Shutting off inflammation: A novel switch on hepatic stellate cells

Potential conflict of interest: Nothing to report. F. J. C. is a Ramón y Cajal Researcher (RYC‐2014‐15242). See Article on Page 1325 Over 100 years have passed since Kupffer's first observations of the "Sternzellen" (star cells) in the liver until Friedman and others1 comprehensively isolated and characterized the currently known hepatic stellate cells (HSCs). Because of their privileged location, HSCs have attracted considerable attention. HSCs, which account for 5%‐8% of the cells in the liver, are located in the perisinusoidal space of Disse, between the endothelial lining and the hepatocytes, a position compatible with a starring role following hepatic insult. Nevertheless, our current knowledge of the biology of HSCs is still controversial, mainly because of the unknown functions of HSCs during hepatic inflammation. Under pathological conditions, and after receiving signals secreted by Kupffer cells (KCs), liver sinusoidal endothelial cells (SECs), damaged hepatocytes, biliary epithelial cells, and immune cells, HSCs orchestrate a response that involves their trans‐differentiation into activated stellate cells. The so‐called myofibroblasts proliferate and newly express several proinflammatory and profibrogenic genes, thus generating a temporary scar at the site of injury to protect the hepatic parenchyma from further damage. Therefore, the intercellular communication between HSCs and the remaining nonparenchymal and parenchymal liver cells turns out to be crucial for the outcome of the acute inflammatory response. Consequently, producer cells release chemical mediators that "switch on" specific regulatory mechanisms in other cells. These "tissue hormones" are collectively termed eicosanoids, comprising prostanoids [prostaglandins and thromboxanes], leukotrienes, and hydroxyl acids and are derived from polyunsaturated fatty acids, particularly arachidonic acid. It has become increasingly clear that prostanoids are particularly important intercellular chemical messengers during inflammation. However, little data are available regarding the role of prostanoid synthesis and function in liver pathophysiology. In an attempt to further characterize the critical role of HSCs during inflammatory liver disease, Fujita et al.3 investigated the actions of prostaglandin D2 (PGD2) on HSCs. PGD2 is the major prostanoid formed in the liver, where KCs and SECs are responsible for 95% of the total PGD2 production.4 PGD2 activity is mainly mediated through the prostanoid receptor DP1, a G‐protein‐coupled receptor which activates adenylate cyclase, increasing the concentration of cyclic adenosine monophosphate (cAMP), activating protein kinase A (PKA) and exerting pro‐ or anti‐inflammatory properties depending on the tissue and the context.5 Moreover, Fennekohl et al.6 first reported that murine HSCs displayed the highest DP1 mRNA levels in all nonparenchymal cells in culture. In the Fujita paper, these data have now been expanded to the protein level and, interestingly, to human HSCs.3 In fact, the authors show that DP1 agonist, BW245C, attenuates tumor necrosis factor (TNF)‐mediated human HSC activation (e.g., α‐smooth muscle actin) in culture via a cAMP‐PKA‐dependent pathway.1Figure 1: Conceptual overview of the modulation of HSC activation by DP1 agonism. Left: The "ON switch." Acute inflammation results in the activation of KCs and the release of proinflammatory cytokines (TNF, IFN‐γ) as well as increased expression of adhesion molecules such as intercellular adhesion molecule 1 (ICAM1) and vascular adhesion molecule‐1 (VCAM1) on SECs, KCs, and HSCs. This allows the infiltration of leukocytes and the ET1‐mediated hepatic sinusoidal vasoconstriction. ET1, partly autocrine‐derived, acts via the endothelin type A receptor (ETA) on HSCs, orchestrating a response that involves inducible nitric oxide synthase‐derived hepatocyte injury and transdifferentiation of HSC into myofibroblasts. Right: The "OFF switch." Binding of KC‐derived PGD2—the major prostanoid formed in the liver—to its receptor DP1 on HSCs attenuates leukocyte infiltration, expression of adhesion molecules, intrahepatic cytokine release, sinusoidal stagnation, and hepatocyte injury. Moreover, agonism of the PGD2/DP1 axis with BW245C, a DP1‐selective agonist, inhibits HSC‐derived production of ET1 and overall HSC activation and contractility.A key finding in their experiments with human HSCs is the molecular mechanism behind the protective action of prostanoid receptor agonism by BW245C. DP1 signaling suppressed the proinflammatory effect of TNF by facilitating the reinduction of inhibitor of κBα, interfering with nuclear factor kappa B (NF‐κB) activation, sequestering nuclear p65 protein in the cytosol, and suppressing C‐Jun N‐terminal kinase (JNK) phosphorylation. Consequently, activation of NF‐κB plays a pivotal role in mediating the TNF‐derived proinflammatory effects by modulating JNK activation in HSCs. Importantly, NF‐κB activation is also required to induce cyclooxygenase 2 (COX‐2), the key inducible enzyme responsible for the production of prostanoids. Thus, the interaction between NF‐κB, COX‐2, and mitogen‐activated protein kinases signaling pathways warrants further research on inflammation induced‐liver fibrosis. In order to evaluate the in vivo role of DP1 agonism in HSC in liver inflammation, Fujita et al.3 used the concanavalin A (ConA) model. This experimental model of hepatitis is characterized by the presence of inflammatory cytokines, including TNF and interferon γ (IFN‐γ), which results in massive liver necrosis accompanied by leukocyte infiltration. The authors found that ConA‐induced hepatitis was exacerbated in the liver of constitutive DP1‐deficient mice. Additionally, the effects of prostaglandin receptor deficiency beyond DP1 were also investigated. COX‐2 KO mice developed severe hepatitis upon ConA treatment, whereas mice deficient in thromboxane receptor or EP3—a receptor for prostaglandin E2—exhibited significantly ameliorated hepatitis.3 More importantly, Fujita et al.3 found that DP1 agonism with BW245C exerts a hepato‐protective action in ConA‐induced hepatitis, confirming their in vitro observations. Following inflammation or injury, leukocytes become tethered to the vessel wall through labile adhesions that permit their "rolling" in the direction of flow. This process is mediated by adhesion molecules such as intercellular adhesion molecule 1 or vascular adhesion molecule 1 expressed both by SECs and activated HSCs.7 Chemoattractants, especially chemokines such as TNF, IFN‐γ, interleukin 1, monocyte chemoattractant protein 1, and cytokine‐induced neutrophil chemoattractant, produced by HSCs and KCs at the site of tissue damage, establish a gradient surrounding the inflammatory area.8 Agonism of the prostainoid receptor DP1, acting presumably on HSC, suppressed the levels of VCAM1 and intrahepatic inflammatory chemokines. Moreover, the finding that PGD2/DP1 signaling inhibits T cell infiltration to the liver parenchyma, confirms previous studies demonstrating that HSCs are directly involved in T cell response,10 reinforcing the immunomodulatory role of HSCs and their likely contribution to hepatic immune tolerance. Furthermore, the hepatic microcirculation is controlled at least in part by the vasoconstricting effects of endothelin.1 Electron, fluorescent, and intravital microscopy have shown that endothelin locally regulates the hepatic sinusoidal microcirculation. In ConA‐induced hepatitis, agonism of DP1 with BW245C decreases HSC production of endothelin, which acts on the endothelin type A receptor. These findings further indicate that morphologically and functionally SECs and HSCs are closely associated in the control of the sinusoidal blood flow, and this mechanism is very likely regulated by prostanoids. However, a crucial point in the article by Fujita et al.3 is the mechanism by which PGD2 or the DP1 agonist BW245C exert their action on HSCs and what, in turn, HSCs produce to achieve the aforementioned protective effects. The authors suggest that BW245C inhibits HSC‐derived production of endothelin 1 (ET1), the major contractile stimulus toward HSC in vivo. This, in turn, inhibits microcirculatory changes in the sinusoid that can initiate the aforementioned cascade. It should be noted that the mechanism triggered by PGD2/DP1 as described by Fujita et al.3 was only observed during acute hepatitis models; furthermore, the relative contribution of prostainoids and their receptors in other animal models where inflammation is a key event to the pathophysiology of the disease (e.g., drug‐induced liver injury, nonalcoholic steatohepatitis, and hepatocellular carcinoma) must be determined before it can be applied in a clinical setting. Perhaps the generation of DP1‐floxed mice specific for HSCs should be a major goal in order to develop a useful preclinical model. Additionally, COX‐2 inhibitors or the blockade of other G‐coupled receptors as DP1 might open new gates for therapeutic intervention to limit the progression of liver fibrosis. The primary message of the article written by Fujita et al.3 is that cytokine production (e.g., TNF, IFN‐γ) by KCs and a number of downstream events can be suppressed by DP1 agonism, which presumably acts on HSCs. Thus, KC‐mediated production of PGD2 via HSCs may represent a means of "shutting off" the acute inflammatory response in the liver. They show that pharmacologic treatment with BW245C, a DP1 agonist, prevents T cell recruitment, sinusoidal constriction and stasis, and inducible nitric oxide synthase‐derived hepatocellular damage. In other words, the hepatoprotective role of prostanoids opens a new field of research for identifying strategies and developing interventions to limit hepatic fibrogenesis.