Expression of Blimp-1 in Dendritic Cells Modulates the Innate Inflammatory Response in Dextran Sodium Sulfate-Induced Colitis
A single nucleotide polymorphism of PRDM1, the gene encoding Blimp-1, is strongly associated with inflammatory bowel disease. Here, we demonstrate that Blimp-1 in CD103+ dendritic cells (DCs) critically contributes to the regulation of macrophage homeostasis in the colon. Dextran sodium sulfate (DSS)-exposed Blimp-1cko mice with a deletion of Blimp-1 in CD103+ DCs and CD11chi macrophages exhibited severe inflammatory symptoms, pronounced weight loss, high mortality, robust infiltration of neutrophils in epithelial regions of the colon, an increased expression of proinflammatory cytokines and a significant decrease in CD103+ DCs in the colon compared with DSS exposed wild-type (WT) mice. Purified colonic macrophages from Blimp-1cko mice expressed increased levels of matrix metalloproteinase 8, 9 and 12 mRNA. WT macrophages cocultured with colonic DCs but not bone marrow-derived DCs from Blimp-1cko produced increased matrix metalloproteinases in an interleukin (IL)-1β- and IL-6-dependent manner. Treatment of Blimp-1cko mice with anti-IL-1β and anti-IL-6 abrogated the exaggerated clinical response. Overall, these data demonstrate that Blimp-1 expression in DCs can alter an innate inflammatory response by modulating the activation of myeloid cells. This is a novel mechanism of contribution of Blimp-1 for the pathogenesis of inflammatory bowel diseases, implicating another therapeutic target for the development of inflammatory bowel disease.
Inflammatory bowel disease (IBD) has a complex pathogenesis dependent on several factors, including genetic susceptibility of the host, commensal bacteria and the host immune system (1). Genome-wide association studies (GWASs) have significantly advanced our understanding of the genetic contribution to IBD, and metaanalyses of GWASs have established >150 susceptibility loci (2, 3, 4). These studies confirmed pathways already identified and also discovered previously unappreciated pathways, raising novel hypotheses about disease pathogenesis (5). However, for most GWAS loci, functional alterations underlying disease susceptibility remain unidentified.
The importance of intestinal microbial flora in disease development in experimental models is being increasingly appreciated. Microbiome-dependent disease development was noted in studies of rodents raised in germ-free conditions or various specific pathogen-free conditions (6,7). Furthermore, different microbiomes can lead to different disease pathogenesis in mouse models of IBD, emphasizing the key role of commensal bacteria in IBD pathogenesis (8). The gut microbiome affects homeostasis or activation of immune cells in the intestine through engagement of pattern recognition receptors, and abnormal expression or activation of pattern recognition receptors can lead to an inflammatory response in the intestine (9, 10, 11, 12). Pattern recognition receptor activation by pathogenic gut microbes often leads to the expression of proinflammatory cytokines, which drive intestinal inflammation (13,14).
Tissue-resident dendritic cells (DCs) and macrophages are key players in controlling intestinal immune responses (15). The complexity of intestinal macrophage and DC subsets is increasingly appreciated; however, the contribution of these populations to intestinal inflammatory conditions is still unclear (16). In addition to the traditional DC markers CD11c and CD103, DCs can be further subdivided into Batf3-dependent CD11blo DCs and Batf3-independent CD11b+ DCs (17). Recently, an additional population of Batf3-independent CD11c+CD11b+ DCs that lack CD103 and CD64 but express chemokine (C-X3-C motif) receptor 1 (CX3CR1) has also been described (18,19). CD103-CX3CR1int DCs are reported to have an immunogenic phenotype (20,21) and to induce the differentiation of TH17 cells (18). Tissue-resident nonmigratory macrophages include a CD11clo and a CD11c+ population, both of which express high levels of F4/80, CD64 and the CX3CR1 chemokine receptor. In the steady state, intestinal DCs and macrophages have been shown to contribute to gut homeostasis through production of interleukin (IL)-10 and induction of regulatory T cells (22,23).
DCs showed the strongest alterations in gene expression among other immune cells, suggesting a role of genetic alterations in IBD (4). PRDM1, the gene encoding B lymphocyte-induced maturation protein-1 (Blimp-1), has been demonstrated in GWASs to have IBD susceptibility single nucleotide polymorphisms (SNPs). Moreover, a recent exome sequencing study identified PRDM1 rare variants that are associated with IBD (24).
In this study, we investigated a pathologic function of Blimp-1 in DCs by using a CD11c-driven Blimp-1 knockout (Blimp-1cko). Blimp-1cko mice exhibited an exacerbated phenotype with a high mortality after dextran sodium sulfate (DSS)-induced colitis. After DSS exposure, colonic Blimp-1ko DCs exhibited an increased production of IL-1β and IL-6, and colonic macrophages exhibited a higher expression of matrix metalloproteinases (MMPs). Increased expression of IL-1β and IL-6 by Blimp-1ko DCs was responsible for the MMP induction in macrophages. Blockade of both IL-1β and IL-6 during DSS exposure mitigated the exacerbated IBD phenotype and mortality in Blimp-1cko mice. Finally, blockade of MMP by an MMP inhibitor reduced colitis development, supporting the hypothesis that increased expression of MMP is responsible for the exacerbated colitis in DCBlimp-1ko mice.
Material and Methods
Human Samples Preparation and In Vitro Differentiation of Monocyte-Derived DCs
Healthy PRDM1 IBD SNP rs6911490 risk allele carriers and nonrisk allele-carrying controls were identified from the genotype and phenotype (GaP) registry at The Feinstein Institute for Medical Research (Manhasset, NY, USA). Fresh blood was collected and total peripheral blood mononuclear cells were purified by gradient centrifugation with Ficoll-Paque (GE Healthcare Bio-Sciences, Piscataway, NJ, USA) at 260g for 20 min without break. Peripheral blood mononuclear cells were collected from the middle layer and washed with Hanks balanced salt solution for three times. CD14+ monocytes were purified with an EasySep Kit (Stem Cell Technologies, Vancouver, Canada), and the purity of CD14+ cells was confirmed by flow cytometry. Purified monocytes were cultured in RPMI medium with 10% fetal bovine serum, penicillin-streptomycin, l-glutamine, 106 U/mL granulocyte-macrophage colony-stimulating factor (GM-CSF) (Peprotech, Rocky Hill, NJ, USA) and 200 U/mL IL-4 (Peprotech) for 7 d to obtain monocyte-derived DCs (Mo-DCs). These studies were performed according to an institutional review board (IRB)-approved protocol.
Mice and Cell Lines
Blimp-1cko mice on a BALB/C background for over 10 generations were bred in The Feinstein Institute for Medical Research animal facility and wildtype (WT) BALB/c mice were purchased from The Jackson Laboratory (Bar Harbor, ME, USA). To generate CX3CR1-GFP on Blimp-1cko background, Blimp-1cko mice were bred with CX3CR1-GFP (The Jackson Laboratory). All mice strains were maintained in a specific pathogen-free facility at The Feinstein Institute for Medical Research.
All the experiments conducted in this study followed the guidelines in the Guide for the Care and Use of Laboratory Animals by the National Institutes of Health (25). The protocol was approved by the committee on the Ethics of Animal Welfare of The Feinstein Institute for Medical Research (protocol number 2009-048). For the survival study, we monitored mice twice a day throughout the experiments to measure weight and monitor development of sickness. We euthanized mice when the weight loss reached >25% of the original weight by CO2. Detailed experimental design and protocol can be found in Supplementary Data.
DSS-Induced Colitis and Preparation of Intestinal Cell Suspensions
The 7- to 10-wk-old mice were administered 4% DSS (36,000–50,000 molecular weight; MP Biomedical LLC, Solon, OH, USA) in drinking water for 7 consecutive days. After 7 d, DSS water was replaced with fresh water to allow the mice to recover. To measure colitis, mice were weighed and stool was collected every other day; the stool and rectum of the mice were also examined for blood. For inhibition assays, anti-IL-1β and IL-6 (100 µg each) together or separately (eBioscience, San Diego, CA, USA), anti-tumor necrosis factor (TNF)α (100 µg) (eBioscience), doxycycline (50 µg) (Sigma-Aldrich, St. Louis, MO, USA) or control IgG (100 µg) were administered by intraperitoneal injection on d 0, 3 and 5 of DSS treatment. To calculate disease activity index (DAI), colitogenic phenotype was investigated and scored: weight loss (0: 0% loss, 1: 1–5% loss, 2: 5–10% loss, 3: 10–20% loss, and 4: >20% loss from baseline); stool consistency (0: normal, 2: pasty/semiformed stool that did not adhere to the anus, and 4: liquid); and bleeding (0: negative hemoccult test, 2: positive hemoccult test, and 4: gross bleeding); DAI = scores of (weight loss + stool consistency + bleeding)/3.
On d 3 or 7 of DSS treatment, mice were sacrificed and their mesenteric lymph nodes (mLNs), small intestines and colons were removed. Colon length and weight was measured. Peyer patches were removed from the intestine. Intestines were cut open longitudinally along their entire length, cleaned and washed by vortexing three times with complete medium (2% fetal bovine serum in Hanks balanced salt solution), and incubated in complete medium containing 1 mmol/L DL-dithiothreitol solution (Sigma-Aldrich) for 20 min while shaking at 37°C and then incubated in 5 mmol/L ethylenediaminetetraacetic acid (Sigma-Aldrich) in complete medium for 50 min while shaking at 37°C. Afterward, the intestines were washed once with complete medium and incubated in 0.3 mg/mL collagenase (C5138; Sigma-Aldrich) in complete medium for 45 min at 37°C.
Antibodies and Flow Cytometry Analysis of Colon Myeloid Subsets
Anti-mouse F4/80-FITC, CD103-PE, CD11b-PE-Cy7, CD11c-APC, CD45-APC-Cy7, MHCII-PB, GR-1-PE-Cy5 and Siglec-H-efluor450 were all purchased from eBioscience. Anti-mouse CD4-FITC was purchased from BD Biosciences (San Jose, CA, USA).
The colonic, small intestine and mLN cells were stained and analyzed by flow cytometry (LSRII; BD Biosciences) or sorted on a cell sorter (FACSAria; BD Biosciences) to look at different populations of myeloid cells as well as isolate cells for RNA extraction and quantitative polymerase chain reaction (qPCR).
TH17 Cell Differentiation
A 48-well plate was coated with 5 µg/mL anti-CD3ε antibody (145-2C11). The next day, total CD4+ T cells (CD45+TCRβ+CD4+) were purified from colon and incubated under TH17 differentiation conditions; 2 µg/mL anti-CD28 (37.51), 10 µg/mL anti-CD4, 10 µg/mL anti-IL-2, 10 µg/mL anti-interferon (IFN)-γ, 5 ng/mL transforming growth factor (TGF)-β and 20 ng/mL IL-6. All the antibodies were purchased from BD Biosciences and recombinant cytokines were purchased from Peprotech. On d 4, cells were restimulated with 100 ng/mL phorbol myristate acetate, 1 µg/mL ionomycin and 20 µg/mL brefeldin A for 6 h.
DCs were prepared either from bone marrow-derived DCs (BM-DCs) differentiated with Flt3L (Peprotech) or purified colonic DCs and stimulated with lipopolysaccharide (LPS) (100 ng/mL) or muramyl dipeptide (MDP) (100 ng/mL) for 24 h. After stimulation, DCs were washed thoroughly and cocultured with macrophages at a 1:5 (DC:macrophage) ratio for 3 d. After coculture, macrophages were purified by using F4/80 staining. A total of 10 µg/mL anti-IL-1β and anti-IL-6 or control IgG were used for the cytokine-blocking experiments.
Gene Expression Analysis
Total RNA was extracted by an RNeasy kit (Qiagen, Valencia, CA, USA), and total RNA was quantified by a Nanodrop™ spectrophotometer (Thermo Scientific [Thermo Fisher Scientific Inc., Waltham, MA, USA]). cDNA was synthesized with iScript™ (Bio-Rad, Hercules, CA, USA). Amplification was performed with a LightCycler 480 (Roche, Indianapolis, IN, USA). Gene-specific primers were purchased from TaqMan gene expression assay (Invitrogen [Thermo Fisher Scientific]), and relative expression of each gene was quantified by using HPRT or Polr2a for normalization.
Cytokines secreted from either human Mo-DCs or mouse colonic DCs were measured by multiplex panel designed and purchased from Meso Scale Discovery (MSD, Rockville, MD, USA) or by enzyme-linked assay (ELISA) (BD Biosciences).
To measure secreted MMP protein, colon macrophages were purified by cell sorter and 2 × 105 cells were cultured in total 200 µL DC-conditioned medium for 72 h. Supernatant was collected and macrophage cell lysate was prepared in 40 µL RIPA buffer with complete protease inhibitor (Roche). Antibodies for MMP-12 (ab25897) and MMP-8 (ab81286) (Abcam, Cambridge, MA, USA) was diluted in 0.5% milk in Tris-buffered saline with 0.05% Tween 20 (TBST) and applied to the membrane for overnight rocking at 4°C. The next day, the membrane was washed with TBST three times, and horseradish peroxidase-conjugated goat anti-rabbit antibody was applied. Chemiluminescent light was exposed to X-ray film and developed.
One-way analysis of variance was applied to determine statistics, and p values <0.05 were considered significant.
All supplementary materials are available online at https://doi.org/www.molmed.org .
Low Blimp-1 Expression and Increased Proinflammatory Cytokine Expression in Mo-DCs from PRDM1 IBD SNP Carriers
Increased Susceptibility of Blimp-1cko Mice to DSS-Induced Acute Colitis
DSS-Induced Mortality Is Not due to Bacterial Spreading or a Different Microbial Community
Appropriate induction of IgA and IgM in mucosal lymphoid tissue is critical for controlling bacterial spreading, and defects can lead to a high mortality in response to DSS administration, as demonstrated in a MyD88ko mouse (27). Therefore, we compared fecal immunoglobulin (Ig) levels and commensal bacterial spreading to peripheral organs in WT and Blimp-1cko mice. There was no significant difference in the level of IgG, IgM or IgA in fecal samples on d 7 of DSS administration (data not shown). In addition to the normal immunoglobulin level, no bacteria were detected in peripheral tissues, such as liver, mLNs and blood of either strain (data not shown). We also wanted to investigate whether a difference in the commensal bacterial community between WT and Blimp-1cko mice might account for the difference in colitis severity. It is known that the bacterial community can affect the development of colitis in this model (8). Fecal pellets were collected from individual WT and Blimp-1cko mice, and the composition of the bacterial community was analyzed by 16s RNA sequencing. There was no significant difference between WT and Blimp-1cko mice. Both groups of mice exhibit one dominant genus, Clostridium (∼80% of classified bacteria), followed by others, Corprococcus and Dehalobacterium (∼5 and 3% each). These data suggest that DSS-induced mortality in DCBlimp-1ko mice is due to increased inflammation in the intestine and irreversible tissue damage rather than systemic bacteremia or a difference in the composition of commensal bacteria.
Increased Expression of Inflammatory Cytokines in Blimp-1cko Mice during DSS Colitis
To identify which myeloid cell population was responsible for the increased proinflammatory cytokine expression, CD103+ colonic DCs and macrophages were isolated on d 3 of DSS treatment and cytokine expression was measured. The difference between WT and Blimp-1cko mice was attributable to DCs rather than macrophages (Figure 4C).
Alterations in Myeloid Cell Populations in Blimp-1cko Mice during DSS-Induced Colitis
An increased expression of proinflammatory cytokines can be due either to their increased expression at the single cell level or to the expansion of cytokine-producing DCs in Blimp-1cko mice. The frequency of DCs and macrophages was calculated as the number of DCs or macrophages normalized to the number of total myeloid cells. No significant difference was observed in either the basal DC or macrophage populations between WT mice and Blimp-1cko mice (data not shown). However, at d 3 of DSS exposure, there was a decrease in DCs and an increase in macrophages in the colon (Supplementary Figure S1A) and an increase in macrophages in mLNs (Supplementary Figure S1B) of Blimp-1cko mice compared with WT mice. This difference was no longer observed at d 7 of DSS treatment (Supplementary Figure S1C). Thus, alteration in infiltrating cells was present only at an early time point of DSS treatment and was an intestine-restricted phenotype. The increased percentage of macrophages might be due to recruitment from the periphery or proliferation of colonic macrophages in response to inflammatory factors presumably derived from DCs.
Increased Expression of MMPs by Macrophages in the Colon
CD103+ Colonic DCs Switch Residential Macrophages into Tissue-Destructive Macrophages
Blocking of IL-1β/IL-6 Reversed the Colitis Phenotypes in Blimp-1cko Mice
DSS-induced colitis is usually a self-limited condition in mice, characterized by acute tissue inflammation in the colon, and initiated by the innate immune system. Our present study highlights a novel role for Blimp-1 in human Mo-DCs and mouse CD103+ DCs in the regulation of the inflammatory response in the gut. Mo-DCs carrying the PRDM1 IBD risk allele exhibited a low level of Blimp-1 and increased production of proinflammatory cytokines on simulation with LPS. Blimp-1 deficiency in CD103+ DCs rendered mice more susceptible to DSS-induced colitis with an increased mortality. Intestinal Blimp-1ko DCs secrete increased levels of proinflammatory cytokines, IL-1β and IL-6, presumably in response to commensal bacteria, resulting in the enhanced activation of macrophages in the colon. Macrophages of Blimp-1cko mice are induced during DSS colitis to express higher levels of MMPs with faster kinetics of expression. Inhibition of IL-1β and IL-6 or MMPs can relieve colitis symptoms, demonstrating that proinflammatory cytokines and MMPs from innate immune cells are a major mechanism of pathology in this mouse model of colitis.
In previous reports, Blimp-1 deficiency was shown to function in different cell types to cause inflammatory diseases. Mice with a T-cell-specific Blimp-1 deletion develop intestinal inflammation (34), probably mediated by a lack of functional T regulatory cells. More recently, Blimp-1 in T cells has been shown to limit Th17 differentiation, a T-cell subset that contributes to intestinal inflammation (35). Mice with an epidermal-specific deletion of Blimp-1 develop chronic skin inflammation with a higher level of granulocyte colony-stimulating factor (G-CSF) and enhanced myelopoiesis (36). This study identifies a protective function of Blimp-1 in innate immune cells and reveals how this protects against intestinal inflammation.
It is well accepted that in the steady state, intestinal myeloid cells are tolerogenic and refractory to stimulation with TLR agonists in contrast to myeloid cells in the blood or in peripheral lymphoid organs (37); therefore, uncontrolled cytokine production by myeloid cells contributes to the pathogenesis of colitis (38,39). Proinflammatory cytokines, mostly secreted from DCs not macrophages, were highly expressed in colonic myeloid cells from Blimp-1cko mice compared with WT mice after DSS exposure, directly regulating the expression of MMPS from macrophages and subsequently increasing tissue damage during the development of colitis. This regulation appears to be tissue specific because the increase in MMPs in macrophages is induced by activated colonic DCs but not BM-DCs.
MMPs are implicated in tissue destruction during inflammation (30). There is a report of increased expression of MMP-1 and MMP-2 in biopsies of patients with ulcerative colitis (40). Moreover, MMP expression correlates with regions of mucosal loss in IBD patients (41). We did not observe significant changes in MMP-1 or MMP-2 (data not shown). Instead, there was an increased expression of MMP-8, -9 and -12 in macrophages isolated from Blimp-1cko mice compared with macrophages from WT mice subjected to DSS colitis. Each MMP has a specific set of target molecules. MMP-8 and MMP-12 can bind to TNFα and chemokines, respectively, and convert the inactive forms into active forms. Therefore, increased expression of MMPs can increase tissue damage directly by matrix degradation and indirectly by activation of proinflammatory cytokines and chemokines, initiating an inflammatory cascade in the mucosa. In fact, the increased infiltration of neutrophils observed in the colon in Blimp-1cko mice may reflect an increased secretion of neutrophil-attracting chemokines from Blimp-1ko DCs and MMPs. Although we observed that an increased secretion of chemokine (C-X-C motif) ligand 1 (CXCL1) in LPS stimulated Blimp-1-deficient DCs in spleen, we did not find the difference in colonic DCs before and after DSS exposure (data not shown). It is possible that the increased MMP-12 cleaves the inactive form of CXCL1 to convert it into an active form of CXCL1 (42) or MMP-8/9 cleaves collagen generating the proline-glycine-proline peptide, which has chemotactic effects on neutrophil (43). Although neutrophil migration is beneficial for killing bacteria, it is presumed that persistent or excessive neutrophil infiltration causes tissue damage, and blocking of neutrophil infiltration can be beneficial in DSS-mediated colitis (44,45). Therefore, MMP-activated chemokines may be one component of the pathogenic inflammatory cascade in Blimp-1cko mice.
While a number of studies have demonstrated the direct activation of adaptive immune cells by DCs, whether DCs can directly regulate innate immune cells has not been studied. The observation we made in this study (that CD103+ colonic DCs upregulate MMP expression in macrophages) is therefore novel. This regulatory activity is tissue specific and agonist dependent. Compared to WT DCs, Blimp-1-deficient DCs secrete increased levels of IL-1β and IL-6 after stimulation with MDP, presumably through activation of the MDP receptor NOD2. We have shown that these cytokines are major contributors to MMP expression in macrophages and to susceptibility to DSS, since administration of neutralizing antibodies to IL-1β and IL-6 improved clinical symptoms of colitis and diminished MMP induction. Currently, we do not know how Blimp-1 deficiency positively regulates MDP-mediated NOD2 activation in colonic DCs. NOD2 is a member of the NLR (NOD-like receptor) family of proteins, which regulate nuclear factor (NF)-κB activation and a subsequent inflammatory cascade. A recent study showed that Blimp-1 can negatively regulate NLRP12/Monarch-1, another member of the NLR family (46). We did not find any report of Blimp-1 regulating NOD2 or NOD2 downstream molecules, but perhaps NOD2 levels are increased in Blimp-1ko DCs. We are currently investigating this possibility. It is also possible that expression of IL-32 is increased in the absence of Blimp-1, since IL-32 synergizes with NOD1/2 ligands to induce IL-1β and IL-6 production (47).
The results of the present study provide several important insights into the function of DCs during intestinal inflammation. First, DC-restricted alterations can increase susceptibility to intestinal inflammation. Blimp-1 deficiency in DCs induces enhanced production of IL-1β and IL-6 on MDP stimulation; this phenotype is present only in colonic DCs but not in BM-DCs. Second, this enhanced cytokine production directly regulates MMP expression in macrophages. Direct cross-talk between DCs and macrophages has not been previously reported or shown to contribute to IBD pathogenesis. The hyperactivated phenotype of intestinal DCs observed in our study is one of the consistently observed phenotypes of DCs from IBD patients (48), suggesting that the Blimp-1 colitis risk allele may exhibit decreased expression in intestinal DCs, a new molecular pathway by lack of Blimp-1 in colitis pathogenesis. We believe that the current study suggests a novel mechanism for PRDM1-determined risk in initiation or disease progression in human IBD and demonstrates that DCs not only modify the function of cells within the adaptive immune response but also modify function of innate immune cells.
The authors declare that they have no competing interests as defined by Molecular Medicine, or other interests that might be perceived to influence the results and discussion reported in this paper.
We thank M Bogunovic for teaching the intestinal myeloid cell isolation technique and G Honig for helping with the assessment of systemic bacteremia. We thank H Borrero and C Colon at the Flow Cytometry core facility.
- 25.Committee for the Update of the Guide for the Care and Use of Laboratory Animals, Institute for Laboratory Animal Research, Division on Earth and Life Studies, National Research Council of the National Academies. (2011) Guide for the Care and Use of Laboratory Animals. 8th edition. Washington (DC): National Academies Press.Google Scholar
- 35.Bankoti R, et al. (2013) B-lymphocyte induced maturation protein 1 (Blimp-1) is required to limit the number of IL17A-producing CD4+ T cells in vivo. (P1139). 190(Meeting Abstracts 1 Suppl):50.12.Google Scholar
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, and provide a link to the Creative Commons license. You do not have permission under this license to share adapted material derived from this article or parts of it.
The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
To view a copy of this license, visit (https://doi.org/creativecommons.org/licenses/by-nc-nd/4.0/)