Elimination of Cancer Stem-Like “Side Population” Cells in Hepatoma Cell Lines by Chinese Herbal Mixture “Tien-Hsien Liquid”

There are increasing pieces of evidence suggesting that the recurrence of cancer may result from a small subpopulation of cancer stem cells, which are resistant to the conventional chemotherapy and radiotherapy. We investigated the effects of Chinese herbal mixture Tien-Hsien Liquid (THL) on the cancer stem-like side population (SP) cells isolated from human hepatoma cells. After sorting and subsequent culture, the SP cells from Huh7 hepatoma cells appear to have higher clonogenicity and mRNA expressions of stemness genes such as SMO, ABCG2, CD133, β-catenin, and Oct-4 than those of non-SP cells. At dose of 2 mg/mL, THL reduced the proportion of SP cells in HepG2, Hep3B, and Huh7 cells from 1.33% to 0.49%, 1.55% to 0.43%, and 1.69% to 0.27%, respectively. The viability and colony formation of Huh7 SP cells were effectively suppressed by THL dose-dependently, accompanied with the inhibition of stemness genes, e.g., ABCG2, CD133, and SMO. The tumorigenicity of THL-treated Huh7 SP cells in NOD/SCID mice was also diminished. Moreover, combination with THL could synergize the effect of doxorubicin against Huh7 SP cells. Our data indicate that THL may act as a cancer stem cell targeting therapeutics and be regarded as complementary and integrative medicine in the treatment of hepatoma.


Introduction
In recent years, there has been increasing evidence supporting the notion that tumors are organized in hierarchical heterogeneous cell populations [1]. The capability to sustain tumor growth resides exclusively in a small proportion of tumor cells termed cancer stem cells (CSCs) or tumorinitiating cells [2], which have the properties of self-renewal, multilineage differentiation capacity, and, most importantly, the tumorigenicity. More importantly, recent researches show that CSCs are much more resistant to conventional cancer therapy than the other bulk cancer cells. CSCs have been considered to be the main cause for the failure of cancer treatment. Searching for the CSCs targeting therapeutics becomes a new strategy and challenge for improving the clinical outcome of cancer patients.
One common method to isolate the CSCs for therapeutics screening is the side population (SP) technique, which is originally used to detect the hematopoietic stem cells in bone marrow. This technique detects the so-called SP cells by dual-wavelength flow cytometry on the basis of the ability of these cells to efflux the fluorescent DNA-binding dye Hoechst 33342 [3]. The SP phenotype is characterized by breast-cancer-resistant protein-1 (BCRP1 or ABCG2), one of 2 Evidence-Based Complementary and Alternative Medicine ATP-binding cassette (ABC) transporters, which is associated with multidrug resistance in many cancers by pumping out the drugs [4]. Since multidrug resistance is an important characteristic of CSCs, it has also been shown that the SP from cancer cells is enriched by CSCs [5]. Thus, SP cells are postulated to be a source of CSCs and represent an important potential target for cancer therapy. Recent work has led to the detection of the SP in a variety of tumor types, including leukemia, glioma, medulloblastoma, hepatoma, as well as breast, prostate, thyroid, colorectal, and ovarian carcinoma [6].
Lots of phytochemicals from fruits, vegetables, and herbs have anticancer activities and represent a promising therapeutic approach for the prevention and treatment of many cancers. The effects of phytochemicals on inhibiting tumor formation are well demonstrated both in vitro and in vivo [7]. Many of these compounds, such as berberine, curcumin, piperine, and cannabinoids [8][9][10], had been reported to eliminate cancer-stem-like cells. Natural products like herbal medicines, which possess evidence of molecular anticancer effects, may be considered as a potential source of therapeutics targeting on CSCs.
The Tien-Hsien Liquid (THL) is a Chinese herbal mixture, which has been used as a complementary anticancer agent for more than 10 years worldwide. It is aqueous preparation of herbal mixture and consists mainly of extracts from 14 Chinese herbs: Cordyceps sinensis (CS), Oldenlandia diffusa (OD), Indigo pulverata levis (IPL), Polyporus umbellatus (PU), Radix astragali (RA), Panax ginseng (PG), Solanum nigrum L. (SNL), Pogostemon cablin (PC), Atractylodis macrocephalae rhizoma (AMR), Trichosanthes radix (TR), Clematis radix (CR), Margarite (M), Ligustrum lucidum Ait (LLA), and Glycyrrhiza radix (GR). The biological activities of these herbs have been reported individually, including antioxidation, immunomodulation, antimutagenesis as well as cytostatic or cytotoxic effects. Recently, THL had been shown to induce apoptosis in many types of cancer cells and activate caspase-8, -9, and -3 in H1299 lung cancer cells [11]. Its effects on targeting PML-RARα and oncogenic signaling pathways in acute promyelocytic leukemia NB4 cells had been demonstrated in our previous study [12]. More recently, its inhibitory effects on the metastasis, angiogenesis, and tumor growth had been reported by Chia et al. [13]. Regarding the crucial role of CSCs in the metastasis and progression of tumors [14], it is worthy and interesting to explore the effects of THL on the elimination of CSCs.
In this study, we separated and characterized cancer stem-like SP cells from human hepatoma cell lines to investigate the effects of THL on CSCs elimination. Our data indicate that THL could eliminate the cancer stem-like SP cells, accompanied with the suppressions of stemness genes expression, colony formation as well as tumorigenicity. These results further elucidate the mechanisms underlying the anticancer effects of this Chinese herbal mixture, which suggests its potential role as complementary medicine for cancer treatment.

Side Population Analysis and Purification Using Flow
Cytometry. The hepatoma cells were detached from the dishes with Trypsin-EDTA (Invitrogen) and suspended at 1 × 10 6 cells/mL in Hank's balanced salt solution (HBSS) supplemented with 3% fetal calf serum and 10 mM HEPES. These cells were then incubated at 37 • C for 90 minutes with 20 μg/mL Hoechst 33342 (Sigma Chemical, St. Louis, MO), either alone or in the presence of 50 μM verapamil (Sigma), which is an inhibitor of verapamil-sensitive ABC transporter. After 90-minute incubation, the cells were centrifuged immediately for 5 minutes at 300×g, 4 • C and resuspended in ice-cold HBSS. The cells were kept on the ice to inhibit efflux of Hoechst dye and 1 μg/mL propidium iodide (BD Pharmingen, San Diego, CA) was then added to discriminate dead cells. Finally, these cells were filtered through a 40 μm cell strainer (BD Falcon) to obtain singlesuspension cells. Cell dual-wavelength analysis and purification were performed on a dual-laser FACS Vantage SE (BD). The Hoechst 33342 was excited by 355 nm UV light and collect blue fluorescence with a 450/20 band-pass filter and red fluorescence with a 675 nm edge filter long pass (EFLP). A 610 nm dichroic mirror short pass (DMSP) was used to separate the emission wavelengths. The propidium iodidepositive dead cells were excluded from the analysis.

Colony Formation of SP and Non-SP Cells.
Freshly sorted SP and non-SP cells were counted, plated in triplicate at Evidence-Based Complementary and Alternative Medicine 3 200 cells per well in 6-well plates, and cultured in the medium described in Section 2.4 for 14 days. After most colonies had expanded to >50 cells, they were washed twice with PBS, fixed in methanol for 15 min, and dyed with crystal violet for 15 min at room temperature to visualize colonies for counting. Colony number and size were scored with the ChemiDoc-XRS imager, using the QuantityOne software package (BioRad, Hercules, CA). The declined colony counts represented the inhibitory effects of THL on colony formation of Huh7 SP cells.

Determining the Cell Viability by Sulforhodamine B (SRB)
Assay. Both the SP and non-SP cells were seeded in 96well plate at a density of 3 × 10 3 cells/well in the medium as described in Section 2.4. After 24 h of culture, cells were treated with drugs as indicated in Figure 6 and Table 1 for 48 h. At harvest, cells were fixed by 10% trichloroacetic acid (TCA). After washing with distilled water, the viable cells were stained by SRB dye at 0.4% (w/v) in 1% acetic acid. The unbound dye was removed by repeated washing with 1% acetic acid and the plates were air-dried. The cellbound SRB dye was subsequently solubilized with 10 mM trizma base, and the absorbance was read on a microplate reader at a wavelength of 570 nm. The absorbance is directly proportional to the cell number over a wide range.

Preparation of Cytoplasmic and Nuclear Proteins.
Cytoplasmic and nuclear extracts of cells were prepared using the Nuclear Extraction Kit (Panomics). Briefly, harvested cells (1 × 10 6 cells/6 cm plate) were washed twice with 5 mL cold 1 × PBS. A 0.5 mL aliquot of Buffer A working reagent (a combination of 0.5 mL 1 × Buffer A, 5 μL DTT, 5 μL protease inhibitor cocktail, and 20 μL 10% IGEPAL) was added to each plate. The plate was transferred to an ice bucket on a rocking platform at 150 rpm for 10 min. Each sample was centrifuged at 14,000 × g for 3 min at 4 • C. The supernatant (cytoplasmic fraction) was removed and the pellet kept on ice. A 75-mL aliquot of Buffer B working reagent (a combination of 0.5 mL 1 × Buffer B, 5 μL DTT, and 5 μL protease inhibitor cocktail) was added to each pellet and vortexed at the highest setting for 10 sec. Each sample was then placed in ice bucket and shook in rocking platform at 150 rpm for 2 h. After centrifugation at 14,000 × g for 5 min at 4 • C, the supernatant (nuclear fraction) was transferred to a new Eppendorf tube for the measurement of the protein concentration of each sample, and was stored at -80 • C.

Western
Blotting. Samples (10 μg) of cytoplasmic or nuclear proteins were size-fractionated electrophoretically by a 10% polyacrylamide SDS-PAGE gel and transferred onto a PVDF membrane using the Bio-Rad Mini-Protean electrotransfer system. The blots were subsequently incubated with 5% skim milk in PBST for 1 h to block nonspecific binding and were probed overnight at 4 • C with the antibodies against total β-catenin (1 : 1000, Cell Signaling), Lamin (1 : 5000, Santa Cruz), and β-tubulin (1 : 5000, Santa Cruz). The membranes were sequentially detected with an appropriate peroxidase-conjugated secondary antibody incubation at room temperature for 1 h. Intensive PBS washing was performed after each incubation step. After the final PBS washing, signals were developed using the ECL (enhanced chemiluminescence) detection system and Kodak X-OMAT Blue Autoradiography Film.

Combination Index (CI) Measurements.
Combination index (CI) between THL and doxorubicin (St. Louis, MO, USA) was obtained by a computer program (CalcuSyn) based on the median effect equation of Chou and Talalay [15]. The CI values below 1 indicate synergistic effects whereas those equal or close to 1 are additive and those above 1 are antagonistic. The analysis used in this study was under the assumption of mutual nonexclusiveness of the mechanism of drug action.

Detection of Side Population in Human Hepatoma Cells.
To determine whether the selected hepatoma cell lines contained SP cells, we stained these cells with Hoechst 33342, which could be actively extruded by verapamil-sensitive ABC transporters. Representative results analysed by flow cytometry were shown in Figure 1. A small percentage of SP cells were found in 1.05% of HepG2, 1.55% of Hep3B, 1.69% of Huh7, 0.81% of PLC/PRC/5, and 1.08% of SK-Hep1 cells, respectively, which were decreased markedly in the presence of verapamil. When preincubated with verapamil for 90 min, the percentage of side population cells shown on the flow cytometer dropped to 0.04% of the total cells (Figure 2(a)). This result is consistent with the reports that Hoechst 33342 exclusion is verapamil sensitive. The SP cells were then collected for the subsequent experiments.

Side Population Cells Have Distinct Stem Cell Properties.
As shown in Figure 2(a), the R2 gate showed lower Hoechst 33342 intensity indicated the SP cells, and the R1 gate showed higher Hoechst 33342 intensity indicated the non-SP cells. Like normal stem cells, the RT-PCR analysis reveals that Huh7 SP cells expressed higher levels of ABCG2, CD133, SMO, β-catenin, and Oct4 mRNA than non-SP cells, suggesting that the SP cells have, at least a part, distinct intrinsic properties of stem cells (Figure 2(b)). After 9 days of culture, most colonies had formed and the number of colonies in SP and non-SP cells was 165 and 55, respectively (Figure 2(c)). The spheroid morphology of SP cells was markedly distinct from the fibroblast-like shape of non-SP cells (Figure 2(d)).
In addition, both the nuclear and cytoplasmic β-catenin protein levels of SP cells were markedly higher than those of non-SP cells. The difference between the nuclear βcatenin levels in SP and non-SP cells was even much higher than that between the cytoplasmic levels ( Figure 2(e)). This phenomenon was consistent with that shown in Figure 2(b) and reflected the cancer stemness of Huh7 SP cells.

THL Decreased Proportion of SP Cells in Human
Hepatoma Cell Lines. To evaluate the effects of THL targeting on hepatoma CSCs, we analyzed its inhibitory effects on side population by using flow cytometry and Hoechst 33342 efflux assays. After 2 days of THL treatment at dose of 2 mg/mL, the proportions of SP cells were reduced from 1.33% to 0.49% in HepG2, 1.55% to 0.43% in Hep3B, and 1.69% to 0.27% in Huh7 cells, respectively, as shown in Figure 3.

THL Suppressed Growth and Colony Formation of Huh7
SP Cells. To further investigate how effective was THL against hepatoma SP cells, the growth and colony formation were measured. As expected, THL dose-dependently inhibited both the proliferation and colony formation of Huh7 SP cells. As shown in Figures 4(a) and 4(b), the cell viability and colony number were significantly reduced from 100 ± 2.3% to 11.9 ± 2.1% and 200 ± 5.3 to 21.3 ± 2.3, respectively, by THL at dose of 2 mg/mL.

Downregulation of Cancer Stemness Genes by THL.
To determine the mechanisms underlying the effects of THL on the elimination of Huh7 SP cells, the expression of several stemness genes that were responsible for stem cell self-renewal, proliferative capacity, or lineage differentiation was examined by RT-PCR. As shown in Figure 5, the mRNA levels of ABCG2 and CD133 were decreased in a dosedependent manner after 2 days of THL treatment. Moreover, the Hedgehog signaling pathway genes such as SMO and its downstream Gli were also significantly downregulated by THL. These results suggested the mechanisms responsible for the eradication of Huh7 SP cells by THL are probably through multiple molecular targeting effects.

The Synergistic Inhibitory Effect of THL and Doxorubicin in SP Cells.
To further investigate the CSC targeting effects of THL, we compared the effects of THL on the growth inhibition of Huh7 SP and non-SP cells. The result showed that THL appeared to preferentially inhibit the proliferation of SP cells ( Figure 6). Next, we studied whether the effect of doxorubicin against Huh7 SP cells could be synergized by combining with THL. By calculation, THL (65 μg/mL) or doxorubicin (250 nM) alone produced only 36% and 5% decrease in the viability of Huh7 SP cells as compared to control, respectively. However, simultaneous treatment with these two drugs resulted in a 63.6% decrease in the viability as shown in Table 1. In addition, the combined index (CI) values of this combination (using mutually nonexclusive model) were all well below 1, indicating the synergistic combination effects of doxorubicin with THL.   in Figure 7(a), at dose of 2 mg/mL, the number of tumor spheres was reduced from 39 ± 1.2 of control to 13.5 ± 2.2 by THL, indicating its inhibitory effects on the self-renewal of Huh7 SP cells. In the xenograft NOD/SCID mice model, the tumorigenicity of THL-pretreated Huh7 SP cells was significantly reduced compared with the untreated SP cells (Figure 7(b)). The untreated Huh7 SP cells formed tumor in 5 out of 5 mice, while the THL-treated SP cells formed tumor only in 2 out of 5 mice at the time of 40 days after SP cells inoculation. In addition, the average final tumor size was reduced from 2.4 ± 0.2 cm 3 to 0.48 ± 0.2 cm 3 , suggesting the inhibitory effect of THL on the tumorigenicity of Huh7 SP cells.

Discussion
The recent discovery of cancer stem cells has great importance both for development of new strategy to combat malignant tumor and evaluation of pitfalls and benefits of current therapeutics. Since the SP isolated from hepatoma cells had been reported to possess high proliferation potential, tumorigenicity, and antiapoptotic properties compared with those of non-SP cells [16,17], we used the SP analysis as a tool to evaluate the effects of THL on elimination of cancer stem cells (CSCs). Besides the method of SP analysis, many researches detected and isolated the CSCs by tracking the specific CSCs surface CD markers that were not expressed on the bulk of the other cancer cells [18]. In addition to the multidrug-resistant efflux pump gene ABCG2, which was responsible for effluxing Hoechst 33342, the sorted SP cells also expressed higher levels of not only CD133, a putative cell surface marker for hepatoma stem cells isolation [19], but also other stemness genes such as SMO, β-catenin Oct-4 as well as clonogenicity than those in non-SP cells. The distinct protein expression and nuclear location of β-catenin in SP cells further reflected their cancer stemness. These results indicated the rationality of using SP cells to evaluate the activities of potential targeting agents on CSCs. Nevertheless, Wu et al. [20] had examined mesenchymal tumors ranging from benign to high-grade sarcomas and found that high aggressive tumors were prone to possess highly proportion of SP cells, which might correlate with the poor prognosis in mesenchymal tumors. In accordance with our point, Wu's study also supported that the SP cells were an important therapeutic target for drug intervention.
Unlike the ABCG2 inhibitor verapamil, THL decreased not only the proportion of SP cells, but also the cell viability and colony formation of SP cells. In addition, THL dosedependently inhibited the expression of important stemness genes such as the putative hepatoma stem cell marker CD133 [19], as well as the Hedgehog signaling pathways components SMO and Gli [21]. CD133/prominin-1, a pentaspan membrane glycoprotein, is an important cancer stem cell surface marker in various solid tumors, including liver cancer [22]. It had been shown that the CD133(+) Huh7 hepatoma cells performed a higher proliferative potential, and tumorigenicity, and lower mRNA expressions of mature hepatocyte markers than the CD133(−) population [19].  Clonogenic assay showing the clonogenicity of SP cells was markedly higher than that of non-SP cells. (d) Cell morphology of freshly sorted SP and non-SP cells at days 0, 3, and 9 after seeding. As shown in the photographs, the differences in the shapes of SP (small and round in shape) and non-SP cells (fibroblast-like) were remarkable. (e) Protein expression and nuclear location of β-catenin in SP and non-SP cells. The SP cells had much higher nuclear β-catenin than the non-SP cells. Similar phenomenon was also observed in the cytoplasmic β-catenin level. Scale bar, 50 μm. Similar results of clonogenicity and stemness genes expression were also obtained in HepG2 and Hep3B SP cells (data not shown). NSP: non-SP cells. The suppressing effects of THL on the CD133 of Huh7 SP cells indicated its potential in targeting CSCs in hepatoma. Hedgehog signaling, a crucial factor in regulating selfrenewal of stem cells, is frequently aberrantly activated in CSCs [23] and thus became a potential target for cancer therapy [23,24]. Recent studies also showed that activation of Hedgehog signaling is critically related to CSCs and EMT (epithelial-mesenchymal transition) features in many types of cancers including colon, gastric, esophagus, liver, and other cancers [25]. The diminished expression of SMO and its downstream Gli by THL further revealed its potential role on the eradication of hepatoma CSCs. The ABCG2, which highly expressed in hepatoma SP cells, is associated with drug efflux related to the resistance of doxorubicin [26] and considered as a cause of poor response of hepatoma patients to this drug. Accordingly, our results showed that THL diminished ABCG2 expression in SP cells and synergized the effects of doxorubicin against Huh7 SP cells. On the other hand, CD133 was shown to confer chemoresistance by activation of the AKT/PKB and Bcl-2 cell survival response in hepatoma cells [27], and suppression of the Hedgehog pathway could also sensitize the hepatoma cells to chemotherapeutic agent [28]. Regarding the profound effects of THL on these two important targets ( Figure 5), the synergistic combination effects might also attribute to the THL-suppressed CD133 and Hedgehog signaling pathway.
Like THL, medicinal plants or phytochemicals are potential sources for therapeutics targeting on CSCs. For example, berberine and cannabinoids had been reported to diminish the cancer stem-like cells in breast and brain cancer, respectively [8,9]. Various agents that directly modulate CSCs had been evaluated in vivo and in vitro [29]. However, the biology of CSCs is extremely complex and accompanied with a considerable crosstalk and redundancy in signaling pathways. Hence, targeting only single molecule or pathway inside CSCs may exert limited benefit for treatment. The combination of CSCs eliminating compounds was thus proposed to enhance the efficacy. Combination of curcumin and piperine was found to further reduce the proportion of breast CSCs than when either drug was used alone [10]. Cotreatment with sulforaphane, a broccoli isothiocyanate, could enhance the sorafenib-mediated elimination of pancreatic CSCs in vitro and synergize its effects on tumor size reduction in vivo [30]. As THL consists mainly of extracts from 14 Chinese herbs and possessing activities of multiple  oncogenic signaling pathways inhibition [12], its significant effects against Huh7 SP cells may be resulted from the synergistic combination effects of each active components contained in this herbal mixture. A recent proposed mechanism to eradicate CSCs is epigenetic modulation by depleting the DNA methyltransferase 1 (DNMT1), which plays a critical role on the aberrant  hypermethylation of DNA. DNMT1 depleting agents such as Decitabine and its analog Azacitidine had been shown to inhibit cultured solid tumor stem-like cells and diminish tumorigenicity [31]. Our previous study had showed that THL could intensively decrease the protein level of DNA methyltransferases 1 (DNMT1), an important enzyme for the aberrantly DNA methylation, in acute promyelocytic leukemia cells [12]. In addition, we also found that the expressions of DNMTs (DNMT1, 3a, 3b) were much higher in Huh7 SP cells compared with those of non-SP cells (data not shown). Therefore, THL might also affect the DNMT1 of Huh7 SP cells, which resulted in its eradication effects on these cells. Hence, the involvement of epigenetic modulation through DNMT1 depletion on THL-induced elimination of Huh7 SP cells is worth of further investigation. Taken together, searching and identifying the potential cancer stem cell targeting agents, either from nature products or synthetic chemical compounds, are believed to provide substantial benefit for curing cancer. As combination effect may be more efficient for CSCs elimination, the traditional herbs or recipes could be considered as a potential source for CSCs targeting therapeutics. Based on the results shown in this study, future clinical trial to evaluate the complementary effects of THL on the relapse free survival of hepatoma patients after surgery is warranted.

Conclusions
We investigated the effects of Chinese herbal mixture Tien-Hsien Liquid (THL) on cancer stem-like side population (SP) in hepatoma Huh7 cells. THL significantly reduced the viability, clonogenicity, and tumorigenicity of Huh7 SP cells. The underlying mechanisms of SP elimination were related to the inhibition on the expression of stemness genes such as ABCG2, CD133, SMO, and Gli. Moreover, combination with THL could synergize the effect of doxorubicin against Huh7 SP cells. Our data indicate that THL may act as a cancer stem cell targeting therapeutics and be regarded as complementary and integrative medicine for the treatment of hepatoma.