SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Hepatocellular carcinoma (HCC) exhibits cellular heterogeneity and embryonic stem-cell–related genes are preferentially overexpressed in a fraction of cancer cells of poorly differentiated tumors. However, it is not known whether or how these cancer cells contribute to tumor initiation and progression. Here, our data showed that increased expression of pluripotency transcription factor Nanog in cancer cells correlates with a worse clinical outcome in HCC. Using the Nanog promoter as a reporter system, we could successfully isolate a small subpopulation of Nanog-positive cells. We demonstrate that Nanog-positive cells exhibited enhanced ability of self-renewal, clonogenicity, and initiation of tumors, which are consistent with crucial hallmarks in the definition of cancer stem cells (CSCs). NanogPos CSCs could differentiate into mature cancer cells in in vitro and in vivo conditions. In addition, we found that NanogPos CSCs exhibited resistance to therapeutic agents (e.g., sorafenib and cisplatin) and have a high capacity for tumor invasion and metastasis. Knock-down expression of Nanog in NanogPos CSCs could decrease self-renewal accompanied with decreased expression of stem-cell–related genes and increased expression of mature hepatocyte-related genes. Overexpression of Nanog in NanogNeg cells could restore self-renewal. Furthermore, we found that insulin-like growth factor (IGF)2 and IGF receptor (IGF1R) were up-regulated in NanogPos CSCs. Knock-down expression of Nanog in NanogPos CSCs inhibited the expression of IGF1R, and overexpression of Nanog in NanogNeg cells increased the expression of IGF1R. A specific inhibitor of IGF1R signaling could significantly inhibit self-renewal and Nanog expression, indicating that IGF1R signaling participated in Nanog-mediated self-renewal. Conclusion: These data indicate that Nanog could be a novel biomarker for CSCs in HCC, and that Nanog could play a crucial role in maintaining the self-renewal of CSCs through the IGF1R-signaling pathway. (HEPATOLOGY 2012;56:1004–1014)

Human hepatocellular carcinoma (HCC), as other solid cancers, exhibits cellular and molecular heterogeneity.1 A subpopulation of cancer cells, often referred to as cancer stem cells (CSCs) or tumor-initiating cells (TICs), are capable of extensive proliferation, self-renewal, and increased frequency of tumor formation. The theory of CSCs proposes a hierarchical organization of cells within the tumor, in which a subpopulation of stem-like cells is responsible for sustaining tumor growth.2, 3 The concept of CSCs has significant clinical implications: CSCs have been shown to be more resistant to chemotherapy and radiotherapy,4 and CSCs possess properties of epithelial-mesenchymal transition (EMT) and are responsible for tumor metastasis.5 Although previous studies have demonstrated that cluster of differentiation (CD)133, CD90, CD13, CD24, and epithelial cell adhesion molecule (EpCAM) were markers for CSCs in HCC, it is still not clear which biomarker represents real CSCs and whether these markers participate in the biology of CSCs in HCC.6-9

Based on global gene-expression profile, several investigators reported that histological poorly differentiated tumors show preferential overexpression of genes normally enriched in embryonic stem cells (ESCs). The ESC-like transcriptional program activated in different human cancers, including HCC, strongly predicts early recurrence, metastasis, and poor survival.10-14 It has been shown that a transcriptional profile for maintaining the self-renewal state of CSCs is more akin to that of ESC than to that of adult stem cells.15, 16 Moreover, key regulators (e.g., octamer-binding transcription factor 4 [Oct4], sex-determining region Y box 2 [Sox2], and Nanog) of ESC identity and their activation targets are more frequently overexpressed in CSCs in different types of cancers.17 Thus, it appears that the key regulators in ESC may also contribute to the pathogenesis of cancers by modulating the self-renewal and differentiation of CSCs. Overexpression of Nanog predicts tumor progression and poor prognosis in colorectal and oral cancers, indicating that Nanog is a key factor regulating human tumor development.18, 19 Transfection of healthy cells with Nanog could induce cell transformation.20 Expression of Nanog was observed in patients with healthy livers and HCC patients,21-23 suggesting that Nanog might be a biomarker of progenitor/stem cells in HCC.21 However, because of a lack of reliable methods, it is still obscure as to what biological role Nanog-expressing cells play in HCC.

In this study, we developed a novel method to isolate Nanog-expressing cancer cells and investigated whether Nanog-expressing cancer cells have characteristics of CSCs. Furthermore, we determined how Nanog cells regulate the self-renewal of CSCs. Our data indicate that Nanog may not only serve as a novel biomarker for CSCs, but may also play a crucial role in maintaining the self-renewal of liver CSCs through the insulin-like growth factor 1 receptor (IGF1R) signal-transduction pathway. Our data may provide new avenues for therapeutic interventions for targeting liver CSCs.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Tissue Samples and Cells.

Fresh tumor specimens were obtained with informed consent from all patients according to protocols approved by the Institutional Review Board of the Southwest Hospital, Third Military Medical University (Chongqing, China). All patients underwent surgical resection of primary HCC at the Institute of Hepatobiliary Surgery, Southwest Hospital, Third Military Medical University. To obtain patient-derived primary HCC cultures of tumor cells, fresh tumor specimens from HCC patients (Supporting Table 1) were cut into small pieces. The single-cell suspension was obtained by mechanical manipulation. The primary culture was established initially in Dulbecco's modified Eagle's medium (DMEM) supplemented with 15% fetal bovine serum (FBS) and was maintained in DMEM supplemented with 10% FBS. Cells were verified for the expression of alpha-fetoprotein (AFP), albumin, and alpha-smooth muscle actin (α-SMA) by immunofloresence (IF) staining. All cells expressed AFP and albumin and did not express α-SMA.

Sphere Formation Assay.

For sphere formation efficiency assay, single NanogPos and NanogNeg cells were sorted and plated into ultra-low attached 96-well plates (Costar, catalog no.: 3474; Corning Inc., Corning, NY). Every well was seeded with 10 cells. Cells were cultured in DMEM/F12 media (Sigma-Aldrich, St. Louis, MO) with B27 supplement (catalog no.: 17504-044; GIBCO, Grand Island, NY), antibiotics, 20 ng/mL of epidermal growth factor (PeproTech Inc., Rocky Hill, NJ), 20 ng/mL of basic fibroblast growth factor (PeproTech), and 10 ng/mL of hepatocyte growth factor (PeproTech). Next, 1% methyl cellulose (Sigma-Aldrich) was added to prevent cell aggregation, and individual sphere derived from a single cell was confirmed. After 4-5 days, equal fresh media was added. Cells were incubated for 2 weeks, and spheres with diameter >75 μm were counted. For serial passaging, spheres were harvested after culturing for 2 weeks and dissociated into single cells with trypsin.

Statistical Analysis.

All data are presented as mean ± standard deviation. When two groups were compared, the Student's t test was used. Analysis of varicance was used for clinical statistical analyses. Kaplan-Meier's method was used for survival analysis. P < 0.05 was considered significant statistically and is marked with an asterisk. P < 0.01 was considered highly significant statistically and is marked with a double asterisk.

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Nanog expression in Cancer Cells Correlates With Poor Prognosis of Patients With HCC.

We first evaluated the expression levels of eight ESC markers (i.e., stage-specific embryonic antigen [SSEA]1, SSEA3, SSEA4, tumor-related antigen [Tra]-1-60, Tra-1-81, Nanog, Oct4, and Sox2) in human HCC cell lines and patient-derived primary HCC cultures of tumor cells. Our data showed that SSEA1 and SSEA3 were not detected in all cells, and SSEA4, Tra-1-60, and Tra-1-81 were only expressed in some cells. However, all HCC cells expressed Nanog, Oct4, and Sox2 (Supporting Fig. 1A). Nanog protein was specifically localized in the nucleus of cancer cells by IF staining (Supporting Fig. 1B). Western blotting analysis showed that Nanog was absent in healthy liver tissues and was higher in HCC tissues than their paired surrounding non-HCC tissues (Supporting Fig. 2; Supporting Table 2). Furthermore, we examined Nanog expression in 12 healthy livers and 59 paraffin-embedded human HCC tissues by immunohistochemistry (IHC). Expression of Nanog could be detected in 49 of 59 cases in HCC tissues (83%), but not in healthy livers (Supporting Table 3). Nanog protein was mainly localized in the cytoplasm of cancer cells. Nuclear accumulation of Nanog was only observed in a small fraction of cancer cells in a few cases (Fig. 1A). No significant association was found between Nanog expression and age, gender, metastasis, liver injury, hepatic fibrosis, and hepatic inflammation (Supporting Tables 4 and 5). Interestingly, we observed that Nanog expression was positively correlated with vascular and capsular invasion (Fig. 1B). Kaplan-Meier's survival analysis showed that the expression of Nanog in HCC significantly correlated with overall and disease-free survival (Fig. 1C). These results indicated that high expression of Nanog correlates with poor prognosis of HCC patients.

thumbnail image

Figure 1. Nanog expression correlates with poor prognosis of patients with HCC. (A) Representative microphotograph of Nanog expression in HCC tissues by IHC analysis. Arrows indicate Nanog expression in the nucleus. (B) Relationship between Nanog expression and pathological analysis. (C) Kaplan-Meier's analysis of overall and disease-free survival of HCC patients respecting to expression level of Nanog protein.

Download figure to PowerPoint

Isolation and Characterization of Nanog-Positive Cancer Cells.

To characterize the behavior of cancer cells expressing Nanog, we isolated Nanog-positive cancer cells with a promoter-reporter strategy. We constructed a lentiviral vector containing human Nanog promoter to drive the expression of the green fluorescent protein (GFP) reporter gene (Lv-PNanog-GFP) (Fig. 2A). After the infection of HCC cells with Lv-PNanog-GFP, we found that GFP was highly expressed in a small percentage of stably infected HCC cell lines and patient-derived primary HCC cells. Fluorescence-activated cell-sorting analysis revealed the percentages of GFP expression ranging from 2.4% to 47.4% (Fig. 2B). Furthermore, GFP-positive and -negative cells were sorted and gene expression was analyzed. NanogHigh and NanogLow cells expressed high and low levels of endogenous Nanog and exogenous GFP (Fig. 2C). NanogNeg cells did not express endogenous Nanog and exogenous GFP. These data indicate that there is a direct correlation between the expression of endogenous Nanog and exogenous GFP. Figure 2D shows that increased levels of Nanog, Oct4, and Sox2 were observed in sorted NanogHigh, as compared with NanogNeg cells. In contrast, there were decreased levels of the mature hepatocyte markers, albumin and glucose-6-phosphatase (G6P), in NanogHigh cells. This result indicated that Nanog-positive cells might have stem-cell–like characteristics.

thumbnail image

Figure 2. Isolation and characterization of NanogPos cancer cells. (A) Schematic of lentiviral vector Lv-PNanog-GFP, in which GFP and Zeocin expression were controlled by human Nanog promoter. (B) GFP expression was measured in HCC cell lines (e.g., Huh7, PLC/PRF/5, and SMMC-7721) and patient-derived primary HCC cells (e.g., T421 and T1216) infected with Lv-PNanog-GFP and was analyzed by flow cytometry. (C and D) GFPHigh, GFPLow, and GFPNeg from Huh7 cells were sorted and subjected to western blotting for the detection of Nanog and GFP expression (C) and to RT-PCR for the detection of stem cell markers (e.g., Nanog, Oct4, and Sox2) and mature hepatocyte markers (e.g., albumin and G6P) (D).

Download figure to PowerPoint

Nanog-Positive Cancer Cells Exhibit Characteristics of CSCs.

Previous studies have demonstrated that cancer stem/initiating cells can form spheroids in vitro in a nonattached culture condition and that CSC can be enriched in this condition.24 NanogPos cells of stably infected Huh7 and PLC/PRF/5 cells remained constant at 4.8% and 3.2%, respectively, after 14 days in conventional culture conditions. Interestingly, approximately 10% cells can form spheres and all spheres exhibited the expression of GFP. The average percentage of NanogPos cells reached 43.0% and 48.5% after 14 days of suspended culture (Supporting Fig. 3A). Furthermore, the high self-renewal capacity of NanogPos cells was confirmed by sphere formation assay. NanogPos cells could form spheres efficiently, but NanogNeg cells rarely formed spheres (Fig. 3A). Interestingly, spheres derived from NanogPos cells were heterogeneous for GFP expression, and spheres derived from NanogNeg cells remained GFP negative (Supporting Fig. 3B). When we recollected and dissociated spheres into single cells and cultured these in suspension conditions, NanogPos cells could form spheres consistently up to the third generation. In vitro passaging of spheres increases the number of sphere-forming cells (Fig. 3A).

thumbnail image

Figure 3. NanogPos cells exhibit characteristics of cancer stem cells. (A) Sphere formation efficiency of sorted NanogPos and NanogNeg cell populations in the first, second, and third generations. (B) Clone formation efficiency of sorted NanogPos and NanogNeg cell populations. (C) Limiting dilution and serial transplantation assay of sorted NanogPos and NanogNeg cells derived from PLC/PRF/5 cells. Sorted NanogPos and NanogNeg cells were injected subcutaneously into SCID mice. (D) Histological and IHC analysis of tumors from NanogPos and NanogNeg cells derived from PLC/PRF/5. Hematoxylin and eosin staining of a subcutaneous tumor (upper) and IHC staining of the tumor with anti-Ki-67 (lower).

Download figure to PowerPoint

Colony formation assay showed that NanogPos cells were able to induce more and larger colonies than NanogNeg cells (Fig. 3B). More than 60% of colonies from NanogPos cells were holoclones, whereas less than 30% of colonies from NanogNeg cell colonies were holoclones (Supporting Fig. 3C). To investigate the tumor-initiating capacity of NanogPos cells, both NanogPos and NanogNeg cells were subcutaneously implanted in severe combined immunodeficiency (SCID) mice. Figure 3C shows that one NanogPos cell could form tumors, whereas tumor formation required 10 NanogNeg cells. Moreover, the weight of tumors derived from NanogPos cells was higher than that derived from NanogNeg cells (Supporting Fig. 3D). IHC for proliferation markers (i.e., Ki-67) showed that tumors derived from NanogPos cells exhibited a higher proliferative rate than NanogNeg cells (Fig. 3D). NanogPos cells derived from established tumors were further sorted and reimplanted into SCID mice. Similarly, 102 NanogPos cells could form tumors, whereas 103 NanogNeg cells could form tumors (Fig. 3C), indicating that NanogPos cells could be passaged serially under in vivo conditions.

Nanog-Positive CSCs Have a High Ability of Invasion and Metastasis and Are Resistant to Chemotherapeutic Agents.

To examine whether Nanog-positive CSCs have a high ability of invasion, we isolated NanogPos and NanogNeg cells from Huh7 and T1216 cells and detected their migratory and invasive abilities. Our data revealed that NanogPos cells displayed more migratory and invasive activity than NanogNeg (Fig. 4A,B). EMT has been shown to endow cancer cells with stem-cell–like characteristics.25 Therefore, we examined the molecular markers of EMT in both NanogPos and NanogNeg cells. As expected, NanogPos cells exhibited the molecular characterization of mesenchyma with decreased expression of E-cadherin and increased expression of vimentin and fibronectin (Fig. 4C and Supporting Fig. 4). These data indicated that the enhanced migratory and invasive capacity of Nanog-positive cancer cells is coincident with EMT phenotype.

thumbnail image

Figure 4. NanogPos cells exhibit the enhanced ability of migration and invasion and are resistant to therapeutic agents. (A and B) Migratory (A) and invasive (B) activities of NanogPos cancer cells were determined by in vitro Boyden chamber assay. (C) Western blotting analysis of EMT markers in NanogPos and NanogNeg cell populations derived from PLC/PRF/5. (D and E) Sorted NanogPos and NanogNeg cells derived from PLC/PRF/5 were treated with 10 and 100 μM of sorafenib (D) or 10 and 50 μg/mL of cisplatin (E) for 48 hours. Percentages of cell survival are shown. (F) Expression levels of drug-resistance relative genes were determined by RT-PCR.

Download figure to PowerPoint

Furthermore, we tested whether NanogPos and NanogNeg cells have different sensitivity to the therapeutic agents, sorafenib and cisplatin. Our data demonstrated that NanogPos cells were more resistant to sorafenib and cisplatin than NanogNeg cells (Fig. 4D,E). Furthermore, multiple drug resistance genes, such as multidrug resistance protein 1 (MDR1), lung resistance protein (LRP), and multidrug resistance-associated protein (MRP), were found to be preferentially expressed in NanogPos cells, as compared with in NanogNeg cells (Fig. 4F). Together, these data indicated that NanogPos cells had a high capacity of tumor invasion and metastasis and were resistant to therapy.

Differentiation of Nanog-positive CSCs Under In Vitro and In Vivo Conditions.

We further evaluated the differentiation potential of NanogPos cells. When NanogPos cells were cultivated in 10% serum-supplemented medium, the proportion of NanogPos cells was reduced to 35.79% at day 7 and to 21.85% at day 14 (Fig. 5A). However, NanogPos cells were not detected in NanogNeg cell-derived progeny, even after 14 days of culture, suggesting that NanogPos cells could differentiate to NanogNeg cells and that NanogNeg cells hardly dedifferentiate to NanogPos cells in conventional culture conditions. Furthermore, we found that there was a decreased expression of stem cell markers (e.g., Nanog, Oct4, and Sox2) and increased mature hepatocyte markers (cytokeratin [CK]7, alpha-antitrypsin, albumin, and tyrosine aminotransferase) after 14 days of culture of NanogPos cells (Fig. 5B). These data indicated that loss of GFP expression in culture of NanogPos cells in conventional conditions was the result of cell differentiation and was accompanied with the down-regulation of stem cell markers. To further investigate differentiation capacity under in vivo conditions, tumors from in vivo implanted NanogPos cells were analyzed for GFP expression. Figure 5C shows that tumors exhibited heterogeneous progeny consisting of NanogPos and NanogNeg cells, with an average of 44.4% of NanogPos cells (range, 16.9%-88.2%) at 2 months after transplantation. These data indicate that differentiation of NanogPos cells also occurs under in vivo conditions.

thumbnail image

Figure 5. Differentiation of NanogPos cells under in vitro and in vivo conditions. (A) In vitro differentiation of NanogPos cells. NanogPos cells were isolated from PLC/PRF/5 by flow cytometry and cultured in conventional conditions. GFP expression was determined by flow cytometry at days 7 and 14. (B) Expression of stem cell and mature hepatocyte markers were measured by real-time PCR before differentiation and 14 days after differentiation. (C) In vivo differentiation of NanogPos cells. Tumors from PLC/PRF/5 NanogPos cells were analyzed for GFP expression after 2 months of transplantation. Percentage of GFP expression analyzed by flow cytometry is indicated in each fluorescence photograph. (D) In vitro single-cell differentiation of NanogPos cells. By IF staining, the progeny from the single NanogPos cell displayed albumin-positive (yellow color) and CK19-positive results (red color).

Download figure to PowerPoint

To avoid the existence of progenitor cells in multicellular differentiation assay, we sorted a single NanogPos cell into a single well of a 96-well plate and cultured this for 2 weeks in conventional culture. The progeny from a single NanogPos cancer cell was heterogeneous for GFP expression and differentially expressed mature hepatocyte markers (e.g., albumin) as well as the cholangiocyte marker, CK19 (Fig. 5D). Single-cell differentiation assay demonstrated NanogPos cancer cells had bidirectional differentiation capacity.

Nanog Regulates Self-Renewal of Liver CSCs.

To figure out whether Nanog participates in the self-renewal of liver CSCs, we performed a Nanog knock-down experiment with a lentivirus-based approach in NanogPos CSCs. Our data showed that both Lv-Nanog-sh1 and Lv-Nanog-sh2 could knock down Nanog expression in NanogPos cells with an inhibitory efficiency of greater than 60% (Supporting Fig. 5A,B). Nanog knockdown could significantly reduce cell proliferation (Fig. 6A), but could not induce cell apoptosis (Supporting Fig. 5C). Furthermore, we found that Nanog knockdown significantly reduced both of the clone formation and sphere formation ability of NanogPos cells (Fig. 6B,C). In addition, the down-regulation of Nanog expression in NanogPos cells could significantly inhibit cell migration (Fig. 6D). Inhibition of Nanog could decrease the expression of GFP (Supporting Fig. 5D). Consistently, upon Nanog knockdown, stemness-associated genes Oct4 and Sox2 were significantly down-regulated, whereas mature hepatocyte markers, such as AFP, CK8, CK18, and transthyretin, were markedly up-regulated (Fig. 6E), indicating that Nanog knockdown resulted in a differentiation of liver CSCs. These results suggested that Nanog is an essential factor for maintaining the self-renewal of CSCs in HCC.

thumbnail image

Figure 6. Knockdown of Nanog expression reduced the self-renewal of NanogPos cells. (A) Cell growth of NanogPos cells from PLC/PRF/5 after being infected with Lv-Nanog-sh1, Lv-Nanog-sh2, and Lv-scramble as the control. (B and C) Inhibition of Nanog decreased sphere (B) and clone formation efficiency (C) of NanogPos cells from PLC/PRF/5. (D) Inhibition of Nanog decreased the migration of NanogPos cells from PLC/PRF/5 and Huh7 cells. (E) Expression of stem cell and mature hepatocyte markers in NanogPos cells from PLC/PRF/5 at 1 week after being infected with Lv-Nanog-sh2.

Download figure to PowerPoint

To further confirm the function of Nanog in maintaining the self-renewal of CSCs in HCC, we restored the expression of Nanog in NanogNeg mature tumor cells by lentiviral vector and investigated their self-renewal capacity. The increased expression of exogenous Nanog was confirmed in NanogNeg cells after infection with Lv-Nanog by real-time reverse-transcriptase polymerase chain reaction (RT-PCR) and western blotting (Fig. 7A). Next, we compared the tumorigenicity of NanogNeg cells with and without Nanog overexpression. After Nanog was up-regulated, the tumor-forming ability of NanogNeg cells was almost recovered to the level of NanogPos cells control (Fig. 7B,C). In addition, the migration ability in NanogNeg cells was increased after the overexpression of Nanog (Fig. 7D). Taken together, these results suggested that Nanog plays an important role in the regulation of self-renewal and the tumorigenicity of liver CSCs.

thumbnail image

Figure 7. Overexpression of exogenous Nanog promotes the self-renewal of NanogNeg cells. (A) Expression of Nanog was determined by RT-PCR and western blotting in NanogNeg cells from PLC/PRF/5 after infection with Lv-Nanog and Lv-PEF1a-DsRed2 as the control. (B and C) Efficiency of tumor formation (B) and tumor weight (C) of NanogNeg cells from PLC/PRF/5 after infection with Lv-Nanog or control lentivector (Lv-PEF1a-DsRed2). NanogPos cells from PLC/PRF/5 served as the positive control. (D) Overexpression of Nanog enhanced the ability of migration in the NanogNeg from PLC/PRF/5.

Download figure to PowerPoint

IGF1R Signaling Participates in Nanog-Mediated Function of CSCs.

To investigate the molecular mechanisms involved in Nanog-mediated self-renewal activity, we performed global gene-expression profiles in purified NanogPos and NanogNeg cells by complementary DNA microarray. The gene-array analysis showed that insulin-like growth factor (IGF)2, IGF2 antisense (IGF2AS), IGF/binding protein (BP)2, and IGF-BP5 were increased in NanogPos cells (Fig. 8A). Therefore, we further investigated whether the IGF pathway is involved in Nanog-mediated self-renewal activity in liver CSCs.

thumbnail image

Figure 8. IGF1R signaling participates in Nanog-mediated self-renewal. (A) Microarray analysis of gene expression in NanogPos and NanogNeg cells from Huh7, T1115, and T1224 cells. Red and green colors depict high and low expression levels, respectively, as indicated by the scale bar. (B) Expression levels of IGF2, IGF1R, IGF2R, and Nanog were determined by RT-PCR in NanogPos and NanogNeg cells from Huh7 and T1224 cells. (C) Expression of IGF1R was determined by western blotting in NanogPos cells from Huh7 cells at 1 week after infection with Lv-Nanog-sh1 and Lv-Nanog-sh2 and Lv-scramble as the control and NanogNeg cells after infection with Lv-Nanog and Lv-PEF1a-DsRed2 as the control. Quantification of IGF1R expression is shown. (D) Expression of Nanog and IGF1R were measured by RT-PCR in NanogPos cells from Huh7 cells at 96 hours after treatment with PPP at a concentration of 0.01 μM. (E) Sphere formation efficiency of NanogPos cells from Huh7 and T1224 cells after treatment with PPP at concentrations of 0.01 and 0.1 μM.

Download figure to PowerPoint

First, we studied IGF-pathway–related gene expression in NanogPos and NanogNeg cells. There was a significant increased expression of IGF2 and IGF1R in NanogPos cells, as compared with in NanogNeg cells (Fig. 8B). To know whether IGF1R was regulated by Nanog, we determined IGF1R expression in NanogPos and NanogNeg cells after the inhibition or overexpresssion of Nanog, respectively. Our data showed that Nanog knockdown in NanogPos cells resulted in the decreased expression of IGF1R and that the overexpression of Nanog in NanogNeg cells resulted in the increased expression of IGF1R, indicating that Nanog could regulate IGF1R expression (Fig. 8C). To further demonstrate the biological function of IGF1R in the self-renewal activity of NanogPos cells, we used IGF1R-specific inhibitor picropodophyllin (PPP) and AEW541 to block the function of the IGF1R pathway. After an evaluation of the dose-dependent effect of PPP and AEW541 on cell survival (Supporting Fig. 6), we chose two doses (0.01 and 0.1 μM) of PPP and one dose (2 μM) of AEW541, which did not affect cell survival. Both PPP and AEW541 could significantly inhibit sphere formation (Fig. 8D and Supporting Fig. 6). Furthermore, we observed that inhibition of IGF1R by PPP could significantly inhibit Nanog expression (Fig. 8E). Taken together, it suggests a feedback loop between the function of Nanog and the IGF1R signal-transduction pathway.

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Nanog expression has been observed not only in embryonic carcinoma and germ cell tumors, but also in other types of tumors, including HCC.21-23 Elevated expression of Nanog was positively associated with late-stage progression and poor prognosis of patients with colon and oral cancers.12, 18 Our data showed that a majority of HCC patients exhibited Nanog expression, particularly in the cases of poor differentiation associated with vascular and capsular invasion and poor survival. Of note, increased expression of Nanog in HCC correlates with a worse clinical outcome, raising the possibility of its utility as a prognostic biomarker. Recent studies have shown that increased Nanog expression was observed in enriched cancer stem cells, and that Nanog plays a critical role in tumor growth and stem cell behavior through the functional glioma-associated oncogene homolog 1/Nanog-p53 network in glioblastoma multiforme.23, 26, 27 Thus, the increased Nanog expression in a fraction of tumor cells raises the possibility that these cells might represent tumor stem/initiation cells in HCC.

To demonstrate this hypothesis, we isolated Nanog-positive cancer cells from stably transduced cancer cells with a lentiviral vector construct, in which GFP expression is controlled by the human Nanog promoter. We demonstrated that exogenous GFP expression is closely correlated with endogenous Nanog expression. The strategy for the isolation of live cells with intacellular markers by lentivector reporters has been successfully applied to isolate mouse and human induced pluripotent stem cells and human CSCs.28, 29 Another advantage of this strategy is that it can monitor the residual undifferentiated pluripotent stem cells, because lentivector-mediated reporter expression is extinguished during differentiation.28 Furthermore, our results show that Nanog-positive cancer cells express stem cell markers, have a high ability of self-renewal and clonogenicity, and have the capacity for differentiation. Moreover, they exhibit an extraordinarily strong ability to initiate tumors. These characteristics are consistent with crucial hallmarks in the definition of CSCs,2, 3 indicating that they represent genuine CSCs in HCC.

We further demonstrated that Nanog-positive CSCs express traditional EMT markers and have a high capacity of invasion and metastasis. In addition, they are resistant to therapeutic agents (e.g., sorafenib and cisplatin). The existence of this subpopulation of cells in HCC may explain that HCC is an aggressive malignancy with vascular and capsular invasion and is resistant to chemotherapy and radiotherapy with poor survival.30, 31

Because Nanog is an important transcription factor, it plays an important role in maintaining stemness in ESC. To know whether Nanog participates in the regulation of liver CSCs, we performed a Nanog knock-down and overexpression experiment to address this issue. Our data showed that the knockdown of Nanog expression in NanogPos CSCs could abolish their self-renewal ability and that overexpression of Nanog in NanogNeg mature tumor cells could restore their self-renewal ability. A recent study has demonstrated that CD24 regulated the capacities of self-renewal and tumor initiation in liver TICs through signal transducer and activator of transcription 3 (STAT3)–mediated Nanog expression.23 Taken together, both these data strongly suggested that Nanog is actively involved in maintaining the self-renewal of liver CSCs.

To further demonstrate the molecular mechanisms involved in Nanog-mediated self-renewal activity, we performed global gene-expression profiles in purified NanogPos and NanogNeg cells from HCC cell lines and patient-derived primary cells. Our data showed that the IGF pathway is one of the most altered gene expressions between NanogPos and NanogNeg cells. Furthermore, we found that the knockdown of Nanog expression in NanogPos decreased the expression of IGF1R and that overexpression of Nanog in NanogNeg cells increased the expression of IGF1R, indicating that Nanog could modulate IGF1R expression. IGF1R-specific inhibitor PPP and AEW541 could significantly deprive the self-renewal ability of NanogPos CSCs. These data are coincident with previous observations that IGF signaling is involved in the biology of CSCs in glioma and colorectal cancer.32, 33 It has been demonstrated that high-level expression of IGF2 is an important autocrine protumorigenic event in HCC and that the activation of IGF2/IGF1R signaling is likely a progression switch by the function that promotes tumor cell dissemination and aggressive tumor behavior.34, 35 Therefore, the high level of IGF2 produced by NanogPos cells may provide autocrine regulation of liver CSCs by acting with IGF1R. Interestingly, we also observed that the inhibition of IGF1R by PPP could significantly inhibit Nanog expression. These data suggest a feedback loop between the function of Nanog and the IGF1R signal-transduction pathway in liver CSCs. Previous studies have demonstrated that the activation of IGF1R plays a critical role in the self-renewal and pluripotency of ESC36 and could enhance the expression of Nanog in ESCs through cross-talk with the Wnt pathway and the phosphorylation of p53.37 Phosphorylated p53 inhibited the translocation of p53 into the nuclei to bind the promoter of Nanog as a negative regulator.37 However, how Nanog controls IGF1R expression is not known. Therefore, the detail mechanisms of the relationship between the function of Nanog and the activation of IGF1R in liver CSCs are still needed for further investigation.

Taken together, we demonstrated that the transcription factor, Nanog, as a prognostic marker for HCC. Using Nanog as a biomarker, we isolated a small subpopulation of NanogPos cells and demonstrated that these cells represent genuine CSCs. Furthermore, our data established the integral role for Nanog in maintaining the self-renewal of liver CSCs through the activation of the IGF-signaling pathway.

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
HEP_25745_sm_SuppInfo.doc281KSupporting Information
HEP_25745_sm_SuppFig1.tif4874KSupporting Information Figure 1
HEP_25745_sm_SuppFig2.tif1202KSupporting Information Figure 2
HEP_25745_sm_SuppFig3.tif7943KSupporting Information Figure 3
HEP_25745_sm_SuppFig4.tif2655KSupporting Information Figure 4
HEP_25745_sm_SuppFig5.tif1693KSupporting Information Figure 5
HEP_25745_sm_SuppFig6.tif1510KSupporting Information Figure 6

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.