IFI16 is involved in HBV-associated acute-on-chronic liver failure inflammation

All patient samples were collected in our hospital from June 2013 to December 2014. Peripheral blood samples were collected from 31 CHB patients and 13 patients with HBV-associated acute-on-chronic liver failure (HBV-ACLF); 16 sex-matched healthy individuals were enrolled as normal controls. Liver tissue samples were collected from 59 CHB patients who had liver biopsies, 17 HBV-ACLF patients who received a liver transplant, and 19 healthy liver transplant donors. CHB patients were defined as those who were HBsAg-positive for more than 6 months and that exhibited symptoms or signs of hepatitis and abnormal hepatic function on occasion or those that had the disease histologically confirmed. By contrast, the diagnosis of HBV-ACLF was based on clinical evidence of either ≥grade 2 hepatic encephalopathy, abrupt and obvious increase in ascites, spontaneous bacterial peritonitis, or hepatorenal syndrome; these clinical criteria were associated with recent development of jaundice [total bilirubin [TBIL] ≥10.0 mg/d or rapidly rising levels of TBIL(TBIL≥1.0 mg/dL/day)] and a prothrombin activity ≤40%. Patients with autoimmune liver diseases or cancer were excluded. All patients were seronegative for markers of hepatitis A, C, D, and E viruses and for HIV virus. The model for end-stage liver disease (MELD) score [14] was used to assess disease severity and prognosis. Alanine aminotransferase (ALT) and aspartate aminotransferase (AST)levels, bilirubin levels, international normalized ratio (INR), creatinine, HBV DNA load, and MELD score were measured. The study protocol was approved by the ethics committee of our hospital and written informed consent was obtained from each subject. The basic characteristics of the enrolled subjects are listed in Tables 1 and 2.

Mouse anti-human IFI16 monoclonal antibody (sc-137,970) was obtained from Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA. Kupffer cell (KC) marker rabbit anti-human CD68 (clone F7.2.38) polyclonal antibody, Natural killer (NK) cell marker rabbit anti-human CD56 (EP2567Y), liver sinusoidal endothelial cell (LESC) marker rabbit anti-human CD299 (EPR11211), dendritic cell (DC) marker rabbit anti-human CD11c (EP1347Y), and hepatic stellate cell (HSC) marker rabbit anti-human SMA (E184) monoclonal antibodies were purchased from ABclonal Biotech Co., Ltd. (Woburn, MA, USA) and Abcam’s RabMAb®technology (Cambridge, UK), respectively. Alexa Fluor® 488 Goat Anti-Mouse IgG and Alexa Fluor® 555 Goat Anti-Rabbit IgG antibodies were obtained from Life Technologies (Grand Island, NY,USA).

PBMCs were purified from EDTA-treated blood using a Ficoll density gradient. For intranuclear IFI16 staining, PBMCs (200 μL) were incubated in 800 μL of RPMI 1640 medium. The cells were treated with Human TruStain FcX, Nuclear Factor Fixation Buffer, and Nuclear Factor Permeabilization Buffer (Biolegend, San Diego, CA, USA) and stained with the IFI16 antibody and Alexa Fluor® 488 Goat Anti-Mouse IgG. Data collection was performed on a BDFACSCantoII (BD Biosciences, San Jose, CA, USA), and data files were analyzed using FlowJo software (FlowJo, LLC, Ashland, OR, USA). Statistical analysis was based on IFI16 staining of at least 10,000 gated cells.

The liver biopsy specimens were considered reliable when the liver specimen length was ≥1.5 cm. The liver biopsy specimens were fixed in formaldehyde, embedded in paraffin, cut into 4-μm sections, and conventionally stained with hematoxylin and eosin. Images were acquired using an Olympus Leica DM4000B microscope. According to the Scheuer scoring system, hepatic inflammation activity grade (G) was divided into G0 (no hepatic necroinflammation) and G1-G4 and hepatic fibrosis stage was divided into F0 (no hepatic fibrosis) and F1-F4. The degree of hepatic inflammation and fibrosis was graded using the modified histology activity index (HAI) [14] described by Scheuer. Liver specimens were interpreted by experienced liver pathologists. Sections of liver tissue were scored in a blinded fashion to assess their histological diagnoses.

Immunohistochemistry was carried out using standard techniques [15]. The expression of IFI16-positive cells was determined by image analysis of the histological sections. Ten sections from each sample were randomly selected, and photomicrographs were obtained under high-power fields (0.625 mm²) and captured for analysis using Image Pro-Plus 6.0 software (Media Cybernetics, Silver Spring, MD, USA). The number of IFI16-positive cells per high-power field was counted and expressed as the mean + the standard error of the mean (SEM).

For immunofluorescence double staining, paraffin-embedded tissue blocks were cut into 3.5-μm sections. After dewaxing and rehydration, antigen retrieval was performed with EDTA (pH 8.0). Then, sections were incubated at 4 °C overnight with primary anti-IFI16 (1:50) and cell type-specific marker antibodies anti-CD68 (1:100), anti-CD11c (1:50), anti-CD56 (1:50), anti-CD299 (1:100), and anti-SMA (1:1000). The sections were then incubated with Alexa Fluor® 488 Goat Anti-Mouse IgG and Alexa Fluor® 555 Goat Anti-Rabbit IgG Antibodies (1:1000) for 1 h. Finally, the sections were incubated with 1 μg/ml Hoechst 33,342 (Sigma-Aldrich, St. Louis, MO, USA) to stain the nuclei. Sections incubated with the PBS control primary antibodies and fluorescently labeled secondary antibodies were used as negative controls. The results were analyzed using fluorescence microscopy (Leica Corp.).

Cells were lysed in radioimmunoprecipitation assay (RIPA) buffer supplemented with protease inhibitor cocktail (Sigma-Aldrich), sonicated, and clarified by centrifugation. Equal amounts of protein were separated by SDS-PAGE and electrophoretically transferred to nitrocellulose membranes. Membranes were incubated with primary antibodies and secondary antibodies conjugated to horseradish peroxidase (KPL, Gaithersburg, MD, USA). Immunoreactive bands were visualized using an ECL western blotting substrate (Pierce Chemical, Rockford, IL, USA). Blots were scanned using FluorChemFC2 software with the AlphaImager system (Alpha Innotech Corp., San Leandro, CA, USA). Figures shown are representative of three or more experiments each.

We first examined the expression of IFI16 in PBMCs from 31 patients with CHB,13 patients with ACLF, and 16 healthy controls using flow cytometry. The number of IFI16-positive cells in patients with ACLF was significantly lower than the number in patients with CHB and in healthy controls (median [25th percentile; 75th percentile], 82.80 [62.53, 90.95] vs 93.00 [87.95, 95.00], and 91.63 [87.95, 95.00], P < 0.001 and P < 0.05; Fig. 1a). There was no difference between CHB patients and healthy controls(P > 0.05). Then, we tested for a relationship between IFI16 expression levels and HBV parameters. There was no significant difference between the HBeAg(+) and HBeAg(−) groups (P = 0.43),and no statistically significant relationships were found between IFI16 expression levels and HBV DNA load (r = 0.104, P = 0.501; Fig. 1 b-c). In addition, we tested for a relationship between IFI16 expression levels and HBV DNA load only in chronic HBV infected patients, there was no statistically significant relationships (r = 0.335, P = 0.065). To further identify whether IFI16 expression contributed to the severity of CHB and ACLF, we evaluated the correlations between IFI16 and clinical parameters, including serum ALT, AST, TBIL, prothrombin time activity (PTA), and INR, in patients with HBV. IFI16 expression levels were positively correlated with PTA (r = 0.336, P = 0.026; Fig. 1g) and negatively correlated with AST (r = − 0.334, P = 0.027), INR (r = − 0.339, P = 0.025), and TBIL (r = − 0.449, P = 0.002) (Figs. 1d–f). No statistically significant relationships were found between IFI16 expression levels and ALT (r = − 0.027, P = 0.861). These results suggest the possibility that aberrant expression of IFI16 might be involved in the development of CHB and ACLF, especially in ACLF patients, and may be closely associated with disease severity.

Furthermore, we evaluated the intrahepatic expression of IFI16 protein in 59 patients with CHB, 17 HBV-ACLF patients who received a liver transplant, and 19 healthy liver transplant donors. To measure the expression of IFI16 in the livers of CHB and ACLF patients, we used immunohistochemistry with anti-IFI16 to probe sections of paraffin-embedded livers. First, we saw different degrees of damage to the liver tissue from CHB and HBV-ACLF patients, as demonstrated by H&E staining (Fig. 2a). IFI16 expression that was observed in the livers of healthy donors and patients with HBV was mainly in non-parenchymal liver cells, especially the inflammatory cells in the livers of HBV patients. Furthermore, intrahepatic IFI16 expression gradually increased in accord with elevations in the grade of liver inflammation (Fig. 2a). Then, a semi-quantitative analysis of IFI16 expression in liver tissues was conducted using the software Image-Pro-Plus 6.0 to measure the number of IFI16-positive cells. The number of IFI16-positive cells in patients with ACLF was significantly higher than that in those with CHB or healthy controls (median [25th percentile; 75th percentile], 435.3 [372.9496.5] vs 199.6 [188.0, 307.1] vs160.8 [142.0, 218.0], both P < 0.0001; Fig. 2b). There was also no difference between the CHB patients and healthy controls (P > 0.05). In detail, the intrahepatic IFI16 expression in patients with inflammation grade 4 was significantly higher than in those with inflammation grade 0,1, or 2 (P < 0.0001; Fig. 2c). Moreover we found the intrahepatic IFI16 expression in patients with fibrosis stage 4 was also significantly higher than in those with fibrosis stage 0,1, or 2 (P < 0.0001; Fig. 2h). To further confirm our results obtained from the semi-quantitative immunohistochemistry, the protein level of IFI16 was analyzed by a western blotting analysis of cell extracts from liver tissues isolated from the different groups. As shown in Fig. 2d, the protein level of IFI16 in the HBV-ACLF group was remarkably higher than in the CHB(P = 0.006) and normal control groups (P = 0.001), and there is no difference between CHB and normal control groups (P > 0.05; Fig. 2e). In particular, this finding was consistent with the semi-quantitative immunohistochemistry levels for all groups. Then, we evaluated the correlations between IFI16 and clinical parameters, including HAI, serum AST, TBIL, PTA, and INR, in patients with HBV. IFI16 expression levels were positively correlated with HAI (r = 0.604, P < 0.001) (Fig. 2f), AST (r = 0.468, P < 0.001), INR (r = 0.562, P < 0.001), and TBIL (r = 0.515, P < 0.001) and negatively correlated with PTA (r = − 0.564, P < 0.001).We also found that there was no significant difference between the HBeAg (+) and HBeAg (−) groups (P = 0.773), and no statistically significant relationships were found between IFI16 expression levels and HBV DNA load (r = − 0.062, P = 0.590; Fig. 2g). Moreover, there was no statistically significant relationships were found between IFI16 expression levels and HBV DNA load only in chronic HBV infected patients (r = 0.169, P = 0.198).

The liver is well-known to be an immunological organ with a predominant innate immune system. The phenotypes of IFI16+ cells in liver tissues were further examined by immune fluorescence double staining. From these analyses, the expression of IFI16 in healthy donors and CHB patients was observed in the nucleus of CD299⁺ endothelial cells, CD68⁺KCs, CD56⁺ NK cells, CD11c⁺DCs, α-SMA⁺HSCs (Fig. 3), but this expression was completely absent from CD299⁺ endothelial cells, CD56⁺ NK cells, CD11c⁺DCs, and α-SMA⁺HSCs in HBV-ACLF patients. To our surprise, we found that IFI16 was expressed in the cytoplasm of CD68⁺KCs in HBV-ACLF patients (Fig. 4).