokines and rescue of monocytes from apoptosis (Shu et al., 1995; Kennedy et al., 1996). Moreover, CD40-CD154 interaction is considered to be important for priming Th1-type cells through production of IL-12 by macrophages and monocytes (Kennedy et al., 1996). Several clinical studies indicate that the CD40-CD154 interaction is involved in the pathogenesis of IBD (Shu et al., 1995; Battaglia et al., 1999; Liu et al., 1999). In the present study, the expression of CD40 on Mac-1-positive cells in intestinal lamina propria and spleen were down-regulated more prominently in GM-IL-10-treated mice than in IL-10-treated mice. Down-regulation of CD40 on Mac-1-positive cells in the spleen may be due to the systemically sustained distribution of IL-10, although it is extremely lower than the distribution of IL-10 in the colon. These results emphasize the importance of locally sustained release of IL-10 from GM-IL-10 for blocking CD40-CD154 pathways systemically and locally.

New treatments for colitis, such as IL-10 gene transfer and administration of engineered bacteria secreting IL-10, were recently introduced (Barbara et al., 2000; Steidler et al., 2000). These treatments are effective for preventing the onset of colitis or reducing the mucosal injury in experimental colitis models. In addition to these treatments, GM-IL-10 should be considered for therapy of IBD, because control and delivery of an optimal dose of IL-10 to local inflamed mucosa can be achieved without remarkably elevating the concentration of IL-10 in the blood. Consequently, IL-10-related side effects in IBD patients can be greatly reduced. Thus, GM-IL-10 might be more suitable for clinical application than other available delivery systems in human IBD.

In conclusion, rectal administration of GM-IL-10 releases IL-10 in the inflamed colonic mucosa continuously and inhibits inflammation more efficiently than IL-10 alone. Clinical trials of these microspheres for patients with IBD should be considered.