Phosphorothioate-modified CpG oligodeoxynucleotide (CpG ODN) induces apoptosis of human hepatocellular carcinoma cells independent of TLR9

Abstract

Toll-like receptors (TLRs) expressed on cancer cells are closely associated with tumor development. In this study, we investigated the biological functions of the TLR9 ligand, CpG oligodeoxynucleotide (CpG ODN), on TLR9 expressed in the cytoplasm of hepatocellular carcinoma (HCC) cells. In vitro, human HCC cell lines were transfected with phosphorothioate-modified oligodeoxynucleotides TLR9 agonist OND M362 and its negative control ODN M362 ctrl, which inhibited the proliferation of HCC cells by inducing apoptosis without altering the cell cycle. Interestingly, ODN M362 and ODN M362 Ctrl displayed a similar proapoptotic effect on HCC, possibly related to phosphorothioate modification of the structure of CpG ODN. Although both of them resulted in the upregulation of the TLR9 receptor, their effect on HCC apoptosis was independent of TLR9. They also upregulated inflammatory cytokines, but did not activate the NF-κB signaling pathway. Finally, the activities of ODN M362 and ODN M362 Ctrl were demonstrated in nude mice inoculated with HCC cells. These findings suggest that the phosphorothioate-modified TLR9 agonist ODN M362, and its control, elicit antitumor activity in HCC cells and may serve as a novel therapeutic target for HCC therapy.

Electronic supplementary material

The online version of this article (doi:10.1007/s00262-014-1518-y) contains supplementary material, which is available to authorized users.

Introduction

Hepatocellular carcinoma (HCC), the most common form of liver cancer, is one of the most malignant tumors found in clinical practice. The latest statistics indicate that there were approximately 6,000,000 new cases of liver cancer worldwide and that it was the fifth most malignant tumor type. Therefore, HCC is a significant public health threat that needs to be addressed.

Treatment for HCC depends on multidisciplinary collaboration. Surgery is the preferred treatment modality for HCC, with other therapeutic options, including radiofrequency ablation, microwave ablation, radiotherapy, molecular targeted therapy and immunotherapy, being reserved for patients too ill to undergo resection [1, 2].

Toll-like receptors (TLRs) are a family of transmembrane receptors that play a key role in the innate and adaptive immune systems [3–5]. Stimulation of TLRs activates signaling pathways, such as NF-κB, MAPKs, JNKs, p38, ERKs and interferon regulatory factor (IRF3, IRF5 and IRF7). Consequently, many agonists of TLRs have been used for immunotherapy [6].

Toll-like receptor 9 is broadly expressed in many types of cancer cells. In addition to immune activation, TLRs appear to play dual roles in the progression of different cancers. Evidences have shown that TLR activation promotes tumorigenesis in a variety of carcinomas by enhancing cell proliferation, angiogenesis and metastasis, and by aggravating immunosuppression [7–14]. Preliminary evidence suggests that TLR9 promotes tumor cell proliferation in prostate cancer cells [15]. In addition, activation of TLR4 expressed in human head and neck squamous cell carcinoma boosts tumor growth and enables tumor cells to escape immune surveillance [16].

Other studies indicated that TLR activation suppresses tumor cell proliferation, induces cell apoptosis and inhibits tumor growth [17, 18]. Stimulation of TLR5 on breast cancer cells, for example, inhibits cell proliferation and exerts a potential antitumor effect on mouse xenografts of human breast cancer [18]. Wang et al. [19] showed that stimulation of TLR9 on lung cancer cells triggered tumor cell apoptosis and arrested tumor growth. Some studies have shown that therapeutic targeting of TLR9 inhibits neuroblastoma cell growth and induces cell apoptosis [20].

Based on these findings, the TLR9 ligand, CpG oligodeoxynucleotide (CpG ODN), has been used for treatment for brain, skin and renal cancer and lymphoma [21]. However, it is currently not known whether CpG ODN can inhibit HCC cell growth. Thus, the present study was undertaken to investigate the biological effects of CpG ODN on TLR9 expressed in the cytoplasm of HCC cells.

Materials and methods

Cell culture and mice

Human hepatoma cell lines HepG2, H7402 and PLC/PRF/5 were purchased from the Chinese Academy of Sciences (Shanghai, China). Cells were maintained in RPMI 1640 (Gibco, Grand Island, NY, USA) supplemented with 10 % fetal bovine serum (FBS) and 1 % penicillin–streptomycin, at 37 °C in a humidified atmosphere containing 5 % CO₂. Five-week-old female athymic (nude–nude) mice and C57BL/6 mice were purchased from HFK Bioscience Co, Ltd (Beijing, China) and maintained at an animal facility under specific pathogen-free conditions.

Transfection

HepG2 and H7402 cells were transfected with CpG ODNs or small interfering RNAs (siRNAs) targeting TLR9, using Lipofectamine 2000 (Invitrogen, Carlsbad, CA), according to the manufacturer’s instructions. The siRNAs were synthesized by RiboBio (Guangzhou, China) and prepared at a concentration of 100 nM. ODN M362 (type C, specific for human/mouse TLR9 with a phosphorothioate backbone), negative control ODN M362 Ctrl, and ODN 2006-G5 (type B, specific for human TLR9 with a phosphodiester backbone) were purchased from InvivoGen (San Diego, CA, USA).

Reverse transcriptase polymerase chain reaction (RT-PCR) assay

Trizol reagent (Invitrogen, Carlsbad, CA, USA) was used to extract RNA according to the manufacturer’s instructions. A micro-volume UV–Vis spectrophotometer determined the concentration and quality of the extracted RNA. M-MLV reverse transcriptase (Invitrogen) was used to reverse transcribe total RNA to cDNA, according to the manufacturer’s instructions. Standard PCRs were performed in a total reaction volume of 25 μL as previously described [22]. The PCR products were electrophoresed on 2 % agarose gels, and the expression level was determined using relative light intensities of bands, analyzed using AlphaEaseFC software (Genetic Technologies, Miami, FL, USA). The primers, listed in Table 1, were synthesized by the Beijing Genomics Institute (Beijing, China).

Table 1.

The synthetic gene-specific primer sets used for (a) RT-PCR, (b) real-time PCR

Target	Sequence 5′ → 3′	Size (bp)	References
a
TLR1	F: CATAACTCTGCTGATCGTCACC R: TGCTAGGAATGGAGTACTGCG	491	[41]
TLR2	F: GATGACTCTACCAGATGCCTCC R: CAGAAGAATGAGAATGGCAGC	754	[41]
TLR3	F: TCCCAAGCCTTCAACGACTG R: TCCTGAAAGCTGGCCCGAAAAC	471
TLR4	F: CTGCAATGGATCAAGGACCA R: TCCCACTCCAGGTAAGTGTT	622	[15]
TLR5	F: CCTTGACTATTGACAAGGAGGC R: TTGTAGGCAAGGTTCAGAACC	717	[41]
TLR6	F: CCAAGTGAACATATCAGTTAATACTTTAGGGTGC R: CTCAGAAAACACGGTGTACAAAGCTG	357
TLR7	F: AGTGTCTAAAGAACCTGG R: CTTGGCCTTACAGAAATG	544
TLR8	F: CAGAATAGCAGGCGTAACACATCA R: AATGTCACAGGTGCATTCAAAGGG	636
TLR9	F: CAACAACCTCACTGTGGTGC R: GAGTGAGCGGAAGAAGATGC	514	[41]
TLR10	F: GAACTGATGACCAACTGCTCC R: GAAGTCTTGATTCCATCACGC	556	[41]
β-Actin	F: GTGGGGCGCCCCAGGCACCA R:CTCCTTAATGTCACGCACGATTTC	539
b
TLR9	F: CTGCCACATGACCATCGAG R: GGACAGGGATATGAGGGATTTGG	123
IFN-α	F: CTCCTTTCTCCTGCCTGAAG R: AAGTGTCTCATCCCAAGTAGC	170	[42]
IFN-β	F: TGCTCTCCTGTTGTGCTTCTCC R:CATCTCATAGATGGTCAATGCGG	222	[42]
IL-6	F: CCACACAGACAGCCACTCAC R: AGGTTGTTTTCTGCCAGTGC	146
TNF-a	F: ATGAGCACTGAAAGCATGATCC R: GAGGGCTGATTAGAGAGAGGTC	217
IL-8	F: TTTTGCCAAGGAGTGCTAAAGA R: AACCCTCTGCACCCAGTTTTC	194
GAPDH	F:GAAGGTGAAGGTCGGAGT R: CATGGGTGGAATCATATTGGAA	155	[42]

Other primers were provided by http://pga.mgh.harvard.edu/primerbank/

Proliferation measurements

The effect of ODN M362 Ctrl and ODN M362 on HCC cell proliferation was determined by a 3-[4,5-dimethylthiazol-2-yl]-2,5 diphenyl tetrazolium bromide (MTT) assay. Cells were plated in 96-well plates at a density of 0.3 × 10⁴ cells/well. After overnight incubation, the cells were transfected with ODN M362 Ctrl or ODN M362 for 72 h, MTT (20 μL of a 10 mg/mL solution; Sigma, St Louis, MO, USA) was added to each well, and the cells were incubated for another 4 h. Transfection of ODN was performed with an ODN (μg) to Lipofectamine 2000 (μL) ratio of 1: 2, and increasing amounts of Lipofectamine 2000 were included as a control. Following centrifugation at 2,500 rpm for 20 min, the supernatant was discarded and 200 μL DMSO was added to dissolve the formazan. Absorbance was read at 570 nm on a scanning multiwell spectrophotometer (Bio-Rad, Hercules, CA, USA).

Cell cycle analysis

Cell cycle analysis was undertaken using fluorescence-activated cell sorting (FACS). Cells transfected with ODN M362 Ctrl or ODN M362 were fixed with 75 % ethanol for 3 h. The cells were washed twice with PBS and stained with 50 μg/mL propidium iodide (PI) and 100 μg/mL RNase I in PBS for 30 min at 37 °C. The cell cycle distribution was then analyzed using FACS.

Detection of apoptosis

Cell apoptosis was detected with an Annexin V-FITC/PI apoptosis detection kit (BestBio, China). Cells were transfected with ODN M362 Ctrl or ODN M362 and incubated for 24 h before analysis. Both the media and the cells were collected and washed with PBS. Annexin V-FITC-positive, Annexin V-FITC and PI-positive cells were counted, and the apoptotic index was calculated.

A One Step TUNEL Apoptosis Assay Kit (Beyotime, China) was also used to detect cell apoptosis. Cell nuclei was stained with 4′,6-diamidino-2-phenylindole (DAPI), and fluorescence was evaluated by fluorescence microscopy.

Quantitative real-time PCR analysis

The amplification of cDNA was performed using real-time PCR with the SYBR Green Master Mix (Toyobo, Osaka, Japan) on an iCycleriQ real-time PCR system (Bio-Rad, Hercules, CA, USA). The GAPDH gene was used for RNA normalization. The primers are listed in Table 1.

Western blot analysis

Cells were solubilized in lysis buffer (BestBio, China) and a protease inhibitor cocktail (BestBio, China). Whole cell extracts were mixed in Laemmli loading buffer, boiled for 5 min and subjected to sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). After electrophoresis, proteins were blocked with 5 % non-fat milk for 2 h, transferred onto nitrocellulose membranes and blotted overnight at 4 °C with anti-NF-κB, anti-p–NF-κB (Cell Signaling Technology, New England BioLabs Inc.), or anti-β-actin (Santa Cruz Biotechnology, CA, USA) (at a dilution of 1:2,000). The membranes were washed with TBST three times and incubated with horseradish peroxidase (HRP)-conjugated second antibody for 1 h. Protein bands were visualized by Immobilon Western Chemiluminescent HRP Substrate (Millipore Corporation, Billerica, USA) and examined with Alpha Ease FC software.

In vivo study

Five-week-old female, athymic (nude–nude) mice were subcutaneously (s.c.) injected with 1 × 10⁷ H7402 cells pretransfected with ODN M362-liposome (0.5 μg/mL), ODN M362 Ctrl-liposome (0.5 μg/mL) or liposome for 24 h in vitro. The mice were killed on day 40, and the tumor weight was calculated. Subsequently, to determine the antitumor activity of ODN M362 and ODN M362 Ctrl, mice subcutaneously inoculated with 1 × 10⁷ H7402 cells received an intratumoral injection of ODN M362-liposome, ODN M362 Ctrl-liposome or ODN 2006-G5-liposome (5.0 μg in 100μL) on day 15. Meanwhile, PBS- and liposome-treated mice were used as controls. The injections were repeated every 4 days until day 55, when the mice were killed. The weight of the tumor was estimated, together with body weight, liver weight and spleen weight. All experimental procedures were performed in accordance with the guidelines for ethics and regulations for animal experiments, as defined by the Department of Pharmaceutical Sciences, Shandong University, China.

Statistical analysis

Statistical analysis was performed using a paired Student’s t tests. Data were presented as mean ± SD. Values of p < 0.05 were considered statistically significant.

Results

TLRs are expressed in human hepatocellular carcinoma cells

To investigate the potential function of a TLR9 agonist on human HCC, the expression levels of TLR1-10 mRNA were detected by RT-PCR. The results clearly demonstrated TLR expression (except expression of TLR7 and TLR8) on human HCC cells. TLR9 mRNA was detected in all three tumor cell lines HepG2, H7402 and PLC/RPF/5 (Fig. 1a). To confirm the activity of CpG ODN M362, mouse splenocytes from C57BL/6 mice were used as positive cells and treated with ODN M362; ODN M362 Ctrl was used as negative control. The results showed ODN M362 was able to activate the immune cells but ODN M362 Ctrl could not (Fig. 1b).

Fig. 1.

Expression of TLR1–TLR10 mRNAs in HCC cells. a Expression of TLR1–TLR10 mRNAs in the three HCC cell lines (HepG2, H7402 and PLC/PRF/5) were detected by RT-PCR. b Freshly isolated splenocytes from C57BL/6 mice were treated with different concentrations of ODN M362 and ODN M362 Ctrl for 24 h, and then CD69 expression levels were analyzed by flow cytometry. Data shown are representative of three independent experiments

The TLR9 agonist ODN M362 and its negative control ODN M362 Ctrl inhibit HCC proliferation and promote HCC apoptosis

MTT incorporation assays were performed to identify the role of TLR9 agonist ODN M362 on human HCC cell proliferation. HepG2 and H7402 cells were transfected with ODN M362 or its negative control ODN M362 Ctrl (0.25–4 μg/mL), respectively, and proliferation was determined at predetermined time points. As shown in Fig. 2a, ODN M362 inhibited the proliferation of HCC cells significantly. Following exposure to 4 μg/mL ODN M362 for 72 h, the inhibition rates were approximately 30 % in HepG2 cells and 80 % in H7402 cells. Unexpectedly, ODN M362 Ctrl displayed similar inhibitory effects as ODN M362 on HCC cells.

Fig. 2.

ODN M362 Ctrl and ODN M362 inhibited the proliferation of HCC cells and induced HCC apoptosis. a HCC cells were transfected with ODN M362 Ctrl or ODN M362, respectively. Cell proliferation was assessed by MTT assay at the indicated times. b HCC cells were transfected with ODN M362 Ctrl or ODN M362 (0.5 μg/mL) for 24 h. Both adherent and non-adherent cells were harvested, and cell cycle distributions were analyzed by flow cytometry. c The statistics for the result shown in b. d HCC cells were transfected with ODN M362 Ctrl or ODN M362 for 24 h, and the apoptosis rates were determined by flow cytometry with AnnexinV/PI staining. e TUNEL analysis of the apoptosis of the tumor cells treated as in a. DAPI was used for nuclear staining. The figure shows representative results of one of at least three independent experiments

To understand the mechanism of suppression of HCC proliferation, H7402 cells were transfected with either ODN M362 Ctrl or ODN M362 at concentrations of 0.5, 1 or 2 μg/mL for 24 h, and the relative distribution of these cells in various phases of the cell cycle was determined using flow cytometric analysis. As shown in Fig. 2b, c, there was no statistical difference in cell cycle between ODN M362-, ODN M362 Ctrl- and liposome-treated cells. However, the apoptotic subpeak was found before the G0/G1 phase in cells treated with ODN M362 Ctrl or ODN M362 (Fig. 2b). Therefore, AnnexinV/PI staining assay was undertaken to further investigate the pro-apoptotic effect of ODN M362 and ODN M362 Ctrl on HCC cells. As shown in Fig. 2d, the exposure of H7402 cells to ODN M362 Ctrl or ODN M362 (0.5–2 μg/mL) for 24 h resulted in a significant increase in the percentage of AnnexinV⁺ apoptotic cells relative to the Lipo control. The degree of apoptosis with ODN M362 Ctrl at concentrations of 2 μg/mL was >50 %. Moreover, as shown in Fig. 2e, TUNEL immunopositive cells were significantly increased after exposure to ODN M362 or ODN M362 Ctrl. These findings indicated that both ODN M362 Ctrl and ODN M362 induced HCC cell apoptosis.

ODN M362 and ODN M362 Ctrl upregulate inflammatory cytokines without activating NF-κB

Quantitative RT-PCR was used to verify whether ODN M362 and ODN M362 Ctrl activated TLR9 in HCC. The results showed that TLR9 mRNA was upregulated by eight to ten fold following stimulation with ODN M362 or ODN M362 Ctrl (Fig. 3a). We also demonstrated a marked increase in IFN-α, IFN-β, TNF-α, IL-6 and IL-8 mRNAs in H7402 cells exposed to ODN M362 or ODN M362 Ctrl (Fig. 3b). The mRNA levels of IFN-β and IL-6 in ODN M362-treated and ODN M362 Ctrl-treated H7402 cells did not show significant differences, whereas in ODN M362-treated H7402 cells, the levels of the other cytokines were higher than that in ODN M362 Ctrl-treated H7402 cells. However, the expression of phosphorylated NF-κB was not obviously changed, which was different from the response of HCC cells to LPS stimulation (Fig. 3c). These data showed that upregulation of TLR9 and inflammatory cytokines by ODN M362 and ODN M362 Ctrl was not accompanied by activation of the NF-κB signaling pathway.

Fig. 3.

ODN M362 Ctrl and ODN M362 were unable to activate the NF-κB signaling pathway. HCC cells were transfected with 0.5 μg/mL of ODN M362 Ctrl or ODN M362. The expression levels of TLR9 (a) and inflammatory cytokine genes (b) were detected by qRT-PCR. c Phosphorylations of NF-κB and NF-κB in cells treated as above were detected by Western blotting. Results are representative of three independent experiments. *p < 0.05; **p < 0.01 compared with Lipo alone

ODN M362 and ODN M362 Ctrl-mediated HCC apoptosis is not dependent on TLR9

Since both ODN M362 and ODN M362 Ctrl increased the expression of TLR9 in HCC, it was necessary to determine whether apoptosis induced by these agents was caused by TLR9 activation. As shown in Fig. 4a, TLR9 expression in H7402 cells was silenced by the TLR9-specific siRNAs, siRNA1 and siRNA3. However, the amount of ODN M362 or ODN M362 Ctrl-induced apoptosis of HCC cells was not influenced by pretreatment with siRNA1 or siRNA3 (Fig. 4b). These findings clearly indicated that ODN M362 and ODN M362 Ctrl-mediated HCC apoptosis was independent of their effects on TLR9.

Fig. 4.

ODN M362 Ctrl and ODN M362 induced HCC apoptosis independently of TLR9 activation. a HCC cells were transfected with 100 nM of TLR9-specific siRNAs for 24 h, and the expression level of TLR9 was detected by qRT-PCR. b HCC cells pretreated with TLR9-specific siRNA1 or siRNA3 were transfected with ODN M362 Ctrl or ODN M362 for 24 h. The apoptosis rate was determined by using the AnnexinV/PI method. The results are representative of three independent experiments. *p < 0.05; **p < 0.01 compared with the negative control (NC)

Phosphorothioate modification was involved in CpG ODN-induced HCC apoptosis

As described above, ODN M362 and its negative control, ODN M362 Ctrl, had similar effects on HCC. Although the nucleotide sequences of ODN M362 and ODN M362 Ctrl are different, both mediators are phosphorothioate-modified derivatives of CpG ODN. To confirm whether phosphorothioate modification was involved in CpG ODN-induced HCC apoptosis, a synthetic phosphodiester CpG ODN (ODN 2006-G5) was selected as a control. As shown in Fig. 5, compared to ODN M362 and ODN M362 Ctrl, the non-phosphorothioate-modified ODN 2006-G5 was unable to induce HCC cell apoptosis. These data indicated that phosphorothioate modification might have contributed to CpG ODN-induced HCC apoptosis.

Fig. 5.

Phosphorothioate modification was involved in CpG ODN-induced HCC apoptosis. HCC cells were transfected with ODN M362 Ctrl, ODN M362 and ODN 2006-G5 (0.5 μg/mL) for 24 h, and the apoptosis rates were analyzed by flow cytometry with AnnexinV/PI staining. The figure shows representative results of one of at least three independent experiments

ODN M362 and ODN M362 Ctrl suppressed the growth of HCC tumor cells in vivo

The anti-HCC effects of ODN M362 and ODN M362 Ctrl were confirmed by in vivo experiments in H7402-bearing nude mice. When palpable tumors had been established after 15 days, the mice were injected with ODN M362-liposome or ODN M362 Ctrl-liposome solution (5.0 μg in 100 μL) intratumorally every 4 days until day 55. Meanwhile, PBS, liposome and ODN 2006-G5 treatment groups were used as controls. As shown in Fig. 6, tumor growth in ODN M362- and ODN M362 Ctrl-treated mice was inhibited by 46.9 and 49.0 %, respectively, compared to controls. However, ODN 2006-G5 did not show a significant inhibitory effect on HCC growth compared with PBS or liposome treatment. Additionally, we found H7402 cells exposed to ODN M362- or ODN M362 Ctrl-liposome in vitro formed no visible tumor after they inoculated subcutaneously into nude mice at 40 days (Figure S1A). Treatment with ODN M362 or its Ctrl had no effect on the ratio of spleen weight to body weight (Figure S1B) or the ratio of liver weight to body weight (Figure S1C).

Fig. 6.

Both ODN M362 Ctrl and ODN M362 inhibited the tumor formation and exerted antitumor activity in human HCC xenografts in nude mice. H7402 cells were inoculated subcutaneously on day 0. ODN M362-liposome, ODN M362 Ctrl-liposome, ODN 2006-G5-lipsome, liposome or PBS was injected intratumorally every 4 days for nine times, starting on day 15. a The growth curves of HCC. b Tumor weight (mean ± SEM) of seven mice is shown. c The survival curves of HCC-bearing mice shown as percentages of the initial number of animals per group. *p < 0.05; **p < 0.01 compared with Lipo alone

Discussion

Toll-like receptor agonists have been widely used for immunotherapy [6]. Recent reports indicate that TLRs are broadly distributed in tumor cells, such as melanoma, colon tumor and breast tumor cells [23, 24]. When used in the adjuvant setting, TLR agonists augment the adaptive and innate immune responses and have direct effects on tumor cells. However, different TLRs have been shown to exhibit either antitumor or tumorigenic activity [16, 18]. In addition, the same TLR can have different effects in different types of tumor [25].

Our results showed that TLRs, apart from TLR7 to TLR8, are present in HCC cell lines HepG2, H7402 and PLC/PRF/5. This finding agreed with the results of a previous study [26].

Phosphorothioated CpG ODN displays enhanced cellular uptake, increased immunostimulatory activity and a prolonged half-life compared with non-phosphorothioated CpG ODN [27, 28]. We, therefore, selected the TLR9 agonist ODN M362 for our experiments. Our results indicated that ODN M362 inhibited the proliferation of HCC cells by inducing apoptosis, independently of any effect on the cell cycle (Fig. 2). This finding is consistent with a previous study, showing that liposome-complexed CpG ODNs induced caspase-dependent apoptotic cell death and suppressed the proliferation of TLR9-expressing neuroblastoma cells [17].

Tanaka et al. [26] showed that the expression of TLR9 on the surface of human HCCs promotes HCC cell proliferation and survival following extracellular ODN M362 stimulation. However, contrary to our findings (Fig. 2), intracellular TLR9 receptors did not affect cell viability when ODN M362 was transfected into HCC cells [26]. Careful comparison of the protocols showed that the concentrations used in our study were significantly lower than those in the previously reported study with ODN M362. We showed that ODN M362 induced HCC cell apoptosis and upregulated the expression of inflammatory cytokine genes in a dose-dependent manner, at concentrations below 0.5 μg/mL. By contrast, high concentrations of CpG ODN (e.g., 5 μg/mL) had no effect on HCC cells (Figure S2). In the study reported by Tanaka et al., the dose of CpG ODN (1 μM; 8 μg/mL) was much higher than the effective dose, and consequently, they were unable to show evidence of HCC cell apoptosis.

Interestingly, both ODN M362 and its negative control sequence, ODN M362 Ctrl, displayed similar proapoptotic effects on HCC (Fig. 2). Although the nucleotide sequences of ODN M362 and ODN M362 Ctrl are different, both mediators are phosphorothioate-modified derivatives of CpG ODN. To confirm whether phosphorothioate modification was involved in CpG ODN-induced HCC apoptosis, we compared the proapoptotic effects of ODN M362, ODN M362 Ctrl and a synthetic phosphodiester CpG ODN (ODN 2006-G5). The results showed that the non-phosphorothioate-modified ODN 2006-G5 was unable to induce HCC cell apoptosis (Fig. 5). These findings suggest that cell apoptosis induced by ODN M362 and ODN M362 Ctrl may be associated with the phosphorothioate modification of CpG ODN. It has also been shown that CpG, non-CpG and TLR9 antagonist ODNs all have the potential to directly co-stimulate mouse and human CD4⁺ T cells through a TLR9- and MyD88-independent mechanism [29]. These findings warrant further investigation.

Consistent with previous findings [30–32], stimulation of TLR9 in HCC cells induced upregulation of TLR9 and inflammatory cytokines, without affecting the NF-κB signaling pathway (Fig. 3). Silencing experiments using a TLR9-specific siRNA were performed to determine whether the apoptosis induced by the TLR9 agonist was dependent on TLR9. Although the siRNA was unable to silence TLR9 expression completely, it produced a sufficient level of inhibition to abrogate the effects of CpG ODNs [33]. We also observed that H7402 cells transfected with siRNA against TLR9 exerted the same sensitivity to TLR9 agonist-induced apoptosis (Fig. 4), indicating that the apoptosis induced by ODN M362 Ctrl and ODN M362 occurred independently of TLR9 stimulation. Recent studies have demonstrated TLR9- and MyD88-independent mechanisms in ODN-stimulated immune cells, including B lymphocytes and neutrophils [34–37]. In some of these studies, CpG and non-CpG ODNs had the same effects in certain cell types, but the underlying mechanism remained unclear.

TLR9 is a broadly expressed receptor in neuroblastoma cells, and mediated antitumor activity in a mouse xenograft model of human neuroblastoma [17]. Consistent with this study, our data showed that administration of ODN M362 in vivo retarded tumor growth in a human HCC xenograft mouse model (Fig. 6). Furthermore, we showed that delivery of CpG ODN directly to the tumor at a lower dose (5 μg) than previously used to activate human leukocytes, prevented enlargement of lymphoid tissue [38–40].

In conclusion, our study showed, for the first time, that phosphorothioate-modified CpG ODN induced human HCC cell apoptosis and exerted an anti-tumor activity against HCC cells in vivo. These findings suggest that phosphorothioate-modified CpG ODN could not only induce the activation of immune system but also plays a key role in tumorigenesis. By employing liver-targeting delivery systems, such as liver-targeting nanoparticles [43], liver-targeting liposomes [44] and PEGylated immuno-lipopolyplexes [45], phosphorothioate-modified CpG ODN might be a potential therapeutic option for human HCC.

Electronic supplementary material

Below is the link to the electronic supplementary material.