Upregulated Seizure-Related 6 Homolog-Like 2 Is a Prognostic Predictor of Hepatocellular Carcinoma

Seizure-related 6 homolog-like 2 (SEZ6L2), which is localized on the cell surface, has been found to be associated with tumor angiogenesis and lung cancer progression. However, the role of SEZ6L2 in hepatocellular carcinoma (HCC) is still unclear. We obtained data from The Cancer Genome Atlas (TCGA) and the Gene Expression Omnibus (GEO) to investigate SEZ6L2 expression and regulation in HCC. Then, HCC tissue samples were collected to verify SEZ6L2 by quantitative real-time polymerase chain reaction (qRT-PCR) and immunohistochemical staining (IHC). Patient information was collected for survival and prognosis analysis. qRT-PCR, IHC, and bioinformatics analysis showed that the SEZ6L2 protein was highly expressed in HCC samples. Clinical data showed that high SEZ6L2 protein expression was correlated with tumor-node-metastasis (TNM) stages (P = 0.046), tumor number (P = 0.016), and tumor size (P = 0.029). Meanwhile, SEZ6L2 overexpression was closely associated with poor overall survival and disease-free survival in HCC patients. Moreover, SEZ6L2 is an independent prognostic predictor for the survival of HCC patients. This study suggests a significant correlation between SEZ6L2 and HCC, which means that SEZ6L2 may potentially serve as a useful prognostic biomarker for HCC patients.


Introduction
Hepatocellular carcinoma (HCC), a malignant cancer, is the most common primary liver cancer worldwide [1]. It is the third most common cause of cancer-related death and has been ranked second in cancer mortality in China since the 1990s [2,3]. The major therapeutic methods used to treat HCC are hepatic resection, liver transplantation, chemotherapy, radiotherapy, and targeted agents [4]. Despite the apparent therapeutic effects of these treatment regimens, the rapid progression of the disease still leads to high mortality in patients [5,6]. In recent studies, new immunotherapy and targeted therapies have achieved good results [7]. Therefore, it is rewarding to find effective gene markers for the early diagnosis and treatment of HCC.
Seizure-related 6 homolog-like 2 (SEZ6L2), which encodes a seizure-related protein localized on the cell surface, is located in the region of chromosome 16p11.2 and is predicted to be a candidate gene for autism spectrum disorders [8,9]. This gene has broad expression in the brain, stomach, and testis and in the adrenal and prostate glands, while it is scarce in other normal tissues (from gene databases). Previous studies have shown that SEZ6L2 regulates neurogenesis and neural differentiation and modulates α-amino-3hydroxy-5-methyl-4-isoxazole propionic acid-(AMPA-) adducin (ADD) signal transduction through control of the phosphorylation of adducin [10,11]. Moreover, a recent study found that AMPA-ADD is overexpressed in lung cancers and can be regarded as a significant prognostic biomarker for lung cancer [12]. However, the expression and clinicopathological significance of SEZ6L2 in HCC remains unclear. Thus, this study aimed to detect the expression of SEZ6L2 in HCC and determine the correlation between clinical value and the SEZ6L2 expression level in HCC.

Patient Information and Clinical Tissue
Samples. All tissue samples and clinical materials for this research, including 16 pairs of fresh-frozen HCC tissues and corresponding adjacent normal liver tissues (ANLT) and 95 formalin-fixed, paraffin-embedded HCC tissue samples, were from HCC patients after hepatic resection without radiotherapy and chemotherapy before surgery at the First Affiliated Hospital of Sun Yat-Sen University during the period from May 2012 to December 2014 [13]. All patients were followed up until June 2018. Information on the clinical characteristics of patients is shown in Table 1. All diagnoses were confirmed by pathology. This study was approved by the Ethics Committees of the First Affiliated Hospital of Sun Yat-Sen University in accordance with the 1964 Declaration of Helsinki (including later amendments) and relevant ethical standards. The clinical samples of patients were collected after obtaining written informed consent.

Quantitative Real-Time Polymerase Chain Reaction.
A total of 16 pairs of RNA from HCC and ANLT were isolated from frozen tissues using RNA plus reagent (TaKaRa, Japan) according to the manufacturer's instructions. Complementary DNA was synthesized by a PrimerScript™ RT Reagent (TaKaRa, Japan) for mRNA according to the manufacturer's instruction. The SEZ6L2 expression level was determined by qRT-PCR using a Power SYBR Green PCR Master Mix (Applied Biosystems). The sequence of genetic specific primers of SEZ6L2 was as follows: SEZ6L2 (forward: 5′-TT CACTGTGATTCGGGCTACC-3′, reverse: 5′-GGGGTTT CACCGTTCCAGG-3 ′ ) and GADPH (forward: TGTGGG CATCAATGGATTTGG, reverse: ACACCATGTATTCC GGGTCAAT) (Servicebio Technology, Wuhan, China). GADPH was used to standardize data as an endogenous control. PCR was performed as follows: 95°C for 30 seconds and then 40 cycles for 5 seconds in 95°C and 30 seconds in 60°C.

Immunohistochemical Staining (IHC).
Ninety-five samples of HCC tissue, which were paraffin-embedded, were sectioned in 4 μm intervals. After deparaffinization, hydration, and blocking, the HCC tissue was incubated with a primary anti-SEZ6L2 antibody (Abcam, ab104194) overnight at 4°C. Then, the staining of HCC was observed under a microscope to evaluate the expression of the target SEZ6L2 protein. The results were analyzed and scored according to the percentage of positive cells. Staining intensity was scored as 0, 1, 2, and 3 for absent, weak, moderate, and strong, respectively. The percentage of positive results was defined as 0, 1, 2, and 3 for no positive results, less than 20%, 20% to 50%, and more than 50%, respectively. Each section was analyzed for a final score by adding the intensity of immunoreaction and the percentage of stained tumor cells and resulted in scores of 1 to 5. The scores of 0 and 1 were thought to have low expression, while the scores of 3, 4, and 5 were thought to have high expression. The microarray data was downloaded from the Gene Expression Omnibus (GEO) datasets (GSE36376, GSE45436, and GSE6764) (http://www.ncbi.nlm.nih.gov/geo). The results were analyzed using Microsoft Excel 15.39, Cytoscape 3.6.1, R 3.6.1, and GraphPad Prism 7.0 software.

Statistical
Analysis. Data were analyzed by SPSS version 20.0 (IBM, USA). The chi-squared test was used to demonstrate the relationship between SEZ6L2 expression and other tumor characteristics. Kaplan-Meier and Cox proportional hazards were used to assess the effect of various prognostic factors on the overall survival (OS) and disease-free survival (DFS) of HCC patients. Differences were considered statistically significant when two tailed P < 0:05.

SEZ6L2 Is Overexpressed in HCC Tissue by Transcription
Factors such as GCNF. We downloaded the mRNA sequence or microarray data sets from TCGA and GEO, respectively, to examine the expression of SEZ6L2 in HCC tissue. These data showed that SEZ6L2 was obviously elevated in HCC tissue compared with normal tissues (Figures 1(a)-1(d)). Then, we analyzed the SEZ6L2 expression level in 16 pairs of freshfrozen HCC tissues and corresponding ANLT using qRT-PCR. The results revealed that SEZ6L2 was upregulated in HCC tissue compared to nontumor sites (Figure 1(e)). Both  In order to explore the reasons leading to the high expression of SEZ6L2 in HCC tissues, we analyzed 407 tumor samples versus 58 normal samples in HCC from the TCGA database. Then, we obtained the differentially expressed genes of these samples and made transcription regulatory network. Transcription factors GCNF, E47, MYOD, ZIC3, and FREAC3 were found to be associated with the upregulation of SEZ6L2 (Figure 2).

SEZ6L2 Protein Expression Level Is Relevant in
Clinicopathological Parameters of HCC. Ninety-five formalin-fixed, paraffin-embedded HCC tissue samples were used to detect SEZ6L2 protein expression using IHC in the search for the association between the SEZ6L2 protein expression and the clinicopathological parameters of HCC. Immunohistochemistry revealed that 31/95 cases (32.6%) had high intensity (>50% of tumor cells) of stain, 20/95 (21.1%) had moderate stain (20-50% of tumor cells), 29/95 (30.5%) had low (<20% of tumor cells) stain, and 15/95 (15.8%) of cases were negative (Figure 3). Table 1 shows the relationship between SEZ6L2 protein expression and the clinicopathological parameters of HCC (low SEZ6L2 expression: negative and low positive group in IHC; high SEZ6L2 expression: high and moderate positive group in IHC). High SEZ6L2 protein expression levels were significantly correlated with tumor-node-metastasis stages (P = 0:046), tumor number (P = 0:016), and tumor size (P = 0:029), while there was no significant correlation with gender (P = 0:186), age (P = 0:612), or AFP (P = 0:984).     (Table 3). Therefore, high SEZ6L2 expression was correlated with OS as well as DFS and predicted poor survival in HCC patients.

Discussion
The SEZ6L2 gene and the correlation between its expression level with its clinical value in HCC patients are still poorly understood. To our knowledge, this is the first report about the relationship between SEZ6L2 expression and clinical prognostic survival in HCC patients. This study shows that SEZ6L2 is overexpressed in HCC tissue, and its expression level is relevant in clinicopathological parameters of HCC. Moreover, high SEZ6L2 expression is associated with poor prognosis in HCC patients.
SEZ6L2 is located in chromosome 16p11.2 and encodes a seizure-related protein that is localized on the cell surface. Previous studies have found that a coding variant in SEZ6L2 has a significant association with autism spectrum disorders [9,14]. Other research has demonstrated that SEZ6L2, a type I transmembrane protein and has Sushi/SCR and CUB domain outside the cellular surface, is highly expressed in the hippocampus and cerebellar cortex, and ataxia occurs in mice lacking members of the SEZ6L2 protein family [10]. Therefore, it can be affected by various drug delivery mechanisms. Many specific drugs such as Afatinib and Rituxan targets on cell surface proteins and have achieved the desired effect [15,16]. Moreover, Hald and Galbo [17] found that SEZ6L2 has extracellular epitopes, which are regarded as useful markers for the detection of pancreatic cell types at specific developmental stages by specific antibodies, and the Sushi and CUB domains are also involved in receptor-ligand interactions and cell adhesion. On the other hand, our results indicate that the SEZ6L2 protein is highly expressed in HCC tumor tissues, suggesting that it may be a potential therapeutic target. However, the role of SEZ6L2 in HCC and its clinical and pathological significance is still unanswered.
In our study, the high expression of the SEZ6L2 gene in HCC patients indicates that these patients may have a poor prognosis. Then, we searched for RNA sequencing data in TCGA (a public repository of comprehensive cancer genome maps) and obtained microarray data from GEO (a public repository that has high-throughput microarrays and nextgeneration sequential functional genome datasets) [18,19]. The results revealed that SEZ6L2 is upregulated in HCC tissue. And then, we investigated fresh-frozen HCC tissues   Disease Markers using qRT-PCR compared with corresponding ANLT, which showed a similar consequence as that mentioned above. According to previous research, SEZ6L2 has been shown to be regulated by transcription factors, such as STAT3, which upregulate the expression of the vascular endothelial growth factor and promote tumor angiogenesis and cancer progression [20,21]. So we did an analysis of the transcription regulatory network; our results indicated five transcription factors could upregulate the SEZ6L2 gene, namely, GCNF, E47, MYOD, ZIC3, and FREAC3; no evidence shows that STAT3 can regulate the expression of SEZ6L2 in HCC. However, the specific regulation mechanism of these transcription factors still needs further study.
To verify our results at the protein level, 95 formalinfixed, paraffin-embedded HCC tissue samples were used to detect SEZ6L2 expression using IHC, which demonstrated that SEZ6L2 had high expression in HCC tissue. Finally, clinical data were collected for Kaplan-Meier analysis, as well as Cox proportional-hazard analysis, and we found that HCC patients with higher SEZ6L2 expression had lower OS and DFS, which means that SEZ6L2 can act as a significant independent prognostic predictor for HCC patients.
This is the first article to prove that SEZ6L2 is upregulated in HCC and can be a significant predictor for survival in HCC patients. A limitation of our research is that we did not study the molecular mechanisms of SEZ6L2 function in HCC. Therefore, in the future, we intend to conduct more in-depth functional and mechanistic studies on mice and cell lines. In conclusion, our study demonstrates that SEZ6L2 is overexpressed in HCC tissue and high SEZ6L2 expression is related to the poor prognosis of HCC. SEZ6L2 can be considered a significant independent prognostic marker and potential therapeutic target for HCC.