EBV-positive NPC cell lines HONE1 and HK1-EBV were kindly provided by Professor Sai Wah Tsao (The University of Hong Kong, Hong Kong, China). HeLa cells were obtained from Professor Hui Li (Wuhan University, Wuhan, China). The EBV-negative HK2 cells were kindly provided by Professor Ling Zheng (Wuhan University).

HONE1 and HK1-EBV cells were maintained in RPMI-1640 medium containing 10% fetal bovine serum (FBS; Gibco-BRL, Gaithersburg, MD, USA) and G418 (400 ng/ml). HeLa cells were cultured in Dulbeccos modified Eagles medium (DMEM) with 10% FBS. HK2 cells were cultured in DME/F-12 containing 10% FBS and hEGF (Sigma-Aldrich, Shanghai, China). All cell lines were cultured at 37̊C with a humidified atmosphere of 5% CO₂. Berberine chloride, cycloheximide (CHX) (both from Sigma-Aldrich, Shanghai, China), acyclovir (ACV; Selleck Chemicals, Shanghai, China), and 12-O-tetradecanoylphorbol-13-acetate (TPA; Sigma-Aldrich) were dissolved in dimethyl sulfoxide (DMSO). Sodium butyrate (SB; Sigma-Aldrich) was dissolved in phosphate-buffered saline (PBS).

Plasmid pSG5-EBNA1 was constructed by ligating the EcoRI-XbaI fragment, which contains the EBNA1 sequence, into the pSG5 vector (YouBio, Shanghai, China). Plasmid pGL3.0-basic was purchased from Promega (Madison, WI, USA). PRL-TK, a plasmid expressing Renilla luciferase, was kindly provided by Professor Deyin Guo (Wuhan University). The sequence of Qp was amplified from the HONE1-strain EBV genome sequence. The primers used in the present study were as follows: Qp forward, 5′-TCAGATCTTATAACGCAGGTCCTG-3′ and reverse, 5′-CGCAAGCTTTGTAAGGATAGCATG-3′ (YouBio). The amplified sequence was digested with BglII and HindIII (NEB, Ipswich, MA, USA). All plasmids were purified through columns (Axygen Scientific Inc., Union City, CA, USA) as described by the manufacturer and confirmed by DNA sequencing.

The viability of the cells was determined by the Cell Counting Kit-8 (CCK-8) assay (Dojindo Laboratories, Kumamoto, Japan). Briefly, HONE1 and HK1-EBV cells were placed into 96-well plates with DMEM at a density of 1×10⁴ cells/100 µl. All cells were treated with controls (DMSO; 0.006%, vol/vol) and increasing concentrations of berberine for 24 and 48 h. Tetrazolium substrates were added into each well (10 µl/well) after different treatments. The plates were incubated at 37̊C for 1 h. The optical density (OD) was assessed at 450 nm with an ELx800 microimmunoanalyser (BioTek Instruments, Inc., Winooski, VT, USA).

Apoptosis was quantified using an Annexin V-FITC/propidium iodide (PI) apoptosis detection kit (MultiSciences, Shanghai, China). Briefly, HONE1 and HK1-EBV cells were placed in 6-well plates and treated with the vehicle control (DMSO; 0.006%, vol/vol) or berberine (50 or 100 µM) for 24 h. Cells were harvested, washed and re-suspended in 500 µl of binding buffer. Then, 5 µl of Annexin V-FITC and 10 µl of PI were added before being analyzed by a Beckman Coulter system (EPICS Altra II; Beckman Coulter, Fullerton, CA, USA).

For transfection, all cells (HONE1, HK1-EBV and HeLa) were transiently transfected using X-tremeGENE HP DNA transfection reagent (Roche, Basel, Switzerland). At 4 h post-transfection, cells were treated with the control (DMSO; 0.006%, vol/vol) or berberine for 44 h before harvesting. CHX (50 µg/ml) was used in experiments as described in the following sections.

In brief, HeLa cells were cultured in 48-well plates (1×10⁵ cells/well) before transfection. After 24 h, pGL3.0-Qp (100 ng/well) and pRL-TK (5 ng/well) were co-transfected into the HeLa cells. Then, the cells were treated with the vehicle control (DMSO; 0.006%, vol/vol) or berberine (25 and 50 µM, respectively) for 24 h, at 4 h post-transfection. Subsequently, the medium was removed and cells were rinsed twice. Total proteins were harvested in 1X lysis buffer (100 µl/well). Luciferase activity was assessed by the GLO-MAX 20/20 system (both from Promega) following the manufacturer's instructions. Renilla luciferase was used to normalize the firefly luciferase activity.

The total RNA was extracted using TRIzol reagent (Invitrogen, Grand Island, NY, USA) according to the manufacturer's instructions. First-strand cDNA was synthesized from the RNA using a reverse transcription kit (Takara, Tokyo, Japan) with random primers. The expression of LMP1 and EBNA1 mRNAs was quantified by CFX96 real-time PCR detection system using a SYBR Premix Ex Taq™ kit (Takara). The relative amounts of LMP1 and EBNA1 mRNAs were normalized to the housekeeping gene GAPDH. The primers were as follows: LMP1 forward, 5′-CTATTCCTTTGCTCTCATGC-3′ and reverse, 5′-TGAGCAGGAGGGTGATCATC-3′; EBNA1 forward, 5′-GGTCGTGGACGTGGAGAAAA-3′ and reverse, 5′-GGTGGAGACCCGGATGATG-3′; GAPDH forward, 5′-ACATCGCTCAGACACCATG-3′ and reverse, 5′-TGTAGTTGAGGTCAATGAAGGG-3′ (YouBio). The quantitative-PCR conditions were as follows: a 30-sec denaturation at 95̊C, followed by 40 cycles of 10 sec at 95̊C, 10 sec at 60̊C, and 20 sec at 72̊C. The specificity of the reaction was controlled by melting curve analysis (65–95̊C, 0.5̊C/sec).

Cells treated under different conditions were harvested in RIPA buffer (Beyotime Institute of Biotechnology, Shanghai, China) supplemented with 0.5 mM phenylmethylsufonyl fluoride (PMSF) and 0.5% cocktail protease inhibitor (Roche) using a micro-scraper. After being sonicated for 15 sec, whole cell extracts were obtained by centrifugation at 12,000 × g for 15 min. The supernatants were transferred to new tubes, and the protein concentration was determined by the bicinchoninic acid method using bovine serum albumin as a standard. Equal amounts of proteins were mixed with 5X loading buffer [250 nM Tris-HCl (pH 6.8), 0.5% BPB, 10% SDS, 50% glycerol, 5% β-mercaptoethanol]. The proteins were subjected to 10% SDS-polyacrylamide gel electrophoresis (SDS-PAGE). The gels were run at 110 V, and then, the proteins were transferred onto a polyvinylidene difluoride (PVDF) membrane (Bio-Rad Laboratories, Inc., Hercules, CA, USA). The membrane was blocked with 5% skim milk in Tris-buffered saline and Tween-20 (TBST) for 1 h at room temperature, followed by incubation with the primary antibody overnight at 4̊C. After 3×15 min washes with TBST, the membrane was incubated with the appropriate secondary antibodies for 1 h.

The primary antibodies used were as follows: GAPDH (cat. no. 10494-1-AP, 1:10,000; Proteintech, Wuhan, China), EBNA1 (cat. no. sc81581, 1:200; Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA), LMP1 (1:4,000; Abcam, Cambridge, UK), BZLF1 (1:500; Dako, Glostrup, Denmark), caspase-3 (cat. no. 9668, 1:1,000), cleaved caspase-3 (cat. no. 9664, 1:1,000), STAT3 (cat. no. 9139, 1:1,000), phospho-STAT3 (cat. no. 4113, 1:1,000, Tyr705), Mcl-1 (cat. no. 14765, 1:1,000) (all from Cell Signaling Technology Inc., Danvers, MA, USA). Secondary antibodies were horseradish peroxidase-conjugated secondary anti-mouse IgG (cat. no. 7076, 1:10,000), anti-rabbit IgG (cat. no. 7074, 1:10,000) (both from Cell Signaling Technology Inc.). Immunoreactivity was detected using the ECL system (Bio-Rad Laboratories). Band gray values were assessed by ImageJ software (National Institutes of Health, Bethesda, MD, USA).

HONE1 and HK1-EBV cells were induced to lytic replication phase by TPA (40 ng/ml) and SB (3 mM). After 3 h, cells were cultured in the absence or presence of berberine for 48 h. EBV virion-associated DNA was collected and amplified. The supernatants of HONE1 and HK1-EBV cells were harvested and filtered with a 0.45 µl filter. Each supernatant was incubated with 2 µl DNase I and 10X DNase I buffer (10 mM Tris-HCl, 0.5 mM CaCl₂, 2.5 mM MgCl₂) at 37̊C. After 60 min, EDTA (2 mM, 20 µl, pH 8.0) was added to inhibit the activity of DNase I. All the samples were treated with proteinase K (0.1 mg/ml) [sample:proteinase K = 1:1 (vol/vol)] for 1 h at 50̊C. The reactions were stopped at 75̊C after 20 min by inhibiting the activity of proteinase K (46). Subsequently, each sample and strand was examined for the sequence of EBNA1 by quantitative-PCR assay using a SYBR-Green PCR kit (cat no. RR420A; Takara Bio). Each quantitative-PCR reaction mixture included 2 µl viral DNA from the prepared sample, 2 µl specific primers (0.2 µM), 12.5 µl 2X SYBR Premix Ex Taq, 0.4 µl ROX reference dye or Dyell and PCR-grade water for a final volume of 25 µl. The primers were as follows: EBNA1 forward, 5′-GGTCGTGGACGTGGAGAAAA-3′ and reverse, 5′-GGTGGAGACCCGGATGATG-3′ (YouBio). The quantitative-PCR conditions were: a 5-sec denaturation at 95̊C, a 20-sec annealing at 60̊C, a 2-sec extension of primers at 72̊C for 45 cycles. Melting curve analysis was performed from 65–95̊C (with 0.1̊C/sec).

All experimental procedures and protocols were approved by the Medical Ethics Committee of Wuhan University. The NOD/SCID mice were purchased and maintained in the Animal Experiment Center of Wuhan University, Animal Biosafety Level-III Laboratory.

Eight-week-old female NOD/SCID mice were subcutaneously inoculated into flanks with 1×10⁷ HONE1 cells suspended in 100 µl of PBS. Seven days later, the mice (five mice/group) were treated with intraperitoneal injection of DMSO (0.006%) or berberine (10 mg/kg, three times a week). Mice were sacrificed after three weeks by cervical dislocation. Tumor size was assessed, and then the tumors were fixed in 10% neutral buffered formalin. Immunohistochemical analysis and hematoxylin and eosin (H&E) staining were performed according to the method described in a previous study (47).

Tumor samples were formalin-fixed and paraffin-embedded and cut into sections of 4-µm thickness. The sections were dewaxed in xylene, rehydrated in ethanol, washed in PBS, and then stained with H&E. After staining, sections were dehydrated through a series of increasing concentrations of ethanol and xylene, and rehydrated in distilled water. After antigen retrieval in a sodium citrate buffer in a microwave oven, the endogenous peroxidase activity was blocked using 0.6% H₂O₂ for 30 min. Sections were incubated overnight at 4̊C with a primary antibody EBNA1 (cat. no. ab8329, 1:1,000; Abcam). After washing in PBS twice, the sample sections underwent detection using the anti-mouse-specific HRP/AEC Detection IHC kit (ab127055; Abcam) according to the manufacturer's instructions.

All results were normalized to the control and are presented as the mean ± standard deviation (mean ± SD) of at least three independent experiments. The statistical significance of the difference was evaluated by Student's t-test using GraphPad Prism (GraphPad Software, La Jolla, CA, USA). Significance was assigned at p<0.05.

To determine whether berberine (Fig. 1A) affects the cell viability of EBV-positive NPC cells, HONE1 and HK1-EBV cells were treated with a vehicle control (0.006% DMSO) or a series of increasing concentrations of berberine (25, 50, 100 and 200 µM) for 24 and 48 h, respectively. Cell viability was determined by CCK-8 assays. As shown in Fig. 1B, berberine inhibited the cell viability of HONE1 cells in a dose- and time-dependent manner. The viability of HONE1 cells was decreased from ~10 to 90% after berberine treatment for 24 and 48 h. The 50% inhibitory concentration (IC₅₀) for HONE1 cells was 101.3 and 56.7 µM with berberine treatment at 24 and 48 h, respectively. Similar results were also found in the HK1-EBV cells treated with a series of increasing concentrations of berberine. The viability of HK1-EBV cells was decreased from 15 to 78% after 24 and 48 h of berberine treatment. The IC₅₀ values were calculated and were 124.5 and 43.1 µM at 24 and 48 h, respectively, in the HK1-EBV cells treated with berberine (Fig. 1C). To further evaluate the effects of berberine in the EBV-negative cell line, HK2 cells were treated with the vehicle control (0.006% DMSO) or berberine for 24 and 48 h. As shown in Fig. 1D, treatment with a series of increasing concentrations of berberine showed a slight inhibition in the proliferation of HK2 cells at 24 and 48 h. These results suggest that berberine inhibits the cell viability of EBV-positive NPC cells efficiently, but not significantly in HK2 cells.

To examine whether berberine induces cell cycle arrest and apoptosis, HONE1 and HK1-EBV cells were treated with the vehicle control (0.006% DMSO) or berberine (50 and 100 µM) for 24 h. Cells were collected and stained with Annexin V-FITC/PI, followed by analyses with flow cytometry. The cell cycle analysis revealed that berberine-treated HONE1 cells underwent a G2 arrest; the percentage of G2 phase cells was 1.41% in the control group, and 16.2% in the berberine-treated group (Fig. 2A). Treatment of HK1-EBV cells with berberine resulted in G1 phase arrest; the percentage of G1 phase cells was 50.9% in the control (0.006% DMSO) and 56.9% in the berberine-treated group (50 µM) (Fig. 2A). In addition, compared to the control group (0.006% DMSO), berberine (50 or 100 µM) induced cell apoptosis in the HONE1 and HK1-EBV cells (Fig. 2B). The results suggest that berberine induced cell cycle arrest and apoptosis in the EBV-positive HONE1 and HK1-EBV cells.

To determine whether berberine alters the expression of EBNA1 in EBV-positive NPC cell lines, HONE1 and HK1-EBV cells were treated with the vehicle control (0.006% DMSO) or berberine (25 or 50 µM) for 48 h. Whole-cell extracts were harvested and subjected to western blotting. As shown in Fig. 3A, the expression levels of EBNA1 in the HONE1 cells were decreased to 36.8 and 6.7%, with berberine treatment at 25 and 50 µM, respectively, when compared to the control group. HK1-EBV cells were also treated with the same concentrations of berberine. When compared to the control group, the expression levels of EBNA1 were not significantly affected in the 25 µM berberine-treated group, but downregulated to 32.4% in the 50 µM berberine-treated group. However, the expression level of another EBV protein, LMP1, was not decreased by berberine in both the HONE1 and HK1-EBV cells. Furthermore, compared to the control group, the expression levels of cleaved-caspase-3 in the HONE1 and HK1-EBV cells were increased while the levels of caspase-3 exhibited no obvious changes with berberine treatment. The different response of the HONE1 and HK1-EBV cells to berberine is possibly due to the different sensitivity of the cell lines to this drug.

In addition, quantitative PCR analysis was used to detect whether berberine affects LMP1 at the mRNA level. HONE1 and HK1-EBV cells were treated with the vehicle control (0.006% DMSO) or berberine (25 or 50 µM) for 24 h, respectively. The total RNAs were collected and subjected to quantitative-PCR. The relative expression of LMP1 was detected by comparison to the GAPDH gene. As shown in Fig. 3B, the mRNA levels of LMP1 in the HONE1 and HK1-EBV cells were not decreased by berberine.

Berberine treatment in a previous study was not observed to affect the viability of EBV-negative HeLa cells (48). Therefore, HeLa cells were used to determine whether berberine specifically decreases the expression of EBNA1 in the present study. In order to determine whether berberine specifically decreases the expression of EBNA1, p-SG5 and pSG5-EBNA1 were transiently transfected into HeLa cells for 4 h, followed by treatment with or without berberine. As shown in Fig. 3C, when compared to the control group, the expression levels of EBNA1 were not significantly affected in the 25 µM berberine-treated group, but downregulated to 32.5% in the 50 µM berberine-treated group.

To further determine whether the decreased expression of EBNA1 was due to possible alterations at the transcriptional levels, HONE1 and HK1-EBV cells were treated with the vehicle control (0.006% DMSO) or berberine (25 or 50 µM). After 24 h, the total RNAs were harvested and subjected to quantitative-PCR. As shown in Fig. 3D, berberine decreased EBNA1 mRNA levels to 81.23% in the 25 µM and 63.17% in the 50 µM berberine-treated groups, respectively, when compared with the vehicle control (0.006% DMSO) group in the HONE1 cells. In the HK1-EBV cells, the EBNA1 mRNA levels were decreased to 85.39% (p<0.05) in the 25 µM and 65.73% (p<0.05) in the 50 µM berberine-treated groups, respectively, when compared with the vehicle control (0.006% DMSO) group. These results demonstrated that berberine inhibited the mRNA levels of EBNA1 in HONE1 and HK1-EBV cells.

EBNA1 transcription is initiated from Qp in EBV-associated NPC cells. According to our previous studies (42,44), the promoters of latent membrane protein 1 (LMP1) and human telomerase reverse transcriptase (hTERT) were studied in HeLa and 293T cells. To determine whether berberine inhibited the activity of EBNA1 promoter, pGL3.0 vector control or pGL3.0-Qp was co-transfected with plasmid pRL-TK into HeLa cells. The HeLa cells were treated with the control (0.006% DMSO) or berberine (25 or 50 µM) for 24 h, at 4 h post-transfection. The whole protein was harvested and the luciferase activity was assessed using the GLO-MAX 20/20 system (Promega). Compared to the control group, the activity of Qp was decreased to 83.4% (p<0.05) and 62.7% (p<0.05) in the 25 and 50 µM berberine-treated groups, respectively (Fig. 3E). The results suggest that the transcriptional activity of EBNA1 was decreased by berberine in the HeLa cell line.

To determine whether berberine decreases the protein half-life of EBNA1, HONE1 and HK1-EBV cells were treated with the vehicle control or berberine, in the presence of CHX. As shown in Fig. 3F and G, berberine decreased the half-life of the EBNA1 protein in the CHX-treated HONE1 and HK1-EBV cells when compared to the vehicle control. The results suggest that berberine decreases the half-life of EBNA1.

The Qp promoter that drives the expression of EBNA1 in EBV-associated tumors is regulated by the JAK-STAT signaling pathway (22,23). STAT3 is essential for the activation of Qp and raises the possibility that aberrant activation of STAT3 may be an essential factor in EBV-associated diseases (22). Since STAT3 represents a potential target for treatment of EBV-associated tumors, it may be beneficial to determine whether the activity of STAT3 is affected by berberine. Therefore, HONE1 and HK1-EBV cells were treated with the vehicle control (0.006% DMSO) or berberine (25 or 50 µM). After 48 h, the total proteins were collected and subjected to western blotting. As shown in Fig. 4A and B, the expression of STAT3 and p-STAT3 were detected in the HONE1 and HK1-EBV cells. Berberine suppressed the expression of p-STAT3 effectively in the HONE1 and HK1-EBV cells. Furthermore, the present study demonstrated that the downregulation of p-STAT3 was associated with the suppressed expression of Mcl-1, which is a downstream survival protein of STAT3. These results suggest that downregulation of EBNA1 by berberine is possibly related to the inhibition of the STAT3 signaling pathway in NPC cells.

To determine whether the berberine-induced decrease in cell viability (Fig. 1B and C) is mainly caused by the inhibited expression of EBNA1, pSG5-EBNA1 or the vehicle control (pSG5) were transfected into HONE1 and HK1-EBV cells. The cells were treated with the vehicle control (0.006% DMSO) or berberine (50 µM) at 4 h post-transfection. After 44 h, the total proteins were harvested and subjected to western blotting. As shown in Fig. 5A and B, the expression levels of EBNA1 were increased by 301.3% in the HONE1 cells, and 368.1% in the HK1-EBV cells, respectively, by transient transfection.

Cell viability was detected after 24 and 48 h. As shown in Fig. 5C and D, the cell viability of both HONE1 and HK1-EBV cells was significantly decreased by berberine. Following the overexpression of EBNA1, the cell viability was increased by 19.8% (p<0.05) in the HONE1 cells and 22.4% (p<0.05) in the HK1-EBV cells after 24 h, respectively, when compared with the control group (berberine-treated control group). After 48 h, compared to the control group (berberine-treated control group), the cell viability was increased by 24.5% (p<0.05) in the HONE1 cells and 27.6% (p<0.05) in the HK1-EBV cells. Overexpression of EBNA1 increased cell viability and decreased the inhibitory effect of berberine. These results suggest that berberine inhibits the proliferation of EBV-positive HONE1 and HK1-EBV cells possibly through a mechanism related to the decreased expression of EBNA1.

To determine whether berberine inhibits lytic replication of EBV, HONE1 and HK1-EBV cells uninduced or induced by TPA and SB were treated with the indicated concentrations of berberine for 48 h. Aciclovir (ACV) was used as a positive control, efficiently decreasing virion production of EBV. After induction, the culture media of HONE1 and HK1-EBV cells were prepared after 48 h incubation and processed for the detection of the EBV EBNA1 fragment. Quantitative-PCR and western blotting were used to detect the effect of berberine on lytic replication of EBV. As shown in Fig. 6A and B, berberine decreased viral genome copies in both uninduced and induced HONE1 and HK1-EBV cells. As expected, ACV, which was used as a positive control, significantly decreased the viral genome copies in the induced HONE1 and HK1-EBV cells. The results indicated that virions were decreased by berberine in a dose-dependent manner. The aforementioned cell lysates were harvested. Western blotting was used to analyze the effect of berberine on the expression of EBV transcriptional activator BZLF1 encoded by the EBV immediate-early (IE) gene in the HONE1 and HK1-EBV cells. As shown in Fig. 6C and D, the expression levels of the EBV transcriptional activator BZLF1 in the HONE1 and HK1-EBV cells were decreased by berberine in both the uninduced and induced HONE1 and HK1-EBV cells. The results suggest that berberine decreased latent and lytic replication of EBV in the HONE1 and HK1-EBV cells.

Since the aforementioned results demonstrate the inhibition of NPC cell proliferation by berberine, the possible effects of berberine in vivo were investigated. To determine whether berberine inhibits the growth of the EBV-positive NPC in NOD/SCID mice, 1×10⁷ HONE1 cells were subcutaneously inoculated in the flanks of the mice. After seven days, a dose of 10 mg/kg of berberine or DMSO was administered via subcutaneous injection three times a week. After three weeks, the mice were euthanized and the volumes and weights of the tumors were assessed. The tumor volumes of the DMSO-treated group were larger than that of the berberine-treated group (98.75±13 vs. 590.5±12.3 mm³, n=5, p<0.05; Fig. 7A). Mice treated with berberine had a significantly lower tumor weight (22.25±7.5 vs. 152.5±10.2 mg, n=5, p<0.05; Fig. 7B). Images of tumors in mice treated with DMSO or berberine are shown in Fig. 7C.

To determine the mechanism by which berberine inhibits tumor growth, the expression levels of EBNA1, BZLF1, p-STAT3 and cleaved caspase-3 in the tumor tissues were analyzed by western blotting. Compared to the DMSO-treated group, berberine downregulated the expression level of EBNA1 significantly and induced tumor cell apoptosis via the caspase-3 pathway. In addition, the expression of BZLF1 was decreased by berberine (Fig. 7D). Then, total RNAs were extracted from tumor tissues. The mRNA levels were analyzed by quantitative-PCR. The mRNA level of EBNA1 in the berberine-treated group was decreased to 61.3% of the control group (Fig. 7E).

Both the vehicle control (DMSO) and the berberine-treated tumor tissue groups were analyzed by H&E staining and EBNA1 immunostaining (Fig. 7F). The tumor tissues of the berberine-treated group were found to have necrosis and a small amount of inflammatory cell infiltration. The number of EBNA1-positive cells in the berberine-treated mice was significantly decreased compared to the control group. As shown in Fig. 7G, the number of EBNA1-positive cells in the tumor tissues of the berberine-treated group were decreased by 40.2% (p<0.05) compared to those of the DMSO-treated group. The results suggest that berberine significantly inhibits the growth of EBV-positive NPC cells in vivo.