|Home | About | Journals | Submit | Contact Us | Français|
Recent genome-wide association studies of pediatric IBD have implicated the 17q12 loci, which contains the eosinophil specific chemokine gene CCL11, with early-onset IBD susceptibility. In the present study, we employed a murine model of experimental colitis to define the molecular pathways that regulate CCL11 expression in the chronic intestinal inflammation and pathophysiology of experimental colitis. Bone marrow chimera experiments showed that hematopoietic cell-derived CCL11 is sufficient for CCL11-mediated colonic eosinophilic inflammation. We show that DSS treatment promotes the recruitment of F4/80+CD11b+CCR2+Ly6Chigh inflammatory monocytes into the colon. F4/80+CD11b+CCR2+Ly6Chigh monocytes express CCL11, and their recruitment positively correlated with colonic eosinophilic inflammation. Phenotypic analysis of purified Ly6Chigh intestinal inflammatory MΦs revealed that these cells express both M1- and M2-associated genes, including Il6, Ccl4 and Cxcl2, and Arg1, Chi3l3, Ccl11 and IL-10, respectively. Attenuation of DSS-induced F4/80+CD11b+CCR2+Ly6Chigh monocyte recruitment to the colon in CCR2−/− mice was associated with decreased colonic CCL11 expression, eosinophilic inflammation and DSS-induced histopathology. These studies identify a mechanism for DSS-induced colonic eosinophilia mediated by Ly6ChighCCR2+ inflammatory monocyte/MΦ-derived CCL11.
The inflammatory bowel diseases (IBD) Crohn’s disease (CD) and ulcerative (UC) colitis are chronic relapsing gastrointestinal (GI) inflammatory diseases that cause substantial morbidity and decreased quality of life. Over 30 years ago, Rutgeerts and colleagues observed that apthous ulcers containing an eosinophilic infiltrate and blunted villi were the earliest endoscopic sign of recurrence in the neoterminal ileum and anastomosis following surgical resection for CD (1). Eosinophils usually represent only a small percentage of the infiltrating leukocytes (2, 3), but their level has been proposed to be a negative prognostic indicator (3, 4). Notably, elevated fecal concentrations of the eosinophil derived granule proteins ECP and EPO were associated with clinical relapse within 3 months in CD (5). Since these initial studies there have been a number of studies suggesting eosinophil involvement in IBD. Elevated levels of eosinophils have been observed in colonic biopsy samples from UC patients and increased numbers of this cell and eosinophil-derived granule proteins MBP, ECP, EPO, and EDN have been shown to correlate with morphological changes to the GI tract, disease severity and GI dysfunction (5–10). Increased numbers of tissue eosinophils with ultrastructural evidence of activation has also been observed in patients with CD (11–13). Consistent with this clinical observation, dextran sodium sulfate (DSS)-induced histopathology is attenuated in mice deficient in eosinophils (14–16).
Genome-wide association studies of pediatric and adult IBD have revealed a number of IBD susceptibility genes associated with innate (CARD15, ATG16L1 and IRGM) and adaptive (IL23R, IL10, IL12B and STAT-3) immunity. Furthermore, a recent investigation identified a significant association between the C-C motif chemokine cluster on 17q12 loci which contains the eosinophil-specific chemokine gene CCL11 and early-onset CD (17). CCL11 is a member of the CC chemokine family (18) and is a relatively potent and specific eosinophil chemoattractant (19–21). CCL11 is constitutively expressed in a variety of tissues that contain eosinophils such as the GI tract and thymus (22–25). Genetic deletion of CCL11 abrogates eosinophil recruitment during eosinophil-associated pulmonary and GI disease, suggesting an important role for CCL11 in eosinophil trafficking during disease (26–29). Consistent with this, clinical investigations by us and others demonstrate increased CCL11 mRNA levels in sputum and intestinal biopsy samples from asthmatic and eosinophilic GI disorder patients. Importantly, the levels of CCL11 positively correlated with tissue eosinophilia (14, 25, 26).
While a link between CCL11 and eosinophils in IBD has been established, the cellular source of CCL11 and molecular regulation of CCL11 expression in experimental colitis is not yet delineated. In the present study, we employed a model of DSS-induced colitis, which features a pronounced CCL11-dependent eosinophilic inflammation, to decipher the molecular regulation of CCL11 in experimental colitis. Performing bone marrow (BM) chimera experiments we show that hematopoietic cell-derived CCL11 is required for DSS-induced colonic eosinophilic inflammation. We show that DSS exposure promotes the recruitment of F4/80+CD11b+CCR2+Ly6Chigh inflammatory monocytes to the colon and that F4/80+CD11b+CCR2+Ly6Chigh colonic monocyte/macrophages (MΦs) positively correlate with colonic eosinophilic inflammation. Ablation of DSS-induced F4/80+CD11b+CCR2+Ly6Chigh monocyte recruitment was associated with decreased intestinal CCL11 expression, colonic eosinophilic inflammation and DSS-induced histopathology. These studies demonstrate that inflammatory monocyte/MΦ-derived CCL11 drives colonic eosinophilic inflammation in experimental colitis.
Male and female, 6- to 8-week-old strain-, age- and weight-matched CCR2−/− (C57BL/6), CCL2−/− (C57BL/6) (Jackson Laboratories, Bar Harbor, ME), C57BL/6, BALB/c, and CCL11−/− (BALB/c) (30), CX3CR1GFP/+ (Jackson Laboratories, Bar Harbor, ME) and Nzeg-eGFP (31) were used. All mice were housed under specific pathogen-free conditions and treated according to institutional guidelines.
DSS (ICN chemicals 40–45 kDa) was administered in the drinking water as a 2.5%–5% (wt/vol) solution for up to 8 days. Disease monitoring and histopathologic changes in the colon were scored as previously described (14).
Immunofluorescence analysis was performed as previously described (14, 32). In brief, frozen sections were fixed in 10% acetone for 10 min, rinsed in PBS, blocked with 3% goat serum/PBS for 2 h at room temperature (RT) and incubated with primary Ab rat anti-mouse F4/80 (5 µg/ml; Ebioscience; San Diego, CA) in 3% normal goat serum/PBS. Sections were incubated with isotype control alone in place of primary Ab as a negative control. After an overnight incubation at 4°C, sections were washed with 0.1% BSA, 0.05% Tween/PBS and incubated with goat anti-rat Alexa Fluor 594 (Invitrogen; Carlsbad, CA) for 2 h at RT. Slides were washed in PBS and counterstained with DAPI/Supermount G solution (Southern Biotech; Birmingham, AL). Images were captured using a Zeiss microscope fitted with Zeiss UPlanApo lenses (×10, ×20 and ×40 magnification) and an AxioCam MRc camera and analyzed with Axioviewer version 3.1 image analysis software (Carl Zeiss; Jena, Germany). Post-acquisition processing (brightness, opacity, contrast and color balance) was applied to the entire image and accurately reflects that of the original.
CCL11, IL-6 and TNF-α levels were measured in the supernatants using the ELISA Duo-Set kit according to the manufacturer's instructions (R&D Systems, Minneapolis, Minn).
Eosinophil levels were quantified by anti-MBP immunohistochemistry as previously described (33).
Colons were excised, flushed with PBS with gentamycin (20ug/mL) and opened along a longitudinal axis. Thereafter, 3-mm2 punch biopsies were excised and incubated for 24 h in 24-well plate with RPMI 1640 supplemented with 10% FCS and antibiotics. Supernatants were collected and kept at −20°C until assessed for cytokines/chemokines by ELISA.
Mouse Hprt, Ccl11, Retlna, Chi3l3, Arg1, Trem1, Ccl22, Ccl17, Cxcl2, Ccl4, Cxcl10, Pdgfb, Il1b, Tnf, Il6, Ccl3 and Il10 mRNA were quantified by real-time PCR as previously described (34). In brief, the RNA samples (1 ug) were subjected to reverse transcription analysis using SuperScript II reverse transcriptase (Invitrogen; Carlsbad, CA) according to the manufacturer’s instructions and quantified using the iQ5 multicolor real-time PCR detection system (Bio-Rad Laboratories; Hercules, CA) with iQ5 software V2.0 and LightCycler FastStart DNA Master SYBR Green I. Primer sets are listed in Supplemental table 1. Gene expression was determined as relative expression on a linear curve based on a gel-extracted standard and was normalized to HPRT amplified from the same cDNA mix. Results were expressed as gene of interest/HPRT ratio.
MΦ populations from the colons of CX3CR1eGFP/+ (C57BL/6) were isolated as previously described (14). In brief, the colon segment of the GI tract was removed and flushed with 20 ml of Ca2+- and Mg2+-free HBSS (CMF-HBSS). The colon was cut longitudinally, placed in CMF-HBSS containing 10% FBS / 5 mM EDTA / 25 mM HEPES and shaken vigorously at 37°C for 30 min. The tissue was cut into 1-cm segments and incubated in digestion buffer containing 2.4 mg/ml collagenase A (Roche Diagnostics; Indianapolis, IN) and 0.2 mg/mL DNase I (Roche; Indianapolis, IN) in RPMI 1640 for 45 min on a shaker at 37°C. Following incubation, the cell aggregates were dissociated by filtering thorough a 19-gauge needle and 70-µm filter and centrifuged at 1200 rpm for 20 min at 4°C. The supernatant was decanted and the cell pellet resuspended in 1% FBS / 5 mM EDTA / PBS, and cells were incubated for 30 min with biotinylated rat anti-mouse CD11b (1 ug/1×106 cells; BD Pharmingen; San Jose, CA) at 4°C. Cells were subsequently incubated with anti-biotin microbeads (Miltenyi; Auburn, CA) for 15 min at 10°C and purified by LS MACS column by positive selection as described by the manufacturer. In brief, 1 mL of cell suspension was added to the LS column, and the column was washed 3 times with 3 mL of 5 mM EDTA / 1%FBS / PBS. CD11b+ cells were removed from the column using a plunger. After washing, CD11b+-selected cells were labeled with rat anti-mouse Ly6C-Alexa-647 (AbD Serotec; Raleigh, NC) and immediately sorted using a FACSAria cell sorter (BD Biosciences; San Jose, CA) for CX3CR1-eGFP and Ly6C. Purity of CX3CR1lowLy6Chigh cells was >95% as assessed by flow cytometry. RNA was isolated using the Qiagen RNeasy micro kit for cDNA synthesis and RT PCR analysis as described above.
Peripheral blood was collected in K2EDTA tubes, and red blood cells were lysed using Red Blood Cell Lysing Buffer (Sigma; St. Louis, MO). Monocytes were enriched using the StemCell Technologies monocyte enrichment kit following the manufacturer’s protocol. Purity was assessed by flow cytometry at >80% F4/80+CD11b+. RNA was isolated using the Qiagen RNeasy micro kit, and cDNA was generated for RT PCR analysis as described above.
Single-cell suspensions were washed with FACS buffer (PBS / 1% FCS) and incubated with combinations of the following Abs: PE anti-mouse F4/80 (clone CI:A3-1; Serotec; Raleigh, NC), PE-Cy7 anti-mouse CD11b (clone M1/70; BD Pharmingen; San Jose, CA), Alexafluor-647 anti-mouse Ly6C (clone ER-MP20; AbD Serotec; Raleigh, NC), FITC anti-mouse CD206 (MR5D3; BioLegend; San Diego, CA), APC anti-mouse CD11c (clone HL3; BD Pharmingen; San Jose, CA), FITC anti-mouse PDL1 (clone MIH6; AbD Serotec; Raleigh, NC), goat anti-mouse CCR2 (polyclonal; GeneTex, Inc; Irvine, CA) followed by donkey anti-goat FITC (Jackson Immunoresearch; West Grove, PA), biotinylated anti-mouse TLR2 (clone mT2.7; Ebioscience; San Diego, CA) followed by Strepavidin-FITC (BD Pharmingen; San Jose, CA), APC anti-mouse CCR3 (clone 83101; R&D Systems; Minneapolis, MN) and PE anti-mouse Siglec F (clone E50-2440; BD Pharmingen; San Jose, CA). The following Abs were used as appropriate isotype controls: FITC rat IgG2a (clone B39-4; BD Pharmingen; San Jose, CA), PE rat IgG2a (clone 53-6.7; BD Pharmingen; San Jose, CA), PE-Cy7 rat IgG2b (clone DTA-1; BD Pharmingen; San Jose, CA) and AlexaFluor 647 rat IgG2a (clone R35-95 BD Pharmingen; San Jose, CA). Cells were analyzed on FACSCalibur (BD Immunocytometry Systems; San Jose, CA), and analysis was performed using Flow Jo software (Tree Star; Ashland, OR).
BM was isolated from Nzeg-eGFP (BALB/c) (31) and CCL11−/− mice (BALB/c). Lethally irradiated [2 doses of 137Cs (475 and 475 rads, 3 hs apart] wild type (WT) or CCL11−/− BALB/c recipients were injected i.v. with 5–10×106 BM cells / mouse. Engraftment was checked by eGFP+ (donor) / eGFP− (recipient) cells from the peripheral blood, mesenteric lymph node (MLN) and colon by flow cytometry. Seven to 8 weeks post irradiation, the mice were administered 5% DSS for 7 days, and colonic eosinophil accumulation was assessed.
Data were analyzed by means of ANOVA, followed by the Tukey posthoc test with GraphPad Prism 5 (San Diego, CA). Data are presented as the mean ± SE. P values of less than 0.05 were considered statistically significant. ND represents below limit of detection.
We have previously reported a pathological role for eosinophils in DSS-induced colonic injury and that eosinophil recruitment into the colon following DSS treatment was dependent on CCL11 (14). With the emerging experimental and clinical data demonstrating an important function for CCL11/eosinophils in IBD, we were interested in defining the cellular source of CCL11 in experimental colitis. To assess if hematopoietic or stromal compartment (SC) expression of CCL11 is sufficient for DSS-induced colonic eosinophilic inflammation we restricted CCL11 expression to either the hematopoietic or stromal compartment (SC) using BM chimeric mice on the BALB/c background. To restrict CCL11 expression to the bone marrow (BM) compartment, we irradiated recipient CCL11−/− mice and reconstituted them with WT BM. In these mice, referred to as SC−BM+, only cells derived from transferred WT BM are CCL11 sufficient (+) whereas the radioresistant MΦs and SC native to the recipient CCL11−/− mice do not express functional CCL11 (−). Conversely, in SC+BM− mice, we restricted CCL11 signaling to only the SC compartment by irradiating WT mice and reconstituting with CCL11−/− BM. As controls, WT or CCL11−/− mice were irradiated and reconstituted with their own BM type (SC+BM+ and SC−BM−, respectively). Seven to 8 weeks following BM reconstitution, chimeric mice were exposed to 5% DSS, and eosinophilic inflammation was evaluated. To facilitate analysis of chimerism in BALB/c mice, we transferred BM from BALB/c Nzeg-eGFP mice to WT and CCL11−/− BALB/c mice. All cells from the Nzeg-eGFP mice are constitutively GFP positive (31) and can be detected by flow cytometric analysis of autofluorescence (AF). In chimeric mice, eGFP+ cells derived from the donor BM are distinguished from any remaining recipient eGFP− cells. The degree of chimerism at 7 weeks as determined by flow cytometry was 98.2% ± 0.2 and 95.4% ± 2.1 for peripheral blood monocytes and eosinophils respectively, and 72.0% ± 0.6 for MLN [CD4+ T-cells] (mean ± SEM; n = 3–4 mice per group) (Supplementary figure S1A–C). Percent reconstitution of colonic MΦs was 66.9% ± 2.8 and 93.3% ± 1.2, respectively, as determined by flow cytometry (Supplementary figure S1D and E). Chimerism was also determined for baseline colonic CD4+ T cells, B cell and eosinophils (Supplementary figure 1D and E). Following the verification of reconstitution, mice were exposed to DSS for 7 days and colonic eosinophil inflammation was quantitated. DSS treatment of SC+BM+ and SC−BM+ mice induced a significant increase in colonic eosinophil levels compared to control-treated mice (Figure 1A and B: SC+BM+ baseline 8.0 ± 1.0 vs. SC+BM+ DSS 17.2 ± 2.9 eosinophils/hpf, p < 0.05; SC−BM+ baseline 3.4 ± 0.1 vs. SC−BM+ 13.2 ± 1.2 eosinophils/hpf, p < 0.05; n=3–5 mice baseline; 7–8 mice DSS). Similarly, DSS treatment of CCL11−/− mice reconstituted with WT bone marrow SC−BM+ induced a 3-fold increase in eosinophil recruitment in the distal colon compared to SC−BM− mice (Figure 1A and B: SC−BM+ 12.4 ± 1.6 eosinophils/hpf vs. SC−BM− 3.8 ± 1.3 eosinophils/hpf, n = 3–4 mice per group). Somewhat not unexpectedly, eosinophils were also significantly increased in SC+BM− mice (Figure 1B). Importantly, DSS-induced colonic eosinophilic inflammation was attenuated in CCL11−/− mice reconstituted with CCL11−/− BM (SC−BM−; Figure 1B). These data indicate that BM-derived CCL11 is sufficient to drive eosinophilic recruitment into the colon during DSS-induced colonic injury.
Following our demonstration that BM-derived CCL11 expression was sufficient to reconstitute DSS-induced colonic eosinophilic inflammation, we were next interested in identifying the hematopoietic source of CCL11 that drove DSS-induced colonic eosinophilic inflammation. We have previously demonstrated CCL11 expression in F4/80+ myeloid cells within the lamina propria of the colon of mice following 7 days of exposure to 2.5% DSS (14). We therefore assessed the relationship between myeloid cell and DSS-induced eosinophilic inflammation. Firstly, we performed flow cytometry analysis on peripheral blood monocytes and colonic MΦs at baseline and following DSS exposure (Figure 2A). Under homeostatic conditions, the peripheral blood was predominantly comprised of F4/80+CD11b+Ly6Chigh myeloid cells whereas the colon consisted of F4/80+CD11b+Ly6Clow and F4/80+CD11b+Ly6Chigh myeloid cells (Figure 2A). The predominant F4/80+CD11b+Ly6Clow myeloid population (>80%) within the colon was CX3CR1highPDL1+TLR-2−CD206+ (Figure 2B), consistent with the resident intestinal MΦ phenotype (35). DSS exposure (5 days) induced a significant influx of F4/80+CD11b+Ly6Chigh monocytes (Figure 2A). Notably, the increase in colonic F4/80+CD11b+Ly6Chigh monocyte/MΦ cell numbers (control 7,497 ± 1565 vs. DSS 39,996 ± 8708 p < 0.01; mean ± SEM; n = 5–6 per group) occurred in the absence of any change in F4/80+CD11b+Ly6Clow MΦ levels (control 55,979 ± 12,490 vs. 42,818 ± 7,190; mean ± SEM; n = 5–6 per group) (Figure 2A). The infiltrating F4/80+CD11b+Ly6Chigh myeloid population was predominantly CCR2+CX3CR1loTLR-2+CD206−CD11c−PDL1− (Figure 2B and Supplementary figure S2). We next assessed the colonic eosinophil population following DSS exposure. We show that eosinophils were a distinct population characterized by FSClow and SSChigh. Confirmation of eosinophil lineage was confirmed as the cells were Siglec F+ CCR3+ double positive. Notably, the colonic eosinophil population was CD11b+ and F4/80+ and CCR2− (Figure 2C).
To directly assess if Ly6Chigh colonic MΦs were a source of CCL11, we purified F4/80+CD11b+Ly6Chigh MΦs from the colon of DSS-treated mice by using CX3CR1eGFP/+ mice. Previous studies have demonstrated that inflammatory (Ly6ChighCCR2+CX3CR1low) and non-inflammatory (Ly6ClowCCR2−CX3CR1hi) tissue MΦs can be distinguished based upon the level of CX3CR1 expression (35). Consistent with this, the resident non-inflammatory F4/80+CD11b+Ly6Clow MΦ population of CX3CR1eGFP/+ mice was CX3CR1high whereas the DSS-induced colonic F4/80+CD11b+Ly6Chigh MΦ population was CX3CR1low (Figure 2B). In conjunction, the CX3CR1low cells within the colon of DSS-treated mice were found to be F4/80+CD11b+Ly6Chigh, consistent with the infiltrating monocyte population (Figure 3A).
Following our confirmation that CX3CR1 and Ly6C expression could distinguish between the two intestinal myeloid sub-populations, we purified CX3CR1lowLy6Chigh cells from the colons of CX3CR1eGFP/+ mice following DSS exposure using flow sorting (Figure 3B). Purity of CX3CR1lowLy6Chigh cells was >95% as assessed by flow cytometry (data not shown). For comparative analyses, we also purified blood CX3CR1lowF4/80+CD11b+Ly6Chigh monocytes from CX3CR1eGFP/+ mice at baseline and on day 5 of DSS (Figure 4A and B). PCR analyses revealed no detectable CCL11 mRNA expression in CX3CR1lowLy6Chigh peripheral blood monocytes at baseline or following DSS (Figure 3C). Ccl11 mRNA expression was induced in the colonic CX3CR1lowLy6Chigh cells following infiltration into the colon during DSS-induced colitis (Figure 3C). Assessment of other genes expressed by the Ly6Chigh myeloid population revealed that recruited F4/80+CD11b+Ly6Chigh MΦs had a similar phenotype to the peripheral blood population as they were positive for Chi3l3, Trem1, Cxcl10, TNFα and Pdgfb mRNA transcripts with no detectable Ccl17 expression (Figure 3C and data not shown). Interestingly, DSS exposure did not significantly influence the Ly6Chigh peripheral blood population as we observed no significant difference in the gene profile between control and DSS-treated Ly6Chigh peripheral blood cells (Figure 3C). However, we detected increased mRNA transcripts for Arg1, Il10, Ccl4, Il1b, Il6, Cxcl2, Retlna and Ccl22 in the Ly6Chigh colonic population compared to the Ly6Chigh peripheral blood population (Figure 3C).
To determine the relationship between the intestinal inflammatory F4/80+CD11b+Ly6Chigh monocyte/MΦs and eosinophil recruitment in D S S-colitis, we quantified colonic F4/80+CD11b+Ly6Chigh MΦ and eosinophil levels following DSS-induced colonic injury. We found a positive correlation between numbers of colonic F4/80+CD11b+Ly6Chigh MΦs and eosinophils (Figure 4; p<0.005). Notably, levels of colonic F4/80+CD11b+Ly6Chigh MΦ did not correlate with colonic neutrophil (F4/80−CD11b+Ly6G+) levels (Figure 4) indicating a specific link between F4/80+CD11b+Ly6Chigh monocyte recruitment and colonic eosinophilic inflammation.
We next assessed the relative contribution of F4/80+CD11b+Ly6Chigh blood monocytes to DSS-induced colonic eosinophilic inflammation and colitis in CCR2−/− mice as CCR2 is important for Ly6Chigh monocyte mobilization from the BM to the peripheral circulation and tissue inflammatory sites (36). Consistent with this, basal homeostatic levels of peripheral blood Ly6Chigh monocytes were 6-fold lower in CCR2−/− mice compared with WT mice (Figure 5A and B). DSS-induced recruitment of F4/80+CD11b+Ly6Chigh MΦs into the colon was attenuated in CCR2−/− mice (Figure 5C and D). Immunofluorescence analyses of F4/80+ cells in the colonic lamina propria of DSS-treated WT mice and CCR2−/− mice revealed a loss of the large F4/80+ infiltrate in the CCR2−/− mice (Figure 5E). This reduction was specific for Ly6Chigh monocytes as recruitment of F4/80−CD11b+ cells (neutrophils) was not impaired (Figure 5F and G). The reduction in intestinal MΦ levels was not due to decreased levels of homeostatic intestinal MΦs, as the levels of resident F4/80+CD11b+Ly6Clow colonic MΦs were comparable between WT and CCR2−/− mice (Figure 5H and I). Collectively, these data indicate that homeostatic resident F4/80+CD11b+Ly6Clow colonic MΦ levels are independent of CCR2 whereas F4/80+CD11b+Ly6Chigh monocyte recruitment to the colon during DSS-induced colitis is CCR2-dependent.
To assess the contribution of F4/80+CD11b+Ly6Chigh monocytes to DSS-induced colonic CCL11 expression and eosinophilia, we quantitated CCL11 and eosinophil levels in the colon of CCR2−/− mice following DSS exposure. Importantly, colonic eosinophil levels and distribution at baseline were comparable between WT and CCR2−/− mice, indicating no role for CCR2 in basal colonic eosinophil recruitment (Figure 6A and B and data not shown). DSS-treatment of WT mice induced a significant increase in colonic eosinophil levels (Figure 6A and B). In contrast, there was no significant increase in eosinophil levels in DSS-treated CCR2−/− mice (Figure 6A and B). Notably, the reduction in colonic eosinophil levels was associated with no significant increase in colonic CCL11 levels in colonic punch biopsies from DSS-treated CCR2−/− mice (Figure 6C: WT baseline 8.7±1.2 pg/mL vs. WT DSS 42.4±9.4 pg/mL, p < 0.05; CCR2 baseline: 12.6±3.1 pg/mL; CCR2−/− DSS: 19.27±4.4 pg/mL; mean ± SEM; n = 9–10 mice per group). These data directly implicate F4/80+CD11b+Ly6Chigh monocyte/ MΦ-derivedCCL11 in the regulation of colonic eosinophilic inflammation in DSS-induced colonic injury.
To assess the effect of ablation of Ly6Chigh monocyte recruitment and colonic eosinophils to DSS-induced colonic injury, we performed histopathological assessment of the colon in WT and CCR2−/− mice. DSS treatment of WT mice induced crypt loss, epithelial erosion and a large inflammatory infiltrate (Figure 7A). The DSS-induced epithelial damage was significantly reduced in CCR2−/− mice compared with WT mice (Figure 7A and B; Histological score of WT 15.7 ± 0.84 vs. CCR2−/− 7.67 ± 0.62, p < 0.001; mean ± SEM; n = 10 per group). Consistent with this observation, CCR2−/− mice displayed less weight loss and delayed development of diarrhea and rectal bleeding (DAI) resulting in decreased DAI score (Figure 7C; DAI of WT 5.5 ± 0.65 vs. CCR2−/− 2.2 ± 0.32, p<0.05; mean ± SEM; n = 4 per group). Attenuation of DSS-induced colitis and recruitment of F4/80+CD11b+Ly6Chigh monocytes to the colon by CCR2 deficiency was associated with decreased production of pro-inflammatory cytokines IL-6 and TNF-α (Figure 7D). Collectively, these studies identify that Ly6Chigh colonic monocyte/MΦs have a pathogenic role in DSS-induced pro-inflammatory cytokine production and histopathology and that recruitment of this cell population is mediated by CCR2-dependent pathways.
CCL2 is a CC chemokine that binds to the CCR2 receptor, and experimental data indicates that this chemokine is important in the recruitment of monocytes and MΦs into inflamed tissues (37–39). Notably, DSS exposure induced a significant increase in colonic CCL2 protein levels (Supplementary figure S4A). To evaluate the relative contribution of CCL2 to DSS-induced Ly6Chigh monocyte recruitment into the colon and disease pathology, we examined CCL2−/− mice. Surprisingly, the levels of colonic F4/80+CD11b+Ly6Chigh monocytes in the colon of DSS-treated CCL2−/− mice were comparable to that of strain- and weight-matched DSS-treated WT mice (Supplementary figure S4B and C). Consistent with our data, recruitment of F4/80+CD11b+Ly6Chigh monocytes into the colon of CCL2−/− mice was associated with DSS-induced weight loss and disease activity (Supplementary figure S4D), disease pathology (Supplementary figure S4E and F) and colonic eosinophil inflammation (Supplementary figure S4G and H). Assessment of Ly6Chigh peripheral blood monocytes and colonic MΦs at baseline revealed comparable levels between WT and CCL2−/− mice, indicating that CCL2 does not contribute to either homeostatic resident intestinal F4/80+CD11b+Ly6Clow MΦ levels or DSS-induced recruitment of F4/80+CD11b+Ly6Chigh monocytes into the colon (Supplementary figure S5A–D).
In the present study, we have investigated the molecular regulation of CCL11 and colonic eosinophilic inflammation in an experimental mouse model of DSS-induced colitis. We demonstrate that reconstitution of CCL11−/− mice with bone marrow-derived CCL11 is sufficient for DSS-induced colonic eosinophil inflammation. We show that DSS exposure promotes the influx of F4/80+CD11b+Ly6Chigh monocytes into the colon, and that recruitment of this cell population positively correlated with colonic eosinophilic inflammation. Purification of the F4/80+CD11b+Ly6Chigh population revealed that these cells express a mixture of M1- and M2- associated genes, including Chi3l3, Retlna, Il10, Il6, Il1b, Trem1, Cxcl2 and Ccl11. Abrogation of F4/80+CD11b+Ly6Chigh monocyte recruitment by genetic deletion of CCR2 was associated with decreased DSS-induced histopathology, CCL11 expression and eosinophil recruitment. These studies indicate that colonic eosinophilic inflammation in experimental colitis is mediated by Ly6ChighCCR2+ inflammatory monocyte/MΦ-derived CCL11.
Flow cytometry analyses identified the presence of two distinct monocyte populations in the peripheral blood of mice characterized by F4/80+CD11b+Ly6Chigh and F4/80+CD11b+Ly6Clow phenotype. We show that the F4/80+CD11b+Ly6Chigh population was CCR2+ and CX3CR1low whereas the F4/80+CD11b+Ly6Clow monocytes were CCR2− and CX3CR1high. The F4/80+CD11b+Ly6Clow blood monocytes are a precursor to resident homeostatic tissue MΦs (35). Consistent with this, under homoeostatic conditions the colon predominantly consisted of F4/80+CD11b+Ly6Clow myeloid cells. Intestinal dendritic cells under homeostatic conditions also express CX3CR1 and F4/80. Analysis of CD11c expression revealed that F4/80+CD11b+Ly6Clow cells consisted of 80% CD11c− and 19.5% CD11c+ cells indicating the presence homeostatic tissue MΦs and DC within the F4/80+CD11b+Ly6Clow population (Supplementary Figure S2). The presence of a F4/80+CD11b+Ly6ClowCD11c+ DC population may have contributed to the gene expression heterogeneity seen within this cell population. The F4/80+CD11b+Ly6Chigh monocytes are an inflammatory monocyte population, and are rapidly recruited into tissues following inflammatory insult. Large infiltrates of Ly6ChighCX3CR1lowCCR2+ blood monocytes have been observed in the peritoneum and injured myocardium following immune stimulation (35, 40). Similarly, we observed the selective increase in the numbers of F4/80+CD11b+Ly6Chigh cells within the colon following DSS exposure. Importantly, the F4/80+CD11b+Ly6Chigh cells within the colon were CD11c− indicating that this population did not contain contaminating DC population (Supplementary Figure S2). Consistent with previous investigations we show that the recruitment of the inflammatory F4/80+CD11b+Ly6Chigh monocytes/MΦs into inflamed tissues was dependent on CCR2 (35). Previous investigations have indicated that CCR2-dependent recruitment of this monocytic population is mediated by CCL2 (37, 41), however the CCR2-mediated recruitment of F4/80+CD11b+Ly6Chigh monocytes/MΦs into the colon was not dependent on CCL2. CCR2 is a promiscuous chemokine receptor binding multiple ligands, including CCL2, CCL8, CCL7, and CCL13 (42), and CCL2, CCL8 and CCL7 have been shown to be increased in the colon in inflamed tissue from IBD patients as well as in the DSS-model of colitis (43, 44).
The demonstration that ablation of F4/80+CD11b+Ly6Chigh monocytes/MΦs into the colon in CCR2−/− mice was associated with a reduction in proinflammatory cytokines (IL-6 and TNFα) indicates that this cell population is an important source of cytokines that drives DSS-induced intestinal inflammation and colitic disease. Consistent with this, PCR analyses of purified Ly6Chigh monocytes/MΦs revealed that these cells expressed high levels of pro-inflammatory markers Trem1 and TLR2, and but not the anti-inflammatory markers CD206 or PDL1. Similar to our observations Platt and colleagues identified a TLR2+ CCR2+ CX3CR1Int Ly6Chi TNFα+ monocyte/MΦ population in the colon following DSS exposure and that this cell type drives DSS-induced pathology (45). We have extended these observations identifying a new cellular function fort the F4/80+CD11b+Ly6ChighTLR2+ population in driving colonic eosinophilic inflammation and histopathology associated with DSS-induced colitis. Furthermore, we have developed a protocol that permits purification of this cell population and demonstrate that the Ly6Chigh colonic monocyte/MΦs express Arg1, Retlna, Chi3l3, Ccl22 (MDC) and Il10. The Arg1 and Chi3l3 (Ym1) mRNA expression in the infiltrating Ly6Chigh cell population is consistent with the previously reported DSS-induced total increase in tissue Arg1 and Ym1 protein (46). The expression of Arg1 and Ym1 suggested that the monocyte/MΦs are alternatively activated, however we also observed evidence of classical activation as the cells expressed the pro-inflammatory cytokines/chemokines Il1b, Il6, Ccl4 and Cxcl2. Surprisingly, TNFα mRNA expression was not increased in Ly6Chigh colonic monocytes compared to peripheral blood monocytes, although protein expression correlated with Ly6Chigh monocyte recruitment. However, it has previously been shown that TNFα mRNA stability and translation regulate its bioreactivity (47–50). Our results suggest that Ly6Chigh colonic monocyte/MΦs have a mixed classical/alternatively-activated phenotype in the acute phase of DSS-induced colitis, further indicating that classical/alternatively-activated phenotypes are part of a spectrum of activation, rather than exhibiting strongly polarized phenotypes that are often seen in vitro (51, 52). Alternatively, the mixed phenotype could suggest the presence of subpopulations of Ly6Chigh colonic monocyte/MΦs within the colon. DSS-induced colitis has been linked with increased expression of the classical/alternative-activation cytokines IL-4/IFNγ (53, 54). In a mouse model of kidney injury, Chi3l3 was expressed in Ly6C+ MΦs in the kidney, but other M2 genes, including Ccl17 and Ccl22 were only expressed at very low levels, while high levels of Cxcl2 and Il1b were expressed, indicating an M1-biased pattern (55). Further investigations are required to determine whether the Ly6Chigh monocyte/MΦ population consists of multiple polarized monocyte/MΦ subsets of different phenotypes or whether the Ly6Chigh monocyte/MΦ population is simply plastic and adapts to exogenous and endogenous stimuli within the microenvironment.
The role of monocyte/MΦs in eosinophil recruitment during chronic inflammatory responses and the relative contribution of CCL11 to this response is not yet fully delineated. Previous studies have demonstrated CCL11 mRNA expression in MΦs in allergen-induced cutaneous biopsies in atopic patients (56), and in bronchial biopsies from asthmatic patients which possess eosinophils (57). Furthermore, a recent study also demonstrated CCL11 expression in lung MΦs from rhinovirus-infected allergic mice with pulmonary eosinophilic inflammation (58). Experimental analyses employing a Nippostrongylus brasiliensis infestation model have demonstrated a role for monocyte/MΦ populations in eosinophil recruitment into peritoneum. Similar to our observations, eosinophil recruitment was associated with Retlna and Chi3l3-monocyte/MΦ expression but was not associated with CCL11 expression (59). We show that during colonic inflammation CX3CR1+Ly6Chigh colonic monocyte/MΦs are the primary cellular source of CCL11 and that this cell population is sufficient to mediate colonic eosinophilic inflammation. These observations suggest that DSS exposure stimulates CX3CR1+Ly6Chigh colonic monocyte/MΦ recruitment in the colon, and that the recruited inflammatory monocyte/MΦ’s subsequently drives eosinophil infiltration via a CCL11-dependent pathway.
We have previously characterized eosinophil infiltration of the colon following DSS exposure and reported that eosinophil levels begin to increase on Day 5 of DSS exposure (14). Assessment of F4/80+CD11b+Ly6Chigh and eosinophil numbers in the colon of mice 3 and 4 days following DSS exposure revealed a significant influx of F4/80+CD11b+Ly6Chigh monocytes/ MΦs prior to commencement of eosinophil recruitment (Supplementary Figure S3). These kinetic data indicate that F4/80+CD11b+Ly6Chigh monocytes/ MΦs are strategically positioned in the colon to regulate DSS-induced eosinophil recruitment. These studies strongly suggest that CCL11 is predominantly monocyte/MΦs derived and is important in colonic eosinophilic recruitment in experimental colitis. We have previously reported an critical role for CCL11 and not the other eotaxin family members CCL24 and CCL26 in the regulation of eosinophilic inflammation in experimental and human IBD (14). Moreover, eosinophil recruitment into the colon during experimental colitis was attenuated in CCL11−/− and not CCL24−/− mice and elevated CCL11 mRNA levels in lesional biopsy samples from IBD patients positively correlated with eosinophil numbers (14). Furthermore, we have previously identified CCL11+ CD68+ monocyte/MΦs in colonic biopsies of pediatric UC patients. The demonstration of a link between colonic eosinophilic inflammation and DSS-induced histopathology suggests a role for eosinophils in the DSS-induced colonic injury. This is consistent with previous demonstration by us and others of a partial role for eosinophils in DSS-induced epithelial histopathology (14–16, 33). Furthermore, this is supported by significant clinical evidence demonstrating increased eosinophils and eosinophil-derived granule proteins in adult UC and CD and a positive correlation between levels of eosinophils and histological score in rectosigmoid colonic biopsy samples from pediatric UC patients (4, 9, 14, 60, 61). We have previously published that mice deficient in eosinophils, are partially protected from DSS-induced colitis (33). Moreover, we observed a 48.1% reduction in histological score between WT DSS vs PHIL DSS (14). Notably, in the present study we observed a 52% reduction in histological score between WT DSS vs CCR2 DSS (histological score WT DSS: 15.7 ± 0.8 vs PHIL DSS 7.7 ± 0.6; n= 9–10 per group; mean ± SEM; p = 0.001; Figure 7). The reduction in histopathology in CCR2−/− mice was associated with reduced F4/80+CD11b+Ly6Chigh monocytes/MΦ recruitment and reduced CCL11 expression and eosinophil infiltration. These data suggest that a significant component of MΦ-mediated DSS-induced histopathology is mediated via regulation of eosinophil function.
BM chimera experiments in mice indicate that bone marrow cell-derived CCL11 is sufficient for DSS-induced colonic eosinophilic inflammation. Moreover, CCL11−/− mice reconstituted with WT BM restored eosinophil recruitment to the colon during DSS-induced colonic injury. As bone marrow reconstitution is not selective for Ly6Chi monocyte-derived CCL11, we cannot rule out the contribution of other BM-derived cell populations; however we have previously demonstrated that DSS-induced CCL11 expression in the colon was restricted to F4/80+ myeloid cells and not F4/80− cells (14). Surprisingly, we observed a significant eosinophilic infiltratein the colon following DSS exposure in WT mice reconstituted with bone marrow from CCL11−/− mice suggesting a role for structural cell-derived CCL11 in colonic eosinophilic inflammation. We postulate that this paradoxical observation is due to the radioresistance of monocyte/MΦs and the effects of radiation of monocyte/MΦs function. Many studies have demonstrated that tissue MΦs are radioresistant (62–64). For example, in the lung, it takes over 90 days for alveolar MΦs to reach ~80% reconstitution (64). Similarly, we found that Ly6Clow colonic MΦs only reached ~60% reconstitution at ~50 days post-irradiation and that ~5% monocyte/MΦs in the colon of DSS-treated mice were recipient-derived cells. Furthermore, irradiation has been shown to induce macrophage oxidative injury (65) as well as alter macrophage activation which may lead to homeostatic resident MΦ involvement in eosinophil recruitment (66). While we cannot rule out a contribution for non-hematopoietic cell derived CCL11 to eosinophil recruitment during DSS-induced colitis, our previous data demonstrating CCL11 expression restricted to F4/80+ cells in the lamina propria of the colon on day 7 of DSS exposure (14) and our observations in CCR2−/− mice indicates that the CCR2-dependent Ly6Chigh monocytes are sufficient to drive eosinophil recruitment in DSS-induced colonic injury.
Recently, a number of unique loci were identified to be associated with early-onset IBD susceptibility including the C-C motif chemokine cluster on 17q12 loci which contains the eosinophil-specific chemokine gene CCL11(17). Clinical and experimental data indicates a strong relationship between eosinophils and the exacerbation and severity to IBD and a pivotal role for macrophages in the augmentation and of the intestinal inflammatory response associated with IBD (67). We provide evidence of a direct pathway involving Ly6Chigh colonic monocyte/MΦ-derived CCL11 in colonic eosinophilic inflammation and histopathology in experimental colitis. These studies provide significant rationale for the assessment of monocyte/MΦ-derived CCL11 in human IBD and the targeting of the monocyte/MΦ:CCL11 pathway as a therapeutic modality for the treatment and prevention of IBD.
A, Representative flow cytometry plots of peripheral blood chimerism of monocytes and eosinophils, B, at 3 weeks post-irradiation. C, Representative flow cytometry plots of MLN chimerism of CD4+ T cells at 5–9 weeks post-irradiation. D, Representative flow cytometry plots of chimerism of hematopoietic subsets in the colon at 7 weeks post-irradiation. E, Quantification of percent donor-derived cells in the colon at baseline and following DSS treatment. Data represents the mean ± SEM of n = 3–4 mice per group.
A, Representative flow cytometry plots of the F4/80+CD11b+Ly6Chigh and Ly6Clow MΦ populations from the colon at baseline (Top panel) and colon following DSS exposure (bottom panel). MΦ were initially gated by SSC vs FSC followed by F4/80+ CD11b+. The double-positive cell populations from baseline and DSS-treated colons were assessed for Ly6C. B,Percentage of CD11c+ F4/80+CD11b+Ly6Chigh and Ly6Clow MΦ populations in the colon at baseline and following DSS. Data represents the mean + SEM of n = 3–4 mice per group. Significant differences (*p<0.05) between groups.
Quantification of Ly6Chigh colonic monocytes/MΦ numbers in the colon on day 0, 3 and 4 of DSS exposure. Colonic lamina propria cells were stained for flow cytometry with F4/80 CD11b and Ly6C, and the number of triple-positive cells were quantitated. I, Data represents the mean ± SEM of n = 3–4 mice per group per time point. Significant differences (*p<0.05; and **p<0.01) between groups. Ns not significant.
A, CCL2 was quantified from supernatants from punch biopsies of WT mice over the course of DSS-induced colitis. B, Flow cytometric analysis of F4/80+CD11b+Ly6Chigh colonic MΦs at baseline and following DSS in WT and CCL2−/− mice. C, Quantification of Ly6Chigh colonic MΦ numbers based on flow cytometry analysis. D, Percent weight loss and diarrhea/rectal bleeding score during the course of DSS treatment. E, Representative sections from baseline and day 7 of DSS were H and E stained. F, Histological score of baseline and DSS-induced WT and CCL2−/− sections. G, Colonic sections from baseline and DSS-treated mice (day 7, 2.5%) were stained for MBP, and eosinophil levels were quantified, H. Data represents the mean ± SEM of n = 3–4 mice per group from duplicate experiments. Significant differences (*p<0.05; ***p<0.001) between groups. Magnification of photomicrographs is ×100.
A, Flow cytometric analysis of F4/80+CD11b+Ly6Chigh peripheral blood monocytes in WT and CCL2−/− mice. B, Quantification of percent Ly6Chigh peripheral blood monocytes in WT and CCL2−/− mice. C, Colonic lamina propria cells were stained for flow cytometry with F4/80 CD11b and Ly6C, and the number of double and triple-positive cells was examined D, Quantification of percent F4/80+CD11b+ and F4/80+CD11b+Ly6Chigh colonic MΦs in WT and CCL2−/− mice at baseline.
We thank Drs Patricia Fulkerson and Debroski Herbert and members of the Division of Allergy and Immunology and Gastroenterology, Hepatology, and Nutrition, Cincinnati Children’s Hospital Medical Center for critical review of the manuscript and insightful conversations. We thank Dr Nives Zimmermann, Victoria Summey and Jeff Bailey from the CCHMC Comprehensive Mouse and Cancer Core facility for expertise and assistance with bone marrow transplantation. We thank Jamie and Nancy Lee for the generous provision of anti-MBP antibody and Dr Klaus Matthaei for the Nzeg-eGFP mice. We would also like to thank Emily Stucke and Heather Osterfeld for animal husbandry and Shawna Hottinger for editorial assistance and manuscript preparation.
Grant support: This work was supported by The Crohn’s and Colitis Foundation of America Career Development Award (S.P.H.), NIH R01 AI073553 (S.P.H.), NIH R01 AI45898 (M.E.R) and American Gastroenterological Association Foundation Graduate Student Research Fellowship Award (A.B.W.).
Writing assistance: A.W. and S.P.H. designed and performed experiments, analyzed and interpreted data and wrote the manuscript. M.E.R provided mice. K.S., R.A. and A.M discussed experimental design and data analysis.