Cirsimaritin inhibits influenza A virus replication by downregulating the NF-κB signal transduction pathway

To determine the antiviral activity of CST, we initially studied the cytotoxicity of CST in MDCK and THP-1 cells by using the CCK assay. The results showed that concentrations of less than 25 μg/ml showed no cytotoxicity in the MDCK or THP-1 cells within an observation period of 48 h (Fig. 1b). CST concentration of 20 μg/ml was selected as the maximum concentration in the following antiviral assays.

The TC₅₀ values of CST and positive control of RBV in MDCK cells calculated by a CCK assay. By CPE assay we measured the antiviral activity of CST against influenza viruses in MDCK cells (Table 1). It showed that CST efficiently inhibited several influenza A virus strains, including A/tianjinjinnan/15/2009(H1N1) and A/JiangXi/312/2006(H3N2). The IC50 values of CST ranged from 5.8 to 11.1 μg/ml.

The antiviral efficacy of CST was also tested by using viral titers. We observed a dose-dependent reduction in viral titers when the MDCK and THP-1 cells were treated with CST after infection (Fig. 2a and b). The results indicated that CST inhibited viral replication. In addition, we evaluated the inhibition ability of CST against the influenza virus in MDCK and THP-1 cells by Western blot and qRT-PCR. Our results showed that CST dose-dependently reduced the amounts of IAV M2 protein and RNA in MDCK cells (Fig. 2c and e). Similarly, CST showed a potent antiviral activity against IAV in the THP-1 cells, as reflected by the dose-dependent decrease in protein expression and M2 RNA (Fig. 2d and f). To further confirm that CST inhibited viral protein synthesis, we analyzed the expression of the viral M2 protein through indirect immunofluorescence assay. In Fig. 2g, CST exhibited a dose-dependent inhibition of the M2 protein expression in MDCK cells. Collectively, CST demonstrated a potent antiviral activity against the influenza virus.

To elucidate the antiviral mechanism of CST, we initially determined whether CST restrained the function of the two IAV envelope glycoproteins (HA and NA), which are required for virus attachment and release [17]. However, the results showed that CST and RBV did not affect the hemagglutinin by the influenza virus (Fig. 3a). The findings also revealed that CST did not exert an inhibitory effect against NA, whereas OC, a known NA inhibitor, displayed significant inhibitory activity (Fig. 3b).

NF-κB is a nuclear transcription factor shown to play an important role in suppressing inflammation, oxidative stress, and host immunity [18]. In addition, accumulating evidence showed that downregulating NF-κB inhibits the replication of diverse species of viruses, including the influenza virus [11]. In this study, we found that CST dose-dependently decreased the level of intracellular p65/NF-κB protein in both infected and uninfected THP-1 cells (Fig. 4a and b).

To further determine whether CST disrupted IAV replication through the NF-κB signal pathway, we investigated the effects of CST on the phosphorylation of p65, a functional subunit of the NF-κB complex in the cytoplasm and nucleus. Initially, we found that CST inhibited the p65/NF-κB protein expression at 30 min of treatment in the nucleus. In addition, CST decreased the p65/NF-κB phosphorylation in the nucleus and cytoplasm (Fig. 4c). Collectively, our results showed that CST inhibits IAV infection by downregulating the NF-κB protein expression and inhibiting NF-κB phosphorylation in the nucleus.

NF-κB is a well-known central regulator of innate and adaptive immune responses. NF-κB also stimulates the expression of enzymes, such as COX-2 and various proinflammatory cytokines, which produce substances contributing to the pathogenesis of inflammatory processes [19, 20]. The H1N1-induced host cell expression of proinflammatory cytokines, including IL-8, IL-10, IL-1β, and TNF-α, has been correlated with disease in H1N1 and H5N1 patients [21]. Therefore, the inhibition of virus-induced cytokine release is also important in the treatment of influenza virus infections. We next examined the expression levels of IL-8, IL-10, IL-1β, and TNF-α in IAV-infected THP-1 cells regardless of the absence or presence of CST (Fig. 5a–d). Compared with the uninfected group, the H1N1-infected groups exhibited a dramatic increase (three folds) in the expression levels of IL-8, IL-10, IL-1β, and TNF-α. After treatment with CST, these cytokine levels were significantly reduced in a dose-dependent manner. In addition, we found that CST significantly reduced COX-2 protein expression in IAV-infected THP-1 cells (Fig. 5e).

To further explain the mechanism of CST’s inhibitory proinflammatory cytokine expression by CST, we then investigated the intervention of CST on the IAV-induced activation of MAPK signaling pathway. In our study, the phosphorylation of MAPKs was tested by lysing the treated THP-1 cells and determining the total proteins by Western blot. The protein levels of phospho-p38 MAPK and phospho-JNK remarkably decreased at 15 and 30 min, respectively, after the CST treatment (Fig. 6). On the contrary, CST showed no significant effect on ERK1/2 phosphorylation at these two time points.

The IAV A/Fort Monmouth/1/1947(H1N1) strain was obtained from the America Type Culture Collection (ATCC). A/tianjinjinnan/15/2009 (H1N1, oseltamivir- resistant), A/JiangXi/312/2006 (H3N2) were kindly provided by Professor Yuelong Shu at the Institute for Viral Disease Control and Prevention, China Centers for Disease Control and Prevention, Beijing, China. The viral stock of this strain was prepared using the method described previously [28].

Madin–Darby canine kidney cells (MDCK) were obtained from ATCC. MDCK cells were grown in minimum essential medium (MEM; Invitrogen, Carlsbad, CA, USA) supplemented with 1% MEM Non-Essential Amino Acid Solution (Invitrogen), 10% fetal bovine serum (FBS; Gibco, Grand Island, NY, USA), and 1% penicillin–streptomycin (10,000 U/mL) (Invitrogen). Human monocytic cells lines (THP-1) were obtained from National Infrastructure of Cell Line Resource, China. THP-1 cells were then cultured in RPMI-1640 Medium (Gibco, Grand Island, NY, USA) containing 10% FBS and antibiotics.

MDCK cell infection was performed as described previously [28]. For THP-1 cells, the maintenance medium was supplemented with 2% FBS and antibiotics.

CST and ribavirin (RBV) were purchased from Sigma–Aldrich (St. Louis, MO, USA). Oseltamivir carboxylate (OC) was purchased from Medchem Express (NJ, USA). CST (5 mg/ml) was dissolved in DMSO. Then, 2 mM stock solutions of RBV and OC were dissolved in the culture medium. These drugs were diluted to a final working concentration in the experiments.

The cytotoxicity of CST in proliferating cells were assayed by the CCK method [29]. In brief, MDCK cells (2.5 × 104 cells per well) were seeded in 96-well plates overnight at 37 °C under 5% CO₂. The media was removed, and serial two-fold dilutions of CST or control drugs were applied for 48 h. THP-1 cells (3 × 104 cells per well) were grown in 96-well plates, and then different concentrations of CST were added in the 96-well plates. After 48 h of incubation, 10 μl of CCK (Transgen, Beijing, China) was added to each well. After incubation at 37 °C for 2 h, the solution absorbance was measured at 450 nm on Enspire (Perkin Elmer, Waltham, MA, USA).

MDCK cells (2.5 × 104 cells per well) were seeded in 96-well plates overnight at 37 °C under 5% CO₂. The cells were infected with influenza virus (100TCID₅₀) for 2 h, and then they were incubated with maintenance medium supplemented with or without CST. After further incubation for 48 h, the cytopathic effect were recorded. Then we calculated the 50% CPE inhibition concentrations (IC₅₀) and the selectivity index (SI) values of CST.

The inhibitory effect of CST on viral attachment was evaluated by the HI inhibition assay [28, 30]. Briefly, 50 μl of CST or RBV in serial two-fold dilutions in saline were mixed with an equal volume of influenza virus suspension and incubated for 30 min at 4 °C. Saline was used as positive control, and the no-virus red blood cell (RBC) was used as the negative control. Afterward, 100 μl of 1.2% chicken RBCs was dispensed to each well and incubated at room temperature for 40 min to monitor for agglutination.

By quantifying the fluorescent product upon the cleavage of 4-methylumbelliferyl- a-d-N-acetylneuraminic acid (MUNANA; Sigma, St Louis, MO, USA) by NA, IAV NA activity was assessed [28, 31]. The reaction system contains 20 μl of sample, 20 μl of enzyme and 60 μl of substrate buffer mix (20 μM MUNANA, 33 mM MES buffer (pH 3.5), 4 mM CaCl₂, and double distilled water). Briefly, neuraminidases derived from influenza A viruses were incubated with diluted drug samples for 60 min at room temperature and then 60 μl neuraminidases substrate was added. The luminescence signal indicating neuraminidase activity was determined on Enspire (Perkin Elmer, Waltham, MA, USA) with excitation wavelength 355 nm and emission wavelength 460 nm before and after incubating for 15 min at 37 °C. The inhibition ratio was calculated according to the equation.

MDCK cells (3 × 105 cells) grown on a coverslip in 12-well plates. The cells were infected with influenza virus (100TCID₅₀) for 2 h with or without CST (5, 10, and 20 μg/ml) or RBV (10 μM). Cells were fixed with 4% paraformaldehyde for 10 min at room temperature at 18 h post-infection (p.i.). After incubating with 0.5% Triton X-100 for 15 min, cells were blocked with 1% bovine serum albumin (BSA) in PBS for 1 h at room temperature. The cells were then incubated with appropriate primary (IAV M2 antibody, Santa Cruz, Dallas, Texas, USA) and secondary antibodies (Alexa-Fluor-488, Transgen, Beijing, China). The nucleus was detected with Hoechst 33,342 (Beyotime, Shanghai, China). Pictures were taken with an Olympus TH4-200 microscope.

Cells were lysed in M-PER mammalian protein extraction reagent containing the Halt Protease inhibitor cocktail (Thermo Fisher Scientific, Waltham, MA, USA), whereas nuclear and cytosolic extracts were prepared using a nuclear and cytoplasmic extraction kit (Beyotime, Beijing, China). Lysates were centrifuged at 12,000 g for 15 min at 4 °C. The supernatants were collected and an equal amount of proteins were subjected to 10% SDS-PAGE. Proteins were detected using antibodies directed against ERK 1/2(1:1000), phosphorylated ERK 1/2 (1:1000), p38 (1:1000), phosphorylated p38 (1:1000), JNK (1:1000), phosphorylated JNK (1:1000), nuclear factor-kappa B subunit p65 (NF-κB p65) (1:1000), phosphorylated nuclear factor-kappa B subunit p65 (NF-κB p65) (1:1000), cyclooxygenase-2 (COX-2; 1:1000), β-actin (1:5000), histone H3 (1:1000) (Cell Signaling Technology, Beverly, MA, USA), IAV M2 (1:400) (Santa Cruz, Dallas, Texas, USA), respectively. Proteins were visualized by an ECL kit (GE Healthcare Life Sciences, Pittsburgh, PA, USA). Membranes were exposed to Biorad GelDoc XR (Bio-RAD, USA).

MDCK and THP-1 cells were infected with 100TCID₅₀ influenza virus for 2 h and cultured in the presence of CST at different concentrations. At 12 h p.i., total RNA was isolated from cells by using RNeasy Mini Kit (Qiagen, Germantown, MD, USA).

The primers used to amplify the IAV M2, IL-1β, TNF-α, IL-8, IL-10, and GAPDH genes (Table 2), were synthesized by Invitrogen Biotechnology Co. Ltd., China. The relative quantification of these genes were carried out with an ABI 7500 Fast real-time PCR instrument (Applied Biosystems, Foster City, CA, USA) by using TransScript II Green One-Step qRT-PCR SuperMix kit (TransGen Biotech, China) with the following procedures: 50 °C for 5 min, 94 °C for 30 s, followed by 35 cycles of 94 °C for 5 s, and 60 °C for 30 s. The relative mRNA amounts of IAV M2, IL-1β, TNF-α, IL-8, as well as IL-10, were then calculated by comparative Ct method after normalizing against the quantity of GAPDH.