Selective upregulation of microRNA expression in peripheral blood leukocytes in IL-10-/- mice precedes expression in the colon

Abstract

IL-10(-/-) mice, an animal model of Th1-mediated inflammatory bowel disease, were screened for the expression of 600 microRNAs (miRNAs) using colonic tissues and PBLs from animals having either mild inflammation or severe intestinal inflammation. The development of colonic inflammation in IL-10(-/-) mice was accompanied by upregulation in the expression of 10 miRNAs (miR-19a, miR-21, miR-31, miR-101, miR-223, miR-326, miR-142-3p, miR-142-5p, miR-146a, and miR-155). Notably, the expression of all of these miRNAs plus miR-375 was elevated in PBLs of IL-10(-/-) mice at a time when colonic inflammation was minimal, suggesting that changes in specific miRNAs in circulating leukocytes may be harbingers of ensuing colonic pathology. In vitro exposure of colonic intraepithelial lymphocytes to IL-10 resulted in downregulation of miR-19a, miR-21, miR-31, miR-101, miR-223, and miR-155. Interestingly, unlike IL-10(-/-) mice, changes in miRNAs in PBL of dextran sulfate sodium-treated mice were minimal but selectively elevated in the colon after pathology was severe. We further show that miR-223 is a negative regulator of the Roquin ubiquitin ligase, Roquin curtails IL-17A synthesis, and the 3' untranslated region of Roquin is a target for miR-223, thus defining a molecular pathway by which IL-10 modulates IL-17-mediated inflammation. To identify additional miRNAs that may be involved in the regulation of Roquin, transcriptome analysis was done using cDNAs from HeLa cells transfected with 90 miRNA mimics. Twenty-six miRNAs were identified as potential negative regulators of Roquin, thus demonstrating functional complexity in gene expression regulation by miRNAs.

FIGURE 1

Colonic tissues from (A) normal BALB/c mice without pathology, (B) grade 1.0 pathology of colon tissue from IL-10^−/− mouse, and (C) grade 4 pathology of colon tissue from IL-10^−/− mouse. (D) Heat Map and Unsupervised Hierarchical Clustering of miRNAs in IL-10^−/− and BALB/c mouse colon samples indicate dysregulation of miRNA expression. The miRCURY LNA microarray miRNA profiling service was used to examine miRNA expression in pooled total RNA samples (3 mice each) from colonic tissue sections from normal BALB/c mice, IL-10^−/− mice with intestinal pathology score = 1 (IL-10ko 1.0), and IL-10^−/− mice with intestinal pathology scores ≥3 (IL10ko ≥3.0). Red color represents an expression level above mean, blue color represents expression lower than the mean. A Delta LogMedian ratio of +/− 1.0 is equal to a fold change of +/− 2.0.

FIGURE 2

Total RNA from IL-10^−/− and BALB/c colon were used for TaqMan qRT-PCR analysis of eleven miRNAs in (A) colonic tissues and (B) PBLs from normal mice (score 0) and IL-10^−/− mice with low pathology (score 1) or high pathology (score ≥3.0). Expression values were calculated by normalizing to U6 snRNA levels in a sample, and recorded as values relative to the expression of normal mice, which was arbitrarily designated as 1.0. Note that miRNAs were elevated in PBLs of IL-10−/− mice with mild intestinal pathology, but were elevated in IL-10−/− mice with severe intestinal pathology. Determination of statistical significance was calculated using Student’s t-test relative to BALB/c control values.

FIGURE 3

(A) In vitro treatment of cIELs from IL-10^−/− mice with severe pathology overnight with recombinant 50 ng/ml rIL-10 resulted in the suppression of 6 of 11 miRNAs relative to cells cultured with PBS (* ≤ 0.05). (B) TargetScan (Release 5.1) analysis of the mouse Roquin (Rc3h1) 3′UTR revealed a potential target site for mmu-miR-223. Alignment of mmu-miR-223 to the conserved site is shown. (C) Intracellular expression of IL-17 and IFNγ in IL-10^−/− cIELs. In vitro IL-10 treatment of cIELs from IL-10−/− mice resulted in significant reduction in (D) IL-17 mRNA expression and (E) IL-17A secretion as determined by ELISA. (F) cIELs from IL-10−/− mice have lower Roquin gene expression. (G) Culture of cIELs from IL-10−/− mice for 24 hours with 50 ng rIL-10 results in an increase in Roquin gene expression. (H-I) Transient transfection of EL4 cells with Roquin-specific siRNA oligonucleotides resulted in suppression of Roquin gene expression and enhanced IL-17 gene expression. Determination of statistical significance was calculated using Student’s t-test.

FIGURE 4

(A) Transient transfection of cIELs with anti-miR-223 resulted in enhanced Roquin expression and suppressed IL-17A expression, whereas transfection with miR-223 suppressed Roquin expression and enhanced IL-17A expression. (B) Results of luciferase assay using the 3′ UTR of Roquin cloned into the pMIR-REPORT luciferase plasmid and transfected into 239T cells for 48 hr with either 30 nM of miR-223 or pre-miR control. miR-223 suppressed luciferase activity relative to non-transfected and pre-miR control levels.

FIGURE 5

Mice were given DSS in the drinking water as described in the Materials and Methods. Representative colonic tissues from (A) day 0, pathology score 0, (B) day 1, pathology score 1 (C) day 2, pathology score 3, and (D) day 7, pathology score 6. (E) Average pathology scores from 3 mice per group. * Statistically-significant difference (p<0.01) compared to days 0 and 1. ▲Statistically-significant difference (p<0.01) compared to day 2.

FIGURE 6

qRT-PCR analysis of the same eleven miRNAs studied for IL-10^−/− mice using (A) PBL and (B) colonic tissues of DSS-treated animals on days 0, 1, 2, and 7. Although some miRNAs were elevated in PBL early during the treatment period (day 1), these were not statistically-significant differences. Five miRNAs (miR-31, miR-223, miR-142-3p, miR-146a, and miR-155) were significantly elevated in the colon when inflammation was severe. Expression values were calculated by normalizing to U6 snRNA levels in a sample, and recorded as values relative to the expression of day 0 (non-DSS-treated) mice, which was arbitrarily designated as 1.0.