Involvement of PI3K/PTEN/AKT/mTOR pathway in invasion and metastasis in hepatocellular carcinoma: Association with MMP-9

Authors


Professor Qian Wang, Department of Hepatobiliary Surgery, the First Affiliated Hospital of Sun Yat-Sen University, Guangzhou 510080, China. Email: wangqian00@hotmail.com

Abstract

Aim:  To investigate the status of Phosphatidylinositol 3-kinase (PI3K)/PTEN/AKT/mammalian target of rapamycin (mTOR) pathway and its correlation with clinicopathological features and matrix metalloproteinase-2, -9 (MMP-2, 9) in human hepatocellular carcinoma (HCC).

Methods:  PTEN, Phosphorylated AKT (p-AKT), Phosphorylated mTOR (p-mTOR), MMP-2, MMP-9 and Ki-67 expression levels were evaluated by immunohistochemistry on tissue microarrays containing 200 HCCs with paired adjacent non-cancerous liver tissues. PTEN, MMP-2 and MMP-9 mRNA levels were determined by real-time RT-PCR in 36 HCCs. The relationships between PI3K/PTEN/AKT/mTOR pathway and clinicopathological factors and MMP-2, 9 were analyzed in HCC.

Results:  In HCC, PTEN loss and overexpression of p-AKT and p-mTOR were associated with tumor grade, intrahepatic metastasis, vascular invasion, TNM stage and high Ki-67 labeling index (P < 0.05). PTEN loss was correlated with p-AKT, p-mTOR and MMP-9 overexpression. Furthermore, PTEN and MMP-2, 9 mRNA levels were down-regulated and up-regulated in HCC compared with paired non-cancerous liver tissues, respectively (P < 0.01). PTEN, MMP-2 and MMP-9 mRNA levels were correlated with tumor stage and metastasis. There was an inverse correlation between PTEN and MMP-9 mRNA expression. However, PI3K/PTEN/AKT/mTOR pathway was not correlated with MMP-2.

Conclusions:  PI3K/PTEN/AKT/mTOR pathway, which is activated in HCC, is involved in invasion and metastasis through up-regulating MMP-9 in HCC.

INTRODUCTION

HEPATOCELLULAR CARCINOMA (HCC) is the fifth most frequent neoplasm worldwide and the third most common cause of tumor-related death.1,2 Despite improvements in treatment modalities during the past few decades, the prognosis of HCC is still very poor because of frequent intrahepatic metastasis and tumor recurrence. Thus, there is an urgent need for further insighting into the molecular mechanisms responsible for the biological behavior of HCC and finding new molecular targets for treatment, which may lead to novel treatment modalities that are more effective and less toxic.

Phosphatidylinositol 3-kinase (PI3K)/phosphatase and tensin homolog deleted on chromosome10 (PTEN)/AKT/mammalian target of rapamycin (mTOR) pathway is involved in many cellular processes including proliferation, differentiation, apoptosis, cell cycle progression, cell motility and tumorigenesis, tumor growth, angiogenesis.3–5 PI3K is a member of intracellular lipid kinase family, which catalyzes the generation of phosphatidylinositol-3,4,5- triphosphate (PIP3) from phosphatidylinositol- 4,5-triphosphate (PIP2). AKT is recruited to the plasma membrane by PIP3 and is then phosphorylated at Thr308 by phosphoinositide dependent kinase 1 (PDK1). For full activation, AKT is subsequently phosphorylated at Ser473 by a proposed PDK-2. Activated AKT translocates to the nucleus and activates mTOR and downstream targets. PTEN is a plasma-membrane lipid phosphatase and tumor suppressor which dephosphorylates PIP3 back to PIP2, inhibiting the activation of Ak,6–8 therefore, PTEN negatively regulates the PI3K/AKT/mTOR pathway. PI3K/PTEN/AKT/mTOR pathway as cancer drug target has attracted broad scientific and clinical interests in recent years,3,9,10 but the activation status of this pathway have not yet been well understood in clinical specimens of HCC.

Invasion and metastasis are characteristic features of HCC and the major cause of treatment failure in patients with HCC.11 Experiments in vitro have shown that PI3K/AKT pathway is involved in HCC cell invasiveness by enhancing MMP-9 expression.12,13 However, few data are available regarding the role of PI3K/PTEN/AKT/mTOR pathway in invasion and metastasis and its correlation with MMP in HCC patients.

To better understand the role of PI3K/PTEN/AKT/mTOR pathway in tumorigenesis and progression in HCC, we investigated the status of this pathway and its correlation with clinicopathological data in a large cohort of patients with HCC in the present study. Furthermore, we analyzed the correlation between PI3K/PTEN/AKT/mTOR pathway and MMP-2 and MMP-9. Thus, this study will have important implications for further elucidating the molecular mechanisms and finding novel therapeutical targets in HCC.

PATIENTS AND METHODS

Patients and tissue specimens

THE COMPLETE COHORT consisted of 200 patients who underwent hepatectomy (n = 191) or orthotopic liver transplantation (n = 9) for HCC at the Department of Hepatobiliary Surgery and Centre of Organ Transplantation, the first affiliated hospital of Sun Yat-Sen University (Guangzhou, China) from July 2004 to March 2007. None of these cases underwent chemotherapy or radiotherapy before surgery. The 200 patients included 183 males and 17 females. The mean age of the patients was 50.14 ± 11.81 years. All of the 200 tumor specimens and paired adjacent non-cancerous liver tissues not less than 20 mm away from the HCC were collected immediately after surgery and fixed in formalin and embedded in paraffin at the Department of Pathology, the first affiliated hospital of Sun Yat-Sen University. Among these 200 cases of HCCs, HCC and paired adjacent non-cancerous liver tissues from 36 patients were also collected and immediately frozen in liquid nitrogen, which were subsequently stored at −80°C in the Tumor Tissue Bank, the first affiliated hospital of Sun Yat-Sen University until used for real-time reverse transcription polymerase chain reaction (RT-PCR) analysis. The routine H&E (hematoxylin-eosin) stained tissue sections from all specimens were reviewed to confirm the diagnosis of HCC and graded using WHO grading system by two pathologists. Tumors were staged according to the 6th edition of American Joint Committee on Cancer (AJCC) TNM classification system. The stage III and stage IV HCCs were combined and designated as stage III in this study because of the small number of cases with stage IV tumors. 36 patients whose samples were collected for real-time RT-PCR analysis were classified as: stage I (n = 9), stage II (n = 12), and stage III(n = 15) according to the 6th edition of AJCC TNM classification system; metastatic (n = 19) and nonmetastatic (n = 17) HCC based on ultrasonographic or computed tomography scan findings, intraoperative examination, vascular invasion, histological observation of the micronodules at the periphery of HCC.14–16 The study was approved by the Committee for the Conduct of Human Research of the first affiliated hospital of Sun Yat-Sen University. Informed consent was obtained from each patient.

Tissue microarray construction

The representative areas of each tumor and paired adjacent non-cancerous liver tissue were carefully selected and marked by a pathologist on the H&E stained slide. The designated areas of each donor block were punched with a tissue cylinder (1 mm in diameter) and precisely arrayed into a new recipient microarray block using a manual tissue arrayer (Beecher Instrument, Silver Spring, MD). Each sample was arrayed in triplicate.

Immunohistochemistry

Paraffin-embedded tissue-microarray blocks, which included HCC and paired adjacent non-cancerous liver tissues, were consecutively sectioned at 4 µm intervals and mounted on polylysine-coated glass slides. Slides were incubated at 60°C for 1–2 h and then deparaffinized and rehydrated. Antigen retrieval was performed in citrate buffer (pH 6.0) in pressure cooker. Endogenous peroxidase activity was blocked by incubation with 3% hydrogen peroxide for 15 min at room temperature. Slides were subsequently incubated with primary antibodies against PTEN (Santa Cruz, CA, USA; 1 : 100 dilution), phosphorylated AKT (p-AKT) (Cell Signalling Technology, Beverly, MA, USA; 1 : 25 dilution), phosphorylated mTOR (p-mTOR) (Cell Signalling Technology, Beverly, MA, USA; 1 : 50 dilution), MMP-2 (Santa Cruz, CA, USA; 1 : 100 dilution), MMP-9 (Santa Cruz, CA, USA; 1 : 50 dilution) and Ki-67 (MIB-1, Santa Cruz, CA, USA; 1 : 100 dilution) overnight at 4°C. After incubation at room temperature for 30 min with biotin-free horseradish peroxidase (HRP) enzyme labeled polymer of EnVision plus detection system (Beijing Zhongshan Golden Bridge Biotechnology Co. Ltd., Beijing, China), the slides were developed with the 3,3'-diaminobenzidine solution followed by counterstaining with hematoxylin. Appropriate positive controls were selected according to the manufacturer's recommendations. Slides processed without the primary antibodies were used as negative control (Fig. 1a).

Figure 1.

Immunohistochemical expression of PTEN, p-AKT, p-mTOR, MMP-2, MMP-9 and Ki-67 in human hepatocellular carcinoma (HCC) (× 200). The negative control (a) was used to show the specificity of the antibody. The positive staining of Ki-67 (b) was localized only in the nucleus, whereas the positive stainings of PTEN (c), p-AKT (d, weak staining intensity; e, moderate staining intensity; f, strong staining intensity), p-mTOR (g, weak; h, moderate; i, strong), MMP-2 (j, weak; k, moderate; l, strong), MMP-9 (m, weak; n, moderate; o, strong) were localized in the cytoplasm of HCC cells. Results are representatives of the immunostainings of the specimens.

Evaluation of immunostaining

Immunohistochemical stainings were semiquantitatively assessed by two independent observers who were blinded to all clinical variables. PTEN expression was categorized as negative (−) or positive (+) because it was difficult to quantify its staining intensity. For levels of p-AKT, p-mTOR, MMP-2, MMP-9 expression, staining intensity was scored as follows: negative (−), weak (1+), moderate (2+) or strong (3+) as previously described.17 The immunoreactivity of Ki-67 (Fig. 1b) was evaluated as Ki-67 labeling index by calculating the percentage of immunopositive cells for each tissue sample.18 Ki-67 expression level was divided into low (≤ 50% cells are positive) and high (> 50% cells are positive) Ki-67 labeling index according to the percentage of immunoreactive cells.

Real-time RT-PCR

Real-time RT-PCR was performed to determine mRNA levels of PTEN, MMP-9 and MMP-2 in HCC and paired adjacent non-cancerous liver tissues. Total RNA was extracted using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions. Two micrograms of total RNA from each sample were subjected to reverse transcription reaction with dNTP, Oligo (dT)18, RNase inhibitor, reaction buffer and Moloney murine leukemia virus (M-MLV) reverse transcriptase (Promega, Madison, WI, USA). Real-time PCR amplifications were undertaken in ABI Prism 7500 SDS (Applied Biosystems, Foster City, USA) using EvaGreen™ qPCR Master Mix (Biotium, Hayward, CA, USA). The primers used were as follows: human PTEN, forward primer, TGGAAAGGGACGAACTGGTG, and reverse primer, CATAGCGCCTCTGACTGGGA; human MMP-2, forward primer, GACAACGCCCCCATACCAG, and reverse primer, CACTCGCCCCGTGTGTTAGT; human MMP-9, forward primer, ACGCAGACATCGT CATCCAGT, and reverse primer, GGACCACAACTCG TCATCGTC; human β-actin, forward primer, GGAGAATGGCCCAGTCCTC, and reverse primer, GGGCACGAAGGCTCATCAT. The optimal PCR condition for MMP-2 and PTEN was 95°C for 4 min followed by 40 cycles of 95°C for 30 s, 60°C for 40 s and 72°C for 30 s. The optimal PCR condition for MMP-9 was 95°C for 4 min followed by 40 cycles of 95°C for 30 s, 58°C for 40 s and 72°C for 30 s. Melt curve analysis was employed at the end of each PCR to confirm the specificity of the PCR product. Each measurement was performed in at least triplicate. Data were analyzed according to the comparative Ct method and expression of target genes was normalized to β-actin expression levels in each sample.

Statistical analysis

SPSS 13.0 for Windows (SPSS, IL, USA) was used to analyze all data. χ2-test or Fisher's exact test was used for comparisons between immunohistochemical results and clinicopathological factors. Spearman's bivariable correlation test was used to determine the relationships among these protein expressions. Student's t-test was used to analyze the differences of PTEN, MMP-2 and MMP-9 mRNA expression between the groups. Pearson's correlation test was used to assess the correlation between PTEN and MMP-2, MMP-9 mRNA expressions. A value of P < 0.05 was considered significant.

RESULTS

Immunohistochemical analysis of PI3K/PTEN/AKT/mTOR pathway components in HCC and correlations between the different proteins

PTEN was localized in the cytoplasm of positive staining HCC cells (Fig. 1c). Positive staining of PTEN was observed in 64% (128/200) of HCC, whereas PTEN was detected in 100% (200/200) of adjacent non-cancerous liver tissues (P < 0.001). Loss of PTEN expression in HCC tissue was associated with higher α-fetoprotein (AFP) (P = 0.001), high tumor grade (P = 0.034), presence of intrahepatic metastasis (P = 0.022) and vascular invasion (P = 0.024), high tumor stage (P = 0.030) and high Ki-67 labeling index (P = 0.045). However, loss of PTEN expression was not correlated with age, sex, liver cirrhosis, tumor size, capsular invasion and viral infection (Table 1).

Table 1.  Correlations between expression of PI3K/PTEN/Akt/mTOR pathway proteins and clinicopathologic variables of patients with HCC
VariablenPTENPp-AKTPp-mTORP
+1+2+3+1+2+3+
Age   0.769    0.274    0.192
 ≤ 50953362 23253512 4325225 
 > 501053966 34262718 6219177 
Sex   0.428    0.437    0.485
 Male18364119 55455627 99393411 
 Female1789 2663 6551 
Etiology   0.399    0.423    0.429
 Noninfection421329 1112145 27870 
 Hepatitis B1455689 39384424 70343011 
 Hepatitis C and other13310 7141 8221 
Liver cirrhosis   0.547    0.436    0.148
 Absence792653 19222315 4414192 
 Presence1214675 38293915 61302010 
Tumor size (cm)   0.871    0.689    0.723
 ≤ 5582038 2014168 311593 
 > 51425290 37374622 7429309 
Serum AFP (µg/L)   0.001    0.438    0.045
 ≤ 20621250 2115206 341981 
 > 201386078 36364224 71253111 
Capsular invasion   0.182    0.336    0.325
 Absence1184771 28324018 5930245 
 Presence822557 29192212 4614157 
Tumor grade   0.034    0.031    0.013
 I29524 14942 18650 
 II1314883 37334021 7231235 
 III401921 69187 157117 
TNM stage   0.030    0.005    0.014
 I611546 2813155 42982 
 II561937 1212248 3012131 
 III833845 17262317 3323189 
Vascular invasion   0.024    0.045    0.037
 Absence1404397 48344018 7633274 
 Presence602931 9172212 2911128 
Intrahepatic metastasis   0.022    0.006    0.038
 Absence1273889 41274613 7724215 
 Presence733439 16241617 2820187 
Ki-67   0.045    0.025    < 0.001
 ≤ 50%1324191 46343616 9128103 
 > 50%683137 11172614 1416299 

P-AKT immunohistochemistry of HCC revealed a specific cytoplasmic staining (Fig. 1d–f). P-AKT was stained in 71.5% (143/200) of HCC, whereas 12.5% (25/200) of adjacent non-cancerous liver tissues exhibited weakly positive (P < 0.001). P-AKT expression showed a positive association with tumor grade (P = 0.031), presence of intrahepatic metastasis (P = 0.006) and vascular invasion (P = 0.045), TNM stage (P = 0.005) and high Ki-67 labeling index (P = 0.025). There were no further associations with other clinicopathological parameters (Table 1).

P-mTOR expression was detected in the cytoplasm of the malignant cells in 47.5% (95/200) of HCC (Fig. 1g–i). However, all adjacent non-cancerous liver tissues were negative for p-mTOR staining (P < 0.001). A significant association was also found between its expression and increased grade (P = 0.325), higher AFP (P = 0.045), and presence of intrahepatic metastasis (P = 0.038) and vascular invasion (P = 0.037), higher TNM stage (P = 0.014) and high Ki-67 labeling index (P < 0.01). No association was found with the other studied variables (Table 1).

In HCC tissues, PTEN protein expression was inversely correlated with p-AKT (r = −0.246, P < 0.01) and p-mTOR (r = −0.141, P = 0.046) protein expression; p-AKT expression was positively correlated with p-mTOR protein expression (r = 0.219, P = 0.002).

Immunohistochemical analysis of MMP-2 and MMP-9 in HCC and correlations with PI3K/PTEN/AKT/mTOR pathway

To investigate whether PI3K/PTEN/AKT/mTOR pathway is associated with MMP-2 and MMP-9 in invasion and metastasis of HCC, we analyzed immunoreactivity of MMP-2 and MMP-9 and their correlations with clinicopathological parameters and PI3K/PTEN/AKT/mTOR pathway. Both MMP-2 and MMP-9 were located in the cytoplasm. MMP-2 and MMP-9 protein expression were detected in 81% (162/200) and 86% (172/200) of HCC (Fig. 1j–l,m–o), 37% (74/200) and 42.5% (85/200) adjacent non-cancerous liver tissues, respectively (P < 0.01). MMP-2 and MMP-9 overexpressions in HCC tissues were correlated with liver cirrhosis, capsular invasion, presence of intrahepatic metastasis, vascular invasion and higher TNM stage. However, there were no marked correlations between their expressions and other clinicopathological features. Moreover, MMP-2 and MMP-9 protein expressions were not associated with high Ki-67 labeling index (Table 2).

Table 2.  Correlation between MMP-2 and MMP-9 expression and clinicopathologic variables of patients with HCC
VariablenMMP-2PMMP-9P
1+2+3+1+2+3+
Age     0.738    0.564
 ≤ 509523162630 12193529 
 > 5010526232927 21183729 
Sex     0.635    0.361
 Male18347355051 32356353 
 Female172456 1295 
Etiology     0.310    0.547
 Noninfection421110156 105189 
 Hepatitis B14536273745 21304945 
 Hepatitis C and other132236 2254 
Liver cirrhosis     0.034    0.018
 Absence7919232017 21122422 
 Presence12130163540 12254836 
Tumor size (cm)     0.065     
 ≤ 55814102311 1492213 
 > 514235293246 19285045 
Serum AFP (µg/L)     0.619    0.057
 ≤ 206218101519 10182113 
 > 2013831294038 23195145 
Capsular invasion     0.005    0.046
 Absence11823272642 26233633 
 Presence8226122915 7143625 
Tumor grade     0.102    0.315
 I2910487 36146 
 II13123293841 19254542 
 III4016699 1161310 
TNM stage     0.037    0.002
 I6124131212 18141712 
 II5612101816 672518 
 III8313162529 9143028 
Vascular invasion     0.042    0.013
 Absence14040283141 28295132 
 Presence609112416 582126 
Intrahepatic metastasis     0.042    0.045
 Absence12739253231 28234333 
 Presence7310142326 5142925 
Ki-67     0.38    0.699
 ≤ 50%13233303435 23245035 
 > 50%681692122 10132223 

There was significant correlation between MMP-9 and PTEN (r = −0.172, P = 0.015), p-AKT (r = 0.354, P < 0.001) and p-mTOR (r = 0.150, P = 0.034) protein expression. However, there was no marked correlation between MMP-2 and PTEN (r = −0.046, P = 0.515), p-AKT (r = 0.104, P = 0.142) or p-mTOR (r = 0.136, P = 0.055) protein expression.

PTEN, MMP-2 and MMP-9 mRNA expressions in HCC and correlations between different mRNA expressions

To validate the effect of PTEN, MMP-2 and MMP-9 in progression of HCC and their correlations, we detected their mRNA levels in HCC and paired adjacent non-tumor tissues by real time RT-PCR. The mRNA levels of PTEN were significantly decreased in HCC tissues compared with paired adjacent non-tumor tissues (P < 0.01) (Fig. 2a). In HCC tissue, PTEN mRNA expression significantly varied with TNM stage (P < 0.01, I vs. II; P < 0.01, I vs. III; P < 0.01, II vs. III) (Fig. 2b). The mRNA levels of PTEN were markedly decreased in metastatic HCC tissues compared with nonmetastatic tissues (Fig. 2c).

Figure 2.

Real-time RT-PCR analysis of PTEN mRNA expression in hepatocellular carcinoma (HCC) and paired adjacent non-cancerous liver tissues (a), HCC with varied TNM stages (b) and nonmetastatic and metastatic HCC (c). **P < 0.01.

However, MMP-2 and MMP-9 mRNA levels were significantly higher in HCC tissues than in paired adjacent non-tumor tissues (P < 0.001) (Figs 3a,4a). MMP-2 and MMP-9 mRNA expressions were significantly up-regulated in later TNM stages (P < 0.001) (Figs 3b,4b). MMP-2 and MMP-9 were markedly higher in metastatic HCC tissues than in nonmetastatic tissues (Figs 3c,4c).

Figure 3.

Real-time RT-PCR analysis of MMP-2 mRNA expression in hepatocellular carcinoma (HCC) and paired adjacent non-cancerous liver tissues (a), HCC with varied TNM stages (b) and nonmetastatic and metastatic HCC (c). **P < 0.01.

Figure 4.

Real-time RT-PCR analysis of MMP-9 mRNA expression in hepatocellular carcinoma (HCC) and paired adjacent non-cancerous liver tissues (a), HCC with varied TNM stages (b) and nonmetastatic and metastatic HCC (C). **P < 0.01.

There was an inverse correlation between PTEN and MMP-2 mRNA expressions (r = −0.30, P = 0.076), but this was not statistically significant. A significant inverse correlation between PTEN and MMP-9 mRNA expression was found in HCCs (r = −0.47, P = 0.004).

DISCUSSION

IN THE PRESENT study, we found that PI3K/PTEN/AKT/mTOR signaling pathway was activated in HCC. Activated PI3K/PTEN/AKT/mTOR signaling pathway was associated with frequent intrahepatic metastasis and vascular invasion, advanced tumor stage and high proliferation index in HCC. Moreover, activation of PI3K/PTEN/AKT/mTOR signaling pathway was correlated with MMP-9 overexpression.

PTEN is a tumor suppressor gene on human chromosome 10q23 and essential for regulating PI3K/AKT signaling pathway.19 PTEN activity is lost by mutations, deletions or promoter methylation in various human malignancies including glioblastoma, melanomas, endometrial carcinoma, prostate, breast, colon, and lung cancers.3,19,20 PTEN loss leads to activation of AKT kinases, which contributes to cell survival, growth, proliferation, angiogenesis, metabolism, migration and invasion.21–23 Introduction of wild-type PTEN gene into these cancer cells can inhibit cell growth, invasion and metastsis.24,25 Therefore, PTEN plays an important role in progression of tumor. Our results showed PTEN loss and activated AKT were associated with poor differentiation, intrahepatic metastasis, vascular invasion, later tumor stage and high proliferation index. In addition, p-AKT expression was inversely correlated with PTEN expression in HCC tissues. Furthermore, quantitative real-time RT-PCR demonstrated PTEN mRNA was significantly down-regulated in HCC tissues compared with adjacent non-tumor tissues; there was a correlation of the down-regulation of PTEN mRNA with tumor TNM stage and metastasis in HCC. These results suggest both PTEN loss and activated AKT are markers of more aggressive HCC and involved in proliferation, invasion and metastasis in HCC.

mTOR, a highly conserved serine/threonine protein kinase, plays an essential role in regulating cell growth and proliferation and its deregulation is associated with tumorigenesis.4 mTOR is one of the downstream targets of AKT. AKT phosphorylates mTOR and activates it. Activation of mTOR (p-mTOR) is an adverse prognostic factor of some cancers such as cervical cancer, biliary tract adenocarcinoma.26,27 In our study, the immunohistochemical results showed p-mTOR expression was correlated with increased grade, higher AFP, intrahepatic metastasis, vascular invasion and high proliferation index. Moreover, p-mTOR expression was inversely correlated with PTEN expression and positively with p-AKT expression. Thus, these results suggest that activated PI3K/PTEN/AKT/mTOR signaling pathway is involved in proliferation, invasion and metastasis in HCC.

Invasion and metastasis are fundamental properties of cancer. Invasion and metastasis of malignant tumor requires degradation of the extracellular matrix (ECM) and the basement membrane. Matrix metalloproteinases (MMPs) are a family of zinc-dependent endopeptidases which play an important role in the proteolytic destruction of extracellular matrix and basement membranes, thereby, they are essential for tumor invasion and metastasis.28,29 MMPs, especially MMP-2 and MMP-9 have been implicated in cancer invasion and metastasis.30 Thus, it is necessary to investigate whether involvement PI3K/PTEN/AKT/mTOR signaling pathway in invasion and metastasis is correlated with MMP-2 and MMP-9 in HCC. The results of immunohistochemical stainings showed both MMP-2 and MMP-9 overexpressions were associated with capsular invasion, vascular invasion, tumor stage and intrahepatic metastasis in HCC. Moreover, quantitative real-time RT-PCR demonstrated up-regulated MMP-2 and MMP-9 mRNA levels were correlated with later TNM stage and metastasis in HCC. These results confirmed MMP-2 and MMP-9 play important roles in HCC invasion and metastasis. Furthermore, we also found MMP-9 significantly correlated with PTEN, p-AKT, and p- mTOR, confirming some of the experimental data in vitro12,13. Thus, we can deduce activated PI3K/PTEN/AKT/mTOR signaling pathway promotes invasion and metastasis in HCC via up-regulating MMP-9. However, the exact mechanisms by which PI3K/PTEN/AKT/mTOR signaling pathway up-regulates MMP-9 need to be further elucidated.

Interestingly, the present study revealed MMP-2 was markedly correlated with invasion and metastasis in HCC, but there was not significant correlation between MMP-2 and PI3K/PTEN/AKT/mTOR signaling pathway. Thus, this result indicates MMP-2 is regulated mainly by other signaling in HCC.

In conclusion, the present study demonstrates PI3K/PTEN/AKT/mTOR signaling pathway is activated in HCC. Furthermore, our results indicate that activated PI3K/PTEN/AKT/mTOR signaling pathway is involved in invasion and metastasis through up-regulating MMP-9 in HCC. Therefore, PI3K/PTEN/AKT/mTOR pathway could serve as a diagnostic and predictive marker of invasion and metastasis and a therapeutical target in HCC. This study further explains the molecular mechanism underlying invasion and metastasis of HCC.

ACKNOWLEDGEMENTS

THIS WORK WAS supported by a grant from the National Natural Science Foundation of China (30672053). We thank the Surgical Laboratory of the First Affiliated Hospital and Key Laboratory of Tropical Disease Control Research of Ministry of Education, Sun Yat-Sen University for providing experimental instruments and equipment.

Ancillary