Identification of the genes chemosensitizing hepatocellular carcinoma cells to interferon-α/5-fluorouracil and their clinical significance

Abstract

The incidence of advanced hepatocellular carcinoma (HCC) is increasing worldwide, and its prognosis is extremely poor. Interferon-alpha (IFN-α)/5-fluorouracil (5-FU) therapy is reportedly effective in some HCC patients. In the present study, to improve HCC prognosis, we identified the genes that are sensitizing to these agents. The screening strategy was dependent on the concentration of ribozymes that rendered HepG2 cells resistant to 5-FU by the repeated transfection of ribozymes into the cells. After 10 cycles of transfection, which was initiated by 5,902,875 sequences of a ribozyme library, three genes including protein kinase, adenosine monophosphate (AMP)-activated, gamma 2 non-catalytic subunit (PRKAG2); transforming growth factor-beta receptor II (TGFBR2); and exostosin 1 (EXT1) were identified as 5-FU-sensitizing genes. Adenovirus-mediated transfer of TGFBR2 and EXT1 enhanced IFN-α/5-FU-induced cytotoxicity as well as 5-FU, although the overexpression of these genes in the absence of IFN-α/5-FU did not induce cell death. This effect was also observed in a tumor xenograft model. The mechanisms of TGFBR2 and EXT1 include activation of the TGF-β signal and induction of endoplasmic reticulum stress, resulting in apoptosis. In HCC patients treated with IFN-α/5-FU therapy, the PRKAG2 mRNA level in HCC tissues was positively correlated with survival period, suggesting that PRKAG2 enhances the effect of IFN-α/5-FU and serves as a prognostic marker for IFN-α/5-FU therapy. In conclusion, we identified three genes that chemosensitize the effects of 5-FU and IFN-α/5-FU on HCC cells and demonstrated that PRKAG2 mRNA can serve as a prognostic marker for IFN-α/5-FU therapy.

Figure 1. Identification of 5-FU-sensitizing genes.

(A) Determination of a 5-FU concentration sufficient to decrease cell viability in HepG2 cells, which were treated with various concentrations of 5-FU for 72 h. Data are shown as means ± standard deviation (SD) (n = 3). (B) An outline of the screening process. The cycle was repeated 10 times, and sequence analysis of ribozymes was performed for gene identification. (C) Assessment of successful screening. HepG2 cells transfected with plasmid DNA recovered from 6 (Rz-C6), 8 (Rz-C8), and 10 (Rz-C10) cycles of screening were treated with the indicated concentrations of 5-FU for 72 h. Data are expressed as mean ± SD. Statistical significance was determined by one-way analysis of variance and Tukey’s HSD method. *P<0.05, compared to control. (D) Expression of the candidate genes in HepG2 cells. mRNA expression levels of each gene were normalized to β-actin. Data are expressed as mean ± SD (n = 3). N.D., not detected. (E) Confirmation of the knockdown efficiencies of siRNA specific to protein kinase, adenosine monophosphate (AMP)-activated, gamma 2 non-catalytic subunit (PRKAG2); transforming growth factor-beta receptor II (TGFBR2); and exostosin 1 (EXT1). The expression levels of each mRNA were normalized to β-actin expression. Data are expressed as mean ± SD (n = 3) (F) Acquisition of resistance to 5-FU by knockdown of the identified genes. HepG2 cells were transfected with siRNA targeting PRKAG2, TGFBR2, or EXT1, followed by treatment with the indicated concentrations of 5-FU for 72 h. Data are expressed as mean ± SD. Statistical significance was determined using Student’s t-test. *P<0.05 and **P<0.01 compared to control.

Figure 2. Enhancing effects of genes on IFN-α/5-FU treatment.

(A–C).Viability of HepG2 cells infected with adenovirus-carrying protein kinase, adenosine monophosphate (AMP)-activated, gamma 2 non-catalytic subunit (PRKAG2) (A), transforming growth factor-beta receptor II (TGFBR2) (B), and exostosin 1 (EXT1) (C) with or without 5-FU. Adenovirus-carrying LacZ served as a negative control. Cell viability was determined using the WST assay at 72 h. *P<0.05 and **P<0.01, between the two groups. (D) Viability of HepG2, HuH7, and HLF cells infected with adenovirus-carrying TGFBR2 and EXT1 with or without agents. Data are expressed as mean ± standard deviation (SD) (n = 3). Statistical significance was determined using Student’s t-test. *P<0.05 between two groups. (E) In vivo studies using NOD/SCID mice subcutaneously transplanted with HepG2 cells. In adenoviruse-administered mice, tumor growth was assessed during IFN-α/5-FU treatment. Data are shown as means ± SD (n = 4–7).

Figure 3. Enhancement of IFN-α/5-FU-induced apoptosis by TGFBR2 and EXT1 overexpression.

(A) Evaluation of nuclear condensation. The arrows indicate cells with apoptosis-specific nuclear condensation and fragmentation. (B) Measurement of intracellular caspase 3/7 activity. After adenovirus infection, cells were treated with the indicated concentrations of IFN-α and 5-FU. Activity was expressed as the fold increase relative to that at 0 h. Data are expressed as mean ± standard deviation (n = 3). Statistical significance was determined using Student’s t-test. *P<0.05 compared to LacZ at 48 h.

Figure 4. Effects of TGFBR2 on 5-FU- and IFN-α/5-FU-induced TGF-β signaling pathway.

(A) 5-FU-induced TGFB1 mRNA expression normalized to β-actin. Data are expressed as mean ± standard deviation (SD) (n = 3). Statistical significance was determined using Student’s t-test. †P<0.05 and ‡P<0.001 compared to control. N.S., not significant. (B) Luciferase reporter assay for TGF-β signaling activation. Data are expressed as mean ± SD (n = 3). Recombinant TGF-β was used as a positive control. The Y-axis is expressed as a logarithmic scale. (C) Western blot analysis of BAX, BCL-2, and BCL-xL in 5-FU and IFN-α/5-FU treatment with or without TGFBR2 overexpression. Each band was quantified by Image J software and normalized to actin.

Figure 5. Effect of EXT1 expression on ER stress with or without 5-FU and IFN-α/5-FU.

(A, B) Real-time reverse transcription-polymerase chain reaction analysis of BiP/GRP78 (A) and CHOP (B) mRNA expression. mRNA expression levels were normalized to β-actin. Data are expresses as mean ± standard deviation (n = 3). Statistical significance was determined using Student’s t-test. *P<0.05, compared to control. (C) Western blot analysis for ATF4, CHOP, BiP/GRP78, and LC3B. Actin was used as an internal control. (D) LC3B expression in HepG2 cells treated with 5-FU in the presence and absence of IFN-α. Cells infected with adenoviruses were treated with 5 µg/mL of 5-FU alone and in combination with 200 U/mL of IFN-α for 48 h. LC3B (green) and nuclei (DAPI, blue) signals were examined by immunofluorescence analysis using confocal microscopy. Treatment with 20 mg/mL of tunicamycin was used as a positive control. (E) Viability of HepG2 cells infected with adenovirus-carrying EXT1 in the presence or absence of TUDCA. Cells were treated with 5 µg/mL of 5-FU alone and in combination with 200 U/mL of IFN-α for 72 h. During this treatment, cells were exposed to TUDCA. Data are expressed as mean ± standard deviation (SD) (n = 3). Statistical significance was determined using Student’s t-test. *P<0.05 between two groups. N.S., not significant.

Figure 6. The correlation between gene expression levels and response to IFN-α/5-FU therapy.

(A) Survival rates of responders and nonresponders treated with IFN-α/5-FU. *P<0.05 between groups. (B) The correlation between protein kinase, adenosine monophosphate (AMP)-activated, gamma 2 non-catalytic subunit (PRKAG2) mRNA and survival period in patients treated with IFN-α/5-FU therapy. Data were analyzed using Spearman’s rank correlation test. (C) The correlation between transforming growth factor-beta receptor II (TGFBR2) mRNA and survival period in patients treated with IFN-α/5-FU therapy. Data were analyzed using Spearman’s rank correlation test. (D, E) Immunohistochemical analysis of PRKAG2 (D) and TGFBR2 (E) expression.