Specimens from 103 thymoma and thymic carcinoma patients, who had undergone thymectomy between 1999 and 2010, were obtained from the Basic Medical College of Xinxiang Medical University (Xinxiang, China). The study was approved by the Ethics Committee of the Basic Medical College of Xinxiang Medical University, and all patients provided informed consent for the use of their samples. All patient characteristics (including gender and age) and tumor clinical data were collected. The patient sample included 68 male and 35 females, with a mean age of 50 years (range, 41–65 years). According to the histological criteria of the WHO classification (16), the thymoma tumor subtypes were as follows: 21 cases of type A; 23 cases of type AB; 12 cases of type B1; 15 cases of type B2 and 9 cases of type B3 thymomas. In addition, 23 cases of stage I–IV thymic carcinomas were identified, according to Masaoka staging (17). The normal thymi of 26 children were used as controls (mean age, 11 years; age range, 8–15 years), representing the ERβ5 expression levels in normal tissues. Survival data, including the overall survival (OS) and progression-free survival (PFS) rates, were recorded. OS was defined as the time (months) between the primary surgical treatment and mortality associated with the thymic tumor. PFS was defined as the interval (months) between the primary surgical treatment and the initial locoregional or distant recurrence.

Monoclonal mouse anti-human ERβ5 antibody (clone 5/25; product code, MCA4676T) was obtained from Bio-Rad Laboratories, Inc. (Hercules, CA, USA). FITC-conjugated goat anti-human IgG antibody was purchased from Santa Cruz Biotechnology, Inc., (Dallas, TX, USA). Horseradish peroxidase-conjugated goat anti-mouse IgG was obtained from Cell Signaling Technology, Inc. (Danvers, MA, USA). The thymic carcinoma cell line, TC1889, and the thymoma cell line, T1682, were established, characterized and verified as previously described (18).

Tissue microarrays (TMAs) were constructed from the paraffin-embedded blocks of the 129 thymic specimens, using a tissue array device (cat. no. BNSW-001_TY9184; Shanghai Outdo Biotech Co., Ltd., Shanghai, China). Representative tumor areas were marked in each paraffin-embedded specimen and at least two areas were sampled. The diameter of the tissue cylinders was 0.6 mm, made using tissue chip drilling apparatus (Shanghai Outdo Biotech Co., Ltd.). For ERβ5 staining, the monoclonal ERβ5 antibody (clone 5/25) was used at a dilution of 1:50, which was demonstrated to be highly specific in immunohistochemical assays performed in this study. Antigen retrieval was performed in 0.01 mol/l sodium citrate buffer (pH 6.0) in the microwave for 15 min. To establish the negative controls, the same procedure was followed, without the primary antibody. A previously described scoring system was adopted (19), the tissue microarrays were digitized and cytoplasmic ERβ5 (cERβ5) or nuclear ERβ5 (nERβ5) immunoreactivity was scored between - and +++ (−, no staining; +, weak staining; ++, moderate staining; +++, strong staining). The percentage of tumor cells displaying staining for cERβ5 or nERβ5 was determined and calculated as the average of six high power fields per specimen. The cases were scored independently by three specialists and discordant results were re-evaluated to reach consensus.

TC1889 and T1682 cells were cultured in RPMI-media containing HEPES supplemented with 10% fetal calf serum and 1% penicillin/streptomycin (Sigma-Aldrich) in an atmosphere containing 5% CO₂ at 37°C. For IIFA, the anti-ERβ5 antibody (clone 5/25) was incubated at a dilution of 1:500 at 4°C overnight. Fluorescein isothiocyanate-conjugated goat anti-human immunoglobulin G (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) was used as the secondary antibody and incubated at a dilution of 1:100 for 1 h at room temperature. Cells were washed with phosphate-buffered saline (J-H Biotechnology Ltd., Shanghai, China) and mounted using 10 mg/ml DAPI (Sigma-Aldrich) in aqueous mountant (Dako North America, Inc., Carpinteria, CA, USA). A fluorescence microscope (Leica DM1000; Leica Microsystems GmbH, Wetzlar, Germany) was used for examination of the samples.

Total RNA was extracted using the RNeasy mini kit (Qiagen, Crawley, UK) with additional purification by centrifugation at 12,000 × g for 15 min through QIAshredder spin columns (Qiagen). The total RNA concentration and purity were calculated using the Nanodrop system (Labtech International Ltd., Lewes, UK). Subsequently, cDNA synthesis was performed using the SYBR® ExScript RT-PCR kit (Takara Biotechnology Co., Ltd., Dalian, China) according to the manufacturer's instructions. Quantitative PCR was performed using the SYBR® Premix Ex Taq™ II kit (Takara Biotechnology Co., Ltd.) with a LightCycler system (Roche Diagnostics, Basel, Switzerland). The primers used were as follows: ERβ5 forward, 5′-CGGAAGCTGGCTCACTTGCT-3′, and reverse, 5′-CTTCACCCTCCGTGGAGCAC-3′; and β-actin forward, 5′-GTGGGGCGCCCAGGCACCAC-3′, and reverse, 5′-CTCCTTAATGTCACGCACGATTT-3′. The reaction volume was 50 µl and comprised the following final quantities/concentrations: 1,000 ng ERβ5 or 100 ng β-actin cDNA, 0.2 µM of each primer, 1 U AmpliTaq Gold DNA polymerase (Life Technologies, Grand Island, NY, USA), 1.5 mM MgCl₂ and 200 µM deoxynucleotide triphosphate. The cycling conditions included a denaturation step at 94°C for 10 min, 45 cycles (for ERβ5) or 25 cycles (for β-actin; RNA input) at 94°C for 45 sec, 55°C for 60 sec and 72°C for 45 sec, and a final extension step at 72°C for 10 min. For each primer, serial dilutions of a standard cDNA were amplified to create a standard curve, and values of unknown samples were estimated relative to this standard curve in order to assess the PCR efficiency. Threshold cycle (Ct) values were collected for β-actin and the genes of interest during log phase of the cycle. Gene of interest levels were normalized to β-actin for each sample [ΔCt = Ct(gene of interest) - Ct(β-actin)]. The samples were resolved on 2% agarose gel and transferred to a nylon transfer membrane (Hybond-N+; GE Healthcare Life Sciences, Chalfont, UK). Samples were then analyzed using ABI PRISM 7000 SDS Software (Applied Biosystems Life Technologies, Foster City, CA, USA).

Mann-Whitney U test, Kruskal-Wallis test and Spearman's rank correlation were performed using the SAS software (version 9.12; SAS Institute Inc., Cary, NC, USA). Associations with OS were initially analyzed by Kaplan-Meier plots (log-rank test). P<0.05 was considered to indicate a statistically significant difference.

The results of the ERβ5 immunohistochemical assay for the 129 cases are listed in Tables I–III. The majority of thymic tumors exhibited ERβ5-positive staining (99.02%; 102/103 cases), with only one case presenting negative staining. In addition, 87.37% (90/103) of the cases were positive for cERβ5, 11.62% of the cases were positive for nERβ5 and 15 cases were positive for nERβ5 and cERβ5 (Table I; Fig. 1 A1-A6). In particular, the thymic carcinomas exhibited strong positive staining, while only 38.46% of normal thymic tissues (10/26; Table I) exhibited positive staining. Furthermore, overexpression of cERβ5 was observed in the thymic tumors compared with normal thymic tissues, and this difference was statistically significant (P=0.001); by contrast, nERβ5 was not overexpressed in thymic tumors (P=0.112; Table I). Similar results were observed in the T1682 and TC1889 cell lines following IIFA (Fig. 1B).

mRNA quantification was conducted during RT-qPCR, and the results were compared against β-actin, which was used as the internal reference gene. The ERβ5 expression levels ranged between 0.468 and 13.292 (median, 5.672; data not shown). In advanced clinical stages, the mean level of ERβ5 gene expression was lower (Fig. 1C).

The association between ERβ5 expression and a range of standard patient characteristics were investigated and are listed in Table II. No positive correlation was detected between cERβ5 expression and patient characteristics, including gender, age and tumor sizes (P=0.245, P=0.514 and P=0.614, respectively). Notably, although no statistically significant association was observed between tumor sizes and cERβ5 expression, the latter was demonstrated to be important in the progression of thymic tumors (Table II).

In addition, no statistically significant association was determined between nERβ5 expression and patient characteristics.

Since cERβ5 was overexpressed in the thymic tumors, a statistically significant correlation was observed between cERβ5 expression and thymoma subtypes (P=0.024; Table III), which presented a particularly strong positive expression in thymic carcinomas. In addition, a statistically significant difference was identified between cERβ5 staining and thymic tumor stages (P=0.003); however, a negative correlation was observed (Table III; R=-0.376).

By contrast, no statistically significant differences were observed between nERβ5 staining and thymomas subtypes or stages (P=0.653; data not shown).

The TMA analysis revealed that cERβ5 staining was significantly associated with improved OS, whereas nERβ5 immunoreactivity was not associated with OS (data not shown). In addition, cERβ5 staining was correlated with improved PFS, which implied that cERβ5 had a negative biological effect in the development of thymic tumors (Fig. 2).