Activity of SHIP, Which Prevents Expression of Interleukin 1ß, Is Reduced in Patients With Crohn's Disease.

BACKGROUND & AIMS: Crohn's disease (CD) is associated with a dysregulated immune response to commensal micro-organisms in the intestine. Mice deficient in inositol polyphosphate 5'-phosphatase D (INPP5D, also known as SHIP) develop intestinal inflammation resembling that of patients with CD. SHIP is a negative regulator of PI3Kp110α activity. We investigated mechanisms of intestinal inflammation in Inpp5d(-/-) mice (SHIP-null mice), and SHIP levels and activity in intestinal tissues of subjects with CD. METHODS: We collected intestines from SHIP-null mice, as well as Inpp5d(+/+) mice (controls), and measured levels of cytokines of the interleukin 1 (IL1) family (IL1α, IL1ß, IL1ra, and IL6) by enzyme-linked immunosorbent assay. Macrophages were isolated from lamina propria cells of mice, IL1ß production was measured, and mechanisms of increased IL1ß production were investigated. Macrophages were incubated with pan-phosphatidylinositol 3-kinase inhibitors or PI3Kp110α-specific inhibitors. Some mice were given an antagonist of the IL1 receptor; macrophages were depleted from ilea of mice using clodronate-containing liposomes. We obtained ileal biopsies from sites of inflammation and peripheral blood mononuclear cells (PBMCs) from treatment-naïve subjects with CD or without CD (controls), and measured SHIP levels and activity. PBMCs were incubated with lipopolysaccharide and adenosine triphosphate, and levels of IL1ß production were measured. RESULTS: Inflamed intestinal tissues and intestinal macrophages from SHIP-null mice produced higher levels of IL1B and IL18 than intestinal tissues from control mice. We found PI3Kp110α to be required for macrophage transcription of Il1b. Macrophage depletion or injection of an IL1 receptor antagonist reduced ileal inflammation in SHIP-null mice. Inflamed ileal tissues and PBMCs from patients with CD had lower levels of SHIP protein than controls (P < .0001 and P < .0002, respectively). There was an inverse correlation between levels of SHIP activity in PBMCs and induction of IL1ß production by lipopolysaccharide and adenosine triphosphate (R(2) = .88). CONCLUSIONS: Macrophages from SHIP-deficient mice have increased PI3Kp110α-mediated transcription of Il1b, which contributes to spontaneous ileal inflammation. SHIP levels and activity are lower in intestinal tissues and peripheral blood samples from patients with CD than controls. There is an inverse correlation between SHIP activity and induction of IL1ß production by lipopolysaccharide and adenosine triphosphate in PBMCs. Strategies to reduce IL1B might be developed to treat patients with CD found to have low SHIP activity.

Arginase activity in alternatively activated macrophages protects PI3Kp110δ deficient mice from dextran sodium sulfate induced intestinal inflammation.

Alternatively activated or M2 macrophages have been reported to protect mice from intestinal inflammation, but the mechanism of protection has not been elucidated. In this study, we demonstrate that mice deficient in the p110δ catalytic subunit activity of class I phosphatidylinositol 3-kinase (PI3Kp110δ) have increased clinical disease activity and histological damage during dextran sodium sulfate (DSS) induced colitis. Increased disease severity in PI3Kp110δ-deficient mice is dependent on professional phagocytes and correlates with reduced numbers of arginase I+ M2 macrophages in the colon and increased production of inflammatory nitric oxide. We further demonstrate that PI3Kp110δ-deficient macrophages are defective in their ability to induce arginase I when skewed to an M2 phenotype with IL-4. Importantly, adoptive transfer of IL-4-treated macrophages derived from WT mice, but not those from PI3Kp110δ-deficient mice, protects mice during DSS-induced colitis. Moreover, M2 macrophages mediated protection is lost when mice are cotreated with inhibitors that block arginase activity or during adoptive transfer of arginase I deficient M2 macrophages. Taken together, our data demonstrate that arginase I activity is required for M2 macrophages mediated protection during DSS-induced colitis in PI3Kp110δ-deficient mice.

SHIP prevents lipopolysaccharide from triggering an antiviral response in mice.

Gram-negative bacterial infections, unlike viral infections, do not typically protect against subsequent viral infections. This is puzzling given that lipopolysaccharide (LPS) and double-stranded (ds) RNA both activate the TIR domain-containing adaptor-inducing interferon beta (TRIF) pathway and, thus, are both capable of eliciting an antiviral response by stimulating type I interferon (IFN) production. We demonstrate herein that SH2-containing inositol-5'-phosphatase (SHIP) protein levels are dramatically increased in murine macrophages via the MyD88-dependent pathway, by up-regulating autocrine-acting transforming growth factor-beta (TGFbeta). The increased SHIP then mediates, via inhibition of the phosphatidylinositol-3-kinase (PI3K) pathway, cytosine-phosphate-guanosine (CPG)- and LPS-induced tolerance and cross-tolerance and restrains IFN-beta production induced by a subsequent exposure to LPS or dsRNA. Intriguingly, we found, using isoform-specific PI3K inhibitors, that LPS- or cytosine-phosphate-guanosine-induced interleukin-6 (IL-6) is positively regulated by p110alpha, -gamma, and -delta but negatively regulated by p110beta. This may explain some of the controversy concerning the role of PI3K in Toll-like receptor-induced cytokine production. Consistent with our in vitro findings, SHIP(-/-) mice overproduce IFN-beta in response to LPS, and this leads to antiviral hypothermia. Thus, up-regulation of SHIP in response to Gram-negative bacterial infections probably explains the inability of such infections to protect against subsequent viral infections.

SHIP-deficient, alternatively activated macrophages protect mice during DSS-induced colitis.

mϕ are heterogeneous in their functions, and although it is clear that inflammatory mϕ contribute to inflammation in IBDs, multiple lines of evidence suggest that M2a mϕ may offer protection during intestinal inflammation. In vivo SHIP-deficient mouse mϕ are M2a so SHIP-deficient mice provide a unique genetic model of M2a mϕ. Based on this, this study tested the hypothesis that SHIP-deficient, M2a mϕ protect mice from intestinal inflammation. The objectives were to compare the susceptibility of SHIP+/+ and SHIP-/- littermates with DSS-induced intestinal inflammation and to determine whether protection was mϕ-mediated and whether protection could be transferred to a susceptible host. We have found that SHIP-/- mice are protected during DSS-induced intestinal inflammation. SHIP-/- mice have delayed rectal bleeding and reduced weight loss, disruption of intestinal architecture, and immune cell infiltration during DSS-induced colitis relative to their WT littermates. Using liposome depletion of mϕ, we found that SHIP-/- mouse protection was indeed mϕ-mediated. Finally, we determined that SHIP-/- mϕ-mediated protection could be conferred to susceptible WT mice by adoptive transfer of M2a mϕ derived ex vivo. This study supports our hypothesis by demonstrating that SHIP-deficient, M2a mϕ are protective in this murine model of acute intestinal inflammation. Adoptive transfer of M2a mϕ to patients with IBDs offers a promising, new strategy for treatment that may be particularly useful in patients who are otherwise refractory to conventional therapies.