SEARCH

SEARCH BY CITATION

Keywords:

  • cancer stem cells;
  • chemoresistance;
  • hepatocellular carcinoma;
  • liver cancer;
  • miRNA;
  • self-renewal;
  • tumor-initiating cells

Abstract

  1. Top of page
  2. Abstract
  3. The cancer stem cell theory
  4. Hepatocellular carcinoma
  5. Identification of liver cancer stem cells
  6. Regulation of liver cancer stem cells
  7. Therapeutic implications and future challenges
  8. Acknowledgments
  9. References

Hepatocellular carcinoma (HCC) is the most commonly diagnosed malignancy of the liver and is the third most frequent cause of cancer death worldwide. Although advances in HCC detection and treatment have increased the likelihood of a cure at early stages of the disease, HCC remains largely incurable because of late presentation and tumor recurrence. Only 25% of HCC patients are deemed suitable for curative treatment, with the overall survival at just a few months for inoperable patients. Additionally, this disease is particularly difficult to treat because of the high recurrence rate, its chemotherapy-resistant nature and the premalignant nature of surrounding cirrhotic liver disease. In the past few years, compelling evidence has emerged in support of the hierarchic cancer stem cell (CSC)/tumor-initiating cell (T-IC) model for solid tumors, including HCC. Understanding the characteristics and function of CSCs in the liver has also shed light on HCC management and treatment, including the implications for prognosis, prediction and treatment resistance. In this review, a detailed summary of the recent progress in liver CSC research with regard to identification, regulation and therapeutic implications will be discussed.


The cancer stem cell theory

  1. Top of page
  2. Abstract
  3. The cancer stem cell theory
  4. Hepatocellular carcinoma
  5. Identification of liver cancer stem cells
  6. Regulation of liver cancer stem cells
  7. Therapeutic implications and future challenges
  8. Acknowledgments
  9. References

There are currently two conflicting views that attempt to explain tumor formation. The classical stochastic model, also referred to as the clonal evolution model,1 proposes that a single cell acquires random and additive genetic mutations, with subsequent clonal selection, to result in the formation of a group of clonal tumor cells. Every cell within the tumor is biologically homogeneous and has an equal potential to generate a tumor. The likelihood of each of these cells becoming a tumor-initiator is governed by a low probability of stochastic mutations. In contrast, the cancer stem cell (CSC) or tumor-initiating cell (T-IC) theory suggests that a tumor comprises a heterogeneous population of cells that form a distinct cellular hierarchy; only a subset of cells within this tumor hierarchy has the ability to self-renew, differentiate into defined progenies and, most importantly, initiate and sustain tumor growth.2 Contrary to the stochastic model, each of the small subset of CSCs in the tumor has a significantly higher probability to become a tumor-founding cell, relative to the non-CSCs that make up the bulk of the tumor. According to this theory, it should be possible to identify and target the cells responsible for tumor initiation and progression because not all of the cells have the same phenotypic and functional characteristics.

Although the idea of CSCs has been around for many years, the work by John Dick and colleagues over a decade ago was the first to demonstrate the critical role of stem cells in hematological malignancies3 and has, as a result, revolutionized the widely held belief of the clonal origin of carcinogenesis. Since then, substantial evidence has emerged to support the notions of tumor heterogeneity and cellular hierarchy within a tumor, not only in the field of hematological cancers but also in solid cancers. Indeed, several pivotal studies have recently provided convincing evidence that these cells do exist in solid tumors of many types, including breast, brain, colorectal, pancreas, liver, melanoma and prostate cancers.4 Studies have now demonstrated that CSCs exhibit many classical properties of both normal stem cells and cancer cells, including the following: (i) a high self-renewal capacity; (ii) an enhanced ability to differentiate and generate heterogeneous lineages; (iii) an increased capacity for self-protection against drugs, toxins and radiation; and (iv) an increased capacity to initiate and sustain tumor growth (Table 1).

Table 1.  Common features of cancer stem cells
• A high self-renewal capacity;
• An enhanced ability to differentiate and generate heterogeneous lineages;
• An increased capacity for self-protection against drugs, toxins and radiation; and
• An increased capacity to initiate and sustain tumor growth.

Despite their resemblance to stem cells, CSCs do not necessarily have to originate from normal stem cells.5,6 Yet, to date, the exact origin of CSCs has not been fully uncovered. The questions of whether a defined subset of cells is destined to become CSCs when they are formed, whether CSCs are transformed from normal stem cells, or whether they originate from more differentiated cells at a lower level of cellular hierarchy (i.e. progenitor cells and mature cells that acquire self-renewal and tumor initiation abilities after genetic lesions) remain to be elucidated for a better understanding of the origin and role of this special subpopulation of cells (Fig. 1).

image

Figure 1. The liver cancer initiation hypothesis based on the cancer stem cell model. Hepatic stem cells or oval progenitor cells have capacities to self-renew and differentiate. Mutations to acquire a tumorigenic potential will convert these normal liver stem/progenitor cells into liver cancer stem cells (CSCs), ultimately giving rise to hepatocellular carcinoma (HCC) and possibly cholangiocarcinoma (CC). Conversely, liver CSCs can also originate from more differentiated liver cells that already have tumorigenic potential but undergo mutations to acquire the stem cell-like capacity to self-renew and differentiate.

Download figure to PowerPoint

Hepatocellular carcinoma

  1. Top of page
  2. Abstract
  3. The cancer stem cell theory
  4. Hepatocellular carcinoma
  5. Identification of liver cancer stem cells
  6. Regulation of liver cancer stem cells
  7. Therapeutic implications and future challenges
  8. Acknowledgments
  9. References

Hepatocellular carcinoma (HCC) accounts for 80–90% of all primary liver cancers. The disease ranks as the fifth most common cancer worldwide and is the third leading cause of all cancer-associated deaths. The prevalence of HCC differs greatly by geographical locations. Eastern countries, such as those of the Asia-Pacific region, and sub-Saharan African regions, have a significantly higher incidence rate than Western countries, such as those of Europe and the United States, although occurrences in the latter have also risen rapidly in recent decades. Chronic hepatitis viral infection (hepatitis B virus [HBV] and hepatitis C virus [HCV]), liver cirrhosis caused by excess alcohol consumption, diabetes and aflatoxin are the major risk factors for HCC development.7

Although advances in HCC detection and treatment have increased the likelihood of a cure at early stages of the disease, HCC remains largely incurable because of the late presentation and tumor recurrence. Only 25% of HCC patients are deemed suitable for curative treatment, with the overall survival at just a few months for inoperable patients. Apart from surgical resection, loco-regional ablation and liver transplantation,8 current treatment protocols include conventional cytotoxic chemotherapy. But due to the highly resistant nature of the disease, the efficacy of the latter regimen is limited. In fact, the emergence of CSC theory lends insight into the explanation of why treatment with chemotherapy often may seem to be initially successful, but eventually results in failure to eradicate the tumor and possibly also in tumor relapse. Commonly used anti-cancer drugs in HCC, such as cisplatin, doxorubicin and 5-fluorouracil, work by targeting the rapidly proliferating and differentiated liver cancer cells that constitute the bulk of the tumor. However, within the tumor a subpopulation of CSCs exists that are more resistant and therefore able to survive and maintain residence after treatment. Thus, such clones of drug-resistant tumor cells grow and self-renew to generate the development and spread of recurrent HCC tumors.

Identification of liver cancer stem cells

  1. Top of page
  2. Abstract
  3. The cancer stem cell theory
  4. Hepatocellular carcinoma
  5. Identification of liver cancer stem cells
  6. Regulation of liver cancer stem cells
  7. Therapeutic implications and future challenges
  8. Acknowledgments
  9. References

Side population (Hoechst 33342 dye staining)

In an attempt to isolate hematopoietic stem cells, Hoechst 33342 dye was initially used for side population (SP) cell sorting in mouse bone marrow cells.9 Stem cells with high levels of expression of adenosine triphosphate (ATP)-binding cassette (ABC) transporters possess the ability to efflux xenobiotic substances, resulting in a low Hoechst staining profile in the SP population. In 2006, Chiba and colleagues extended the application of SP cell sorting to identify cancer stem-like cells in HCC.10 SP cells were detected in two HCC cell lines (Huh7 and PLC/PRF/5 cells) as a minute population comprising less than 1% of the total. The sorted SP cells, when compared with non-SP counterparts, were characterized by a higher proliferative potential, anti-apoptotic property, upregulated expression in “stemness” genes and higher tumorigenic potential. As few as 1 × 103 SP cells were adequate for tumor formation in NOD/SCID mice, and tumorigenicity could still be maintained in serial transplantation; in contrast, as many as 1 × 106 non-SP cells failed to initiate tumor formation. Criticism has challenged the use of SP cell sorting as a method to define liver CSCs as non-SP cells. In particular, transporter protein-expressing cells are likely to suffer from the toxicity of Hoechst 33342 dye and cannot grow normally, resulting in the apparent differential properties observed in these functional experiments.11

CD133

After the early attempt using SP cell sorting, significant efforts have been made to further characterize and delineate CSCs in HCC. Attention has been drawn to CD133 as an important liver CSC marker in the past six years; CD133+ HCC cells were first suggested to represent a potential CSC subpopulation by Suetsugu and colleagues. These authors found that a sorted CD133+ subpopulation from a Huh7 cell line possessed higher proliferative and tumorigenic potential, and expressed lower levels of mature hepatocyte markers, such as glutamine synthetase and cytochrome P450 3A4, when compared with their CD133- counterparts.12 Similar findings were obtained by another group of researchers who isolated a CD133+ fraction from a SMMC-7721 cell line, and this population of cells demonstrated an enhanced clonogenicity in vitro and tumorigenicity in vivo.13

Our research group has also pioneered work on the identification and characterization of liver CSCs using the CD133 surface phenotype. It is generally believed that normal stem cells and CSCs share similar properties that regulate both self-renewal and differentiation processes. A severe partial hepatectomy model was employed to study the role of normal stem cells during liver regeneration in the hope of finding clues that may assist understanding of mechanisms that regulate self-renewal and differentiation in CSCs. With over 70% of the mouse liver removed, the expression of prominin-1, the homolog of human CD133 in mice, was found to be significantly upregulated during early liver restoration.14 CD133 was also found to represent a small subpopulation in human tumor tissue, and it was absent in normal liver tissue.

The subsequent analysis of CD133 expression in human liver cell lines revealed a positive correlation between CD133 expression and tumorigenic potential in vivo. Sorted CD133+ and CD133- fractions from HCC cell lines (i.e. PLC8024, Huh7 and HepG2) were then subjected to functional analyses in vitro and xenograft transplantation in vivo to study the exhibition of properties representative of both stem cells and cancer cells. The CD133+ cells were found to be more tumorigenic than the CD133- cells, as evidenced by a greater colony-forming ability, higher proliferative potential and the ability to initiate tumor formation. Moreover, the CD133+ cells were characterized by properties of normal stem/progenitor cells, including the increased expression of “stemness”-associated genes and the abilities to self-renew and differentiate into non-hepatocyte lineages.14 Recently, our studies have been extended with the use of HCC clinical specimens, and a similar phenomenon has been observed.15 The clinical significance of CD133 in HCC was also similarly reported by Song and colleagues.16

Aldehyde dehydrogenase activity

Aldehyde dehydrogenase (ALDH), a molecular metabolic mediator, was first identified as conferring resistance to cyclophosphamide in normal hematopoietic stem/progenitor cells.17 Recent studies have suggested that high ALDH activity can confer chemoresistance in CSCs.18–20 In colon cancer, higher ALDH activity has been observed in EpCAMhigh/CD44+ colon CSCs.21 ALDH was also found to be able to predict a poor clinical outcome in CSC-driven breast cancer patients.19 In the subsequent analysis of the characterization of liver CSCs marked by a CD133 phenotype, our group identified ALDH to be preferentially expressed in the CD133+ population in HCC, and the use of a combination of these markers was shown to more accurately define liver CSCs.20 A hierarchical organization of cells that differentially express CD133 and ALDH exhibit an ascending tumorigenic potential in the order of CD133+ALDH+ > CD133+ALDH- > CD133-ALDH-.20

CD90

In the following year, another CD surface protein was used for the identification of liver CSCs. Yang and colleagues found a significant positive correlation of CD90 expression with tumorigenicity and metastatic potentials in the panel of liver cell lines tested.22,23 In the clinical specimens, all of the tumor tissues and almost all of the blood samples contained a CD45-CD90+ subpopulation. The CD45-CD90+ cells isolated from both the tumor tissues and blood samples was shown to initiate and maintain tumor formation when injected intrahepatically into SCID/Beige mice in the first and the subsequent serial transplantation experiments.22,23 The existence of a CD45-CD90+ population in blood samples from HCC patients suggests the presence of CSCs in the systemic circulation.

CD44

The use of multiple stem cell markers in the characterization of CSCs from solid tumors was first reported in breast cancer: CD44+CD24−/low.24 In HCC, CD44 is also an important marker used in combination with other CSC markers to better define the surface phenotype of liver CSCs. For the aforementioned markers, including CD133 and CD90, cells co-expressing either of these markers and CD44 present a more aggressive phenotype than cells with a positive expression of either CD133 or CD90 alone. Yang et al. showed that CD44+ cells developed tumor nodules in immunodeficient mice faster than CD44- cells, whereas lung metastases were only observed in immunodeficient mice transplanted with CD90+CD44+ cells.23 In another study, CD44 was found to be preferentially expressed in a CD133+ population in four HCC cell lines, including Huh7, SMMC-7721, MHCC-LM3 and MHCC-97L. CD133+CD44+ cells exhibit enhanced abilities to form tumors, are chemoresistant and express a higher level of “stemness”-associated genes, as compared with their CD133+CD44- counterparts.25

EpCAM

EpCAM is present in the embryonic liver, bile duct epithelium and proliferating bile ductules in the cirrhotic liver, but it is absent in normal adult hepatocytes.26 An elevated expression of EpCAM was first identified in premalignant hepatic tissues; therefore, this surface protein was suggested to be an early biomarker for HCC.27 Following a cDNA microarray analysis on a clinical cohort of primary HCC tissue, EpCAM+ HCC was linked with the gene signature and the molecular pathway of hepatic progenitor cells, whereas genes expressed in EpCAM- HCC cells were associated with mature hepatocyte functions.26 These two subtypes of HCC were further stratified into four groups with features resembling different hepatic lineages, and they showed prognostic differences based on the expression of alpha fetoprotein (AFP).26,28 EpCAM+ HCC cells have also been shown to be highly invasive and tumorigenic, in comparison with their EpCAM- counterparts.28

CD13

Recently, a novel cell surface marker, CD13, was identified for potentially dormant and semi-quiescent CSCs in HCC. Haraguchi et al. found that CD13 was commonly enriched in an SP population sorted from Huh7, PLC/PRF/5 and Hep3B cells by gene expression microarray analysis. CD13 was selected as a putative marker to enrich the semi-quiescent liver CSCs because of its predominant distribution during the G1/G0 phase.29 This result suggested that CD13+ cells represent the dormant or slow-growing population that is believed to account for the chemoresistant capacity in HCC.

Researchers then assessed the tumorigenic potential of CD13 with two other liver CSC markers, CD133 and CD90. The results clearly showed that the CD13+CD133+ and CD13+CD90- fractions in Huh7 and PLC/PRF/5 cell lines, respectively, initiated tumor formation effectively in limiting-dilution and serial transplantation assays. CD13+ cells have also been shown to be highly chemoresistant to doxorubicin and 5-fluorouracil treatment, and the CD13+ population was found to be enriched after chemotherapy. This observation was explained by low levels of intracellular ROS in the CD13+ fraction that protected the cells from DNA damage and induced apoptosis via a ROS scavenger pathway. It is noteworthy that the enrichment of the CD13+ population near the fibrous capsule after treatment is compatible with the fact that tumor relapse usually takes place near that region, which could be a potential protective niche for the maintenance of CSCs.29

Regulation of liver cancer stem cells

  1. Top of page
  2. Abstract
  3. The cancer stem cell theory
  4. Hepatocellular carcinoma
  5. Identification of liver cancer stem cells
  6. Regulation of liver cancer stem cells
  7. Therapeutic implications and future challenges
  8. Acknowledgments
  9. References

Pathways mediating therapeutic resistance

Most cytotoxic therapies used for cancer therapy disrupt mitosis or damage DNA to induce cell death in highly proliferative tumor cells. If tumor growth is driven by CSCs, this can explain why current therapies that have been developed largely against the rapidly dividing bulk of tumor cells are only transiently, if at all, able to shrink the primary tumor but are unable to provide a lasting cure for the disease. The chemo- and radio-resistant nature of these residual CSCs could partially explain tumor relapse in advanced or aggressive tumors. Indeed, there have been several studies implicating CSCs as being particularly resistant to conventional chemo- and radiation therapies in a variety of different cancers. Specifically for HCC, there are currently four original articles that have documented the molecular mechanism by which CD133+, EpCAM+ or CD90+ liver CSCs mediate chemoresistance. Ultimately, the chemoresistance displayed by these CSCs as a result of metabolic alterations or increased drug efflux, such as the expression of aldehyde dehydrogenase 1 (ALDH1) or ABC (ATP-Binding Cassette) transporters, highlights the need for the development of CSC chemotherapy-sensitization techniques and compounds that will allow these resistant populations to be eradicated to prevent a recurrence of disease.

AKT/PKB and Bcl-2 dysregulation in CD133+ liver CSCs

Following our identification of a liver CSC population marked by a CD133 surface phenotype in 2007, we extended these studies to examine both the sensitivity of these cells to the chemotherapeutic agents, doxorubicin and 5-fluorouracil, and the possible mechanistic pathway by which resistance may be regulated. Sorted CD133+ cells from HCC cell lines and a xenograft mouse model survived chemotherapy in increased proportions relative to differentiated CD133- counterparts through a dysregulated AKT/PKB and Bcl-2 pathway. CD133+ liver CSCs showed a significantly elevated expression level of key players in the pathway, including phospho-AKT (serine 473), phospho-Bad (serine 136) and Bcl-2. When cultured in the presence of the two drugs, the expression of each of these proteins persisted at higher concentrations and for a longer period time when exposed to a fixed concentration of the drug in CD133+ liver CSCs. Interestingly, the survival proteins, Bcl-2 and phospho-AKT (serine 473), were found to co-localize with CD133, as demonstrated by dual-color immunofluorescence. To further confirm the importance of the AKT/PKB pathway in conferring chemoresistance in CD133+ liver CSCs, we observed that treatment of the cells with an AKT1 inhibitor significantly reduced the expression of survival proteins in the pathway. Further, treatment of CD133+ liver CSCs with the AKT1 inhibitor, along with doxorubicin or 5-fluorouracil, led to the complete inhibition of the preferential survival effect induced by CD133+ liver CSCs.30

Oncostatin M-induced differentiation in EpCAM+ liver CSCs

Recently, Yamashita and colleagues have also identified a mechanism by which EpCAM+ liver CSCs are rendered sensitive to 5-fluorouracil chemotherapy. The receptor of oncostatin M (OSM), an interleukin 6-related cytokine that is known to induce the differentiation of hepatoblasts into hepatocytes, was detected in the majority of EpCAM+ liver CSCs. Based on this finding, the authors investigated the effect of OSM on EpCAM+ liver CSCs. The treatment of these cells with OSM enhanced the hepatocytic differentiation of EpCAM+ liver CSCs by inducing Stat3 activation, as determined by a decrease in the stem-cell related gene expression and a decrease in EpCAM, alpha fetoprotein and cytokeratin 19 protein expression, which was concomitant with an increase in albumin protein expression. Further, OSM-treated EpCAM+ liver CSCs showed enhanced cell proliferation with an expansion of the EpCAM- non-CSC population. The combination of OSM treatment with 5-fluorouracil, which eradicated the EpCAM- non-CSCs, dramatically increased the number of apoptotic cells in vitro and suppressed tumor growth in vivo, when compared with either OSM or 5-fluorouracil treatment alone. Findings from the study suggest that OSM could be effectively used for the differentiation and active cell division of OSM receptor-positive EpCAM+ liver CSCs and that the combination of OSM and 5-fluorouracil can efficiently eliminate HCC by targeting both CSCs and non-CSCs subpopulations.31

Oct4-AKT-ABCG2 dysregulation

Using chemotherapeutic drugs to select drug-resistant cancer cells in HCC, Wang et al. have demonstrated that chemoresistant cells display CSC-like features, including increased self-renewal ability, increased cell motility, resistance to multiple chemotherapeutic drugs, enhanced tumorigenic potential and elevated expression of CD90+ cells. In addition, the expression of Oct4, a transcription factor essential in embryonic stem cells, was also strongly upregulated in the chemoresistant HCC cell subpopulation. The authors demonstrated that Oct4 plays a role in cancer cell chemoresistance through the following findings: (i) chemoresistant cancer cells displayed an enhanced expression of Oct4 through gene demethylation processes; (ii) the overexpression of Oct4 significantly increased, whereas the knockdown of Oct4 reduced the drug resistance of liver cancer cells in vitro and in vivo; and (iii) the overexpression of Oct4 induced the activation of TCL1, AKT and ABCG2 to mediate the chemoresistance. The importance of the Oct4-TCL1-AKT-ABCG2 pathway in mediating the observed chemoresistance was further validated when the addition of a PI3K/AKT inhibitor abolished this effect.32

ABCB5 dysregulation

Cheung et al. have found that the growth factor, granulin-epithelin precursor (GEP), regulated chemoresistance in liver cancer cells through modulation of the expression of the ABCB5 drug transporter. Specifically, chemoresistant HCC cells that expressed GEP had increased levels of ABCB5, whereas suppression of ABCB5 sensitized the cells to doxorubicin treatment and apoptosis. Most interestingly, HCC cells that expressed GEP and ABCB5 were also found to co-express the liver CSC markers, CD133 and EpCAM. Conversely, blocking ABCB5 reduced the expression of CD133 and EpCAM. The expression levels of GEP and ABCB5 were increased in liver cancer cells, as compared with non-tumor liver tissue from patients with cirrhosis or hepatitis, or normal liver tissue. ABCB5 expression was also associated with a higher recurrence rate in patients with HCC who had undergone curative partial hepatectomy.33

Dysregulated self-renewal pathways in liver CSCs

The maintenance of CSCs involves regulatory pathways that are known to be involved in stem cell maintenance and self-renewal and pluripotency, which include Bmi-1, Wnt/β-catenin, transforming growth factor-β (TGF-β), Notch and Sonic hedgehog. Thus, new therapeutic strategies targeting signaling pathways that are involved in the self-renewal of CSCs and which also block differentiated cancer cells have been suggested. In HCC, the disruption of a number of these pathways has also been implicated in liver CSCs.

Bmi-1

Bmi-1 belongs to a family of polycomb group (PcG) proteins that are highly conserved throughout evolution and are known to be vital transcriptional repressors, contributing to epigenetic chromatin modifications during stem cell self-renewal programs and tumor development. The forced expression of Bmi-1 was shown to promote the self-renewal of hepatic stem/progenitor cells and contribute to malignant transformation,34 and the aberrant upregulation of Bmi-1 was found to play a particularly important role in liver CSCs identified by CD133+ and CD90+ expression.14,15,22,23 Chiba et al. performed a more detailed study on the critical role of Bmi-1 in the maintenance of CSCs with the SP phenotype in HCC cell lines. The knockdown of Bmi-1 completely abolished the self-renewal and tumorigenic potential of SP cells.35 Results from the same study indicated that Bmi-1 expression was also tightly correlated with the CSC phenotype represented by CD133+ HCC cells because altering Bmi-1 expression resulted in a similar change in the maintenance of a CD133 subpopulation in liver cancer cells.35

Wnt/β-catenin signaling

The Wnt/β-catenin signaling pathway plays a critical role in the proliferation, self-renewal and differentiation of stem cells in many tissue types. Disruption of WNT signaling results from both genetic and epigenetic changes and is associated with a wide range of cancer types, especially colon cancer and liver cancer.36 Indeed, the activation of WNT signaling has been demonstrated in different prospectively isolated liver CSC subpopulations, and the elevated expression of WNT and its downstream mediators have been reported in CD133+ and EpCAM+ liver CSCs.14,15,26,37 EpCAM has been shown to be one of the direct transcriptional targets of Wnt/β-catenin signaling in HCC, and the RNAi-mediated knockdown of EpCAM has resulted in a decrease in the self-renewal, tumorigenicity, migration and drug resistance of HCC cells.28

Transforming growth factor-β

The TGF-β family plays a vital role in the control of proliferation and cellular differentiation in both stem cells and cancer cells. Mishra and colleagues have shown that impaired TGF-β signaling by the activation of interleukin-6 (IL-6) in hepatic stem/progenitor cells can contribute to altered differentiation patterns and thus, HCC development.38 Similar results were found in EpCAM+ liver CSCs because the targeting of the pathway using the indirect modulation of IL-6/STAT3 was found to be effective for the eradication of EpCAM+ liver CSCs.31 In addition, Rountree and colleagues also provided evidence to show that TGF-β can regulate the expression of CD133+ liver CSCs through an inhibition of the expression of DNA methyltransferases, DNMT1 and DNMT3beta, and the subsequent demethylation of the CD133 promoter.39

miRNA regulation

miRNAs are a class of small, non-coding RNAs that function as important regulatory molecules by negatively regulating gene and protein expression at the post-transcriptional level. miRNAs have been implicated in the control of a wide variety of cellular processes, including differentiation and pluripotency. The aberrant expression of this class of molecules has also been found to contribute considerably to cancer development and the progression and generation of metastasis, whereas its expression has also been correlated with the stage of the tumor and the prognosis for cancer patients.40,41 Recently, there has been increasing evidence in support of a role of miRNA in the regulation of CSC,42 suggesting the possibility of an miR-directed therapy to correct CSC dysregulation (miR-directed CSC eradication). Currently, there are two studies that document the role of miRNAs in the regulation of EpCAM+ and CD133+ liver CSCs. More results are expected from this emerging and exciting field in CSC research.

miR-181 in EpCAM+ in liver CSCs

Evidence of miRNA regulation in liver CSCs was first demonstrated by Wang and colleagues; in their study, they found the expression of miR-181 family members (i.e. miR-181a, miR-181b, miR-181c and miR-181d) to be significantly elevated in EpCAM+AFP+ liver CSCs. The forced expression of the miR-181 family members led to a significant enrichment of the EpCAM+ subpopulation. Overexpressed miR-181s levels are also important in the enhanced tumorigenic potential of EpCAM+ liver CSCs. More importantly, the miR-181 family was found to be critical in maintaining the “stemness” of EpCAM+ liver CSCs, in part, by targeting an inhibitor of Wnt/β-catenin signaling (NLK) and two hepatic transcriptional regulators of differentiation (CDX2 and GATA6). The targeting of NLK by miR-181s can constitutively activate the Wnt pathway. It thereby enhances the self-renewal ability of EpCAM+ liver CSCs. Conversely targeting the activation of the differentiation of CDX2 and GATA6 by miR-181s can maintain EpCAM+ liver CSCs in their undifferentiated state.43,44

miR-130b in CD133+ liver CSCs

More recently, studies from our group have also demonstrated a similar finding, whereby liver CSCs are regulated by dysregulated miRNA expression. By comparing the miRNA profiles of CD133+ and CD133- cells isolated from HCC primary tumors and experimental cell lines, significantly elevated miR-130b expression was identified in CD133+ liver CSCs. miR-130b was found to be preferentially expressed in CD133+ spheres derived from HCC clinical samples and in chemotherapy-treated unsorted spheres enriched for CD133. Functional studies found that miR-130b was required for self-renewal, tumorigenicity and chemoresistance. CD133- cells overexpressing miR-130b displayed enhanced proliferation, superior resistance to chemotherapeutic agents, elevated expression of stem cell-associated genes, enhanced tumorigenicity in vivo and greater potential for self-renewal in serial passages than control cells transduced with the empty vector alone. Conversely, the antagonization of miR-130b in CD133+ cells was shown to result in the opposite effect. Furthermore, the increased amount of miR-130b paralleled a reduction in TP53INP1, a known miR-130b target. The silencing of TP53INP1 in CD133- cells enhanced both self-renewal and tumorigenicity in vivo. Thus, our findings suggested that miR-130b regulates CD133+ liver CSCs by silencing TP53INP1.15

Angiogenesis

In addition to resistance to chemo- and radiation therapies, CSCs seem to be particularly adept in stimulating angiogenesis to promote tumor growth and increase the overall tumor aggressiveness before and after therapy. In fact, recent clinical studies have shown enhanced antitumor cell effects when anti-angiogenic therapy is combined with radiation or chemotherapy, suggesting that possibly radioresistance, chemotherapy resistance and angiogenesis in CSCs work in concert to initiate tumor recurrence in advanced or aggressive tumors. Given the evidence for the CSC dependence on tumor vasculature, combining radiation therapy or chemotherapy with anti-angiogenic therapies has promise in possibly mediating targeted anti-CSC effects in the promotion of prolonged recurrence-free survival.

In HCC, there are currently two original articles that have documented a link between liver CSCs and angiogenesis. The first report, by Yang et al., found that high expression levels of hepatic stem/progenitor cell biomarkers, such as cytokeratin 19, ABCG2, CD133, nestin and CD44, are related to tumor angiogenesis and are indicative of high tumor recurrence and poor prognosis of surgically resected HCC.45 More recently, through a systematic comparison of the gene expression profiles between sorted CD133 subpopulations in HCC cells by genome-wide microarray analysis, we have also found that CD133+ liver CSCs produce much higher levels of IL-8 than the CD133- non-CSC population. Further, CSC-mediated IL-8 production leads to increased self-renewal ability, amplified endothelial tube formation in vitro and enhanced tumorigenicity in vivo. Moreover, we have also provided evidence that the preferential expression of IL-8 in CD133+ liver CSCs is mediated through a neurotensin-activated mitogen-activated protein kinase (MAPK)-signaling cascade (Tang et al., unpubl. data, 2011 [manuscript submitted]).

Therapeutic implications and future challenges

  1. Top of page
  2. Abstract
  3. The cancer stem cell theory
  4. Hepatocellular carcinoma
  5. Identification of liver cancer stem cells
  6. Regulation of liver cancer stem cells
  7. Therapeutic implications and future challenges
  8. Acknowledgments
  9. References

The identification of novel therapeutic targets for HCC treatment has begun in earnest in the field of basic liver cancer research. Although there has been a significant improvement in the detection and treatment of early stage HCC, the disease remains largely incurable because the current therapeutic regimen is unable to provide a lasting cure for patients with advanced HCC. Recent findings in the identification (Table 2) and characterization of liver CSCs have lent insight and offered great promise for developing better therapeutic strategies against the disease.

Table 2.  Cell surface markers of liver cancer stem cells
MarkerCell line/primary tumorReference
Side populationCell line10
CD133+Cell line13
CD133+Cell line14
CD133+CD44+Cell line25
CD133Cell line/primary tumor15
ALDH activityCell line20
CD90+Cell line/primary tumor22,23
EpCAM+Cell line/primary tumor28
CD13+Cell line/primary tumor29

CD90+CD44+ HCC cells, as discussed previously, possess a high tumorigenic capacity.23 Researchers who have characterized this subpopulation of cells have also examined the potential benefits of targeting CD44 via a neutralizing antibody approach. The systemic administration of anti-human CD44 antibodies in immunodeficient mice, formed by the intrahepatic inoculation of CD90+ liver CSCs, suppressed tumor nodule formation in the liver and metastatic lesions in the lung.23 Furthermore, the administration of CD44 antibodies was also shown to induce apoptosis in both CD90+ and CD90- cells in vitro.23 In addition to CD44, CD133 has also been suggested as a putative therapeutic target in HCC.46 Using a murine anti-human CD133 antibody conjugated to the cytotoxic drug, monomethyl auristatin F, Smith et al. found that the antibody-drug conjugate was able to productively induce the inhibition of CD133+ liver CSC-driven cancer cell growth both in vitro and in vivo.46 The granulin-epithelin precursor (GEP), which has been suggested to play a role in liver cancer cell chemoresistance,33 has also been identified as a potential target for antibody therapy.47 Indeed, anti-GEP monoclonal antibody treatment has resulted in the inhibition of tumor growth in immunodeficient mice, decreased serum GEP levels and reduced tumor angiogenesis.33

The recent work by Haraguchi et al. on the study of CD13+ liver CSCs has also demonstrated that CD13 inhibition by a CD13-neutralizing antibody could elicit cellular apoptosis and inhibit the proliferation of CD13+ liver CSCs-driven HCC. Further, when the CD13 inhibitor, ubenimex, is used in conjunction with the chemotherapeutic drug, 5-fluorouracil, a greater tumor regression was observed than when either agent was used alone.29 Apart from antibody-targeted therapy, the recent discovery by Lee and colleagues showed that lupeol, a phytochemical present in fruits and vegetables, could target CD133+ liver CSCs by inhibiting their self-renewal and tumorigenic capacity.48 In addition, lupeol was able to sensitize HCC cells to chemotherapeutic agents (doxorubicin and cisplatin) through the phosphatase and tensin homolog (PTEN)-AKT-ABCG2, pathway. The combination of lupeol, doxorubicin and cisplatin was found to exert a synergistic effect on tumor suppression, allowing the use of a lower dosage of conventional chemotherapeutic drugs, which may have cytotoxic effects when used at high concentrations.48

As our understanding of CSC grows, new drug discoveries are also underway with the anticipation of attaining the complete eradication of cancer. Recent studies have highlighted the importance and necessity of exploring the susceptibility of CSCs to existing therapies in combination with the disruption of key “stemness” pathways controlling self-renewal, chemoresistance and angiogenesis through molecular-targeted therapy, as conceptualized in Figure 2. Other novel and important directions for effective therapies may include the disruption of the tumor niche essential for CSC homeostasis and the depletion of CSCs by forced differentiation. However, more work is still required to advance our knowledge on the role of CSCs in tumor hierarchy and to design more effective and specific anti-CSC therapy. Overall, the current state of knowledge strongly indicates the advantage of targeting CSCs to improve the limited efficiency of existing therapies, and it has provided an important framework for the development of novel therapeutic regimens with the ultimate hope of bringing long-term clinical benefits to the patient.

image

Figure 2. Liver cancer stem cell properties and mechanisms of regulation. A simplified summary of our current understanding of how liver cancer stem cells are characterized and regulated in terms of tumor initiation, self-renewal, chemoresistance and pluripotency.

Download figure to PowerPoint

Acknowledgments

  1. Top of page
  2. Abstract
  3. The cancer stem cell theory
  4. Hepatocellular carcinoma
  5. Identification of liver cancer stem cells
  6. Regulation of liver cancer stem cells
  7. Therapeutic implications and future challenges
  8. Acknowledgments
  9. References

We thank members of our laboratory for helpful discussion. Work in our laboratory is partially supported by grants from the Sir Michael and Lady Kadoorie Funded Research into Cancer Genetics, Research Grant Council Collaborative Research Fund (HKU1/06C, HKU7/CRG/09 and HKU5/CRF/08), National Key Sci-Tech Special Project of Infectious Diseases (2008ZX1002-022) and The University of Hong Kong Strategic Research Theme in Cancer.

References

  1. Top of page
  2. Abstract
  3. The cancer stem cell theory
  4. Hepatocellular carcinoma
  5. Identification of liver cancer stem cells
  6. Regulation of liver cancer stem cells
  7. Therapeutic implications and future challenges
  8. Acknowledgments
  9. References