Oral Administration of the Immunomodulator JBT-3002 Induces Endogenous Interleukin 15 in Intestinal Macrophages for Protection against Irinotecan-mediated Destruction of Intestinal Epithelium¹

Irinotecan (CPT-11), a semisynthetic derivative of camptothecin, is a potent chemotherapeutic agent that is increasingly being used for the treatment of drug-refractory colorectal cancer (1, 2, 3). Unfortunately, the utility of irinotecan is limited by induction of severe diarrhea (2, 3). We and others have ascribed this toxicity to structural changes in intestinal architecture that resulted from disordered epithelial cell differentiation and apoptosis (4, 5).

Gastrointestinal mucosal integrity is regulated by several growth factors and cytokines (6, 7), such as transforming growth factor-β (8) and IL-2³(9, 10). IL-15 is a novel cytokine that shares the IL-2 receptor β- and γ-chains for signal transduction and, hence, many of the biological activities of IL-2 (11, 12, 13, 14). In vitro studies have shown that IL-15 can be mitogenic for intestinal epithelial cells (7, 15), and recent reports have demonstrated that administration of rIL-15 can protect rats from intestinal toxicity induced by 5-fluorouracil (16) or irinotecan (17). IL-15 is expressed by intestinal epithelial cells and mononuclear cells in the lamina propria (7), and IL-15 transcription is known to be increased by macrophage-activating microbial agents such as LPS or Bacillus Calmette-Guérin (18). These findings suggest that IL-15 may have biological activities that promote recovery of epithelial integrity after wounding caused by chemotherapeutic drugs.

We have reported that the incubation of monocytes/macrophages with a new synthetic lipopeptide, JBT-3002 (N-acylated derivative of Ψ-amino-C1-C3-alkanesulfonic acid), can induce production of monocyte-derived cytokines (19, 20, 21). Moreover, the p.o. administration of JBT-3002 to mice overcomes irinotecan-induced intestinal toxicity by an unknown mechanism (4). The purpose of this study was to determine whether up-regulation of endogenous IL-15 in the lamina propria prevents irinotecan-induced injury of the intestinal mucosa in mice given p.o. JBT-3002.

Specific pathogen-free BALB/c mice were purchased from the Animal Production Area of the National Cancer Institute (Frederick, MD). The mice were maintained according to institutional guidelines in facilities approved by The American Association for Accreditation of Laboratory Animal Care, in accordance with United States Department of Agriculture, Department of Health and Human Services, and NIH regulations and standards.

Eagle’s MEM, Ca²⁺-/Mg²⁺-free HBSS, and fetal bovine serum were purchased from M. A. Bioproducts (Walkersville, MD). JBT-3002 was obtained from Jenner Biotherapies (San Ramon, CA). Irinotecan, produced by Yakult Honsha Co. (Tokyo, Japan), was used according to manufacturer’s instructions. Simian rIL-15 was obtained from R&D Systems (Minneapolis, MN). Murine rIFN-γ (specific activity, 10⁷ units/mg) was obtained from PharMingen (San Diego, CA). Salmonella LPS was purchased from Sigma Chemical Co. (St. Louis, MO). All reagents used in tissue culture (except for LPS) were free of endotoxin as determined by the Limulus amebocyte lysate assay (Associates of Cape Cod, Woods Hole, MA).

CT-26 murine colon carcinoma cells, syngeneic to BALB/c mice (4), and IEC-6 normal rat intestinal epithelial cells obtained from the American Type Culture Collection (Manassas, VA) were grown as monolayer cultures in Eagle’s MEM supplemented with 5% fetal bovine serum, vitamins, sodium pyruvate, l-glutamine, and nonessential amino acids. The adherent monolayer cultures were incubated at 37°C in a humidified atmosphere containing 5% CO₂ in air. PEMs were collected by peritoneal lavage of BALB/c mice given an i.p. injection of 1.5 ml of thioglycolate broth (Baltimore Biological Laboratories, Cockeysville, MD) 4 days before the harvest (4). One h after plating, the nonadherent cells were removed by washing with HBSS. At that time, >98% of adherent cell populations were macrophages according to morphology, phagocytic criteria, and positive staining with antibodies against scavenger receptor (22). All cultures were free of Mycoplasma, reovirus type 3, pneumonia virus of mice, K virus, encephalitis virus, lymphocyte choriomeningitis virus, ectromelia virus, and lactate dehydrogenase virus (assayed by M. A. Bioproducts, Walkersville, MD).

To prepare tumor cells for inoculation, we harvested CT-26 cells in exponential-growth phase by a brief exposure to a 0.25% trypsin-0.1% EDTA solution. The cell suspension was pipetted to produce a single-cell suspension, washed, and resuspended in HBSS. Cell viability was determined by trypan blue exclusion, and only single-cell suspensions of >90% viability were used for injection. Tumor cells (1 × 10⁴/0.05 ml of HBSS) were injected into the spleens of BALB/c mice after laparotomy under methoxyflurane anesthesia. The incision was closed in one layer with wound clips (4). This protocol resulted in a 100% incidence of liver metastasis. The mice were euthanized when the control group became moribund. The volume of tumors (TV) in the spleens was calculated using the following formula: TV = L (mm) × W² (mm²)/2, where L and W represent the length and the width of the tumor mass, respectively. The livers were harvested and placed in Bouin’s solution for 24 h before fixation in 10% buffered formalin. The number of experimental metastases on the surface of the liver was determined using a dissecting microscope. When the number of metastases exceeded 100, we assigned a value of >100.

Because our recent data have suggested that JBT-3002 may induce activation of macrophages by interacting with micellar JBT-3002 (21), for this study, we used soluble JBT-3002 instead of JBT-3002 encapsulated in multilamellar vesicles-liposomes (4). JBT-3002 was suspended in HBSS at 1 mg/ml, sonicated for 5 min, and stored at 4°C. The solution was mixed on a vortex machine prior to each experiment. JBT-3002 suspended in 0.2 ml of HBSS was administered by mouth using a rigid, curved feeding tube (4). In a preliminary study, we determined the biological effective dose of oral JBT-3002. BALB/c mice were, therefore, given p.o. JBT-3002 at doses at 0.0001, 0.001, 0.01, 0.1, or 1.0 μg per dose (per mouse) once a day for 3 consecutive days. Seven days later, the mice were given i.p. injections of 100 mg/kg irinotecan once a day for 4 days. The mice were monitored for morbidity until day 14. The administration of irinotecan (CPT-11) alone produced severe diarrhea and corresponding loss of body weight. In contrast, all of the mice given 0.01, 0.1, or 1.0 μg/dose of JBT-3002 prior to the irinotecan survived with minimal to no diarrhea or reduction in body weight. In all of the following studies, we treated mice with 0.05 μg of JBT-3002 (2.5 μg/kg).

The small intestines of control and treated mice were everted, rinsed in water, and immediately fixed overnight at 4°C in a solution containing 3% glutaraldehyde plus 2% paraformaldehyde in 0.1 m cacodylate buffer. The samples were then treated with 1% buffered osmium tetroxide followed with 1% thiocarbohydrazide for 10 min and then with 1% aqueous osmium tetroxide for 10 min. The samples were rinsed with distilled water, treated with 1% aqueous tannic acid for 30 min, and rinsed again with distilled water; they were then incubated for 12 h with 1% aqueous uranyl acetate. The samples were then washed with distilled water three times, dehydrated in increasing concentrations of ethanol, and placed in hexamethyldisilazane for 10 min according to Adams et al. (23). The samples were dried overnight under vacuum, and the specimens were mounted directly on double-stick carbon tabs and then mounted on aluminum specimen mounts. The samples were coated with platinum/palladium alloy and examined in a Hitachi S520 scanning electron microscope at an accelerating voltage of 5 kV.

Mice were injected i.v. with 0.2 ml of saline containing 250 μg BrdUrd 1 h before they were killed (24). Samples from the small intestines were fixed in 10% buffered formalin. Immunohistochemical staining was performed by the immunoperoxidase technique, as described previously (4). Paraffin sections (5 μm thick) were dewaxed in xylene and then rehydrated in graded ethanol. For S-phase labeling with BrdUrd, the sections were incubated with 2 n HCl for 30 min at 37°C. For labeling with IL-15, the sections were incubated with Pepsin Reagent (Biomeda Corp., Foster City, CA) for 20 min at room temperature. The slides were rinsed twice with PBS, and endogenous peroxidase was inactivated by treatment with 3% hydrogen peroxide in methanol for 10 min. Nonspecific reactions were blocked by incubating the sections in a solution containing 5% normal horse serum in PBS. The sections were incubated overnight at 4°C with a 1:50 dilution of monoclonal mouse anti-BrdUrd antibody (Becton Dickinson, Mountain View, CA) or a 1:100 dilution of goat polyclonal anti-IL-15 antibody (Santa Cruz Biotechnology, Santa Cruz, CA). The samples were then rinsed three times with PBS and incubated at room temperature for 1 h with the appropriate dilution of peroxidase-conjugated secondary antibodies. Secondary antibodies were antimouse IgG1 antibody (PharMingen) at 1:100 dilution for BrdUrd and anti-goat IgG antibody (Jackson ImmunoResearch, West Grove, PA) at 1:200 dilution for IL-15. After three rinses with PBS, the sections were incubated for 5 min with diaminobenzidine substrate (Research Genetics, Huntsville, AL). The sections were then rinsed with distilled water and counterstained with Gill’s hematoxylin (Sigma). The number of BrdUrd-positive cells in the intestines was determined in multiple ×200 fields using a light microscope. The results are presented as the number of positive cells per 0.25 mm².

PEMs (3 × 10⁴) and IEC-6 cells (1 × 10⁴) were resuspended in 0.5 ml of serum-free MEM, plated into chamber slides (Nunc, Inc., Naperville, IL), maintained for 72 h, and fixed by immersing the slides into ice-cold acetone for 5 min, followed by rinsing in PBS. Immunostaining for IL-15 was performed by the technique described above.

Total RNA was extracted from the small intestines, PEMs, and IEC-6 cells using a Tri Reagent (Molecular Research Center, Inc., Cincinnati, OH). First-strand cDNA was synthesized from 1 μg of RNA using a First Strand Synthesis Kit (Pharmacia, Piscataway, NJ) in 33 μl of reaction mixture, according to the manufacturer’s instructions. The synthesized first-strand cDNA (1 μl) was amplified by PCR in a final volume of 50 μl containing 10 mm Tris-HCl, 3 mm MgCl₂, 50 mm KCl, 0.01% gelatin, 200 mm dNTP, 50 pmol of each primer, and 2.5 units of an AmpliTaq Gold Taq polymerase (Perkin Elmer, Foster City, CA). The IL-15 primers used were derived from full-length cDNA sequence for murine (25) and rat IL-15 (GenBank accession no. U69272). Sequences used were sense 5′-GTTCTGGATGGATGGCAGCT-3′ and antisense 5′-CTGTTTGCAAGGTAGAGCACG-3′. PCR thermocycles consisted of 1 min at 94°C, followed by 30 s at 58°C and 45 s at 72°C. Before the first cycle, a denaturation step for 8 min at 95°C was included, and after 30 cycles, the extension was prolonged for 7 min at 72°C.

The number of cycles of PCR amplification used was first determined by amplifying cDNA from LPS-stimulated PEMs through 20–35 cycles. The PCR product was resolved by electrophoresis on a 1.2% agarose gel, denatured in 1.5 m NaCl and 0.5 m NaOH, neutralized in 1.5 m NaCl and 1.0 m Tris-HCl (pH 7.0), blotted onto nylon filters (DuPont Co., Boston, MA) by overnight passive transfer, and UV cross-linked with 120,000 μJ/cm² using a UV Stratalinker 1800 (Stratagene, La Jolla, CA). DNA was then hybridized with a ³²P-labeled cDNA probe, which was made by amplifying cDNA from LPS-stimulated PEMs using an inner sense primer, 5′-CCATCTCGTGCTACTTGTG-3′, and the antisense primer as described above. Hybridizations were performed as described previously (26). Nylon filters were washed three times at 55–60°C with 30 mm NaCl, 3 mm sodium citrate (pH 7.2), and 0.1% SDS (w/v). Signals were developed for 30 min on Hyperfilm-MP (Amersham, Buckinghamshire, England) and quantified on an LKB Ultrascan XL laser densitometer (Pharmacia LKB Biotechnology, Uppsala, Sweden).

To ensure that equivalent amounts of cDNA were used in each reaction, we also performed PCR for glyceraldehyde-3-phosphate dehydrogenase. The primers used were: sense, 5′-TTCACGACCATGGAGAAGGCTG-3′; and antisense, 5′-CGTTGAGCTCAGGGATGACCTT-3′. The PCR product was electrophoresed through a 1.2% agarose gel and visualized with ethidium bromide.

Tumor cells (CT-26) and normal intestinal epithelial cells (IEC-6) suspended in 0.1 ml of EMEM were seeded into 38-mm² wells of 96-well microculture plates (Falcon Plastics, Oxnard, CA) at a density of 2000 cells per well and incubated for 96–120 h with different concentrations of rIL-15 in the absence or presence of IC₅₀ irinotecan (8 μg/ml for CT-26 cells and 2 μg/ml for IEC-6 cells). MTT (1 mg/ml in PBS) was added to the cultures at 0.025 ml/well during the final 2 h of incubation. The medium was removed, and the cells were lysed in DMSO. The conversion of MTT to formazan by metabolically viable cells was monitored by a microtiter plate reader at 570 nm (27). Percentage cytostasis was calculated by the following formula: cytostasis (%) = (B − A)/B × 100, where A is the absorbance of treated cells and B is the absorbance of the control cells.

The significance of the differences in in vitro data was analyzed by the unpaired Student’s t test (two-tailed). The significance of the in vivo differences was analyzed by the Mann-Whitney U test and Student’s t test.

In the first set of experiments, we examined whether the p.o. administration of soluble JBT-3002 could protect mice against lethal diarrhea induced by high-dose irinotecan. Mice were given intrasplenic injections of CT-26 cells on day 0. Groups of mice were then given three daily p.o. administrations of 0.05 μg of JBT-3002 starting on day 3. Seven days after tumor cell injection, groups of mice were given four consecutive daily i.p. injections of 100 mg/kg irinotecan. The mice were euthanized and necropsied when the control mice became moribund (on day 21). The results are shown in Table 1. The size of the local spleen tumors and the incidence and median number of experimental liver metastases were significantly reduced in mice treated with irinotecan. However, these mice had a >15% loss in body weight associated with diarrhea. Seven of these 10 mice died within 10 days after the last injection of irinotecan. At necropsy, we found that the small intestines of these mice were shortened and had thickened walls. The p.o. administration of JBT-3002 without irinotecan significantly reduced the size of spleen tumors and the median number of experimental surface liver metastases, albeit to a lesser degree than that found with irinotecan treatment. The p.o. administration of JBT-3002 prior to i.p. irinotecan produced significant therapeutic effects (P < 0.0001). Most important, this combination therapy was not associated with significant loss in body weight, and none of the mice in this group died.

To examine the architectural changes in the intestinal mucosa, we harvested tissues from the ileum 4 days after the final treatment with irinotecan and processed them for examination by scanning electron microscopy. Fig. 1 shows that irinotecan-treated mice had severe damage of the mucosal surface of the ileum characterized by stubby villi and patches bare of villi. In mice given p.o. JBT-3002 without irinotecan, the villi were taller than normal. In mice treated with JBT-3002 prior to injections of irinotecan, the mucosa was thoroughly covered with villi with an appearance similar to that of control mice.

Tissues from these intestinal samples were processed for histopathological analysis (Fig. 2). H&E staining revealed severe morphological changes consisting of loss of villi, epithelial vacuolation, and inflammatory cell infiltrate in the ileum of mice treated with irinotecan alone. In mice treated with JBT-3002 without irinotecan, the intestinal epithelium exhibited mild hyperplastic changes. Of note, in mice treated with JBT-3002 prior to injections of irinotecan, the morphology of the ileum was similar to that found in normal mice. Enlargement of the crypts and lamina propria was also observed in this group.

Proliferation of epithelial cells was evaluated by immunohistochemical staining for S-phase labeling with anti-BrdUrd antibody (Fig. 2). The number of BrdUrd-positive cells in mice given irinotecan alone (mean, 24 ± 7 cells per 0.25 mm²) was significantly lower than the number in control mice (mean, 78 ± 13 cells per 0.25 mm²; P < 0.01). In mice given JBT-3002 without irinotecan, the number of BrdUrd-positive cells in the crypts was significantly increased (mean, 110 ± 11 cells per 0.25 mm²; P < 0.05). Notably, mice treated with JBT-3002 prior to injections with irinotecan had a nearly 6-fold increase in the number of dividing epithelial cells (mean, 153 ± 20 cells per 0.25 mm²; P < 0.01).

In the next set of experiments, we examined whether the expression of IL-15 correlated with alterations in the mucosal epithelium. Immunohistochemistry, using antibodies against IL-15, depicted only a faint staining in intestinal mucosa of mice given irinotecan alone. In contrast, the ileum of mice treated with JBT-3002 before injections of irinotecan exhibited intense staining with the anti-IL-15 antibody (Fig. 2). The expression of the IL-15 protein was most prominent in cells associated with the lamina propria.

To examine expression of mRNA specific for IL-15, total RNA was isolated from the ileum 4 days after the final treatment with irinotecan and analyzed by reverse transcriptase-PCR. The results, shown in Fig. 3 A, closely agreed with those obtained by immunohistochemistry. A nearly 90% reduction in IL-15 transcripts was observed in the ileum from mice treated with irinotecan alone. The ileum of mice treated with JBT-3002 prior to injections with irinotecan had an 1.8-fold increase in expression of IL-15 mRNA, which decreased by 20% after 7 days (data not shown).

Because both lamina propria mononuclear cells and intestinal epithelial cells can release IL-15 (7), we determined whether exposure of these cells of JBT-3002 would lead to enhanced expression of IL-15. PEMs and IEC-6 cells were cultured for 6 h with medium (control) or medium containing 10 ng/ml JBT-3002. Total RNA was isolated from the PEMs and IEC-6 cultures and the expression of IL-15 mRNA was analyzed by reverse transcriptase-PCR (Fig. 3,B). Activation of PEMs with rIFN-γ (10 units/ml) and LPS (10 ng/ml) for 6 h produced a 2.4- and 7.1-fold increase in expression of IL-15 mRNA, respectively. Treatment of PEMs with JBT-3002 (10 ng/ml) induced a 4.2-fold increased expression of IL-15 mRNA. IEC-6 cells constitutively expressed IL-15 mRNA, but treatment with the stimuli did not enhance expression of the cytokine. The level of gene expression closely correlated with secretion of IL-15 protein as determined by an immunoperoxidase assay (Fig. 3 C).

In the final set of experiments, we examined whether IL-15 can protect intestinal epithelial cells from toxicity produced by irinotecan. IEC-6 cells were cultured for 96–120 h with rIL-15 in the absence or presence of irinotecan at IC₅₀ (2 μg/ml). Cell viability was measured by the MTT assay (27). Cultivation of IEC-6 cells with only rIL-15 (1.5–100 ng/ml) did not promote proliferation (data not shown). The presence of 12 ng/ml IL-15 in medium containing 2 μg/ml irinotecan (IC₅₀) decreased irinotecan-mediated cytostasis (Fig. 4,A). The cultivation of CT-26 cells for 96 h with rIL-15 (1.5–100 ng/ml) alone did not promote proliferative responses (data not shown). In contrast to IEC-6 cells, rIL-15 did not protect CT-26 tumor cells against irinotecan-induced toxicity (Fig. 4 B).

We have recently reported that the p.o administration of JBT-3002, a new synthetic macrophage-activating agent encapsulated in phospholipid liposomes (19, 20, 21), can prevent irinotecan-induced toxicity to the intestinal epithelium of mice, thus allowing irinotecan dose intensification to raise the therapeutic efficacy against liver metastasis produced by CT-26 murine colon carcinoma cells (4). Although efficient in situ activation of macrophages is achieved by the encapsulation of immunomodulators in phospholipid liposomes (28), our recent data show that JBT-3002 suspended in aqueous solution to form micelles produces superior activation of macrophages (21) and that the in vivo use of this micellar formulation of JBT-3002 requires only a 1/20 equivalent dose of liposome-encapsulated JBT-3002.

Topoisomerase I inhibitors such as irinotecan are highly S phase specific; therefore, diarrhea induced by high-dose irinotecan may result from direct toxicity against dividing epithelial cells in the intestinal crypts (29). Indeed, four consecutive daily administrations of 100 mg/kg irinotecan severely disrupted the intestinal architecture, producing a nearly 3-fold decrease in proliferating crypt cells as measured by BrdUrd labeling. In contrast, p.o. administration of JBT-3002 prior to injections with irinotecan increased cell division in the crypts and preserved epithelial organization in the villi. These findings suggest that JBT-3002 can prevent chemotherapy-induced injury of the intestinal mucosa by affecting epithelial stem cell integrity (4).

Recent studies have demonstrated that a variety of growth factors and cytokines regulate the proliferation of intestinal epithelial cells (6, 7). For example, IL-2 has been shown to modulate proliferation of both IEC-6 cells and rat primary intestinal epithelial cells under in vitro conditions (9). IL-15 is a novel cytokine that shares the IL-2 receptor β- and γ-chains for signal transduction and, hence, many of the biological activities of IL-2 (11, 12, 13, 14). In addition, the α-chain specific to the IL-15 receptor transduces IL-15 signals that contribute to epithelial organization by enhancing tight junction formation (15). Unlike IL-2, IL-15 is produced by intestinal epithelial cells and lamina propria macrophages (7). These results show that tissues from the intestines of mice given p.o. JBT-3002 before injections of irinotecan overexpressed the IL-15 gene. Immunohistochemical analyses revealed that lamina propria leukocytes rather than epithelial cells were responsible for this overexpression. These findings were further confirmed by in vitro experiments in which PEMs but not IEC-6 cells exposed to JBT-3002 had evidence of increased expression of IL-15 mRNA and protein. Exposure of macrophages to LPS is known to enhance expression of IL-15 (18). Because JBT-3002 shares many intracellular signaling pathways for cytokine expression with LPS, it is not surprising that exposure of macrophages to JBT-3002 leads to enhanced expression of IL-15 (21). Studies have shown that the binding of nuclear protein to the nuclear factor κB binding site is required for transcriptional activation of the IL-15 gene in LPS-stimulated macrophages (30). Recent data show that administration of human rIL-15 can protect rats from intestinal toxicity produced by 5-fluorouracil (16) or irinotecan (17). Thus, the p.o. administration of JBT-3002, which activates lamina propria macrophages to produce IL-15, may well be involved in the prevention of irinotecan-induced injury to the intestinal mucosa.

Several lines of evidence suggest that intestinal epithelial cells, including IEC-6 cells, possess IL-15 receptor (9, 10, 15). Nevertheless, the mechanism by which IL-15 affects intestinal integrity is still unknown. Treatment of IEC-6 cells with IL-15 at concentrations of 1.5–100 ng/ml did not promote proliferation. However, IL-15 decreased irinotecan-mediated cytotoxicity, suggesting that IL-15 may promote the recovery of epithelial integrity after wounding. IL-15 may function by up-regulating other cytokines such as transforming growth factor-β, which is involved in epithelial cell migration (8). In that case, IL-15 would have an indirect effect on recovery from intestinal injury.

At higher concentrations (250 ng/ml), IL-15 stimulates the proliferation of intestinal epithelial tumor cells, such as Caco-2 (7). In this study, however, IL-15 (tested at 1.5–100 ng/ml) did not stimulate the division of CT-26 colon cancer cells or protect these tumor cells from irinotecan-mediated cytotoxicity. These differences have a tremendous therapeutic benefit, and indeed, treatment of mice with oral JBT-3002 followed by i.p. irinotecan significantly reduced the size of spleen tumors and eradicated liver metastases. IL-15 has also been shown to induce activation of CTLs and natural killer cells (11, 31). JBT-3002 activates macrophages to become tumoricidal and produce inflammatory and immunomodulatory cytokines (4, 19, 20, 21). JBT-3002 is also a potent activator of inducible nitric oxide synthase in macrophages, resulting in production of high concentrations of nitric oxide (19, 20) shown to induce apoptosis in tumor cells to which activated macrophages bind (32, 33). The release of IL-15 by JBT-3002-activated macrophages may, therefore play a role in antitumor activity.

In summary, we have demonstrated that p.o. administrations of JBT-3002 prevents irinotecan-induced dose-limiting intestinal toxicity. One of the mechanisms for this protection may involve the production of endogenous IL-15 by activated macrophages within the lamina propria. The p.o. administration of JBT-3002 therefore allows the use of high-dose irinotecan that could produce superior therapeutic results against colon cancer liver metastases.

We thank Kenneth Dunner, Jr., for expert microscopy, Walter Pagel for critical editorial review, and Lola López for expert preparation of this manuscript.