Co-expression of LASS2 and TGF-β1 predicts poor prognosis in hepatocellular carcinoma

Longevity assurance homolog 2 of yeast LAG1 (LASS2) has been reported to act as an important tumor suppressor in the development of human cancers. However, little is known about the prognostic value of LASS2 in hepatocellular carcinoma (HCC) . In the present study, we analyzed correlation between LASS2 and TGF-β1 levels, and evaluated their prognostic values in HCC patients. We first analyzed the expression of LASS2 and TGF-β1 in two independent cohorts (test cohort: 184 HCC patients; validation cohort: 118 HCC patients) using immunohistochemistry (IHC). Kaplan-Meier survival and Cox regression analyses were executed to evaluate the prognosis of HCC. The results of IHC analysis revealed a positive correlation between the expression of LASS2 and TGF-β1. HCC Patients with low expression of LASS2 and TGF-β1 had shorter overall survival (OS) and time to recurrence (TTR) than patients with high expression of LASS2 and TGF-β1. Furthermore, combination of LASS2 and TGF-β1 was an independent and significant risk factor for OS and TTR. In conclusion, low expression of LASS2 and TGF-β1 contributes to the aggressiveness and poor prognosis of HCC, and may represent a novel prognostic biomarker for HCC patients.


Results
The relationship between LASS2 and TGF-β1 in HCC patients. We enrolled a test cohort consisting of 184 HCC patients and a validation cohort of 118 HCC patients, who had undergone curative resection between 2008 and 2013. Using immunohistochemical staining, we examined expression of LASS2 and TGF-β 1 in the two independent cohorts, and analyzed their coexpression. The observed LASS2 staining pattern in HCC tissues was both plasma membranous and cytoplasmic. The level and extent of staining varied from weak focal to extensive strong expression. For TGF-β 1 staining, TGF-β 1 was primarily localized in the cytoplasm of tumor cells, which were producing TGF-β 1 (Fig. 1A,C). Based on analysis of integrated optical density (IOD) value, we found that a significant positive correlation between expression of LASS2 and TGF-β 1 was observed in the test cohort (Pearson's correlation, r = 0.512, n = 184, p < 0.05, Fig. 1B). Similarly, the positive correlation between LASS2 and TGF-β 1 was also observed in the validation cohort (Pearson's correlation, r = 0.496, n = 118, p < 0.05, Fig. 1D).
Low expression of LASS2 and TGF-β1 significantly correlated with progression and poor prognosis of HCC in the test cohort. Among the 184 HCC patients of test cohort, 64 (34.8%) and 100 (54.3%) cases had high expression of LASS2 and TGF-β 1, respectively. We performed further analyses to determine the clinicopathological significance of LASS2 and TGF-β 1 in HCC. The expression level of LASS2 was negatively correlated with the tumor size, tumor differentiation, and TNM stage (Table 1). In addition, we also found that TGF-β 1 expression was correlated with age and TNM stage (Table 1).

Univariate and multivariate analyses of prognostic variables for HCC patients in the test cohort.
To evaluate whether the expression levels of LASS2 and TGF-β 1 were independent prognostic value for HCC patients in the test cohort, univariate and multivariate analyses using a Cox regression model were applied. As shown in Table 2, LASS2 level, TGF-β 1 level, and the LASS2/TGF-β 1 combination was responsible for the OS and TTR of HCC patients. In addition to the above factors, serum AFP, TNM stage, child-pugh, tumor size, tumor number, microvascular invasion, and tumor differentiation were also correlated with OS and/or TTR (Table 2).  Factors showed significance in univariate analysis were then subjected to multivariate Cox proportional hazards analysis. Despite neither LASS2 nor TGF-β 1 alone were independent prognostic markers for OS and TTR, the LASS2/TGF-β 1 combination was an independent prognostic marker for both OS and TTR in the test cohort. These results indicated that the LASS2/TGF-β 1 combination had a potent prognostic value compared with LASS2 or TGF-β 1 alone.  (Fig. 5). LASS2/TGF-β 1 combination was associated with OS (p = 0.001) and TTR (p = 0.001) by univariate analysis ( Table 2) and was an independent predictor for both OS (p = 0.009) and TTR (p = 0.025) by multivariate analysis (Table 3).

Discussion
Hepatocellular carcinoma is one of the most aggressive cancers featured with high rate of metastasis and mortality. Although several molecular biomarkers have been reported to have clinical significance for predicting HCC prognosis [29][30][31] , clinical and molecular features to be used as prognostic parameters in clinical practice are still substantially needed. Critical oncogenes or tumor suppressors involved in different stages of growth, cell cycle progression, disease initiation and responses to environmental stimuli provide important clues to identify prognostic biomarkers of cancer. LASS2 is the gene identified from a human liver cDNA library by our laboratory and interacts with proteolipid subunit of vacuolar H + ATPase (VPL) 7 . A growing amount of research has reported that LASS2 mainly acts as a tumor suppressor and is closely associated with a variety of tumor progression, including   prostate 9 , breast 10 , bladder 11 and liver cancer 12 . LASS2 can induce apoptosis of prostate cancer, inhibit growth and invasion of breast cancer cell in vitro through interacting with vacuolar ATPase 32,33 . Our previous studies have shown that low expression of LASS2 is associated with poor prognosis in patients with breast cancer 10 , and deletion of LASS2 is associated with a high risk of spontaneous or DEN-induced HCC in hepatocyte-specific LASS2-KO mice models 13,14 . Recent research also revealed that LASS2 was the target of miR-9, and involved in the chemoresistance of bladder cancer 11 . In addition, it has been reported that decreased expression of LASS2 is associated with worse prognosis in meningiomas 34 . However, the prognostic significance of LASS2 expression in HCC is still unclear. In current study, we found that the expression of LASS2 level was negatively correlated with the tumor size, tumor differentiation, and TNM stage. Moreover, the Kaplan-Meier survival analysis showed that the OS and TTR of HCC patients with low LASS2 expression were shorter than those with high LASS2 expression. These results suggested that LASS2 may act as a critical tumor suppressor in HCC progression.
Our previous study showed that knockout of LASS2 significantly upregulated the expression of TGF-β 1 in DEN-induced liver tumourigenesis of mice 13 . This clue strongly prompted us to further study the correlation between LASS2 and TGF-β 1 in HCC patients. Interestingly, we were surprised to find that a significant positive correlation between expression of LASS2 and TGF-β 1 was observed in HCC patients. This discrepancy may be mainly explained by the following points: 1) In the process of cancer development, TGF-β 1 plays biphasic functions in tumor tumorigenesis, having a growth inhibitory effect at early stages, but at later stages enhancing the malignant conversion. This contradictory results may be due to the function switch of TGF-β 1 in different stages of carcinogenesis, which is still largely unclear since now. 2) Although the animal models of DEN-induced HCC were used as a tool to test preventive treatments for HCC, the mice model of DEN-induced hepatocarcinogenesis and the actual process of human liver cancer are still different. For example, our study was performed mainly on hepatitis B virus (HBV)-related HCC, which is a dominant etiology of Asian HCC patients. According to the statistics, 157 of 184 (85.3%) HCC patients had positve hepatitis B surface antigen (HBsAg) in the test cohorts of 3) The correlation between LASS2 and TGF-β 1 was firstly observed in a hepatocyte-specific LASS2-KO transgenic mouse model, which was generated using the Cre/LoxP system for site-specific excisional DNA recombination. The particular mouse model also might lead to its inconsistent with clinical results. However, the underlying mechanism remains to be further investigated.
TGF-β 1 particularly achieves its tumor suppressive effect through the inhibition of proliferation and the induction of apoptosis or senescense in the early stage-tumors 18,19 . Consistent with a tumor-suppressor role for TGF-β 1, analysis of clinical samples and tumor-derived cell lines suggested that it is a common occurrence that reduction in epithelial responsiveness to TGF-β 1 during carcinogenic progression in many tissues 35 . It was reported that TGF-β 1 can induce p53-independent and p16-independent, and reactive oxygen species (ROS)-dependent senescence arrest in well-differentiated HCC cells 18 . In our study, most HCC patients of our study were well and moderate differentiation. According to the statistics, 182 of 184 (98.9%) and 114 of 114 (100%) HCC patients were well and moderate differentiation in the test cohort and validation cohort, respectively. The Kaplan-Meier survival analysis also showed that the OS and TTR of HCC patients with low TGF-β 1 expression were shorter than those with high TGF-β 1 expression.
In the present study, we provided the evidence that high level of LASS2/TGF-β 1 expression predicted a favorable OS and TTR rate for HCC patients. The 1-, 3-and 5-year OS rates of HCC patients with high level of LASS2 or TGF-β 1 expression were remarkably higher than those of HCC patients with low levels of LASS2 or TGF-β 1 expression, respectively. Accordingly, the 1-, 3-and 5-year TTR rates of HCC patients with high level of LASS2 or TGF-β 1 expression were significantly lower than those of HCC patients with low levels of LASS2 or TGF-β 1 expression, respectively. As combination of multiple markers might yield more information for predicting clinical outcome of HCC patients, combination of LASS2 and TGF-β 1 expression were therefore used as a predictor of clinical outcome. In both test cohort and validation cohort, HCC patients with both high expression of LASS2 and TGF-β 1 had the most favorable OS and TTR rates, whereas those patients with both low expression of LASS2 and TGF-β 1 had the poorest OS and TTR rates. Besides, we used univariate and multivariate analyses to further study the prognostic value of LASS2 and TGF-β 1, the results showed that combination of LASS2 and TGF-β 1 was significant and independent prognostic factor for HCC patients, but either LASS2 or TGF-β 1 was not.
In conclusion, our study demonstrated that high levels of LASS2 expression and TGF-β 1 expression have a positive relationship with prognosis of HCC patients. The combination of LASS2 and TGF-β 1 was more sensitive than either of them alone with respect to OS and TTR, and could be used as a new independent prognostic marker of HCC patients.

Materials and Methods
Patients and tissue samples. This  were used to verify tumor recurrence. No patient received either radiotherapy or chemotherapy before the surgery, and no other cancers co-occurrence. The overall survival (OS) was defined as the length of time between the surgery and death or the last follow-up examination. The time to recurrence (TTR) was calculated from the date of tumor resection until the detection of tumor recurrence, death or the last observation. Tumor stage was defined according to the American Joint Committee on Cancer (AJCC 2010, 7th edition) TNM staging system 36 . The grade of tumor differentiation was assigned by the Edmondson-Steiner grading system. Micrometastases were defined as tumors adjacent to the border of the main tumor that was only observed under the microscope 37 .
For the use of clinical materials for research purposes, prior patients' consents and approval were obtained from the Ethics Committee of Renji Hospital, Shanghai Jiao Tong University School of Medicine and EHBH of the Second Military Medical University. Written informed consent was obtained from all subjects for publication of this study. All experiments were performed in accordance with approved guidelines of Shanghai Jiao Tong University School of Medicine.
Immunohistochemistry and Scoring. HCC tumor tissues were resected from the patients, and then were fixed with 10% formalin and embedded in paraffin before prepared for tissue microarray. Briefly, 1-mm cores were taken from intratumoral tissue and serial sections (4-μ m thick) were placed on slides coated with 3-aminopropyltriethoxysilane. The immunohistochemistry analysis was conducted as described previously 30 . Immunostaining scores were independently evaluated by two pathologists who were blinded to the clinical outcome. The sections were incubated with primary goat anti-human LASS2 polyclonal antibody (clone sc-65102, 1:80, Santa Cruz Biotechnology) 38 and rabbit anti-human TGF-β 1 polyclonal antibody (18978-1-AP, 1:100, Proteintech) 39 at 4 °C overnight, and then secondary antibody within 30 minutes. The average sum of integrated optical density (IOD) of each sample was calculated using ImageJ software.
Statistical Analysis. Differences among variables were assessed by χ 2 analysis or two-tailed Student t test.
Kaplan-Meier analysis was used to assess survival. Log-rank tests were used to compare survival of patients between subgroups. Multivariate analyses were performed by multivariate Cox proportional hazard regression model. Data were presented as mean ± SEM. Differences were considered to be statistically significant for p < 0.05.