Multidrug resistance in hepatic cancer stem cells: the emerging role of miRNAs

Abstract

There is a fast growing body of evidence that shows several miRNAs are essential to the key features of cancer stem cells (CSC) in hepatocellular carcinoma. However, the function of the miRNAs in different hepatic CSCs and their role in multidrug resistance mechanisms, in particular in those related to the CSC marker ABCG2, is still poorly understood. This limitation is mainly due to the heterogeneity of hepatocellular carcinoma tissues, different CSC markers and high number of deregulated miRNAs, both in primary tumor sites as well as in the circulation. The identification of CSC-related miRNAs and the modulation of hepatocellular carcinoma multidrug resistance would provide a systemic approach in the management of the disease.

Keywords:

• cancer stem cell
• cell marker
• hepatocellular carcinoma
• miRNA
• multidrug resistance

One of the main obstacles in the management for hepatocellular carcinoma (HCC), in particular for the intermediate-advanced stages, is the multidrug resistance (MDR). Despite rapid advance in both basic and clinical research, the prognosis of liver cancer is still poor; it ranks second in the most common cancer-related deaths worldwide [1].

Even still debatable, the cancer stem cells (CSC) are considered the origin of the HCC. Besides their intrinsic properties in tumor-initiation and self-renewal, drug resistance is one of the key features of this cell population. This phenomenon is mainly due to the presence of the ATP-binding cassette (ABC) transporters protein family, in particular the breast cancer resistance protein (BCRP/ABCG2/MXR). It is highly expressed in the CSC and has been widely used to identify the side population (SP) of progenitor/stem cells, mainly based on its functional activity in pumping out dye Hoechst 33342 out of the cells, thus maintaining intracellular non-toxic concentration.

Recent studies had identified several molecules involved in the modulation of the properties of the hepatic CSC with focus on the miRNAs. A simple search in the PubMed database using keywords as HCC and miRNA showed an impressive increase of the number of publications reporting hundreds of deregulated miRNAs [2], while the pace of the research on HCC and CSC is by far slower. Even though the molecular pathways regulating the CSC remain under studies, a growing body of evidence highlights the role of a number of miRNAs in the regulation of tumor initiation, self-renewal, pluripotency metastasis and drug resistance.

A microarray screening conducted by Li et al. identified 78 deregulated miRNAs in SP cells obtained from rat HCC cells compared to normal fetal liver cells. In particular miR-10b, miR-21, miR-470*, miR-34c-3p and let-7i* were over-expressed in the SP of the HCC, while miR-200a* and miR-148b* were under-expressed [3]. A recent study showed that in SP of human HCC cells, miR-9 precursor and miR-194 were decreased, compared to their non-SP counterpart [4]. Down-regulations of tumor-suppressor miRNAs targeting the ABCG2 (miR-181a, miR-142-3p, miR-145) have been reported, but their role remains to be fully clarified in the CSC [5,6]. In glioma stem cells, the function of ABCG2-mediated miR-145 loss led to increased cell proliferation and invasion [7].

Due to the heterogeneity of the hepatic CSC with different CSC markers and their clinical significance, the correlation between a specific CSC marker and its miRNA profile can be variable. In the CSC population CD133+, a preferential over-expression of miR-130b and let-7 family members was correlated with drug resistance [8]. By virtue of their multiple functions, miR-130 and let-7 members did not only regulate the drug resistance but also other key aspects of the CSC by targeting SOCS1, Caspase-3 and TP53INP. The down-regulation of miR-150 was associated with cell renewal via a modulation of the downstream target c-Myb [9] (comprehensively reviewed by the group of Ma et al. [10]). The correlation between let-7 expression and MDR was noticeable only in the CSC, while in the whole HCC cell population, the role of the let-7 family members was ambiguous, suggesting a fine regulation and a targeting specificity for different cell populations [11]. In gastric cancer, the association between miR-130b and MDR was supported by the ability of this miRNA to inhibit the expression of RUNX3 [12], a tumor-suppressor gene down-regulating MRP1, MDR1 and Bcl-2 [13].

In the CSC population CD90+, the miR-548c-5p was related with cellular proliferation and migration. Its transfection promoted apoptosis through the regulation of the expression of β-catenin, Tg737, Bcl-2, Bcl-XL and Caspase-3. In this CSC population, the expression of miR-145 was also significantly decreased compared to CD90 cells [14]. In the CSC population EpCAM+, the miR-181 was up-regulated and its family members were highly expressed in embryonic liver and in isolated hepatic stem cells. miR-181 could target hepatic transcriptional regulator of differentiation. Interestingly, miR-181 expression was found to be higher in EpCAM+CD133+ than double-negative or single-positive cells. [15]. It is noteworthy that the stemness properties of CSC EpCAM is closely associated with the activated Wnt/β-Catenin signaling. Recent data showed a low expression of miR-148a in aggressive stem cell-like subtype of HCC. Since this miRNA is related to HCC growth and metastasis through Wnt signaling, including by targeting EpCAM and CD90 [16], we believe that Wnt/β-catenin could be a good link between several important miRNAs and MDR. Of notice is the observation of the activation of ABCG2 expression through the β-catenin pathway.

As mentioned above, the significance of a specific miRNA can be associated with a specific CSC marker, but their role in ABCG2-related MDR is yet to be fully understood in HCC. Even though it needs to be validated, many reports had shown a positive correlation between CSC surface marker protein and its functional ABCG2 activity, correlating CSC marker with drug resistance. In hepatic CSC, it had been demonstrated that the ABCG2 was preferably expressed in CD90 and CD133 cells [17,18].

Recent studies showed that hepatic CSCs were present not only in the primary tumor but also in the circulation. It had been reported that EpCAM circulating cells expressing ABCG2 and CD133 might be used as an independent prognostic indicator for HCC recurrence [19]. The circulating CSC CD133-related miR-130b was significantly reduced after surgery, showing that this miRNA originated from the tumor [20]. As mentioned above, the miR-130b had been reported to be preferentially expressed in chemoresistant CSC CD133+ and was found to be up-regulated in HCC compared to its adjacent non-tumoral tissues [8].

The characterization of the hepatic CSC miRNA signatures represents a further step in the identification of blood-based biomarkers for circulating CSC. The identification of CSC-related miRNA would provide a special hint in the classification and assessment of the origin of the diverse HCCs for a potential treatment and management of the disease.

Given the relevance of the MDR phenomena in HCC, partially sustained by the CSC, we highlight the importance in discovering miRNA regulatory pathways (e.g., the activation of β-catenin signaling) with the perspective to use specific ABCG2-related miRNA mimics as a supportive therapeutic option, especially in intermediate–advanced stage of the disease. Even though further study in this complex and fascinating field is needed, several important aspects must be considered. First, we need to assess and validate the sensitivity of the CSC identification. As matter of fact, the detection of CSC protein and mRNA in tissue and circulation, and the isolation of CSC validated in xenograft model might differ among different laboratories, limiting the reproducibility of data to a more general clinical scenario. We also need to define a target miRNA (or several miRNAs) that may represent heterogeneous CSC populations as well as can modulate the MDR of the CSC. In particular, because of high stability of the miRNAs in the circulation, it would be important to define a circulating miRNA that can be used as a non-invasive potential biomarker. Furthermore, the characterization of the miRNA in an isolated CSC population from blood of HCC patients will be crucial in the profiling of miRNA-related CSC.

The rapid advances in technologies may provide us with a positive approach to solve the problem in a translational approach. Since the number of open questions largely bypasses the number of solved issues, let us put our cells in culture and get back to work.

Financial & competing interests disclosure

The authors were supported by an in house grant from the Fondazione Italiana Fegato, Italy. D Pascut is funded by the PORFESR 2007-2013 of the region Friuli Venezia Giulia grant. B Anfuso is funded by the fellowship of the Fondazione Italiana per la Ricerca in Epatologia (FIRE). The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.