Elucidating the molecular mechanisms of tumor progression and tumor prognosis in human HCC remains largely unexplored. In the current study, we identified 6 candidate miRNAs targeting AKTs using an improved prediction protocol with two steps: (1), 5 different programs for miRNA target prediction were used due to their various algorithms and we identified the overlap which was the consensus among 3 out of 5 different programs; (2), we selected miRNA-target interactions (MTIs) with low hybridization energies. These two steps were performed to minimize false positive and false negative results. According to the enrichment analysis on GO items and KEGG pathways, we found that the most common functions of the experimentally validated targets of candidate miRNAs targeting AKTs included focal adhesion, cell cycle, p53 signaling pathway, mTOR signaling pathway, apoptosis, VEGF signaling pathway, which cover all the hallmarks of HCC, providing a convincing evidence that these candidate miRNAs may have a definitive impact on hepatocarcinogenesis. In order to improve our understanding of this candidate miRNAs on HCC malignant progression, we constructed the protein interaction network of their validate targets. Following the network topological analysis, we found that AKT1 and its interaction with mTOR respectively had the highest node-betweenness and edge-betweenness. Many previous studies have recognized highly connected nodes (hubs) as the most important points in the network. Very recently, growing reports have put forward a complementary notion that is bottlenecks as nodes/edges with a high betweenness centrality (i.e., network nodes/edges which have many ‘shortest paths’ going through them). In fact, bottlenecks represent key connector proteins/interactions with surprising functional and dynamic properties [
20]. On this context, AKT1-mTOR may be the most important interaction in our miRNAs-regulated protein interaction network. Thus, we further performed a systematic experiments to validate the regulatory effect of a candidate miRNA targeting AKT1-miR-149 on AKT1-mTOR axis, and its involvement in hepatocarcinogenesis and clinical progression of HCC.
MiR-149, located on chromosome 2 [
21], has been reported to play controversial roles in the progression of various types of human cancers. It is downregulated in several cancer cells and functions as a tumor suppressor by targeting oncogenes. For example, Chan et al. [
22] reported that miR-149 could suppress breast cancer cell migration/invasion and metastasis by targeting GIT1; Cheng et al. [
23] observed the decreased expression of miR-149 in clear cell renal cell carcinoma; Ke et al. [
24] found that the expression of miR-149 was downregulated in lung cancer and miR-149 could inhibit EMT by targeting FOXM1; Pan et al. [
18] indicated that miR-149 may play a tumor suppressive role in the proliferation and invasion of glioma cells via blockade of the AKT1 signaling. In contrast, miR-149 is also upregulated in several cancer cells and functions as a oncomiRs by targeting certain tumor suppressive genes. For example, miR-149 was highly expressed in nasopharyngeal carcinoma and promoted malignant progression by suppressing the expression of its target gene, Smad2 [
25]; Its oncogenic role was also demonstrated in melanoma in which miR-149 upregulation caused the downregulation of GSK3- and upregulation of Mcl-1, leading to apoptotic resistance [
26]. According to miRTarBase (Release 4.5: Nov. 1, 2013;
http://mirtarbase.mbc.nctu.edu.tw/), E2F1 and MYBL2, together with AKT1, are validated targets for miR-149. Chen and colleagues reported that the upregulation of E2F1 protein might associate with worse outcomes in patients with HCC [
27]. Frau and colleagues also indicated that MYBL2 upregulation could induce fast growth and progression of premalignant and malignant liver, through cell cycle deregulation and activation of genes and pathways related to tumor progression [
28]. Nakajima and colleagues identified MYBL2 as a probable transcriptional target of E2F1 in HCC and as a useful biomarker for diagnosis and an attractive target for molecular therapies useful to treat HCC [
29]. In the present study, our data showed the downregulated expression of miR-149 in HCC cell line and clinical specimens compared with the normal liver cell and matched adjacent nonneoplastic liver tissues, respectively. In addition, our functional study showed that AKT1 was a downstream target and effector of miR-149. Enforced expression of miR-149 and knockdown of AKT1 both inhibited the cell proliferation and tumorigenicity of HCC cell line in vitro. Strikingly, miR-149 overexpression did not inhibit the cell proliferation and tumorigenicity of HCC cells in vitro when AKT1 was knocked down. Then, our investigation of the downstream effectors of miR-149-AKT1 signaling axis showed that enforced expression of miR-149 and knockdown of AKT1 both could impair the activation of AKT1 and mTOR. Our data here provide the first and direct evidence that miR-149 has an tumor suppressive role in HCC cells by regulating AKT1/mTOR signaling, which was similar with the previous findings of Pan et al. [
18] on glioma cells. More importantly, our data showed that the expression level of miR-149 was associated with tumor stage of HCC patients. Advance tumor stage samples have lower miR-149 expression compared to early tumor stage samples. These findings were similar to the data of Wang et al. on gastric cancer [
30]. Furthermore, we also identified miR-149 as an independent prognostic factor for both 5-year disease-free survival and 5-year overall survival of HCC patients.