|
|
||||||||
Published online before print October 4, 2005
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Pediatrics, Case Western Reserve University School of Medicine, Cleveland, Ohio
1 Correspondence: Case Western Reserve University School of Medicine, Department of Pediatrics, Rm. 737, Rainbow Babies and Childrens Hospital, 11100 Euclid Avenue, Cleveland, OH 44106. E-mail: mld19{at}cwru.edu
| ABSTRACT |
|---|
|
|
|---|
and interleukin-6 when cocultured with autologous colon T cells than in cocultures with syngeneic or allogeneic spleen T cells. These data suggest that in the IBD colon, DC-T cell interactions may create conditions with an abundance of proinflammatory cytokines, which favor the inflammatory state.
Key Words: T helper 1 cytokines inflammatory bowel disease intestinal inflammation
| INTRODUCTION |
|---|
|
|
|---|
Small numbers of DC are distributed throughout the body, including many sites in the intestine [13 14 15 ]. In the normal intestine, they sample luminal bacteria for presentation of antigen to T cells [16 , 17 ]. These cells play a significant role in maintaining immune tolerance to food particles, endogenous microbes, and other gut antigens [18 ]. Recent evidence indicates that cytokines produced by DC in the intestine may be important in IBD pathogenesis [1 , 19 ]. To address the problem of the immunological events occurring in the IBD gut, it is essential to create a model that mimics the cellular interactions occurring between cells at this site. In the literature, detailed, functional studies of gut lamina propria DC (LPDC) are lacking, primarily as a result of the difficulty in isolating these cells in sufficient numbers with high purity. In fact, a recent study performed to evaluate the function of colon DC has used limited numbers of cells or semipurified colonic DC populations [20 ]. We have adopted a novel approach to decipher the cellular mechanism of DC-T cell stimulation in the severe combined immunodeficient (SCID) mouse model of IBD. This model of IBD is established by the transfer of CD45RBhi CD4+ syngeneic donor T cells into SCID mouse recipients [21 , 22 ], which subsequently develop colitis in response to endogenous gut flora.
Using freshly isolated and purified colon DC and spleen DC, we investigated the ability of these cells to stimulate T helper cell type 1 (Th1) and Th2 cytokines in coculture with T cells, as well as their potential to induce T cell proliferation. We now demonstrate that the T cell-reconstituted SCID mouse model of colitis yields approximately a tenfold increase in DC numbers in the colon lamina propria in comparison with normal mice. When colon LPDC were used to stimulate autologous colon T cells retrieved from the same tissue of IBD mice, significantly higher interferon-
(IFN-
) and interleukin (IL)-6 levels were observed above the levels released with stimulation of LPDC-allogeneic T cells.
Our data support the hypothesis that the limited numbers of lamina propria mononuclear cells (LPMC) in unmanipulated control mice, combined with the rarity (low frequency) of DC in this population, limit the likelihood of significant DC-T cell interaction in vivo and hence, do not favor IBD pathology. We propose that the increased presence of these cells at a site in which they typically exist at low frequency provides conditions for persistent, robust interaction with T cells, generating substantial amounts of proinflammatory mediators, which disrupt the balance from tolerance to inflammation. The stimulation of colon T cells with colon DC and the cytokines elicited provides a crucial environment favoring the immune pathogenesis of IBD in the colons of mice.
| MATERIALS AND METHODS |
|---|
|
|
|---|
Adoptive transfer of T cells and induction of colitis in SCID mice
Colitis/IBD was induced in SCID mice as described previously [21
, 22
]. Briefly, spleen cells were prepared from donor BALB/c mice and labeled with Thy1 (midiMACS, Miltenyi Biotec, Auburn, CA) magnetic cell sorter (MACS) beads for the enrichment of T cells on MACS columns according to the manufacturers instructions. Recovered T cells were monoclonal antibody (mAb)-stained with CD4 phycoerythrin (PE; L3T4 GK.1.5, PharMingen, San Diego, CA) and CD45RB fluorescein isothiocyanate (FITC; C363.16A, PharMingen) on ice for 30 min. Cells were sorted on a FACSAria sorter (Becton Dickinson, San Jose, CA), collecting the 35% most double-positive T cells. SCID mice were reconstituted with 0.71 x 106 T cells by intravenous injection of cells in 400 µl phosphate-buffered saline (PBS) in the lateral tail vein.
Clinical observation of mice
Mice were weighed on a weekly basis and observed for signs of diarrhea. When soft stools were evident, mice were monitored every other day. Diarrhea was induced in T cell-reconstituted SCID mice within 512 weeks after T cell transfer. On sacrifice, colons and spleens were harvested for isolation of cells. Gross pathology revealed enlarged and thickened colons in colitic mice as compared with BALB/c or SCID mice controls. In addition,
90% of colitic SCID mice exhibited splenomegaly.
Isolation and purification of LPDC
Digestion of colon tissue was performed on pooled colons from colitic SCID mice or BALB/c mice using a modification of an earlier method [23
]. Briefly, each colon was flushed with 30 cc Hanks balanced salt solution (HBSS; Ca2+-, Mg2+-free), cut longitudinally, and the inner surface cleaned with a cell scraper. After mincing with scissors, release of epithelial cells from tissue was performed in HBSS supplemented with 5% fetal bovine serum (FBS; Gibco-BRL, Life Technologies, Grand Island, NY), EDTA (0.05 mM), Hepes buffer (0.6 mM), penicillin/streptomycin/fungizone (1.2%, Cambrex Bioscience, Walkersville, MD), and 15 µg/ml dithiothreitol at 37°C for 15 min. Samples were filtered through a 100-µm filter to remove epithelial cells from the tissue. A further disaggregation of tissue was performed in the HBSS buffer and then in RPMI medium (Gibco-BRL, Life Technologies), supplemented with 10% FBS, penicillin (100 units/ml)/streptomycin (100 µg/ml, Gibco-BRL, Life Technologies), and Hepes buffer, each time incubating tissue for 15 min. After filtering, tissue was digested in a solution of collagenase D (Boehringer Mannheim, Germany, 1.75 mg/ml) and DNase (Boehringer Mannheim, 0.05 mg/ml) in RPMI medium for 50 min at 37°C with intermittent pipetting. The suspension containing released LPMC was filtered, collected, and washed twice in RPMI medium with supplements. CD11c+ LPMC were fractionated with magnetic beads based on CD11c expression, and selection was performed on MACS columns (Miltenyi Biotec). Any residual epithelial cells were excluded from the recovered cell fraction in the final step of purification, in which cells were stained with CD11c PE mAb and sorted on a Becton Dickinson FACSAria to obtain a purity of typically >98% CD11c+ cells.
Isolation of spleen DC
Spleens were harvested from mice, pooled, and a single-cell suspension prepared by conventional means. In addition, to maximize the yield of DC from spleen tissue, the spleen capsules were cut into fine pieces, and collagenase digestion was performed as above for the isolation of LPDC. In some experiments, the entire spleen preparation was incubated for an additional 10 min in collagenase/DNase solution at 37°C to ensure that spleens and colons received enzyme exposure. We did not notice any variations in results, whether or not this additional collagenase treatment of spleens was included. Spleen CD11c+ cells were enriched on MACS columns, followed by cell sorting on a Becton Dickinson FACSAria, as described above for LPMC. The morphology of DC was determined on Giemsa-stained cytocentrifuge preparations.
DC phenotypes
To determine the expression levels of cell-surface antigens on DC, immunostaining was performed using a panel of fluorescent-labeled mAb. Briefly, DC were preincubated with the Fc receptor for immunoglobulin G (IgG)III/II antibody (CD16/CD32, PharMingen) for 10 min to block nonspecific antibody staining. Cells were resuspended at 25 x 105 cells in 100 µl staining buffer (PBS, 1% FBS, 0.1% sodium azide) in flow tubes. Cells were subsequently stained directly with mAb against CD11c PE (HL3) and CD80 (B7-1; 16-10A1), CD86 (B7-2; GL1), CD40 (3/23), CD8a (Ly-2, 53-6.7), CD11b (membrane-activated complex 1), CD4 (L3T4, H129.19), B220 (CD45R; RA3-6B2), or anti-major histocompatibility complex (MHC) class II, I-Ad (AMS-32.1; all FITC-conjugated, PharMingen). The appropriate species-specific IgG isotype controls were included in each experiment. Staining was performed for 30 min at 4°C, and cells were washed twice in staining buffer as before and resuspended in PBS. At least 10,000 events were acquired on a FACScan (Becton Dickinson), and analysis was performed using Cell Quest software.
T cell purification
LPMC were prepared from colons as described above. In experiments where T cells were required from this site, 40% of the LPMC was set aside for DC purification, and 60% was used for T cell purification. T cells were purified from LPMC or from spleen cell populations. In each case, unfractionated spleen cells or LPMC were purified using the CD4 MACS MicroBeads and purification column system (MACS, Miltenyi Biotec). This routinely resulted in a cell enrichment of
95% CD4+ T cells from spleen cells. The purity of T cells from LPMC by this method was less efficient but was improved to 99% by including an additional step of cell labeling with CD4 PE mAb and flow sorting on a Becton Dickinson FACSAria. In experiments where comparisons were made between spleen and colon T cell responses, T cells from both sources were flow-sorted to obtain similar purity.
Stimulation of carboxyfluorescein diacetate succinimidyl ester (CFSE)-labeled T cells with LPDC
T cells from syngeneic BALB/c and allogeneic C57BL6/J mice were purified from spleens using the CD4+ MACS system and were adjusted to a concentration of 7.5 x 106/ml. CFSE [24
, 25
] was added at a final concentration of 0.5 µM in 5% FBS/PBS for 7 min. Excess CFSE dye was removed by washing cells three times in 5% FBS/PBS medium. Cells were adjusted to a concentration of 2 x 106/ml (100 µl vol) in RPMI medium with supplements and cultured with DC (1x106/ml, 100 µl vol) in 96-well flat-bottomed plates in a 5% CO2-humidified incubator for a period of 6 days of CFSE dye incorporation into dividing daughter cells. Cells were harvested, washed twice in flow staining buffer, and mAb-stained with CD4 peridinin chlorophyll protein (PerCP; L3T4; RM4-5). Dead cells were gated out, and greater than 25,000 events were collected in the live gate for analysis using Cell Quest software.
Cytokine determination in supernatants
DC were cultured with T cells as described above to determine cytokine release on stimulation. Cell culture supernatants were centrifuged for the removal of cells and stored at 70°C. Cytokine concentration in supernatants was measured by enzyme-linked immunosorbent assay (ELISA) for IL-10 and IFN-
using paired antibodies (PharMingen). For bioactive IL-12 p70, IL-12 p40/IL-23, and IL-6, Duoset ELISA kits DY419, DY499, and DY406, respectively, were used according to the manufacturers instructions (R&D Systems, Minneapolis, MN).
Statistical analysis
Analysis of data was performed by Students t-test. A P value of <0.05 was indicative of significant differences.
| RESULTS |
|---|
|
|
|---|
Colonic DC from mice with IBD differ in morphology compared with spleen DC
The most frequently used DC identification marker for mouse cells is the CD11c antigen. Light scatter plots of sorted CD11c+ cells from colonic LPMC and spleen cell suspensions from colitic mice revealed different patterns of cellular distribution. Colonic CD11c+ cells are more variable in size and are generally larger than spleen CD11c+ cells (Fig. 1A
and 1B
). Colonic DC had large nuclei, which were, in some cases, irregular-shaped, surrounded by a large volume of cytoplasm, with or without dendrites at their cell surfaces, as demonstrated by visual analysis of Giemsa-stained cells (Fig. 1A
, inset). The morphology of these cells was characteristic of the classic mature and immature myeloid DC [27
, 28
]. Spleen DC were predominantly smaller, and nuclei occupied almost the entire volume of these cells (Fig. 1B
, inset). DC were rarely present in the LPMC preparations of BALB/c and unmanipulated SCID mice, as a result of the rarity of DC in healthy colons (see Fig. 2
), but when observed, they had a similar appearance to those of the colonic DC of colitic mice (data not shown). There were no visual differences between splenic DC of BALB/c and colitic SCID mice (data not shown).
|
|
15% CD11c+ cells (Fig. 2A
). In contrast, BALB/c LPMC typically stained less than 3% of CD11c+ cells (Fig. 2B)
. LPDC of colitic mice expressed high levels of CD80, MHC class II I-Ad, and CD11b (Fig. 2A)
with lower levels of CD8
, CD40, CD86, and B220 (data not shown). mAb staining of CD11c+-sorted LPMC revealed a population of cells, which predominantly expressed CD11b (Fig. 3A
and 3B
), the latter, a marker usually found on myeloid DC [27
, 28
].
|
Stimulation of T cells by colonic DC results in production of Th1 cytokines
The cytokine balance in the intestine is a critical, determining factor between tolerance and immunity. We tested the ability of DC from colitic mice to stimulate spleen T cells by quantifying the cytokines released in DC-T cell cocultures. As the exogenous bacterial antigens that drive T cell-mediated inflammation in IBD are unknown, we performed allogeneic stimulation of T cells, a well-studied system used for investigating the immunostimulatory potential of DC. We focused on a range of proinflammatory cytokines, which are relevant to the IBD process or to inflammation in general, including IL-12, IFN-
, and IL-6 [29
30
31
32
33
]. We also investigated the release of IL-10, a cytokine that ameliorates IBD in this model and is also known to be important in intestinal homeostasis [26
, 34
35
36
]. DC stimulation of allogeneic C57BL6/J CD4+ spleen T cells (B6T) cells in culture resulted in the production of significant amounts of IFN-
, which peaked at Day 6 (Fig. 4A
). There was spontaneous production of IL-6 in DC cultures in the absence of T cells, but in two of four experiments, IL-6 release was enhanced in allogeneic DC-T cell cocultures above basal levels (Fig. 4B)
. IL-12 p40/IL-23 was also produced on allogeneic stimulation but to a lesser extent (Fig. 4C)
. IL-10 was only detected in one of our four coculture experiments and reached a level of 366 pg/ml (data not shown). IL-12 p70 was not released on DC-T cell stimulation (not shown).
|
with both models (Fig. 5A
). Low levels of IL-6 and IL-12 p40/IL-23 were also produced but did not achieve significance (Fig. 5B
and 5C) . In addition, we tested spleen DC for their ability to stimulate T cells, as they displayed a distinct phenotype and morphology from LPDC (Figs. 1
and 3)
and thus, may have the ability to stimulate the release of different levels of cytokines. Spleen DC from colitic mice stimulated the production of higher levels of IFN-
than released in allogeneic and syngeneic LPDC-T cell cocultures (Fig. 5A)
.
|
production by colon T cells of IBD mice
In our prior assays, we observed the ability of colonic DC to stimulate IFN-
production in syngeneic and allogeneic splenic T cell cocultures, the most potent response being noted with allogeneic cells (Figs. 4
and 5)
. When autologous T cells from the diseased colons of sick mice were cocultured with DC obtained from the same tissue, there was a more rapid (not shown) and more robust release of IFN-
(Fig. 6A
). The level of IFN-
released in these autologous cocultures was significantly higher than in colon DC-allogeneic T cell cocultures (P=0.01). Thus, DC from colons of IBD mice are capable of inducing an even greater Th1 response than observed in an allogeneic system. IL-6 production in autologous colonic DC-T cell cocultures was also enhanced significantly above that obtained in colon DC-allogeneic B6T cell cocultures (Fig. 6B)
. No significant increase was observed for IL-12 p40/IL-23 release. The levels of this cytokine were low in all cases. As measured by cytokine production, all T cells used in these studies appeared to be in a nonactivated state, with little or no spontaneous cytokine release (Fig. 6) .
|
|
| DISCUSSION |
|---|
|
|
|---|
Because of the ability of DC to direct T cell cytokine and proliferative responses, the interaction of these cells in the colon can be a determining factor of the magnitude of local inflammation and of disease activity. The relationship of these two cell types in IBD has not been addressed adequately as a result of practical limitations in isolating these two cell populations from the lamina propria. Using the CD45RBhi CD4+ T cell-reconstituted SCID model of colitis, our studies investigated several characteristics of colon DC, including the ability of these cells to induce responses in allogeneic T cells and more specifically, cytokine secretion profiles resulting from autologous colon DC-colon T cell stimulation. These data represent one of the first accounts to solidify the role of DC in promoting IBD pathogenesis.
In normal mice, colon DC may contribute to the regulation of immunologic homeostasis and the maintenance of tolerance in the gut. Limited numbers of LPMC combined with the low percentages of these cells in unmanipulated mice limit DC-T cell interaction and cytokine production and potential adverse pathology in these mice. The recovery of significantly increased numbers of LPMC from the colons of colitic mice in comparison with normal mice is of importance to the disease process, because of the increased likelihood of enhanced cellular interaction, T cell expansion, stimulation of the release of damaging Th1 cytokines, and recruitment of cells. Furthermore, the increased presence of CD11c+ cells to provoke a T cell response in these mice creates an environment in which T cells can elaborate a variety of cytokines and contribute to immune imbalance in the gut.
The DC isolated from the colons of sick mice displayed high levels of CD80 and MHC class II expression, which is consistent with the phenotype of mature/activated DC, which when expressing high levels of costimulatory molecule CD80 and MHC class II I-Ad, may be more efficient at presentation of antigens encountered and may easily trigger T cell activation. Increased costimulatory molecule expression on colon DC has also been observed by others using the SCID mouse model of colitis, as well as in other animal models of colitis and on DC in human IBD [1 , 19 , 40 , 41 ]. In contrast to the colonic DC, which were mostly CD11c+/CD11b+, DC isolated from the spleens of IBD mice were mainly of the CD11c+CD11b subset and were morphologically similar to the plasmacytoid DC subset [42 ]. The obvious difference in morphology and in phenotype of colon and spleen DC may indicate that these cells are of different subsets. Alternatively, colon DC and spleen DC may be of the same lineage but at a different stage of differentiation, because of the microbial antigens to which these cells respond in the spleen or colon.
The allogeneic coculture assay is used widely as a measure of DC stimulatory capacity, particularly where the stimulating antigen is unknown. Stimulation of allogeneic T cells with DC from this IBD model resulted in the release of significantly high levels of IFN-
in supernatants. However, when colon DC were used to stimulate autologous colon T cells, the levels of IFN-
released exceeded those detected in allogeneic cocultures. Furthermore, a robust cytokine response in autologous cocultures occurred as early as Day 1, suggesting that IFN-
may be critical to early inflammation at this site. The potent response in this system may be a result of the presence of microbial antigens that copurify with the colon DC. The presence of high levels of IFN-
has been associated with the production of other Th1 cytokines such as tumor necrosis factor-
and IL-12 and worsening of IBD in mouse colitis [30
, 43
].
IL-6 is also associated with inflammation in some disease models and in human and mouse IBD [29 , 33 ], and we noted significant production of IL-6 in our in vitro coculture autologous model. Interaction of IL-6 with the IL-6 receptor induces phosphorylation of the signal transducer and activator of transcription (STAT)-3 protein, which binds directly to target genes [44 ]. Activation of STAT-3 is associated with severe colitis in mice as with other diseases [32 , 45 ]. Overproduction of IL-6 resulting from colon DC-colon T cell interaction may be a contributing factor to colitis.
IL-23 is a cytokine composed of the IL-12 p40 subunit and a cytokine-like subunit related to IL-12 p35, p19. IL-23, like IL-12, induces proliferation and IFN-
production by human T cells, but unlike IL-12, IL-23 primarily stimulates memory rather than naïve T cells [46
, 47
]. It has also been implicated in multiorgan inflammation, including the intestinal tract [48
]. Recent studies reported that this cytokine was expressed in the terminal ileum of mice but not in the colon [13
]. Colon DC-colon T cell stimulation released low but detectable levels of IL-12 p40/IL-23 in our system.
Several reports indicate that IL-10, a Th2 cytokine, is a characteristic product of intestinal DC from normal mice, where it contributes to intestinal homeostasis. It is associated with the improvement of mouse IBD and in general, with down-regulation of Th1 immune responses [26
, 34
35
36
, 49
]. We did not observe significant levels of IL-10 in DC-T cell cultures, possibly as a result of the fact that the high levels of IFN-
released in these cocultures may contribute to the suppression of IL-10 production. The absence of IL-10 on the interaction of two critical cell types in the IBD colon, combined with the up-regulation of cytokines such as IFN-
, may provide an environment that favors immune reactivity at this site.
Chemoattractants and inflammatory mediators such as IFN-
released by constant stimulation of DC-T cells in the IBD gut may be a cue for the migration of other cells to the site. In addition, as we report, DC possess the ability to induce local stimulation and proliferation of T cells. This too may contribute to the cellular increase and resulting inflammation in the colon of colitic mice.
This is the first report about the ability of colon DC from an inflamed gut to stimulate autologous T cells from the same site and elaborate proinflammatory cytokines. We provide novel evidence that cells isolated from the colons of IBD mice induce a Th1 cytokine response, which is significantly greater than those obtained even in an allogeneic system. Our data suggest that colon DC activity may be crucial to the chronic inflammation observed in IBD. It is not certain at this stage whether cytokines such as IFN-
produced by this mechanism are responsible for initiating T cell proliferation in the gut, but it is most likely a factor in the immune-mediated damage at this site.
Although the presence of DC in spleens was less frequent than in the colons of IBD mice, the cytokine and proliferative responses induced by spleen DC suggest that DC, at sites distant from the gut, may contribute to the systemic component of inflammation in this model. In addition to the morphologic and phenotypic variations of these spleen and colon DC, spleen DC were potent stimulators of T cells, and colon DC were moderately immunostimulatory. This is consistent with recent studies in which semipurified colon DC of normal mice stimulated only a weak, allogeneic, mixed leukocyte reaction in Terasaki plate cultures [20
]. Other investigators reported that spleen DC of normal mice had a greater T cell stimulatory ability than liver DC [50
]. Thus, it is possible that there is an inherent, immunogenic potency in spleen DC in comparison with those from other sites. We have shown too that spleen DC induced higher levels of IFN-
in cocultures than obtained with LPDC-T cell cocultures. Furthermore, our own unpublished observations indicate that spleen DC, when stimulated with certain microbial antigens, release higher levels of IFN-
than colon DC (M. L. Drakes, unpublished observations). Future studies will determine the ability of cells from this systemic site to contribute to the overall disease state.
Earlier investigators have focused on DC-T cell interactions in the lymph nodes of mice as a mechanism of IBD inflammation [51 ]. Using both cell types isolated from the colons, we have highlighted the unique ability of colon DC of colitic mice to stimulate autologous T cells and suggest that in the gut, these cells play a crucial, immunoregulatory role in IBD pathogenesis.
| ACKNOWLEDGEMENTS |
|---|
Received June 24, 2005; revised July 29, 2005; accepted August 4, 2005.
| REFERENCES |
|---|
|
|
|---|
, MIP-3ß, and secondary lymphoid organ chemokine J. Exp. Med. 191,1381-1394
expression by T cells J. Exp. Med. 187,1225-1234
production and type 1 cytokine responses Immunity 4,471-481[CrossRef][Medline]
This article has been cited by other articles:
![]() |
S. Takenaka, E. Safroneeva, Z. Xing, and J. Gauldie Dendritic Cells Derived from Murine Colonic Mucosa Have Unique Functional and Phenotypic Characteristics J. Immunol., June 15, 2007; 178(12): 7984 - 7993. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. L. Drakes, S. J. Czinn, and T. G. Blanchard Regulation of Murine Dendritic Cell Immune Responses by Helicobacter felis Antigen. Infect. Immun., August 1, 2006; 74(8): 4624 - 4633. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |