Pentraxin-3 modulates lipopolysaccharide-induced inflammatory response and attenuates liver injury

PATIENTS

Samples from 54 patients with clinical, analytical, and histological features of AH were selected. These cases are part of our historic cohort of patients with AH and controls with cirrhosis admitted to the Liver Unit of the Hospital Clinic of Barcelona from July 2009 to January 2015. The study was approved by the Ethics Committee of the Hospital Clinic of Barcelona, and informed consent was obtained from all patients before inclusion. Inclusion and exclusion criteria have been described.4

IN VITRO STUDIES

Human primary HSCs were isolated from control patients and cultured as described.32 For gene expression analysis by quantitative PCR, total RNA was isolated using the RNeasy Mini Kit (Qiagen) per the manufacturer's protocol.

EFFECT OF PTX3 IN HUMAN MONOCYTE INFLAMMATORY RESPONSES AND POLARIZATION

Buffy coats, provided by the Blood and Tissue Bank (Barcelona, Spain), were obtained from healthy blood donors following the institutional standard operating procedures for blood donation and processing. Peripheral blood mononuclear cells were isolated from donors as described.33 The percentage of CD14⁺ cells (peripheral blood monocytes) routinely obtained was 5% (±3%).

EXPERIMENTAL MOUSE MODELS OF LIVER DAMAGE

Experimental models were performed using male C57BL/6 mice 6-12 weeks old (Charles River, l'Arbresle, France). All animal studies were approved by the Ethics Committee of the University of Barcelona. To induce chronic liver injury, mice were treated with CCl₄ injected intraperitoneally at a dose of 0.5 mL/kg twice a week; control mice were injected with corn oil. In order to study the effects of acute-on-chronic liver injury, we performed a model combining the damage of chronic CCl₄ with administration of LPS (Sigma-Aldrich), mimicking the outcome of endotoxemia in the situation of chronic liver disease.34

The role of PTX3 during acute-on-chronic liver injury was investigated in vivo in chronic CCl₄ (0.5 mL/kg twice a week for 2 weeks)–treated mice. Intraperitoneal administration of recombinant PTX3 (rPTX3; 5 mg/kg body weight) was performed 2 hours before intravenous injection of LPS (2.5 mg/kg); control mice were injected intraperitoneally with corn oil and intravenously with vehicle. Mice were sacrificed 24 hours after LPS injection, and blood, livers, lungs, and kidneys were removed for subsequent analysis.

See the Supporting Information for more details.

PTX3 EXPRESSION IS INCREASED IN EXPERIMENTAL MODELS OF CHRONIC AND ACUTE-ON-CHRONIC LIVER INJURY

PTX3 is rapidly induced after injury in several tissues; therefore, we evaluated the expression of PTX3 in animal models of chronic liver disease and acute-on-chronic liver injury. Administration of CCl₄ to mice, a well-established animal model of chronic liver injury and fibrosis, induced a time-dependent increase in PTX3 liver expression (Fig. 1A). Liver cirrhosis is frequently associated with increased gut permeability and endotoxemia with elevated circulating levels of LPS.2, 3 Therefore, we evaluated the effect of LPS administration on PTX3 expression. Interestingly, while the administration of LPS to healthy animals did not induce PTX3 expression (Fig. 1B), LPS infusion in CCl₄-treated animals, a model of acute-on-chronic liver injury, induced a marked increase in hepatic PTX3 expression (Fig. 1C). Interestingly, Tlr4 hepatic expression was increased in CCl₄-treated animals but not further increased by LPS stimulation (Supporting Fig. S1). These results suggest that preexisting liver injury may be necessary to mediate LPS-induced PTX3 expression.

In order to elucidate the cell source of PTX3 in chronic liver injury, we evaluated PTX3 expression by immunohistochemistry. PTX3 expression in mice with CCl₄-induced chronic liver injury plus LPS was mainly localized at inflammatory areas and sinusoidal cells (Fig. 1D). Furthermore, we assessed Ptx3 gene expression in different liver cell sorted populations, namely hepatocytes, HSCs, macrophages, neutrophils, and T cells isolated from healthy mice and those with acute-on-chronic liver injury. Interestingly, while the main producing cell type in healthy livers was neutrophils, macrophages and particularly HSCs were the main cell types expressing PTX3 in injured livers (Fig. 1E). In order to evaluate the purity of sorted populations, we evaluated the gene expression of Alb, Mpo, Cd3e, F4/80, and Col1a1 in each of the populations obtained, which showed high enrichment (Supporting Fig. S2).

PTX3 ENHANCES HSC ACTIVATION AND IS EXPRESSED IN ACTIVATED HSCs

HSCs are the main cell type producing PTX3 in mouse models of liver injury. Therefore, we assessed the expression and effects of PTX3 in human HSCs, the main cell type responsible for the wound-healing response of the liver.35 We first evaluated the expression of PTX3 in human quiescent HSCs freshly isolated from normal livers, followed by their activation in culture. Upon culture activation, HSCs strongly increased the expression of PTX3 compared to quiescent HSCs (Fig. 2A). Moreover, proinflammatory mediators such as tumor necrosis factor alpha (TNF-α), interleukin 1β (IL-1β), and LPS enhanced PTX3 expression in human primary activated HSCs compared to vehicle (P < 0.05) (Fig. 2B).

Stimulation of human primary activated HSCs with rPTX3 induced an up-regulation of HSC activation marker and extracellular matrix genes (COL1A1, ACTA2, CTGF, LOX) (Fig. 2C) but did not significantly change the expression of proinflammatory genes (TNF-α, chemokine [C-C motif] ligand 20 [CCL20], MCP-1) (Fig. 2C). Moreover, rPTX3 induced extracellular signal–regulated kinase 1/2 and AKT phosphorylation in HSCs, suggesting that PTX3 induces intracellular signaling in HSCs (Fig. 2D). To further investigate the biological effect of PTX3 on HSCs, we performed a wound-healing assay with primary HSCs. As a positive control, cells were stimulated with platelet-derived growth factor subunit B, a main mitogen for HSCs. Incubation with rPTX3 increased HSC migration and wound closure (Fig. 3A). Next, we evaluated the effects of silencing PTX3 in the LX2 human HSC cell line by small interfering RNA (siRNA). Transfection of HSCs with an siRNA targeting PTX3 reduced PTX3 mRNA and protein expression compared to scramble siRNA (Fig. 3B). HSCs treated with PTX3 siRNA significantly reduced the expression of activation genes such as COL1A1, ACTA2, and LOX but did not reduce the expression of inflammatory mediators (Fig. 3C). Results from recombinant PTX3 stimulation and silencing of endogenous PTX3 expression suggest that both exogenous as well as endogenous PTX3 induce HSC activation. In order to elucidate if PTX3 might exert its role on HSCs through Fc gamma receptors (FcγRs), we investigated their expression in primary human HSCs. Gene expression of FcγR-1 and FcγR-2 was not detected in cultured HSCs. Expression of FcγR-3, the FcγR with the highest affinity for PTX3,22 was assessed by flow cytometry, with no expression shown in cultured HSCs (Supporting Fig. S3A). Taken together, these results indicate that PTX3 is expressed in HSCs and involved in their activation, reinforcing its potential role in the liver wound-healing response.

PTX3 MODULATES LPS-INDUCED LIVER INJURY AND INFLAMMATION

PTX3 was highly expressed in LPS-induced acute-on-chronic liver injury. Therefore, we explored the potential role of PTX3 in LPS-induced inflammation in the context of chronic injury. First, we evaluated the effect of PTX3 in high precision-cut liver slices from mice treated with CCl₄. Liver slices were incubated in the presence or absence of rPTX3 and stimulated with LPS. rPTX3 alone did not induce any significant effect; however, incubation with rPTX3 was able to attenuate the expression of inflammatory mediators such as Ccl20, Ccl5, and Mcp-1 induced by LPS (Fig. 4A). Moreover, incubation with rPTX3 also reduced lactate dehydrogenase (LDH) and aspartate aminotransferase levels in tissue culture supernatant (Fig. 4B). Taken together, these results suggest that PTX3 may protect against LPS-induced liver damage by attenuating the expression of proinflammatory cytokines and hepatocellular damage.

Next we evaluated the role of PTX3 in vivo in an acute-on-chronic hepatic injury model of CCl₄-induced liver injury plus LPS (Fig. 5A). Treatment with rPTX3 before LPS stimulation substantially reduced liver damage as assessed by LDH, aspartate aminotransferase, and alanine aminotransferase serum levels (Fig. 5B). In addition, PTX3 treatment reduced the inflammatory response by down-regulating Ccl20, Mcp-1, Ccl5, IL-1β, Tnf-α, Cxcl2, and Nos-2 gene expression (Fig. 5C). Moreover, macrophage M1 inflammatory mediators (i.e., Mcp-1, IL-1β, Tnf-α) but not M2 markers (Mrc1, Arg1, and IL-4) were reduced in PTX3-treated animals (Fig. 5C). Moreover, histological evaluation showed a significant reduction of neutrophil and macrophage liver infiltration in mice treated with rPTX3 (Fig. 5D). This reduction in recruitment was not associated with changes in Fcγ receptor expression between experimental groups (Supporting Fig. S3B). It is important to note that chronic liver injury and endotoxemia are associated with extrahepatic organ injury.6 In this regard, rPTX3 administration significantly reduced the expression of LPS-induced inflammatory mediators and neutrophil infiltration in the kidney and the lung (Supporting Fig. S4A-C). Together, the results from ex vivo and in vivo experiments suggest that PTX3 administration exerts a hepatoprotective effect and modulates LPS-induced liver injury and inflammation.

PTX3 REGULATES INFLAMMATION BY TARGETING MONOCYTES AND LIVER MACROPHAGES

Our data suggest that PTX3 is able to attenuate inflammatory responses to LPS in liver as well as the inflammatory cell infiltrate. However, because no direct anti-inflammatory effect was observed on HSCs, we analyzed whether PTX3 could be targeting macrophages. First, we evaluated whether PTX3 was able to block the TLR4-LPS response in peripheral blood monocytes isolated from healthy donors. Interestingly, while rPTX3 clearly reduced the early response to LPS by decreasing TNF-α and IL-1β production, it increased the production of CCL5 (Fig. 6A). Second, we evaluated the effect of rPTX3 in human liver macrophages. As shown in monocytes, stimulation of liver macrophages with rPTX3 reduced LPS-induced TNF-α and IL-1β secretion in cell culture media and induced a marked increase in CCL5 production (Fig. 6B). It has been shown that in macrophages PTX3 modulates the inflammatory response to infection by Aspergillus conidia by binding to MD2, the TLR4 accessory protein.26 We therefore studied the ability of PTX3 to interfere with MD2 activity in a widely used cellular system. We transfected human embryonic kidney (HEK-293) cells stably expressing TLR4 or TLR4 plus CD14 with MD2. rPTX3 did not significantly affect the LPS response in TLR4 and TLR4/CD14 human embryonic kidney cells transfected with MD2, suggesting that in this system PTX3 may not prevent an MD2 contribution to LPS activation through TLR4 and TLR4–CD14 complexes (Supporting Fig. S5).

Next, we evaluated whether PTX3 influenced the polarization of monocytes into macrophages. As control stimuli, we treated macrophages with interferon/LPS to induce an M1 phenotype (CD80-positive); IL-4 (M-IL-4) to induce an M2a phenotype (CD2-positive and CD206-positive); and IL-10 (M-IL-10) to induce M2c (CD163-positive and CD36-positive).36 As observed by comparison with all the different stimulations, rPTX3 modulated the expression of polarization markers in a similar way as IL-10: it induced an increase in CD163 and CD36 expression, suggesting that it may enhance the acquisition of an IL-10-like macrophage phenotype. Conversely, the use of human albumin as a negative control protein did not substantially alter the macrophage phenotype (Fig. 6C). Moreover, we also evaluated the potential of PTX3 to induce M2 phenotype in liver macrophages. Human liver macrophages stimulated with rPTX3 showed an increase in expression of M2 markers CD206, CD163, and Mertk and did not induce M1 marker CD80 (Fig. 6D).

Taken together, these results suggest that PTX3 may modulate the LPS signaling response in monocytes and liver macrophages.

PTX3 IS UP-REGULATED IN PATIENTS WITH ACUTE-ON-CHRONIC LIVER DISEASE AND ASSOCIATED WITH DISEASE SEVERITY

AH is an acute-on-chronic condition characterized by liver injury, fibrosis, and important inflammatory responses, in which endotoxemia is associated with a bad outcome.2, 3 Therefore, PTX3 liver expression and plasma levels were evaluated in patients with alcoholic cirrhosis and AH.

PTX3 gene expression was increased in liver tissue from alcoholic patients with cirrhosis and strongly up-regulated in AH patients (P < 0.01) (Fig. 7A), consistent with PTX3 expression in experimental models of liver injury (Fig. 1A,B).37, 38 Plasma levels of PTX3 were also increased in compensated alcoholic patients with cirrhosis and further increased in patients with both mild and severe AH (Fig. 7B). Interestingly, PTX3 hepatic gene expression and peripheral plasma levels showed a positive correlation, suggesting that the liver may be an important source of circulating PTX3 (Fig. 7C). PTX3 staining in alcoholic cirrhosis was localized in septal areas with fibrosis and inflammation (Fig. 7D). We next assessed the association of PTX3 plasma levels with clinical parameters and scores of disease severity. We found a positive correlation of PTX3 plasma levels with the Model for End-Stage Liver Disease (MELD) (P < 0.001) and age, serum bilirubin, international normalized ratio, and serum creatinine (ABIC) scores (P < 0.01) (Fig. 7E). Moreover, PTX3 plasma levels positively correlated with Maddrey score (P < 0.01) and procalcitonin but not with ultrasensitive C-reactive protein serum levels and the hepatic venous pressure gradient. Importantly, PTX3 plasma levels correlated with LPS serum levels (Fig. 7F), indicating a potential association of PTX3 expression with endotoxemia. However, PTX3 plasma levels did not correlate with IL-10, suggesting that PTX3 may not be associated with IL-10 expression and immune exhaustion (Supporting Fig. S6C). Additionally, PTX3 levels were higher in patients with infection at admission and in those who developed in-hospital infections compared to noninfected patients (Supporting Fig. S6B). Next, we evaluated the association of PTX3 plasma level with 6-month mortality. PTX3 serum levels showed an area under the receiver operating characteristic curve of 0.76 (95% confidence interval 0.69-0.89; P = 0.002) for predicting 6-month mortality. The Kaplan-Meier analysis showed that PTX3 levels >10 ng/mL were associated with a higher mortality at 180 days (Fig. 7G). This cutoff showed a sensitivity of 78% and a specificity of 60%. Finally, to evaluate the potential prognostic impact of PTX3 on 180-day mortality, we fitted a bivariate Cox regression (hazard ratio) analysis, adjusting PTX3 plasma levels for AH severity-scoring systems at admission (e.g., ABIC score, MELD, and Lille model). PTX3 plasma levels were independently associated with 180-day mortality after adjusting for the well-validated scoring systems (P < 0.05 for all cases; Supporting Table S2).

These results suggest that PTX3 is strongly associated with disease severity, and its expression may be enhanced as a result of acute-on-chronic liver injury and endotoxemia.