Background
Worldwide, hepatocellular carcinoma (HCC) is the fifth most common cancer and the second leading cause of cancer death in men, and the seventh most common cancer and the sixth leading cause of cancer death in women[
1]. At present, surgical resection is the first choice for treatment of HCC, but the long-term prognosis remains unsatisfactory; the recurrence rate is high, owing to the lack of an effective adjuvant therapy[
2]. Fan et al. reported that the presence of circulating liver cancer stem cells (CSCs) is closely related to HCC recurrence and metastasis after resection[
3]. Recent studies have suggested that antiangiogenic therapy can stimulate tumor invasion and metastasis[
4,
5], and the mechanism of this stimulation may also be related to CSCs[
6].
CSCs have the ability to self-renew, to differentiate into defined progeny, and, most importantly, to initiate and sustain tumor growth, and they play a key role in tumor progression, metastasis, and recurrence[
7]. CSCs also show resistance to chemotherapeutics and radiation[
8]. In HCC, CD133
+ cells exhibit liver CSC–like properties, such as high clonogenicity, tumorigenicity, and resistance to radiation[
9‐
13]. In addition to CD133, EpCAM[
14], CD44[
13] and CD90[
15] have also been used as markers for the identification of HCC CSCs. Tang et al. reported that CD133 expression overlaps extensively with expression of EpCAM and CD44, which suggests that stem cells marked by CD133/EpCAM and possibly CD133/CD44 may have similar characteristics and regulatory mechanisms[
11]. Tang et al. also showed that CD90 is expressed at very low levels in HCC cell lines and clinical specimens. Other studies have shown that the presence of CD133
+ CSCs in HCC patients after surgery is correlated with early recurrence and poor prognosis[
16,
17]. Two studies have reported that hepatocyte nuclear factor 4 alpha and bone morphogenetic protein 4 can promote the differentiation of CD133
+ HCC stem cells, inhibition of self-renewal, and resistance to chemotherapy[
18,
19]; and these results suggest that inducing CSC differentiation is a promising approach to the treatment of HCC. However, the use of differentiation-inducing drugs for HCC has not been well explored.
A number of studies have shown that the Hedgehog (HH) signaling pathway plays an important role in CSC self-renewal and inhibition of tumor differentiation in various cancers[
20‐
23]. Targeting HH signaling has been shown to inhibit self-renewal of melanoma CSCs
in vitro and to decrease their tumor initiation ability
in vivo[
24]. In a clinical study, advanced basal cell carcinoma patients were found to benefit from inhibition of HH signaling[
25]. In HCC,
GLI1 expression is significantly correlated with tumor size and TNM stage;
GLI1 expression is high in patients with early recurrence and poorer overall survival[
26]. Therefore,
GLI1 may be a useful target for the prevention of HCC recurrence after surgery.
Arsenic trioxide (As
2O
3) induces differentiation and apoptosis of acute promyelocytic leukemia (APL) cells by binding to the PML-RARα oncoprotein[
27], and treatment with As
2O
3 has been shown to increase remission rates and prolong survival in patients[
28]. Some studies have shown that As
2O
3 inhibits the growth and metastasis of HCC by inducing apoptosis[
29,
30], but a clinical trial showed that treatment with As
2O
3 is not effective for patients with advanced HCC[
31]. There have been no reports of As
2O
3 preventing HCC recurrence or inducing differentiation of HCC CSCs.
We hypothesized that As2O3 could induce differentiation of CD133+ HCC cells through the HH-GLI pathway, and also investigated whether As2O3 could inhibit HCC recurrence and prolong survival time after hepatectomy in a mouse model.
Discussion
The CSC theory is based on the fact that CSCs exhibit properties similar to those of normal stem cells, such as self-renewal, the ability to produce heterogeneous progeny, and the ability to divide in an unlimited way, giving rise to high tumorigenicity, chemotherapy and radiation resistance, metastasis, and cancer recurrence after therapy[
8]. Therefore, CSCs have the potential to be effective targets for preventing relapse after resection, including in HCC. Therapy involving induced differentiation of CSCs could lead to the cells losing their self-renewal ability and could induce terminal differentiation. To date, the best example of the clinical use of differentiation-inducing therapy is treatment of APL with all-trans retinoic acid, which enhances the effects of chemotherapy and significantly improves patient outcome[
37]. Targeting induction of CSC differentiation may be a powerful therapy for HCC. CD133
+ HCC cells are considered to be CSCs because of their higher tumorigenicity[
10] and lower expression of mRNAs for mature hepatocyte markers[
9]. As
2O
3, which is an FDA-approved agent for the treatment of APL, can induce APL cells differentiation. Our results revealed that CD133 expression and the percentage of CD133
+ cells were dramatically decreased in both CD133
+ Huh7-wt and cd133
+ Hep3B-wt cells treated with As
2O
3. But real-time PCR analysis revealed that after the CSCs were treated with As
2O
3 for 48 hours, a cluster of hepatocyte marker genes—including phosphoenolpyruvate carboxykinase, glucose-6-phosphatase, cytochrome P450, glutamine synthetase, biliverdin reductase, aldolase B, apolipoprotein c III, apolipoprotein A I, 4-hydroxyphenylpyruvate dioxygenase, and glycogen synthetase 2—were not re-expressed (data not shown). Furthermore, we found that As
2O
3 treatment decreased the expression of some stemness genes, which are important for the maintenance of CSC self-renewal ability and tumorigenicity[
13,
38]. We also found that high-dose As
2O
3 reduced CSC proliferation, whereas low-dose As
2O
3 had no such effect. CSCs are considered to be more resistant to chemotherapeutic agents than non-CSCs, and consistent with this, low-dose As
2O
3 has been found to inhibit Huh7-wt cell proliferation[
39]. In short, low-dose As
2O
3 may induce HCC CSCs differentiation into non-CSCs,which is very important for the clinical application of As
2O
3 because of its toxicity.
It should be possible to treat cancer by differentiation-inducing therapy targeted at CSCs. If CSCs could be induced to differentiate, then their malignant potential could be controlled[
34]. We have shown that As
2O
3 induced HCC CSC differentiation
in vitro, and thus we postulate that As
2O
3 may be useful for inhibiting recurrence of HCC after resection and may prolong survival
in vivo. The HCC recurrence rate was decreased by systemic administration of As
2O
3, and survival time was significantly prolonged. However, in a phase II clinical trial, patients with advanced HCC did not benefit from As
2O
3 treatment[
31], and we believe that the effective treatment of advanced HCC must involve a combination of differentiation-inducing therapy and conventional chemotherapies to eradicate the tumor mass, and this belief is partly supported by the results of Wang et al.[
13]. Due to the toxic effects of As
2O
3 limits its clinical application, we found that low-dosage of As
2O
3 can induce differentiation of HCC CSCs and inhibit recurrence, so the application of As
2O
3 to prevent recurrence of HCC patients after surgery may be an effective way. We also found CTCs were decreased by As
2O
3 treatment. Because the presence of CTCs is associated with metastasis to distant organs and is strongly associated with poor overall survival[
40], we suggest that As
2O
3 could decrease metastasis and prolong survival. This suggestion could be tested with other models.
The HH-GLI pathway is an important mechanism for determining embryonic pattern and regulating cell fate[
11]. This pathway has been implicated in several types of tumors[
41] and plays an important role in the differentiation of CSCs[
20‐
23]. Differentiation of CSCs by inhibition of the HH-GLI pathway may be a promising therapeutic strategy for human tumors. Currently, all the therapeutics in clinical development that function by inhibiting the HH-GLI pathway are targeted at
SMO[
41]. However, in some types of tumors, specifically those in which increased
GLI expression or activation is induced in a
SMO-independent manner,
SMO inhibitors are ineffective, and inhibitors that modulate
GLI may be useful[
42‐
44]. In HCC, expression of
GLI1 mRNA has been reported to adversely affect recurrence and survival of patients with HCC after resection[
26], which is consistent with our results, As
2O
3 may induce CSC differentiation by down-regulating the expression of
GLI1, thus inhibiting recurrence and prolonging survival.
Methods
Cell culture and reagents
Unmodified Huh7 and Hep3B cells (Huh7-wt and Hep3B-wt) were obtained from American Type Culture Collection. Cells from both cell lines were transfected with GFP (Huh7-GFP and Hep3B-GFP, respectively) as described previously[
45]. Huh7-LV-shGLI1 and Hep3B-LV-shGLI1 cells were obtained by infecting CD133
+ Huh7-wt and CD133
+ Hep3B-wt with a lentiviral vector encoding short hairpin RNA for
GLI1 (sc-37911-V, Santa Cruz Biotechnology, Santa Cruz, CA, USA) to silence its expression. As controls, Huh7-LV-shNon and Hep3B-LV-shNon cells were obtained by infection of CD133
+ Huh7-wt and CD133
+ Hep3B-wt with a different lentiviral vector (sc-108080, Santa Cruz Biotechnology). Huh7-LV-oeGLI1 and Hep3B-LV-oeGLI cells were obtained by infecting CD133
+ Huh7-wt and CD133
+ Hep3B-wt with lentiviral vectors that overexpress
GLI1 (Genechem, Shanghai, P.R. China). Huh7-LV-oeNon and Hep3B-LV-oeNon cells were also used as controls (Genechem). All cell lines were cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum.
As2O3 (SL Pharm, Beijing, P.R. China) was dissolved at a concentration of 0.05 mg/ml in phosphate buffered saline (PBS) for the in vitro study. For the in vivo study, As2O3 was dissolved in normal saline.
Cell isolation by fluorescence-activated cell sorting
Huh7-wt and Hep3B-wt cells were labeled directly with phycoerythrin-conjugated anti-human CD133/1 antibody (Miltenyi Biotec, Gladbach, Germany) according to the manufacturer’s instruction and were sorted by fluorescence-activated cell sorting to obtain CD133+ cell subpopulations; a sorted cell purity exceeding 90% was deemed acceptable for in vitro experiments.
Cell proliferation assay
Cell proliferation was counted by a CCK8 assay (Dojindo, Tokyo, Japan). Three thousand CD133
+ cells were seeded in 96-well culture plates. After adhesion for 24 hours, the cells were treated with As
2O
3 at a final concentration ranging from 1 to 6 μM, and the treated cells were cultured for another 24 or 48 hours. Cells that were not exposed to As
2O
3 were used as controls. The cell proliferation assay was carried out as previously described[
5]. The optical density values were read by an enzyme-linked immunosorbent assay reader at 450 nM.
Real-time polymerase chain reaction analysis
Real-time PCR analysis was carried out as described as elsewhere[
5]. The following primers were used for amplification of human genes: CD133[
18], forward 5′- ACATGAAAAGACCTGGGGG-3′, reverse 5′- GATCTGGTGT CCCAGCATG-3′; β-actin, forward 5′-GCTCTGCAGACTTCAGACCA-3′, reverse 5′-GGCCGGACTCATCGTACTCCTGC-3′.
Western blotting assay
The procedure used for Western blotting assay is described elsewhere[
5]. Primary antibodies included anti-CD133/1 (Miltenyi Biotec), anti-GLI1 (Santa Cruz Biotechnology), and anti-β-actin (Kangcheng Technology, Shanghai, P.R. China).
Immunofluorescence
To assess the distribution and changes of CD133 in HCC CSCs after treatment with As2O3, we stained CD133+ Huh7-wt and CD133+ Hep3B-wt cells for visualization by means of an immunofluorescence assay. Briefly, the cells were seeded on slides, allowed to adhere for 24 hours, treated with As2O3 at a final concentration ranging from 1 to 6 μM, and then cultured for another 24 or 48 hours. Cells that were not exposed to As2O3 were used as controls. The slides were washed with PBS and fixed with 4% paraformaldehyde. The cells were incubated with primary antibody to CD133/1 (Miltenyi Biotec) overnight at 4°C after being blocked with 5% bovine serum albumin, and then goat anti-mouse tetramethyl rhodamine isothiocyanate-conjugated secondary antibody was added before staining with 4′,6-diamidino-2-phenylindole. Fluorescence images were visualized with an inverted fluorescence microscope (Olympus, Melville, NY, USA). For a negative control, primary antibodies were replaced with PBS.
Flow cytometry analysis
The percentages of CD133+ cells were determined by flow cytometry to evaluate the expression of CD133. Isolated CD133+ HCC cells were incubated with or without As2O3 for 5 days. The cells were trypsinized, washed, and resuspended in PBS. Then the cells were incubated with phycoerythrin-conjugated anti-CD133/1 antibody (Miltenyi Biotec) on ice for 30 min and analyzed by means of flow cytometry (BD Biosciences, USA).
Culture of HCC CSC spheres
The self-renewal capability of HCC cells was evaluated by testing their sphere-formation ability. Specifically isolated CD133
+ HCC cells treated with 4 μM As
2O
3 treatment for 5 days and untreated cells were cultured in a methyl cellulose–based medium in low-adherent 96-well culture plates (Corning, Corning, NY, USA) under serum-free conditions and supplemented with 20 μg/ml insulin, 20 μg/ml epidermal growth factor, and 10 μg/ml basic fibroblast growth factor (RD Systems, Minneapolis, MN, USA) according to a previously published procedure[
46]. Single-cell suspensions of 100 CD133
+ HCC cells were seeded, and epidermal growth factor, basic fibroblast growth factor, and insulin were added every second day for 2 weeks.
Tumorigenic capacity of HCC CSCs
The tumorigenic capacity of 4 μM As2O3-treated and untreated CD133+ HCC cells for 5 days were analyzed in an NOD/SCID mouse xenograft model. Male NOD/SCID mice (4–6 weeks old) were injected subcutaneously in the lateral flanks with 5 × 104 CD133+ Huh7-wt cells and 2 × 105 CD133+ Hep3B-wt cells. Tumorigenic capacity was evaluated after the cells had been allowed to undergo implantation for 7 weeks.
Animal model
Male BALB/c nu/nu nude mice (Shanghai Institute of Materia Medica, Chinese Academy of Science, Shanghai, P.R. China) were housed in laminar-flow cabinets under specific pathogen-free conditions and used when they weighed approximately 20 g and were 4–6 weeks old. The experimental protocol was approved by the Shanghai Medical Experimental Animal Care Committee. The animal model was established as previously described[
47]. Briefly, Huh7-wt, Hep3B-wt, Huh7-GFP, and Hep3B-GFP cells (1 × 10
7) were subcutaneously inoculated into the right flanks of the nude mice. After 3–4 weeks, a piece of non-necrotic tumor tissue about 2 mm in size was orthotopically implanted into the liver. On the fourteenth day after implantation, the tumor was excised, with the length between incisional margin and tumor edge being ≥ 2 mm.
As2O3 treatment and grouping
The in vivo experiment involved 2 groups: a recurrence rate group, which used the Huh7-GFP and Hep3B-GFP cell lines, and a survival observation group, which used the Huh7-wt and Hep3B-wt cell lines. In the recurrence rate group, treatment was started on the fifth day after resection. The mice were randomly assigned to 2 groups (n = 12 for each group): one group received a daily intraperitoneal dose of 1.5 mg/kg As2O3, and the other group received an equal volume of normal saline (control group). The treatment was continued for 2 weeks, and after another 3 weeks, the mice were anesthetized and orbital blood was obtained. The mice were then sacrificed by cervical dislocation to determine the recurrence rate.
In the survival group, treatment was started on the fifth day after resection. The mice were randomly assigned to 2 groups (n = 8 for each group): one group received a daily intraperitoneal dose of 1.5 mg/kg As2O3, and the other group received an equal volume of normal saline (control group). The treatment was administered for 2 weeks, then stopped for 1 week to allow the mice to recover, and then administered for an additional 2 weeks; then the median survival was determined.
Detection of CTCs
CTCs in the peripheral blood of mice in the recurrence group were counted by flow cytometry, with GFP as a marker. The protocol was carried out as previously reported[
15]. The GFP-positive cells were gated with processed blood from a mouse that had not been subjected to the xenograft procedure.
PCR microarray analysis of stemness gene expression and pathway exploration
Statistical analysis
Values for continuous variables are expressed as means ± SDs and were compared by means of the unpaired 2-tailed Student t test, unless otherwise specified, Multiple comparisons of Kaplan-Meier curves were made using the log-rank test. All statistical tests were performed using SPSS for Windows (ver. 120.0, SPSS, Inc). P < 0.05 (2-tailed) was considered to indicate statistical significance.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
ZKZ designed the study, established the animal model, carried out the Western blotting assays, performed the statistical analysis, and drafted the manuscript. ZQB and ZQB participated in the design of the study, data analysis, and drafting the manuscript. SHC participated in the design of the study and helped to draft the manuscript. AJY, CZT, ZXD, LL, ZYY, BY, and KLQ helped to acquire experimental data. TZY conceived the study, participated in its design and coordination, and helped to draft the manuscript. All authors read and approved the final manuscript.