Introduction

Esophageal cancer is one of the most prevalent types of malignant tumor in both developed and developing countries according to the latest statistics on trends in cancer incidence and mortality rates. Approximately 73 % of all new cases occurred in countries with low or medium levels of human development; moreover, 49 % of all new cases occurred in China [1]. There are two main types of esophageal cancer: esophageal squamous cell carcinoma (ESCC) and esophageal adenocarcinoma (EA). Both types are more common in men than in women. In general, ESCC is the most common histologic type in Central and East Asia, accounting for 90 % of cases, whereas EA is more prevalent in Western countries [2]. Surgical resection is the main curative treatment option for patients with early-stage disease, whereas the latest chemotherapy agents and chemotherapy regimens provide a better survival benefit for patients with late-stage cancer. However, existing therapies are insufficient for complete tumor eradication. Although extensive research has been conducted on the pathogenesis and progression of ESCC, it is still difficult to predict the clinical outcome of patients with ESCC. Therefore, new prognostic markers and therapeutic targets are urgently needed for better patient classification and the development of individualized therapy regimens.

Sirtuins are promising proteins belonging to the family of class III NAD+-dependent histone deacetylases, which have been implicated in multiple functions including gene expression regulation, cell aging, and tumorigenesis through epigenetic modifications [3]. The mammalian sirtuin family includes seven proteins, Sirtuin-1–7 (SIRT1–7), that exhibit different cellular localization, activity, and function. Aberrant expression of SIRT1 has been noted in many solid tumors [4–7], where it is found to accelerate tumor progression. For example, SIRT1 can silence some tumor suppressors, thus promoting cancer development [8–10]. In addition, SIRT1 serves as a driver of tumorigenesis by activating the PTEN/PI3K/AKT signaling pathway, leading to activation of cell proliferation and inhibition of apoptosis [11]. However, contradictory reports have revealed that SIRT1 does not always function in carcinogenesis. Increasing evidence suggests decreased expression of SIRT1 in bladder and ovarian cancers [12, 13], favoring a role for SIRT1 as a tumor suppressor. Indeed, SIRT1 has been reported to suppress the growth of colon cancer [14], breast cancer [15], and pancreatic cancer [16]. In light of these findings, we speculate that SIRT1 plays a dual role in tumor development, acting as a tumor suppressor or promoter. Despite abundant literature on this protein, surprisingly little is known about the expression and prognostic significance of SIRT1 in human ESCC.

In this study, we performed immunohistochemical (IHC) analysis of the expression of SIRT1 protein in ESCC cell lines and tissue samples and investigated the correlation of SIRT1 protein expression with clinical features and survival outcomes of ESCC patients. We provide the first report of increased expression of SIRT1 protein in ESCC cell lines and tissue samples. Our observations implicate SIRT1 as a predictive biomarker for diagnosis and prognosis in ESCC. Furthermore, our research provides a rationale for clinical exploration of drugs that selectively inhibit SIRT1 as a promising targeted therapy for ESCC.

Materials and methods

Patients and tissue samples

Between July 2006 and October 2008, 86 pairs of primary ESCC and adjacent normal tissue samples were obtained from patients who had undergone surgery in the Cardiothoracic Surgery Department of Jinling Hospital (Jiangsu, China). The deadline for follow-up was September 2013 and the end point of our survival analysis was overall survival. None of the patients had received radiation therapy or chemotherapy before surgery. All patients enrolled in our study received standard postoperative adjuvant therapy according to the NCCN guidelines for esophageal cancer. Clinical and pathologic characteristics, including age, gender, operation time, tumor stage, and related pathologic data, were collected retrospectively from the patient records. The research protocol was reviewed and approved by the Ethical Committee and Institutional Review Board of Jinling Hospital, and written informed consent was obtained from each patient included in the study.

Cell lines

A human normal esophageal epithelial cell line (HEEC) and four human ESCC cell lines (Kyse30, Kyse70, Eca109, Ec706) were purchased from the Cell Bank of Shanghai Institute of Cell Biology (Chinese Academy of Medical Sciences, Shanghai, China). Kyse30, Kyse70, Eca109, and HEEC cell lines were cultured in RPMI-1640 medium (Gibco, USA) supplemented with 10 % fetal bovine serum (Gibco) and 1 % penicillin-streptomycin solution. The Ec9706 cell line was expanded in Dulbecco’s modified Eagle’s medium (Gibco) with 10 % fetal bovine serum and 1 % penicillin-streptomycin. Cells were cultured at 37 °C in a 5 % CO2-humidified incubator.

Immunohistochemistry

Primary ESCC tissues and adjacent normal tissues obtained after surgery were subjected to IHC analysis according to the routine protocol. In brief, formalin-fixed and paraffin-embedded specimens were sectioned at 3- to 4-μm thickness for immunohistochemistry. The samples were deparaffinized with xylene and rehydrated through a series of decreasing concentrations of ethanol to water. A high-temperature antigen retrieval method was performed with citrate buffer solution (Maixin Bio, China), and the slides were immersed in 100 μl of 3 % hydrogen peroxide for 10 min at 37 °C to block endogenous peroxidase activity. After washing with phosphate-buffered saline (PBS), the sections were incubated with 5 % bovine serum albumin (BSA; Sigma-Aldrich, USA) for 30 min, followed by incubation with a monoclonal mouse anti-SIRT1 antibody (1:200, Abcam, UK) at 4 °C overnight. After washing with PBS, the sections were incubated with secondary antibody for 30 min at 37 °C. The slides were stained with diaminobenzidine as the color reagent and hematoxylin as a counterstain for nuclei. PBS was used as a negative control for the staining reactions. All slides were dehydrated in increasing concentrations of ethanol and xylene. Finally, all sections were mounted with neutral gum.

The IHC sections were scored by three pathologists independently to ensure interobserver agreement. Each section was scored according to the intensity and percentage of positive cells. Staining intensity was scored as follows: 0 (negative), 1 (weakly positive), 2 (moderately positive), and 3 (strongly positive). The percentage of SIRT1-positive cells was also graded according to four categories: 1 point for 0–25 % positive cells, 2 for 26–50 % positive cells, 3 for 51–75 % positive cells, and 4 for more than 75 % positive cells. Overall scores ≤3 were defined as negative; scores >3 were defined as positive.

Real-time quantitative RT-PCR

Total RNA was extracted from cultured cell lines and tissue samples using TRIzol reagent (Takara, Japan) following the manufacturer’s protocol. Subsequent synthesis of complementary DNA (cDNA) was carried out in 5× All-In-One RT MasterMix (ABM, Canada) with incubation at 25 °C for 10 min followed by 42 °C for 15 min. Real-time quantitative reverse transcription PCR (RT-PCR) was performed with a StepOne Sequence Detection System (Applied Biosystems, USA) with cycling conditions of 10 min at 95 °C followed by 35 cycles of 15 s at 95 °C and 1 min at 60 °C. Each reaction was carried out in a total volume of 20 μl consisting of 10 μl of EvaGreen 2× qPCR MasterMix (ABM, Canada), a 300 nM concentration of each primer (Takara), 10–100 ng of cDNA, and nuclease-free H2O. The PCR primers used for SIRT1 were 5′-CCCAGAACATAGACACGCTGGA-3′ and 5′-ATCAGCTGGGCACCTAGGACA-3′. The housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was amplified as an endogenous control for RNA normalization using the primers 5′-GCACCGTCAAGGCTGAGAAC-3′ and 5′-TGGTGAAGACGCCAGTGGA-3′. Expression data were collected and analyzed with ABI StepOne software (Applied Biosystems).

Western blot analysis

The level of protein expression was analyzed by Western blotting. In brief, tumor and adjacent normal tissues and cells were homogenized in ice-cold lysis buffer (KeyGEN Whole Cell Lysis Assay, China) according to the manufacturer’s instructions, and total protein concentrations were measured using the KeyGEN BCA Protein Quantitation Assay. The lysates were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred onto polyvinylidene fluoride membranes. Membranes were blocked by 5 % non-fat milk in Tris-buffered saline with Tween-20 (TBST) and incubated with primary antibody against SIRT1 or antibody against GAPDH as a control (Cell Signaling Technology, USA) at 4 °C overnight. The membranes were incubated with horseradish peroxidase-labeled secondary antibody for 120 min at 37 °C and washed with TBST. Band signals were detected with an enhanced chemiluminescence system (Thermo Scientific, USA) and exposure to Kodak X-ray film. Quantification of Western blots was performed by multiplying the area and intensity of each band using ImageJ software.

Statistical analysis

Data analysis was performed using SPSS software version 18.0 (SPSS Inc., USA). Survival time was defined from the date of surgery to the latest follow-up or the date of death. Student’s t test was used to analyze real-time quantitative RT-PCR data from cell lines and tissues. Chi-square test and Fisher’s exact test were used to analyze the correlation between SIRT1 expression and clinical parameters of patients. Kaplan-Meier survival curves were conducted with log-rank tests to determine any significant relationships between clinical features and outcomes. Cox proportional hazards regression model was used for multivariate analysis. The results are expressed as mean ± SD of three independent experiments. Results were considered statistically significant for P < 0.05.

Results

SIRT1 expression in ESCC tissues and cell lines

First, we detected SIRT1 expression in four ESCC cell lines (Kyse30, Kyse70, Eca109, and Ec9706) and a human normal esophageal epithelial cell line (HEEC). As shown in Fig. 1a, b, the relative expression of SIRT1 mRNA and protein was significantly higher in most ESCC cell lines than in normal epithelial cells. Furthermore, we investigated SIRT1 expression in 30 paired samples of ESCC tissues and adjacent normal tissues. Real-time quantitative RT-PCR analysis showed that the expression levels of SIRT1 mRNA in tumor tissues were significantly upregulated compared with matched normal specimens (Fig. 1c). Consistent with these data, SIRT1 protein expression in tumor tissues was also notably increased, as shown in Fig. 1d. These results indicated aberrant expression of SIRT1 in ESCC tissue and a potential contribution of SIRT1 to ESCC development.

Fig. 1
figure 1

Relative expression levels of SIRT1 in ESCC cell lines and tissues. a Real-time quantitative RT-PCR and b Western blot analysis were performed to investigate the expression of SIRT1 in different ESCC cell lines and human normal esophageal epithelial cell line. Most ESCC cell lines exhibited higher expression levels of SIRT1 compared with the normal esophageal epithelial cell line. c Real-time quantitative RT-PCR analysis of the mRNA levels of SIRT1 in 30 pairs of ESCC tissues and matched adjacent normal tissues. The results are presented as −ΔΔCt changes in tumor tissues normalized to adjacent normal tissues. SIRT1 mRNA was significantly overexpressed in ESCC tissues compared with normal tissues. d Western blot analysis of SIRT1 protein expression in paired ESCC tissues and adjacent normal tissues. Stronger expression levels of SIRT1 were observed in most tumor tissues compared with matched adjacent normal tissue as controls. Representative data are shown. Results are expressed as mean ± SD of three independent experiments. *P < 0.05, **P < 0.01

Full size image

Clinicopathologic features of patients

After screening the medical records of all patients, 86 cases (22 females and 64 males) were enrolled in our study. Among the 86 patients, 58 patients died and 28 survived until the time of the last follow-up (September 2013). The age of the patients ranged from 41 to 81 years, and 41 patients were younger than 65. According to the patient records, 41 patients had a smoking history and 42 had an alcohol addiction. All patients had undergone curative resection of ESCC. Sixteen tumors were located in the upper thoracic, 47 in the middle thoracic, and 23 in the lower thoracic region. In addition, 47 cases had lymph node metastasis. With respect to histologic grade, 22 cases were well differentiated, 43 were moderately differentiated, and 21 were poorly differentiated. According to criteria of the Union for International Cancer Control (UICC) TNM Classification of Malignant Tumors 7th edition, the study included 10 patients in stage I, 45 in stage II, 31 in stage III, and none in stage IV.

Correlation between the expression of SIRT1 and clinicopathologic features of patients with ESCC

SIRT1 protein is normally located in the nucleus and shuttled into the cytoplasm when required to act on targets [17]. However, in cancer cells, SIRT1 protein is located primarily in the cytoplasm [18]. Consistent with these findings, as shown in Figs. 2 and 3, SIRT1 protein was detected mainly in the cytoplasm of ESCC tissues, especially in the tumor cells, and only occasionally in the nucleus.

Fig. 2
figure 2

Positive and negative expression of SIRT1 in ESCC and adjacent normal tissues by immunohistochemistry. Representative images are shown (×400). Note that SIRT1 was expressed not only in tumor tissues but also in adjacent normal tissues

Full size image
Fig. 3
figure 3

Immunohistochemical analysis of SIRT1 staining intensity in ESCC tissues. a No staining, 0; b weak staining, 1+; c moderate staining, 2+; d strong staining, 3+ (×400)

Full size image

Among 86 ESCC samples, 54 (62.8 %) were positive for SIRT1 protein and 32 (37.2 %) were negative. It is noteworthy that the expression levels of SIRT1 protein were significantly higher in tumor tissues than in matched adjacent normal tissues. Details of the correlation between SIRT1 protein expression levels and clinicopathologic characteristics of ESCC are shown in Table 1. Statistically significant correlations were noted between SIRT1 protein expression level and TNM stage and lymph node status (P = 0.035 and P = 0.044). However, there was no obvious relationship between SIRT1 protein expression and patient age, gender, smoking and alcohol history, or tumor location, diameter, and histologic differentiation. Furthermore, the expression of SIRT1 protein correlated with the survival outcome of ESCC patients (P = 0.02); patients with positive SIRT1 expression had a shorter survival time than those with negative expression. In conclusion, our findings indicated that increased expression of SIRT1 protein was associated with ESCC development and progression.

Table 1 Correlation between SIRT1 expression and clinicopathological features
Full size table

Correlation of SIRT1 expression and patient survival in ESCC

For the 86 patients enrolled in this study, analysis of the total survival curve by the life table method revealed that the 5-year overall survival rate was 31.3 % and the median survival time was 31 months (Fig. 4). Kaplan-Meier survival analysis revealed an inverse correlation between SIRT1 protein expression and overall survival time. The SIRT1-negative group exhibited a significant survival advantage over the SIRT1-positive group (P = 0.001, Fig. 5), with a longer median survival time (62 vs. 22 months). A similar result was obtained for the 5-year survival rate, which was 18.1 % in the SIRT1-positive group compared with 53.1 % in the SIRT1-negative group.

Fig. 4
figure 4

Overall survival curves of patients in ESCC by life table analysis. The 5-year overall survival rate was 31.3 % and the median survival time was 31 months

Full size image
Fig. 5
figure 5

Relationship between survival prognosis of ESCC patients and SIRT1 expression. Overall survival curves for ESCC patients with positive or negative SIRT1 expression. Patients with positive SIRT1 expression exhibited worse survival (P = 0.001)

Full size image

We analyzed other clinicopathologic features with respect to overall survival. As summarized in Table 2, a series of factors, including age, gender, tumor location, differentiation, and clinical stage, were entered into the analyses to assess the impact on survival of ESCC patients. The log-rank test revealed that gender, lymph node status, grade of differentiation, and clinical stage were significantly correlated with survival of ESCC patients (P = 0.012, P < 0.001, P < 0.001, P = 0.001, respectively, Fig. 6a–d). Other clinicopathologic parameters such as age, tumor diameter, smoking and alcohol intake history, and tumor location had no obvious association with survival time.

Table 2 Univariate analysis and multivariate analysis for prognostic factors in ESCC patients
Full size table
Fig. 6
figure 6

Survival analyses of ESCC patients by the Kaplan-Meier method. Overall survival curves are shown according to gender (P = 0.012), lymph metastasis (P < 0.001), differentiation (P < 0.001), and TNM stage (P = 0.001)

Full size image

To examine whether these patient features correlated with clinical prognosis, we carried out a multivariate analysis by the Cox proportional hazards model. The significant variables mentioned above for univariate analyses were further analyzed to determine whether they could serve as independent predictors for ESCC. The results presented in Table 2 showed that SIRT1 expression, clinical stage, lymph node status, and differentiation functioned as independent prognostic factors for ESCC patients (P = 0.043, P = 0.007, P = 0.033, and P < 0.001, respectively).

Discussion

The cancer burden of esophageal cancer is currently growing at an alarming pace, necessitating urgent implementation of efficient strategies to eradicate the disease. ESCC accounts for the majority of cases of esophageal cancer in China and exhibits both high morbidity and high mortality. Despite great improvements in combination treatment with surgical approaches and radiochemotherapy, the general outcome of ESCC remains poor due to obscure symptoms, late presentation, and early metastasis. Patients with ESCC have different clinical courses and prognosis. Some patients with early-stage disease suffer recurrence or metastasis soon after surgery, whereas some with advanced-stage cancer achieve a long overall survival time. Therefore, traditional therapeutic strategies based on TNM stage and tumor differentiation cannot meet the needs of individual patients. Further explorations of molecular mechanisms and biomarkers related to invasion and metastasis are necessary to provide guidance for clinical therapy.

Sirtuins are a class of proteins that play important roles in life processes. There is accumulated evidence that sirtuins perform the dual functions of either promoting cancer cell survival or facilitating cell death. Of the sirtuin family members, SIRT1 is the most studied and is known to have a dual role in cancer development. To date, the role of SIRT1 in the prognosis of ESCC remains poorly understood.

Therefore, in this study, we performed an immunohistochemical analysis of SIRT1 expression in ESCC tissues and cell lines. Our results showed upregulated expression of SIRT1 in ESCC tissues compared with matched normal controls. Consistent with this observation, SIRT1 protein and mRNA expression were significantly higher in ESCC cell lines with the exception of Kyse30, a well-differentiated cell line. We consider that the decreased expression of SIRT1 in Kyse30 is due to the heterogeneity of this cell line. Moreover, patients with lower SIRT1 expression exhibited more favorable clinical outcomes than those with high expression. We further demonstrated that SIRT1 expression was closely associated with TNM stage and lymph node status, indicating that higher SIRT1 expression could lead to disease metastasis. Consistent with these findings, Han et al. found that SIRT1 was overexpressed in brain metastasis tissues of non-small cell lung cancer (NSCLC) compared with the primary NSCLC tissues [19]. Moreover, SIRT1 expression in NSCLC cell lines was tightly related to cell migration ability. These observations support a role for SIRT1 as a potential biomarker of metastasis.

It is well known that alcohol consumption and tobacco use are two major causes of esophageal squamous cell carcinoma. According to an Australian study, 49 and 32 % of ESCC cases are attributed to smoking and heavy alcohol consumption, respectively [20]. However, in our study, no significant difference between patients with or without a history of smoking and alcohol intake was observed. This might be due to the relatively small sample size. Moreover, we only detected SIRT1 expression in tumor tissues before treatment, and it would be better if the expression of SIRT1 in lymph nodes and distant metastasis could be determined at the same time.

Additionally, we found that SIRT1 expression was significantly linked to poor prognosis after surgical resection, which is consistent with data from a previous study [21] in which SIRT1 expression was associated with higher tumor stage, poorer differentiation status, and poor prognosis in NSCLC, and decreased expression of SIRT1 enhanced the chemosensitivity to cisplatin. Similar roles of SIRT1 have also been reported in lung adenocarcinoma [22] and gastric cancer [23]. Our results suggest that SIRT1 might be a significant predictive indicator of prognosis for ESCC patients.

Several sirtuin activators and inhibitors have been developed or are currently under development, and such agents may be suitable for distinct therapeutic purposes [24]. Many of these small chemical compounds have been reported to modulate sirtuin protein expression and have been proposed for therapeutic use against neurodegenerative disease and cancer [25]. To date, in addition to nicotinamide, a physiologic sirtuin inhibitor, several novel and specific inhibitors have been reported including splitomicin, tenovins, AKG2, cambinol, salermide, and suramin. Previous studies have shown that tenovin-6 and cambinol have an antitumor effect through targeting SIRT1 and SIRT2 [26, 27]. Cambinol also exhibited potent activity against Burkitt lymphoma cell lines by hyperacetylation of BCL6 and p53 and inhibited the growth of Burkitt lymphoma xenografts [28]. Furthermore, the combination of salermide, a SIRT1 inhibitor, and cholera toxin B induced apoptosis of MCF-7 and MRC-5 cell lines and might serve as an anticancer agent [29]. Tenovins, inhibitors of SIRT1 and SIRT2, have been tested in murine cancer models. Tenovin was reported to decrease tumor growth in vivo at single-digit micromolar concentrations in a p53-dependent manner as a single agent [26]. As a result, inhibitors and activators of sirtuins have potential therapeutic efficacy as new drugs against tumors with aberrant SIRT1 expression. Given the complex roles of the sirtuin family, further efforts are needed to develop more selective sirtuin modulators to treat cancers or other diseases.

Conclusion

In summary, to our best knowledge, this is the first report of upregulated SIRT1 expression in ESCC tissues and cell lines and its implication for prognosis in patients with ESCC. In addition, this research provides evidence supporting the application of sirtuin modulators in clinical targeted therapy against ESCC. Further studies are necessary to characterize the role of SIRT1 in ESCC.