The common lymphatic endothelial and vascular endothelial receptor (CLEVER-1; also known as FEEL-1 and stabilin-1) is a recycling and intracellular trafficking receptor with multifunctional properties. Here we demonstrate for the first time increased endothelial expression of CLEVER-1/stabilin-1 at sites of leucocyte recruitment to the inflamed human liver including sinusoids, septal vessels and lymphoid follicles in inflammatory liver disease and tumour-associated vessels in hepatocellular carcinoma. We used primary cultures of human sinusoidal endothelial cells (HSEC) to demonstrate that CLEVER-1/stabilin-1 expression is enhanced by hepatocyte growth factor but not by classical proinflammatory cytokines. We then showed that CLEVER-1/stabilin-1 supports T cell transendothelial migration across HSEC under conditions of flow with strong preferential activity for CD4 FoxP3+ regulatory T cells. CLEVER-1/stabilin-1 inhibition reduced Treg transendothelial migration by 40% and when combined with blockade of ICAM-1 and vascular adhesion protein-1 (VAP-1) reduced it by more than 80%. Confocal microscopy demonstrated that 60% of transmigrating Tregs underwent transcellular migration through HSEC via ICAM-1 and VAP-1 rich transcellular pores in close association with CLEVER-1/stabilin-1. Thus CLEVER-1/stabilin-1 and VAP-1 may provide an organ-specific signal for Treg recruitment to the inflamed liver and to hepatocellular carcinoma.

Fas-mediated mechanisms of apoptosis are thought to be involved in the bile duct loss that characterizes diseases such as primary biliary cirrhosis (PBC). We have previously shown that activation of CD40 on hepatocytes can amplify Fas-mediated apoptosis; in the present study, we investigated interactions between CD40 and Fas in biliary epithelial cells (BEC). We report that the bile ducts in PBC liver tissue frequently express increased levels of Fas, Fas ligand (FasL), and CD40 associated with apoptotic BEC. The portal mononuclear infiltrate contains CD40L+ve T cells and macrophages, thereby demonstrating a potential mechanism for CD40 engagement in vivo. Primary cultures of human BEC also expressed Fas, FasL, and CD40 but not CD40L protein or mRNA. Activation of CD40 on BEC using recombinant CD40L increased transcriptional expression of FasL and induced apoptosis, which was inhibited by neutralizing antibodies to either Fas or FasL. Thus, CD40-induced apoptosis of BEC is mediated through Fas/FasL. We then investigated the intracellular signals and transcription factors activated in BEC and found that NF-kappaB and AP-1 were both activated after CD40 ligation. Increased functional NF-kappaB was seen early after CD40 ligation, but returned to baseline levels after 4 h. In contrast, the rapid up-regulation of AP-1 was sustained over 24 h. This study provides further functional evidence of the ability of CD40 to induce Fas/FasL-dependent apoptosis of liver epithelial cells supporting the importance of cross-talk between tumor necrosis factor (TNF) receptor family members as an amplification step in apoptosis induction. Sustained activation of AP-1 in the absence of NF-kappaB signaling may be a critical factor in determining the outcome of CD40 engagement.

Liver dendritic cells (DCs), which may orchestrate the liver's unique immunoregulatory functions, remain poorly characterized. We used a technique of overnight migration from pieces of normal human liver and skin to obtain tissue-derived DCs with minimal culture and no additional cytokine treatment. Liver and skin DCs had a monocyte-like morphology and a partially mature phenotype, expressing myeloid markers, MHCII, and co-stimulatory molecules; but only the skin DCs contained a population of CD1a+ cells. Overnight-migrated liver DCs activated naïve cord blood T cells efficiently. Liver DCs produced interleukin (IL)-10 whereas skin DCs failed to secrete IL-10 even after stimulation and neither skin nor liver-derived DCs secreted significant amounts of IL-12p70. Compared with skin DCs, liver DCs were less effective at stimulating T-cell proliferation and stimulated T cells to produce IL-10 and IL-4 whereas skin DCs were more potent stimulators of interferon-gamma and IL-4. Monocyte-derived DCs were down-regulated after culture with liver-conditioned media, suggesting that local microenvironmental factors may be important. Thus we show for the first time clear tissue-specific differences in nonlymphoid DCs. Although it is not possible to conclude from our data whether liver DCs are more regulatory, or skin DCs more proimmunogenic, the ability of liver DCs to secrete IL-10 may be important for regulating local immune responses within the liver in the face of constant exposure to gut antigens.