miR-361-3p Regulates Liver Tumor-initiating Cells Expansion and Chemo-resistance

Increasing evidence shows that liver tumor-initiating cells (T-ICs) closely associated with the progression, metastasis, recurrence and chemo-resistance of hepatocellular carcinoma (HCC). However, the underlying mechanism for the propagation of liver T-ICs remains unclear. Here we show that miR-361-3p is upregulated in liver T-ICs. Knockdown of miR-361-3p impairs the self-renewal and tumorigenicity liver T-ICs. Conversely, forced miR-361-3p expression enhances the self-renewal and tumorigenicity liver T-ICs. Mechanistically, miR-361-3p directly targets SOX1 via binding its 3'-UTR in liver T-ICs. Moreover, miR-361-3p knockdown hepatoma cells are more sensitive to cisplatin or sorafenib treatment. Clinical cohort analysis demonstrates that miR-361-3p low HCC patients are benefited from TACE (transcatheter arterial chemoembolization) or sorafenib treatment. In conclusion, our findings revealed the crucial role of the miR-361-3p in liver T-IC expansion and TACE or sorafenib response, rendering miR-361-3p an optimal target for the prevention and intervention in HCC.


Background
With nearly 700,000 new cases per year, hepatocellular carcinoma (HCC) ranks as the sixth most common cancer in the world (1). However, the early symptoms of HCC are not obvious. Most HCC patients are diagnosed at advanced stage (2). Due to high grade invasive HCC can develop to life threatening metastases and less than 20% patients with Portal vein thrombus-only have 2-year overall survival (3). Currently, most of drugs which used for clinical HCC patients were proved to be disappointed.
So, the underlying mechanism of HCC initiation and progression needs to be deeply explored.
Tumor-initiating cells (T-ICs) or cancer stem cells (CSCs) are a subgroup of cancer cells, which have the self-renewal ability and tumorigenesis capacity (4,5). T-ICs or CSCs were rst discovered in blood system diseases (6). Later studies also con rmed the existence of T-ICs or CSCs in solid tumors, including breast cancer, liver cancer and glioma (7)(8)(9). Currently researches suggest that T-ICs or CSCs are responsible for the progression, metastasis, recurrence and chemo-resistance of cancers (10)(11)(12). Therefore, identi cation of the underlying mechanisms governing T-ICs or CSCs propagation may lead to the discovery of promising therapeutic strategies for cancer patients.
MicroRNAs (miRNAs) are a class of small noncoding RNA molecules that contain approximately 22 nucleotides (13). It was reported to be essential for many physiological processes, such as cellular homeostasis, development, differentiation, cell survival and death, tumor initiation and progression (14,15). miR-361-3p is a newly discovered miRNA, and its function and mechanism of action in biological processes and diseases are not completely understood. Previous studies found that miR-361-3p promotes human breast cancer cell viability by inhibiting the E2F1/P73 signaling pathway (16). Moreover, miR-361-3p regulates ERK1/2-induced EMT via DUSP2 mRNA degradation in pancreatic ductal adenocarcinoma (17). However, the function of miR-361-3p in liver T-ICs is unclear.
In this study, we for rst nd that miR-361-3p is highly expressed in liver T-ICs. Next, by using loss-offunction analysis and gain-of-function analysis in liver T-ICs, we demonstrate that miR-361-3p promotes the self-renewal capacity and tumorigenicity of liver T-ICs. Further mechanism study reveals that miR-361-3p directly targets SOX1 in liver T-ICs. Interestingly, miR-361-3p knockdown hepatoma cells are more sensitive to cisplatin or sorafenib treatment. Clinical cohort analysis demonstrates that miR-361-3p low HCC patients are bene ted from TACE or sorafenib treatment. Altogether, we discover that miR-361-3p promotes the expansion of liver T-ICs via interacting with SOX1.
It was well accepted that CD24, CD133 and EpCAM were liver T-ICs markers (18)(19)(20). Therefore, we isolated CD24, CD133 and EpCAM positive HCC cells by ow cytometry. As expected, the expression of miR-361-3p was dramatically increased in CD24, CD133 and EpCAM positive HCC cells compared with their negative HCC cells ( Figure. 1A-C). Spheroid culture of cancer cells is a routine approach to enrich T-ICs. We observed that expression of miR-361-3p was signi cantly upregulated in the self-renewing spheroids compared with the attached cells ( Figure. 1D). In serial passages of Huh7 or HepG2 spheroids, miR-361-3p expression was gradually increased (Figure. To evaluate the potential role of miR-361-3p in liver T-ICs, Huh7 and HepG2 cells were infected with miR-361-3p overexpression virus. The overexpression effect of miR-361-3p was determined by real-time PCR ( Figure. 2A). Spheroids formation is a routine approach to assess self-renewal ability. We found that Huh7/HepG2 miR-361-3p mimic cells formed much more spheres than control cells ( Figure.  miR-361-3p knockdown inhibits liver T-ICs self-renew and tumorigenesis. To further explore the role of miR-361-3p in liver T-ICs, Huh7 and HepG2 cells were infected with miR-361-3p interference virus. The knockdown effect of miR-361-3p was determined by real-time PCR (Figure. 3A).
Spheroids formation is a routine approach to assess self-renewal ability. We found that Huh7/HepG2 miR-361-3p knockdown cells formed much less spheres than control cells ( Figure To elucidate mechanism underlying miR-361-3p-mediated liver T-ICs expansion, TargetScan and miRbase were used to predicted the potential targets of miR-361-3p in liver T-ICs. Bioinformatics analysis found that miR-361-3p has a putative binding site in SOX1 mRNA 3'-UTR (Figure.  We next explored whether miR-361-3p was involved in the regulation of chemo-resistance of HCC. As expected, miR-361-3p expression was markedly upregulated in cisplatin-resistance or sorafenibresistance xenograft ( Figure. 5A). Consistently, miR-361-3p expression was also signi cantly increased in cisplatin-resistance or sorafenib-resistance HCC cell lines ( Figure. 5B C). Moreover, the sensitivity of cisplatin or sorafenib was increased in miR-361-3p knockdown HCC cells compared with control HCC cells ( Figure. 5D). We also observed that interference of miR-361-3p sensitized HCC cells to undergo cisplatin-induced or sorafenib-induced cell apoptosis ( Figure. 5E-H). Furthermore, Kaplan-Meier analysis revealed the survival bene ts in adjuvant TACE-treated or sorafenib-treated HCC patients with low miR-361-3p levels ( Figure. 5I J).

Discussion
Hepatocellular carcinoma (HCC) is the sixth most common cancer in the world, accounting for 90% of human liver cancer (21). The incidence of liver cancer is rising due to various factors such as hepatitis, alcoholic fatty liver, nonalcoholic fatty liver and a atoxin (22,23). Hepatectomy and liver transplantation are commonly used in patients with early stage HCC (24). However, most patients with advanced HCC are contradicting for surgery, and the survival bene t of TACE or targeted drug sorafenib is limited (25).
Therefore, it is necessary to further clarify the development of HCC. In this study, we con rmed for the rst time that miR-361-3p was highly expressed in liver T-ICs, which enhanced liver T-ICs self-renewal and tumorigenesis ability. Our clinical data showed that miR-361-3p could be used to predict TACE and sorafenib response in HCC patients.
MiRNAs, a small non-coding RNA molecule (containing about 22 nucleotides), regulates RNA silencing and post-transcriptional of gene expression (26). miRNAs act as oncogenes or tumor suppressors in tumors dependent on special conditions. miR-361-3p has been reported upregulated in various human tumors, including breast cancer, oral squamous cancer and pancreatic cancer (16,17,27). However, miR-361-3p functions in liver T-ICs have never been investigated before. In current studies, we found that miR-361-3p was upregulated in liver T-ICs. Moreover, we found that miR-361-3p knockdown inhibited liver T-ICs self-renew and tumorigenesis. Conversely, forced miR-361-3p expression promoted liver T-ICs self-renew and tumorigenesis.
SOX1, belonging to SRY (sex determining region Y)-box (SOX) family proteins, is evolutionarily conserved in many species and participated as a key regulator of neural cell fate determination and differentiation (28,29). Previous studies have demonstrated that SOX1 plays an essential role in liver cancer progression (30). In addition, SOX1 was also reported to be involved in the regulation of cancer stemness (31). We hereby found that miR-361-3p knockdown increased SOX1 mRNA and protein expression in liver T-ICs. Moreover, we found that miR-361-3p directly regulates SOX1 expression via interaction with its 3' UTR. In addition, a signi cant negative correlation was identi ed between SOX1 and miR-361-3p expression in clinical samples of HCC. SOX1 siRNA could diminish the self-renewal ability and liver T-ICs frequency between miR-361-3p knockdown HCC cells and control cells.
For advanced HCC patinets, TACE (transcatheter arterial chemoembolization) and the targeted drug sorafenib are the major means of treatment for a long time. However, only a small proportion of patients respond to sorafenib or TACE, and the majority of patients not only have no curative effect, but also have serious side effects (32,33). Therefore, elucidating the mechanism of TACE and sorafenib resistance has become an important link to prolong the survival time of HCC patients. Due to the different responses to sorafenib and TACE in different HCC patients, how to select biomarkers to predict drug reactivity has become a key issue to improve the clinical e cacy of sorafenib and TACE. In the present study, our nding revealed that miR-361-3p knockdown of HCC cells is more sensitive to cisplatin or sorafenib treatment. The TACE or sorafenib cohort analysis further indicated that a low miR-361-3p level in HCC patients can serve as a reliable predictor for TACE or sorafenib response.

Conclusion
We demonstrate that miR-361-3p is upregulated in liver T-ICs, which in turn promotes the self-renewal and tumorigenicity of liver T-ICs. In addition, miR-361-3p promotes liver T-ICs expansion through directly regulating SOX-1. Moreover, miR-361-3p knockdown HCC cells are more sensitive to cisplatin or sorafenib treatment. These data provide insight into the miR-361-3p as a potential therapeutic target against liver T-ICs and a potential predictor for TACE or sorafenib treatment of HCC patients.

Materials And Methods
Cell Culture.
Patients and specimens HCC tissues were obtained from patients who underwent liver resection in the Eastern Hepatobiliary Surgery Hospital (EHBH). Samples were frozen in liquid nitrogen immediately after surgical resection for further RNA extraction. A total of 70 patients received TACE therapy after surgery for primary HCC at EHBH from 2010-2015 were included in Cohort 1. Detailed clinicopathological features of these patients are described in supplementary Table 1. A total of 68 patients received adjuvant sorafenib therapy after surgery for primary HCC at EHBH from 2011-2015 were included in Cohort 2. Detailed clinicopathological features of these patients are described in supplementary Table 2. Another group of 50 HCC specimens were used for analyzing the correlation between SOX1 and miR-361-3p expression. Patient informed consent was obtained and the procedure of human sample collection was approved by the Ethics Committee of EHBH.

Reverse transcription polymerase chain reaction (RT-PCR).
Total RNA was extracted from the cells using Trizol reagent (Invitrogen, 15596-018). Total cDNAs were synthesized by ThermoScript TM RT-PCR system (Invitrogen, 11146-057). The mRNA amount presented in the cells was measured by semi quantitative RT-PCR. PCR conditions included 1 cycle at 94 °C for 5 minutes, followed by up to 40 cycles of 94 °C for 15 seconds (denaturation), 60 °C for 30 seconds (annealing) and 72 °C for 30 seconds (extension). The sequences of primers used was listed in Supplementary Table 3.

Luciferase reporter assay.
A 300-bp fragment of the SOX1 3'UTR containing the conserved miR-361-3p-binding sites was inserted into a luciferase reporter plasmid. The SOX1 3'UTR mutant luciferase plasmid contained changes in potential miR-361-3p-binding base sequence "CUGGGGG" to "GACUCUA". HCC cells were transfected with SOX1 3'UTR luciferase reporter in combination with the pRL-TK vector (Promega, E2241) as an internal control. The dual luciferase assay kit was purchased from Promega (0000060417). The luciferase activities were determined using a luminometer (Wallac 1420 Victor 2 multilabel counter system) as described in previous studies (34).
Western Blotting Assay.
The cells collected by cell lysis buffer, then disposed like we described before (35). Equal aliquots of cell extracts were separated on SDS-polyacrylamide gels. The proteins were then transferred to PVDF membranes (Bio-Rad), blocked, and probed with the antibodies. Primary antibody-bound proteins were detected by using an alkaline phosphatase-linked secondary antibody and an ECF Western Blotting system (Amersham). The densitometric analyses of the protein bands vs. the individual loading controls were performed using the ImageQuant 5.2 software (GE Healthcare). The primary antibodies used were listed in supplementary Table 4. The results shown were representative one of three independent experiments.
Spheroid formation assay HCC cells were digested and then cultured in a 96-well ultra-low attachment (300 cells per well) and cultured in DMEM/F12 (Gibco) media, supplemented with 1% FBS, 20 ng/mL bFGF and 20 ng/mL EGF for 7 days. The total number of spheres was counted under the microscope (Olympus).
The CSC proportions were analyzed using Poisson distribution statistics and the L-Calc Version 1.1 software program (Stem Cell Technologies, Inc., Vancouver, Canada) as previously described (36).
In vivo limiting dilution.
After two months, tumors formation was evaluated.
Huh7 miR-361-3p sponge and control cells were treated with cisplatin (10 µg/ml) or sorafenib (10 µM) for 48 hours, followed by staining with Annexin V and 7-AAD for 15 minutes at room temperature in the dark.