Herein, we provide several lines of evidence for a key inhibitory role for PIR-B on macrophage function during DSS-induced colonic injury. We show that Pirb-/- mice have increased susceptibility to DSS-induced colitis. The increase in disease severity was linked with elevated macrophage-associated pro-inflammatory cytokine production. In vitro analysis of E. coli-stimulated BM- and thioglycolate-derived macrophages revealed a link between PIR-B deficiency, elevated proinflammatory cytokine production and MAPK (particularly ERK1/2 and p38) and FosB activation. Adoptive transfer experiments confirmed a critical role for PIR-B in the negative regulation of macrophage function in DSS-induced colitis. Finally, and relevant to human disease, we demonstrate expression of PIR-B human orthologues, ILT-2 and -3 on mononuclear cells within the LP including CD68+ macrophages and intestinal epithelial cells.
We demonstrate that DSS exposure of mice deficient in Pirb
leads to exaggerated cytokine production and intestinal disease. PIR-B is expressed on various cell populations including macrophages, neutrophils, mast cells, eosinophils, DC- and B-cells 28
. Notably, PIR-B negatively regulates neutrophil and eosinophil function 19, 29
. Given that these cell types have a role in the pathophysiology of DSS-induced colitis 16, 30, 31
, we cannot rule out that hyper-activated neutrophils and eosinophils may contribute in part to the increase disease severity in Pirb-/-
mice. However, our adoptive transfer experiments demonstrate that loss of PIR-B expression and negative regulation of macrophage activation is sufficient to enhance susceptibility to DSS-induced colitis. Interestingly, transfer of WT BM-macrophages into Pirb-/-
mice delayed disease progression and had a protective effect. There are two possible explanations for these observations. First, WT BM-macrophages could be displacing Pirb-/-
macrophages in the colon and thus reducing the pro-inflammatory environment and disease pathology; alternatively, WT BM-macrophages in Pirb-/-
mice possess anti-inflammatory activity and negatively regulate Pirb-/-
macrophage activation and suppress DSS-induced colitis. Previous studies have demonstrated that myeloid-specific STAT-3 deficient mice develop spontaneous entero-colitis 32
. Susceptibility to spontaneous entero-colitis was linked with a loss of anti-inflammatory IL-10:Stat3 signaling in macrophages 32
. The involvement of altered IL-10:Stat3 signaling in exaggerated Pirb deficient macrophage proinflammatory cytokine production is currently under investigation.
The ligands of PIR-B are not yet fully delineated 27
. Initial studies suggested that MHC class I (H-2) molecules are ligands for PIR-B; however, recent studies indicate interaction with S. aureus
(Gram-positive) and E. coli
. We show that stimulation of WT and Pirb-/-
macrophages (resident or inflammatory) with E.coli
-derived LPS induced an equivalent increase in IL-6 and IL-1β production and MAPK/NFκB phosphorylation (data not shown). Whereas, activation of Pirb-/-
macrophages with E. coli
increased proinflammatory cytokine production and MAPK activation. These studies indicate that PIR-B may not be directly activated by TLR-4 ligands such as LPS, but rather inhibits TLR-mediated activation in response to bacterial stimulation. Consistent with this concept, TLR-2 agonist PAM3
stimulation of WT and Pirb-/-
macrophages led to equivalent cytokine production, however activation with whole S. aureus,
which exerts its proinflammatory effects primarily via TLR-2, led to exaggerated IL-6 and TNF-α production 14
. While further investigation is required to define the bacterial cell wall components that bind and/or activate PIR-B, recent studies suggest PIR-B may possess scavenger receptor-like binding activity towards bacteria.
TLR ligands and proinflammatory cytokines including IL-1β and TNFα are thought to activate macrophage MAPK (p38, ERK1/2 and JNK) and NFκB pathways, promoting gene expression and cytokine production, leading to cytokine-mediated inflammation and IBD 33
. Consistent with this, both the MAPKs (p38, JNK and ERK1/2) and NFκB are significantly activated in the inflamed colonic mucosa of IBD patients 34
. Our mechanistic analysis revealed that PIR-B negatively regulates bacterial-induced macrophage activation of MAPK-, primarily ERK1/2- and p38- and to lesser extent the NFκB-phosphorylation. Pharmacological blockade of MAPK activation, specifically p38, improved disease activity and histology disease score in a DSS-model of colitis35
. Unexpectedly, we observed that Pirb-/-
macrophages have a substantial defect in JNK phosphorylation and hallmark target of JNK, c-Jun. Previous studies have shown that Pirb-/-
myeloid dendritic cells are of immature phenotype and are unable to upregulate MHC-II molecules, a process that is dependent on JNK and c-Jun stimulation 15
. Notably, the observed elevated IL-6, TNF-α and IL-1β in Pirb-/-
macrophages following bacterial stimulation occurred in the absence of heightened JNK:c-Jun activation. These data suggest that increased activation of p38 and ERK1/2 in the absence of PIR-B is sufficient to regulate increased macrophage activation. Similarly, PECAM-1, another inhibitory receptor has also been shown to have divergent inhibitory effects on downstream signaling intermediates (inhibits IκB, and JNK; enhances ERK activation) in activated macrophages 36
. The dependency of global attenuation of the JNK:c-Jun pathway or negative regulation of DSS-induced ERK1/2- and p38-activation by PIR-B on exacerbation of colitis is yet to be defined.
The inhibitory activity of PIR-B has been associated with the recruitment of SHP-1 and -2 following activation 12, 23, 24
. We demonstrate that stimulation of macrophages with E. coli,
induced tyrosine phosphorylation of PIR-B and recruitment of the phosphatases SHP-1 but not -2. These findings clearly indicate that PIR-B is activated upon ligation with E. coli
and suggest that SHP-1 may be involved in the negative regulation of ERK1/2 and p38 by PIR-B. SHP-1 can directly and indirectly negatively regulate MAPK kinase (ERK and JNK) activation37
. Nitric oxide-induced dephosphorylation of ERK1/2 in rat vascular smooth muscle cells was associated with SHP-1 interaction and activation. Notably, ERK1/2 dephosphorylation was attenuated by a protein phosphatase 1 (SHP-1) inhibitor 38
. Furthermore, SHP-1 dephosphorylates vascular endothelial growth factor-induced ERK phopshorylation in endothelial cells 39
. In contrast to PIR-B, SIRP-1α, another ITIM-bearing receptor, inhibits LPS/TLR-4-mediated signaling primarily through sequestering SHP-2 but not -1 40
suggesting that different inhibitory receptors may utilize divergent intracellular phosphatases to elicit their inhibitory effect.
A limitation of these in vitro macrophage analysis is that all studies were conducted on BM-derived or thioglycolate-elicited macrophages and not intestinal macrophages. We would have preferred to use intestinal macrophages, however purification of intestinal macrophages is technically challenging, often complicated by significant cellular contamination and insufficient cell yields for in vitro analyses. Consistent with our in vitro analyses, we demonstrate PIR-B expression on intestinal macrophages and show that in vivo transfer of BM-macrophages leads to elevated cytokine production and exacerbates DSS-induced colitis suggesting that PIR-B negatively regulates intestinal macrophage activation.
We demonstrated the expression of PIR-B human orthologues ILT-2/LIR-1 and ILT-3/LIR-5 in the colonic mucosa of normal healthy control and UC patients. ILT-2/CD85J was primarily expressed by CD68+
mononuclear cells within the lamina propria, whereas ILT-3/CD85K was localized to the crypt epithelium. The differential expression of PIR-B human orthologues suggests that these molecules may negatively regulate both hematopoietic and non-hematopoietic cell function. The gastrointestinal tract is a tightly regulated immune organ that possesses suppressive immune mechanisms that prevent uncontrolled proinflammatory reactions towards enteric flora. We speculate that ILT-2/CD85J and ILT-3/LIR-5 may play a role in the maintenance of intestinal mononuclear cell and epithelial cell immune quiescence or non-responsiveness. We did not observe differences in ILT-2/LIR-1 or ILT-3/LIR-5 expression in the colonic mucosa of normal and UC patients. Consistent with this observation, levels of LIRs (ILT-1, ILT-4 and ILT-5) on monocytes from rheumatoid arthritis (RA) patients were comparable to that of sex- and aged-matched control subjects 41
. The mechanism of LIRs modulation of inflammation remains unclear, however it is postulated that LIRs regulate the threshold for activation of inflammatory cells and determine the severity of inflammation 42
via the relative balance of activating or inhibitory LIR's expressed on a particular cell. Thus assessment of expression patterns of ILT/LIR family members in IBD will be important in defining their relative contribution to human gastrointestinal homeostasis including the exacerbation and contraction of the intestinal inflammatory response and mucosal recovery.
In summary, our results define a key role for PIR-B in the regulation of inflammatory macrophage activation during colonic inflammation predominantly by negative regulation of bacterial induced-ERK and -p38 activation. We confirm that the human homologues of PIR-B are expressed on both immune and epithelial cells in the inflamed and normal colon. These data highlight inhibitory receptors such as PIR-B as novel targets for suppression of macrophage functions in inflammatory settings such as IBD.