MicroRNA Signatures for circulating CD133-positive cells in hepatocellular carcinoma with HCV infection

Aim Molecular characterization of the CD133+ stem cells associated with hepatocarinogensis through identifying the expression patterns of specific microRNAs (miRNAs). Methods We investigated the expression pattern of 13 miRNAs in purified CD133+ cells separated from the peripheral blood of healthy volunteers, chronic hepatitis C (CHC), liver cirrhosis (LC) and hepatocellular carcinoma (HCC) patients a long with bone marrow samples from the healthy volunteers and the LC patients using custom miScript miRNA PCR array. Results The differential expression of the 13 studied miRNAs in CD133+ cells separated from the HCC patients' peripheral blood compared to the controls revealed that miR-602, miR-181b, miR-101, miR-122, miR-192, miR-125a-5p, and miR-221 were significantly up regulated (fold change = 1.8, 1.7, 2, 5.4, 1.6, 2.9 & 1.5 P value = 0.039, 0.0019, 0.0013, 0.0370, 00024, 0.000044 &0.000007 respectively). As for the HCC group compared to the CHC group; miR-602, miR-122, miR-181b, miR-125a-5p, and miR-192 were significantly up regulated (fold change = 13, 3.1, 2.8, 1.6 & 1.56, P value = 0.01, 0.001, 0.000004, 0.002 & 0.007 respectively). Upon comparing the HCC group to the LC group; miR-199a-3p, miR-192, miR-122, miR-181b, miR-224, miR-125a-5p, and miR-885-5p were significantly up regulated (fold change = 5, 6.7, 2.3, 3, 2.5, 4.2 & 39.5 P value = 0.001025, 0.000024, 0.000472, 0.000278, 0.000004, 0.000075 & 0.0000001 respectively) whereas miR-22 was significantly down regulated (fold change = 0.57 P value = 0.00002). Only, miR-192, miR-122, miR-181b and miR-125a-5p were significant common miRNAs in CD133+ cells of the HCC group compared to the other non-malignant groups. Conclusion We identified a miRNA panel comprised of four miRNAs (miR-192, miR-122, miR-181b and miR-125a-5p) that may serve as a molecular tool for characterization of the CD133+ cells associated with different stages of hepatocarinogensis. This panel may aid in developing a new target therapy specific for those CD133+ cells.


Results
The differential expression of the 13 studied miRNAs in CD133+ cells separated from the HCC patients' peripheral blood compared to the controls revealed that miR-602, miR-181b, miR-101, miR-122, miR-192, miR-125a-5p, and miR-221 were significantly up regulated (fold change = 1.8, 1.7, 2, 5.4, 1.6, 2.9 & 1.5 P value = 0.039, 0.0019, 0.0013, 0.0370, 00024, 0.000044 &0.000007 respectively). As for the HCC group compared to the CHC group; miR-602, miR-122, miR-181b, miR-125a-5p, and miR-192 were significantly up regulated (fold change = 13, 3.1, 2.8, 1.6 & 1.56, P value = 0.01, 0.001, 0.000004, 0.002 & 0.007 respectively). Upon comparing the HCC group to the LC group; miR-199a-3p, miR-192, miR-122, miR-181b, miR-224, miR-125a-5p, and miR-885-5p were significantly up regulated (fold change = 5, 6.7, 2. Introduction Hepatocellular carcinoma (HCC) is the most common primary liver cancer and the second leading cause of the cancer-related death in the world. Also, It ranks as the fifth and seventh common cancer in men and women respectively [1]. Egypt is one of the countries who has the highest prevalence of hepatitis C virus (HCV) worldwide estimated by (21.9%) among adults in 1995-1996 [2]. CD133+ is a transmembrane glycoprotein that known as an important cell surface marker of the putative normal liver stem cells "oval cells" as well as cancer stem cells (CSCs) [3][4][5][6]. In line with this, the increased expression of CD133+ was found to be correlated with advanced stages of HCC [7,8]. A special attention is given to the naive circulating CD133+ cells and its clinical significance in HCC prognosis. We have previously showed that circulating CD133+ cells were significantly higher in the blood samples of chronic hepatitis and HCC patients compared to the healthy controls; revealing the prognostic role of CD133+ cells in those patients [9].
MicroRNAs (miRNAs) are small, non-coding RNAs that can regulate carcinogenesis and cancer development through their complementary binding to mRNAs of oncogenes or tumor suppressors genes [10,11]. miRNAs are involved in the regulation of proliferation and selfrenewal of the CSCs and they also promote the differentiation to determine the fate of stem cell [12]. Both normal stem cells (NSCs) and CSCs are regulated by miRNAs which accounts for the importance of the identifying the most common and unique miRNAs expression patterns between CSCs and NSCs.
The purpose of this study is to assess the expression patterns of 13 miRNAs in the naive circulating CD133+ stem cells associated with different stages of hepatocarcinogensis on top of HCV infection. Those miRNAs have been previously reported in literature to play a role in hepatocarcinogensis machinery. Also, this study is considered as a complementary study to our previously published paper which illustrated the expression pattern of those miRNAs in the serum samples of CHC, LC and HCC patients on top of HCV infection as well as healthy controls [13,14]. and all were seronegative for HBV and HCV antibodies. The CHC and LC patients were diagnosed by clinical examination, abdominal ultrasound, laboratory investigations and liver biopsy. The diagnosis of HCC patients was based on computed tomography (CT) and elevated AFP level and confirmed by histopathology. The criteria for inclusion were as follows: a) adult patients of both sexes who were positive for HCV antibodies (EIAgen HCV Ab (v.4) kit code: 071064, adaltis. Milano Italy) as well as HCV RT-PCR (artus HCV RT-PCR Kits CE, cat no. 4518265, QIAGEN GmbH, QIAGEN Strasse 1, D-40724 Hilden); b) newly diagnosed HCC cases without prior chemotherapy. The criteria for exclusion were as follows: a) patients with positive hepatitis B surface antigen (HBsAg) and was confirmed by PCR; b) patients who received previous treatment for HCC. The study was approved by the Institutional Review Boards (IRB) of the National Cancer Institute, Cairo University, which was in accordance with the guidelines of 2004 Declaration of Helsinki. A written informed consent was obtained from all participants prior to enrollment in the study (Organization No.IORG0003381. IRB NO. IRB00004025).

Collection of clinical specimen
Ten milliliters of venous peripheral blood were collected from all enrolled subjects in Cell-Save blood collection tubes (Immunicon Inc., Huntingdon Valley, PA, United States) containing a cellular preservative and EDTA. Similarly, 10 ml of heparinized BM aspirate were collected from LC patients and normal volunteers during their regular treatment regimen and follow up. Briefly, patients' skin was cleaned with 70% ethanol at the iliac crest with a boring movement. Then, needles were passed perpendicularly into the cavity of the ileum at a point just posterior to anterior superior iliac spine or 2 cm posterior and 2 cm inferior to the anterior superior iliac spine to aspirate the BM (Salah and Klima New Delhi 110055, Delhi, India). The mononuclear cells were isolated from freshly collected peripheral blood and BM samples using RBC's lysis buffer. Cells were washed twice using phosphate buffer saline and counted using a hemocytometer.

CD133+ cells' sorting
CD133 + cells were isolated using MACS kit (Miltenyi Biotec, Germany) according to manufacturer's instructions. The mononuclear cells were incubated for 30 min. at 4˚C after labeling with 100μl CD133/1 magnetic microbeads and 100 μl FcR. Then, the cells were sorted twice using the autoMACS™ separator (Miltenyi Biotec, GermanyF) to increase their purity. The cells were washed twice with phosphate buffered saline (pH 7.2); supplemented with 0.5% bovine serum albumin and 2mM EDTA, and centrifuged at 300 xg for 10 min at 4˚C. For confirmatory set, CD133+ cells were sorted using the BD FACSCalibur™ (BD Biosciences, USA) platform that allows performing both cell analysis and cell sorting in a single bench top system.

Flow cytometry analysis
The sorted cells were incubated with phycoerythrin (PE)-conjugated anti-CD133/2 for 30 min at 4˚C. (Miltenyi Biotec, Bergisch Gladbach, Germany). Then, they were treated with FcR blocking reagent (Miltenyi Biotec). The labeled cells were analyzed with Flowcytometer Epics1 Elite Coulter system using the CELL Quest software (Becton-Dickinson Immunocytometry Systems, San Jose, CA, USA). A nonspecific isotype control was used in each sample. All antibodies were of IgG1k isotype. Results were expressed as a specific percentage of positive markers, calculated by subtracting the nonspecific fluorescence of the isotype control form the specific fluorescence of the monoclonal antibody.

RNA extraction and cDNA synthesis
Total RNA including miRNA was extracted using RNeasy mini kit (Qiagen cat. no. 217004) and QIAzol lysis reagent according to manufacturer's instructions. The quality of the RNA was estimated using 2000\2000c nanodrop (Thermo scientific, USA). RT2 miRNA First Strand Kit (Qiagen) was used to prepare cDNA from 500 ng RNA. Reverse transcription was carried out in a final reaction volume of 20 μl that were incubated for 60 min at 37˚C followed by 5 min at 95˚C.

miRNAs selection and target prediction
The selection of the miRNAs was based on a previous review of literature along with using miR2Disease database; a comprehensive resource of miRNA deregulation in various human diseases (http://www.mir2disease.org). While, miRTaBase database; the experimentally validated miRNA-target interaction database (mirtarbase.mbc.nctu.edu.tw) was used to predict downstream targets of miRNAs.

Confirmatory reverse transcription and Quantitative Real-Time Polymerase Chain Reaction (qRT-PCR)
CD133 cells were sorted from blood samples obtained from eight HCC patients and eight control subjects using flowcytometry with cell sorter. Total RNA was reverse transcribed for sorted CD133 cells using miScript II RT kit (Qiagen). Reactions were incubated at 37˚C for 1 hr followed by inactivation of the reaction by incubation at 95˚C for 10 min. For miRNA expression Confirmatory Real-time PCR was carried out for the four miRNAs (miRNA-122, miRNA181b, miRNA-125a-5p, miRNA-192) using primers specific for these miRNAs, Table 1, 1 μL of diluted RT product was used (equivalent to 3 ng) as template in a 10 μL PCR reaction containing 1X SYBR Green master mix (Qiagen), 200nMmiRNA specific forward primer, and 200 nM universal primer. The conditions for qRT-PCR were as follows: 95˚C for 10min, followed by 40 cycles of 95˚C for 15s and 50˚C for 30 s and 70˚C for 30 s. All the RT-qPCR reactions were performed on 7500 fast real-time PCR system (Applied Biosystems). All samples were analysed in duplicate.

Statistical analysis
Data analysis was performed using the supplied software "http://www.sabiosciences.com/ pcrarraydataanalysis.php"; based on Student's t-test of the replicate (2^(-ΔCt) values for each miRNA in the tested groups and the control group. P-values less than 0.05 are considered statistically significant. Fold-change values >1 imply an up-regulation while fold-change values <1 imply a down-regulation. The SPSS software package (version 15) was used to analyze the clinical data. Continuous variables were expressed as median and range while categorical variables were expressed as percentages. Comparisons between groups were analyzed by one way ANOVA for the continuous variables and by the χ2 or Fisher's exact test for the categorical variables whenever possible.

The demographic features and the clinical data of the studied groups
The demographic and the clinical features of the different studied groups were shown in Table 2. The age differed significantly between the studied groups (P value < 0.05) with a trend for increasing ages with disease progression. However, there was no significant difference between the LC and the HCC groups regarding the age. Also, no significant difference was observed regarding the gender between the studied groups with male predominance in the groups with HCV-related liver disease except for those in the CHC group. The level of AST was significantly higher in the LC and CHC groups compared to the other groups (P value < 0.05). However, levels of total bilirubin and AST were significantly higher in the LC group compared to the other groups (P value < 0.05, 0.01 respectively). Albumin was significantly lower in the LC group compared to the other groups (P value < 0.05) whereas; AFP was significantly higher in the HCC group compared to the other groups (P value < 0.01).

CD133+ subpopulation purity in the different studied groups
Prior to performing the custom miScript miRNA PCR array, the purity of sorted CD133+ cells was tested in all studied groups by flowcytometry. The purity of CD133+ subpopulation ranged from 67 to 76% in the different studied groups as shown in Fig 1.

The differential expression of miRNAs in CD133+ cells purified from the peripheral blood (PB) of the different studied groups
Among the 13 studied miRNAs, only miR-101 was significantly up regulated (fold change = 1.69 P value = 0.016) while miR-122 was significantly down regulated (fold change = -1.74 P value = 0.0114) upon comparing CD133+ cells of the CHC group to those of the control group as shown in S1 The heat map with dendograms for the differential expression of 13 miRNAs in CD133+ cells of the three studied groups compared to the control group indicates co-regulated miR-NAs across all studied groups along with multi group plot provided both a bar chart and line graph representation (with optional error bars) are shown in Fig 3. Among the 13 studied miRNAs in CD133+ cells of the HCC group compared to the CHC group; miR-602, miR-122, miR-181b, miR-125a-5p and miR-192 were significantly up regulated (fold change = 13, 3.  4B. The differential expression of 13 miRNAs in CD133+ cells of the HCC group compared to the other studied groups was illustrated in Fig 4C and Fig 4D. It has been shown that miR-122, miR-192, miR-125a-5p and miR-181b were significantly common up regulated miRNAs in CD133+ cells of the HCC group (fold change = 5, 6.7, 2.3 & 39.5, P value = 0.001025, 0.000024 & 0.0000001; respectively) up on comparing CD133+ cells of the malignant group (HCC group) to the other non-malignant groups (CHC, LC and control groups) as shown in S6 Table.  The histograms representing the differential expression of miRNAs in CD133+ cells purified from peripheral blood of all studied groups were shown in Fig 5. The possible mechanistic role of significant overlapping miRNAs in CD133+ cells of HCC with their downstream targets as well as the relationship between miRNAs and HCV were illustrated in Table 3 and Fig 6. The differential expression of miRNAs in CD133+ cells purified from the bone marrow (BM) of the different studied groups The differential expression of the studied miRNAs in CD133+ cells separated from BM to those separated from PB of the control group revealed that only 3 out of 13 miRNAs were significantly up regulated; miR-122, miR-221 and miR-125a-5p (fold change = 1.  The histograms representing the differential expression of miRNAs in CD133+ cells purified from bone marrow of the LC and control groups were shown in Fig 8.

Confirmatory set results
The differential expression of the studied miRNAs in CD133+ of the HCC group compared to the control group showed up regulation of miR-181b, miR-192, miR-125a-5p and miR-122 in concordance with the our previous data Table 4, Fig 9. Moreover, there is no significant difference in the miRNA expression between both methods used for CD133+ cell sorting.

Discussion
Cancer stem cells (CSCs) represent a subset of cells that exist within the tumor. They are responsible for resistance, survival and recurrence after the treatment. CD133+ has drawn a significant attention as a critical liver CSC marker [15]. There is evidence that miRNAs are significantly involved in self-renewal and differentiation of CSCs [16]. They are responsible for regulating the stemness characteristics through modulation of tumor-suppressive and oncogenic pathways [17].
In this study, we aimed to differeniate between the naive circulating CD133 + cells obtained from patients at different stages of HCV-associated liver disease and the CD133 + cells of the healthy controls at the molecular level through comparing the expression patterns of 13 miRNAs.
Our results revealed that miR-181b, miR-602, miR-101, miR-122, miR-192, miR-125a-5p miR-221 and miR-22 were significantly up regulated up on comparing CD133+ cells of the HCC group (PB) to those of the control group (PB). MiR-181b which belongs to miR-181s family; plays an important role in liver injury and cancer development [18]. Significant up regulation of miR-181b was in concordance with a previous report by Ji et al. who found that multiple members of the miR-181 family, including miR-181b are up-regulated in the liver CSC subset marked with EpCAM+ AFP+ [19]. Similarly, our previous study addressed that miR-181 was significantly up regulated in sera of the HCC group when compared to the LC group [14]. Moreover, Meng et al identified the differential expression of miR-181 family members in an Oct4 + CD133 + liver CSC subset. They also identified a critical role for miR-181s in Twist-driven metastasis in this Oct4 liver CSCs through targeting RAS-association domain family 1, isoform A (RASSF1A), tissue inhibitor of metalloproteinase 3 (TIMP3) and nemo-like kinase (NLK) [18]. Furthermore, miR-181s was found to maintain stemness by directly targeting GATA-binding protein 6 (GATA6) caudal type homeobox 2 (CDX2) and nemo-like kinase (NLK) which is a potent inhibitor of β-catenin resulting in activation of the Wnt/β-catenin pathway in EpCAM + liver CSCs [19].
Our data regarding the significant up regulation of miR-602 was in agreement with yang et al. who identified the oncogenic function of miR-602 in HCC through targeting the tumor suppressive function of RASSF1A [20]. On the other hand, miR-101 was reported to inhibit tumor aggressiveness through targeting EZH2 which is an epigenetic regulator of cell survival, proliferation, CSC phenotype and function, in various cancers including endometrial cancer [21] prostate cancer [22], pancreatic cancer [23]. On contrary to these data, we found that miR-101 was significantly up regulated. The possible explanation is that individual miRNA may act as oncomir or tumor suppressor according to its organ specificity.
Besides, miR-122 and miR-192 are the most two abundantly expressed miRNAs in the human liver [24]. miR-122 represents 70% of the total miRNAs in the normal liver [25,26]. In the primary HCC tissues as well as hepatoma cell lines, miR-122 was found to be down-regulated conceiving its role as a tumor-suppressive miRNA [27,28].It targets many oncogenic genes such as disintegrin and metalloprotease 17 (ADAM17) [29], Bcl-w [30], Wnt1 [31] and cyclin G1 (CCGN1) [32]. The significant up regulation of miR-122 in the circulating CD133 cells purified from HCC patients associated with HCV infection may be attributed to that CD133 cells serve as a reservoir for HCV [33] which in turn stimulates the expression of miR-122 in favor of promoting its genome replication and translation [34]. Also, the exogenous expression of miR-122 supports HCV replication and propagation in the non-permissive cell line [35]. Similarly, our previous study reported that miR-122 was significantly up regulated in the serum samples of HCC patients associated with HCV infection when compared to the healthy controls [14]. As for miR-122, miR-192 also enhances HCV replication [36]. Moreover, Yan-Chun et al., reported that miR-192-5p promotes the proliferation and metastasis of HCC expressing CD133 and CD90 cell surface markers through targeting semaphorin 3A (SEMA3A) [37].
Additionally, miR-125a is frequently down-regulated in HCC and it inversely correlated with aggressiveness and poor prognosis of the tumor. Ectopic expression of miR-125a may down-regulate MMP11 and VEGF in vitro and in vivo supporting its role as tumor suppressive miRNA; inhibiting the proliferation, invasion and metastasis of HCC [38]. Moreover, miR-125a inhibits the invasion of HCC cells through regulating the PI3K/AKT/mTOR pathway [39]. However, the significant up-regulation of miR-125a-5p in this study came in agreement Table 3

. Functional annotation of miRNAs and their target genes in CD133+ cells of HCC. miRNA
Target Gene annotation

miR-122
Bcl-w Inhibits apoptotic pathway ADAM17 Process many receptors as EGFR, TNFR-I and TNFR-II Notch receptors. It plays a prominent role in notch pathway activation

Wnt-1
Reacts with frizzled family receptor results in activation of wnt/β-catenin.
CCNG1 p53 and p73 activate the CCNG1 during the transcription steps that in turn suppress p53 family proteins leading to cell survival activation.

miR-192
SEMA3A Inhibits the Rac-PAK pathway responsible for regulation of actin dynamics; resulting in cytoskeleton remodeling.

miR-181b
TIMP3 Exerts its anti-angiogenic effect via a direct interaction with vascular endothelial growth factor (VEGF) receptor-2 resulting in inhibition of proliferation, migration and tube formation of endothelial cells (ECs).

GATA6
The transcription factor GATA6 directly regulates expression of leucine-rich repeat containing G protein-coupled receptor 5 (LGR5) involved in the differentiation pathway.
CDX2 CDX2 is (caudal type homeobox 2) is an important transcription factor in the differentiation pathway with an early report by Guo et al. who proved that miR-125a alone was capable of increasing the number of hematopoietic stem cells in vivo via targeting multiple proapoptotic genes e.g. Bak1 [40]. In addition, ERBB2 and ERBB3 were negatively regulated by miR-125a and miR-125b. In CD133+ cells, the down-regulation of ERBB2 due to the effect of miR-125 might lead to more polarized morphology. Therefore, miR-125a plays an anti apoptotic role and participates in remodeling of cytoskeleton and dedifferentiation [41]. miR-221 is up regulated in 70%-80% of HCC samples with the maximum divergence affecting several cancer pathways [42]. Also, miR-221 stimulates the onset of tumors, promotes tumor progression and shorts the time to death in the mouse model of liver cancer [43]. miR-  On comparing CD133+ cells of the HCC group (PB) to those cells of the control group (PB); miR-122, miR-192, miR-885-5p, miR-224, miR125a-5p, miR-181b and miR-199a-3p were significantly up-regulated whereas miR-22 was significantly down-regulated. Similarly,  and miR-224 were up regulated in the sera of HCC and LC patients compared to the normal controls [49]. Moreover, Chen et al. showed that miR-224 starts to be up regulated from the precancerous stage and its up regulation persists throughout the HCC development [50]. miR-224 plays an important role in cell proliferation, migration, invasion, and prevention of apoptosis in HCC through its binding to target genes, such as CDC42, CDH1, PAK2, BCL-2, and MAPK1 suggesting its role as an oncomir [51,52].
Down regulation of miR-199a-3p in HCC has been reported in various previous studies [53 -56] supporting its function as a tumor suppressor miRNA. On the contrary, miR-199a-3p   . The significant down regulation of miR-22 was concordant with our previous data which reported its down regulated level in the sera of HCC patients compared to LC patients [14]. Also, Zang et al., reported that the decreased level of miR-22 was associated with poor clinical outcome of liver cancer patients [62]. Several studies addressed that miR-22 over-expression targeting PTEN, p21, and p53 leads to reduction of cell growth, invasion, and metastasis in several cancer types [62-65]; supporting its tumor suppressive function. Furthermore, the proliferation of CD133+ is greatly affected by the miRNAs expression level itself. Such relation was reported by Zhang et al who showed that miR-150 over expression lead to a significant reduction of CD133+ cells, accompanied by significant inhibition of cell growth and tumorsphere formation. So, miR-150 may be involved in liver CSC self-renewal, potentially via modulation of the downstream target c-Myb [66].
To the best of our knowledge, this the first study showed the differential miRNAs expression pattern of CD133+ cells obtained from BM versus CD133+ cells obtained from PB of the same participant. We found that miR-122, miR-221 and miR-125a-5p were significantly up- regulated on comparing CD133+ cells separated from BM to those separated from PB of the control group. However, miR-885-5p, miR-224 and miR-199a-3p were significantly up-regulated whereas miR-22, miR-101 and miR-29b were significantly down-regulated in the CD133 + cells separated from BM compared to those separated from PB of the LC group. On comparing the differential miRNAs expression pattern of CD133+ cells separated from BM of the LC group to those separated from BM of the control group vs. CD133+ cells separated from PB of the LC group to those separated from PB of the control group we found that miR-122, miR-192, miR-199a-3p, miR-125a-5p and miR-224 were common significantly down regulated miR-NAs. Further study will be needed to investigate the role of those miRNAs in the differentiation of CD133+ cells.

Conclusion
These results have provided the evidence that the miRNAs differentially expressed between the CD133+ cells associated with HCC group associated with HCV infection and CD133+ cells of the other non-malignant studied groups. Those miRNAs are highly involved in stemness, proliferation, and remodeling of cytoskeleton and apoptosis blocking. These findings broaden our knowledge about the molecular characterization along with phenotypic characterization of CD133+ cells for stem cell-based therapies. However, there are some limitations in this study as phenotypic characterization with the double stained and clonogenic assays will be needed to prove the carcinogenic properties of the stem cells harboring CD133+ surface marker in HCC.
Supporting information S1 Table. The