Human β-Defensin 2 Mediated Immune Modulation as Treatment for Experimental Colitis

Abstract

Defensins represents an integral part of the innate immune system serving to ward off potential pathogens and to protect the intestinal barrier from microbial encroachment. In addition to their antimicrobial activities, defensins in general, and human β-defensin 2 (hBD2) in particular, also exhibit immunomodulatory capabilities. In this report, we assessed the therapeutic efficacy of systemically administered recombinant hBD2 to ameliorate intestinal inflammation in three distinct animal models of inflammatory bowel disease; i.e., chemically induced mucosal injury (DSS), loss of mucosal tolerance (TNBS), and T-cell transfer into immunodeficient recipient mice. Treatment efficacy was confirmed in all tested models, where systemically administered hBD2 mitigated inflammation, improved disease activity index, and hindered colitis-induced body weight loss on par with anti-TNF-α and steroids. Treatment of lipopolysaccharide (LPS)-activated human peripheral blood mononuclear cells with rhBD2 confirmed the immunomodulatory capacity in the circulatory compartment. Subsequent analyzes revealed dendritic cells (DCs) as the main target population. Suppression of LPS-induced inflammation was dependent on chemokine receptor 2 (CCR2) expression. Mechanistically, hBD2 engaged with CCR2 on its DC target cell to decrease NF-κB, and increase CREB phosphorylation, hence curbing inflammation. To our knowledge, this is the first study showing in vivo efficacy of a systemically administered defensin in experimental disease.

Keywords: IBD; antimicrobial peptides; host defense peptides; innate immunity; β-defensins.

Figure 1

hBD2 reduced the pro-inflammatory effect of LPS and Pam3CSK4 in human primary PBMC's. PBMC's were challenged with LPS (20 pg/ml) or Pam3CSK4 (0.3 μg/ml), respectively, and treated with various concentrations of hBD2 (100, 40, 10, or 1 μg/ml). Release of (A) IL-1β and (B) TNF-α. Results are presented as mean ± SEM, (A) and (B) n = 150–194. Statistical analysis was performed by one-way ANOVA with Bonferroni post-test.

Figure 2

hBD2 reduced the pro-inflammatory effect of LPS in human primary PBMC's of ulcerative colitis patients. PBMC's were challenged with LPS (20 pg/ml) and treated with 10 μg/mL hBD2. Release of of (A) IL-1β, (B) TNF-α, (C) INF-y, (D) IL-23, and (E) IL-12p70 is shown in % normalized to LPS. Data are presented as mean ± SEM (n = 4) and analyzed by unpaired t-test.

Figure 3

Cytokine production of DC's was affected by hBD2 in a TLR- and CCR2-dependent manner. Human mo-DC's and murine BM-DC's were treated with LPS (10 μg/ml) alone or co-incubated with various concentrations of hBD2 (100 μg/ml or 10 μg/ml). Murine BM-DCs were additionally treated with pertussis toxin or the CCR2 inhibitor RS prior to stimulation with LPS or a cytokine cocktail containing TNF-α (0.2 mg/ml), IL-6 (0.2 mg/ml), and IL-1β (0.2 mg/ml). Release of TNF-α in (A) human Mo-DC's and in (B) murine BM-DC's was quantified by ELISA. Release of TNF-α (C,E,G) and IL-10 (D,F,H) in murine BM-DCs was quantified by LEGENDplex. Results are presented as mean ± SEM, n = 3. Statistical test used is one-way ANOVA with Bonferroni post-test.

Figure 4

hBD2 modulates NF-κB and CREB phosphorylation. BM-DC's were incubated for 60 min with LPS (10 μg/ml) or a cytokine cocktail containing TNF-α (0.2 mg/ml), IL-6 (0.2 mg/ml), and IL-1β (0.2 mg/ml) alone or in combination with hBD2 (100 μg/ml). BM-DCs were additionally pretreated with pertussis toxin or the CCR2 inhibitor RS prior to stimulation. The cells were stained with CD11c, MHCII, and CD86 antibodies followed by intracellular staining against p-NF-κB or p-CREB and analysis by flow cytometry. Statistical analysis of (A,B) p-NF-κB, (C,D) MHCII, (E,F) CD86, and (G,H) p-CREB staining. (I) Shows histograms of FACS analysis. Results are presented as mean ± SEM, n = 6. Statistical test used is one-way ANOVA with Bonferroni post-test.

Figure 5

hBD2 ameliorated the outcome of DSS colitis in vivo. Colitis was induced by adding 2% DSS into the drinking water. On day 8 DSS was removed from the drinking water. Mice were treated either once a day s.c. with 0.1 mg/kg hBD2 or intraperitoneally with an anti-TNFα antibody (300 μg/mouse) on day 1, 4, and 8. (A) Weight change of mice during the experiment, (B) development of DAI, (C) histology score from the colon of the mice at the end of the experiment, and (D) representative images from the colon of differentially treated mice. Results are presented as mean ± SEM, control group n = 9 and treated groups n = 10. Appropriate statistical comparison are shown within the graph by a Kruskal-Wallis-test for non-parametric data with a Dunn's post-test.

Figure 6

hBD2 significantly improved TNBS colitis in vivo. Colitis was induced by a single dose of TNBS into the colon. Mice were then treated s.c. with different doses of hBD2 (0.1 mg/kg) or Prednisolone (10 mg/kg) once a day and monitored for 7 days. (A) Development of mouse weight during the experiment. After euthanization, the colon of the mice was examined for (B) colon weight, (C) macroscopic abnormalities, and (D) microscopic evidence of inflammation. (E) Representative images from the colon of differentially treated mice are shown. Results are presented as mean ± SEM, control group n = 14, prednisolone group n = 14, and hBD2 group n = 15. Appropriate statistical comparisons are shown within the graph by a Kruskal-Wallis-test for non-parametric data with a Dunn's post-test.

Figure 7

Protective effect of hBD2 in T cell transfer colitis. Colitis was induced by transferring CD4⁺/CD25⁻ T cells from WT mice to SCID mice. Development of colitis was observed for 94 days. Daily treatment with hBD2 (0.1 mg/kg or 1 mg/kg, s.c.) and Dexamethasone (0.3 mg/kg, i.p.) started 7 days after the transfer. (A) Weight change of mice during the experiment was monitored as well as (B) development of clinical symptoms as DAI. (C) Alteration of stool consistency was assessed as stool score. At the end of the experiment (D) colon weight was measured and (E) activity of myeloperoxidase (MPO) was quantified. (F) Representative images from the colon of differentially treated mice are shown. Results are presented as mean ± SEM, n = 6 (sham treated animals), n = 10 (vehicle and dexamethasone mice) and n = 11 (hBD2 mice). Appropriate statistical comparisons are shown within the graph by a Kruskal-Wallis-test for non-parametric data with a Dunn's post-test.