Massive but selective cytokine dysregulation in the colon of IL-10-/- mice revealed by multiplex analysis

Abstract

IL-10-deficient mice develop enterocolitis due to a failure of cytokine regulation; however, the full scope of that response remains poorly defined. Using multiplex analysis to quantify the activity of 23 regulatory and effector cytokines produced by colonic leukocytes, we demonstrate a vast dysregulation process of 18 cytokines in IL-10-/- mice from 7 to 27 weeks of age. Of those, IL-12p40, IL-6, granulocyte macrophage colony-stimulating factor, IFN-gamma, IL-13 and monocyte chemoattractant protein-1 (MCP-1) had the highest single correlations with pathology (r = 0.7766-0.7016). Importantly, there were strong associations (r = 0.7071-0.9074) between those cytokines and as many as 10 additional cytokines, indicating a high degree of cytokine complexity as disease progressed. IL-17 was notable in that it was produced at high levels by colonic leukocytes from IL-10-/- mice with pathology ranging from mild to severe, though it was not produced by healthy IL-10-/- mice lacking pathology. Tumor necrosis factor alpha (TNFalpha) by itself displayed only a modest association with pathology (r = 0.6340), ranking sixth lowest, though it cross-correlated strongly with the synthesis of 12 other cytokines, implying that the destructive effects associated with TNFalpha may be due to interactions of multiple cytokine activities. IL-23 expression did not correlate with pathology, possibly suggesting that IL-23 is involved in the initiation but not the perpetuation of inflammation. Four cytokines (IL-2, IL-3, IL-4 and IL-5) remained negative in IL-10-/- mice, demonstrating that cytokine dysregulation was not universal. These findings emphasize the need to better understand cytokine networks in chronic inflammation and they provide a rationale for combining immunotherapies in the treatment of intestinal inflammation.

Fig. 1

(A) CELs from IL-10-/- and normal mice consist of both TCRαβ and TCRγδ T cells, comprised of both CD4⁺ and CD8⁺ populations. The ICOS activation marker is expressed on CELs from an IL-10-/- mice with colonic pathology that included rectal prolapse. (B) Histopathological analysis of colonic tissue from IL-10-/- mice demonstrating (a) grade 1, (b) grade 2, (c) grade 3, and (d,f) grade 4 pathology. (e) Tissue from an IL-10-/- mouse at 20 wks of age with no pathology.

Fig. 2

(A) Pathology scores of all 25 IL-10-/- and 6 control mice; the line in the graph demarcates the highest level of pathology scoring in BALB/c control animals. (B) Comparison of pathology scores for all IL-10-/- mice and control mice. (C) Comparison of pathology scores of female and male IL-10-/- mice. (D) Purity of CELs calculated as the number of CELs relative to the total number of extracted cells. (E) Number of CELs recovered from colonic tissues of IL-10-/- and control mice by age. (F) Spearman correlation analysis of number of CELs from IL-10-/- mice as a function of pathology score.

Fig. 3

Analyte activities for IL-1α, IL-1β, IL-6 and IL-9 from CELs of IL-10-/- and control mice. Values are mean scores of duplicate assays; bars indicate the range of duplicate values in pg/ml.

Fig. 4

Analyte activities for IL-12p40, IL-12p70, IL-13, and IL-17 from CELs of IL-10-/- and control mice. Values are mean scores of duplicate assays; bars indicate the range of duplicate values in pg/ml.

Fig. 5

Analyte activities for eotaxin, G-CSF, GM-CSF, and IFN-γ from CELs of IL-10-/- and control mice. Values are mean scores of duplicate assays; bars indicate the range of duplicate values in pg/ml.

Fig. 6

Analyte activities for KC/CXCL1, MCP-1, MIP-1α, and MIP-1β from CELs of IL-10-/- and control mice. Values are mean scores of duplicate assays; bars indicate the range of duplicate values in pg/ml.

Fig. 7

(Top Panels) Analyte activities for RANTES and TNFα from CELs of IL-10-/- and control mice. Values are mean scores of duplicate assays; bars indicate the range of duplicate values in pg/ml. (Bottom Panels) Q-realtime PCR comparing gene expression of IFN-γ, TNFα, and IL-6. Data are mean values ± SEM of two samples expressed as fold increase in gene expression of IL-10-/- mice over normal control mice for tissues from the ascending and descending colon regions.

Fig. 8

Spearman correlation analysis comparing pathology score as a function of analyte level for the 18 analytes listed in Table 2.

Fig. 9

IL-12p19 activity of IL-10-/- and control mice. (A) Spearman correlation analysis of IL-12p19 activity from CEL culture supernatants as determined by ELISA and compared to pathology scores for all IL-10-/- and control animals. Note the lack of correlation of IL-23p19 with pathology (r = -0.0623). (B) Comparison of total average IL-23p19 activity of IL-10-/- mice and control mice, indicating a lack of statistical difference in analyte activity. (C) Grade 3 histopathology of tissue from the descending colon of an IL-10-/- mouse used for analysis in panels D and E. (D) Comparison of conventional PCR-amplified IL-23p19 activity of IL-10-/- and control mice using RNAs obtained from the ascending and descending colon regions. (E) Q-realtime PCR results comparing combined gene expression of ascending and descending colon regions from IL-10-/- and control mice, indicating a lack of statistical significance in IL-23p19 activity.