A total of 58 formalin-fixed paraffin-embedded MB samples and 4 fresh MB surgical samples and matched adjacent normal human cerebellums were gathered from the Neurosurgery Department of the First Affiliated Hospital of Sun Yat-Sen University (Guangzhou, China) between June 2002 and March 2014. The inclusion of patients in the study was unbiased and depended exclusively on the availability of sufficient tumor material and clinical follow-up data. Table I summarizes the clinical information corresponding to the samples. The following information was assessed: General information (age and sex) and residual tumor size. The MB patients were divided into two groups on the basis of the metastatic phase, age and extent of resection. Infants (≤3 years of age), patients with residual tumor (≥1.5 cm²) following neurosurgery and patients with leptomeningeal dissemination at presentation belonged to the high-risk group; otherwise, patients belonged to the standard-risk group (18). The World Health Organization (WHO) histological subtype (19), metastatic status and differentiation level was evaluated by a pathologist following hematoxylin and eosin staining using the following procedure: 4-µm thick serial sections were fixed for 20 min at 98°C with sodium citrate buffer (10 mM sodium citrate, 0.05% Tween-20, pH 6.0), and then stained with hematoxylin (Beijing Dingguo Changsheng Biotechnology Co., Ltd., Beijing, China) for 45 sec and 1% eosin for 10 sec at room temperature. The evaluation was performed with a CX31 microscope (abbe condenser with blue filter) and corresponding Image Analysis system software (CellSens Standard, version1.16; both from Olympus Corporation, Tokyo, Japan). A follow-up investigation was performed for all 58 patients, and December 2014 was the final date of follow-up for the present study. The present study was approved by the Ethics Committee of Sun Yat-Sen University and written informed consent was obtained from all subjects.

Each tissue block was cut into 4-µm thick serial sections on aminopropyltriethoxysilane-coated glass slides, and fixed as described above. Then the sections were incubated with anti-SPOPL antibody (1:50 dilution; catalog no. 191175; Abcam, Cambridge, UK) overnight at 4°C in a humidified chamber. Next, the slides were processed using a ChemMate EnVision/horseradish peroxidase kit (Dako; Agilent Technologies, Inc., Santa Clara, CA, USA) for 30 min at room temperature, which was followed by processing with diaminobenzidine for visualization. All sections were counterstained with Mayer's hematoxylin (Beijing Dingguo Changsheng Biotechnology Co., Ltd.) for 30 sec at room temperature. Normal human cerebellum was used as a control.

Immunohistochemical staining evaluation was performed based on the proportion and intensity of positively stained tumor cells (20,21). The scoring was as follows: 0, 0–5; 1, 6–25; 2, 26–50; 3, 51–75; and 4, >75%. The staining intensity was scored as one of the following four grades: 0, negative; 1, weak; 2, moderate; and 3, strong. The final score for each section was the product of the percentage and intensity score. The SPOPL protein expression level was categorized as lower (final score <4) and higher (final score ≥4). Two pathologists separately scored all immunohistochemical staining. The pathologists were not familiar with the patient information. The mean scores were the final score in each case. All evaluation was performed with the microscope and image analysis software as described above.

The normal human astrocytes (NHAs) were obtained from the American Type Culture Collection (Manassas, VA, USA) and human medulloblastoma Daoy, D283 and D341 cell lines were from the Institute of Basic Medical Sciences of the Chinese Academy of Medical Sciences (Beijing, China). The Daoy and D283 cell lines were cultured in Dulbecco's modified Eagle's medium (Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA). The D341 cell line was cultured in RPMI-1640 medium (Gibco; Thermo Fisher Scientific, Inc.), and NHA was cultured in Astrocyte medium (#1801; ScienCell Research Laboratories, Carlsbad, CA, USA) with astrocyte growth supplement and 10% fetal bovine serum (Thermo Fisher Scientific, Inc.). All cells were incubated at 37°C in a humidified incubator in the presence of 5% CO₂.

Total RNA was isolated using TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. cDNA was generated using a First-Strand cDNA Synthesis kit (Thermo Fisher Scientific, Inc.). Real-time PCR was performed with SYBR-Green qPCR kit on the OpenArray® Real-Time PCR Platform (both from Thermo Fisher Scientific, Inc.), and expression levels of the SPOPL gene were calculated by the 2^−ΔΔCq method (22) and normalized to GAPDH (internal control). The PCR program was: 10 min at 95°C, followed by 40 cycles of 15 sec at 95°C and 60 sec at 60°C. Relative mRNA levels of SPOPL were measured in relation to NHA and normal human cerebellum as positive control. Primers were designed using the Primer Express software (version 2.0 software; Applied Biosystems; Thermo Fisher Scientific, Inc.), and the primer sequences were as follows: SPOPL forward, 5′-GCTGGAGTCGTAACTCGGAAG-3′ and reverse, 5′-CCCTATCTCCCGCTCCTAAAC-3′; GAPDH forward, 5′-GACTCATGACCACAGTCCATGC-3′ and reverse, 5′-AGAGGCAGGGATGATGTTCTG-3′. The RT-qPCR analysis was performed at least three times.

The NHA, Daoy, D283 and D341 cells were lysed in lysis buffer (catalog no. P0013B; Beyotime Institute of Biotechnology, Haimen, China) on ice and then centrifuged at 12,000 × g for 10 min at 4°C. The protein contents in the supernatants were determined using a Bicinchoninic Acid protein assay kit (Pierce; Thermo Fisher Scientific, Inc.). Equal amounts of protein lysates (50 mg) were separated by SDS-PAGE (10% gels) and transferred on to polyvinylidene difluoride membranes (EMD Millipore, Billerica, MA, USA). The membranes were blocked with 1% bovine serum albumin for 2 h at room temperature, and then incubated with anti-SPOPL (catalog no. 191175; dilution, 1:1,000; Abcam, Cambridge, UK) and anti-β-actin primary antibody (catalog no. AF7018; dilution, 1:1,000; EarthOx Life Sciences, Millbrae, CA, USA) overnight at 4°C. Then, horseradish peroxidase-conjugated secondary antibody was added for 2 h at room temperature. Enhanced chemiluminescence by ECL substrate (Pierce; Thermo Fisher Scientific, Inc.) was used for detection. Experiments were performed at least twice.

SPSS software (version 16.0 for Windows; SPSS Inc., Chicago, IL, USA) was used for statistical analysis. A χ² test was used for comparisons between groups. P<0.05 was considered to indicate a statistically significant difference. The overall survival (OS) time was measured (months) from the date of diagnosis to the date of mortality or the last follow-up prior to study termination. The Kaplan-Meier estimator method and Cox's regression were used for survival analysis.

Western blotting results demonstrated that the SPOPL protein level was significantly decreased in the MB cell lines compared with the NHA cells (Fig. 1A). RT-qPCR results additionally confirmed that the SPOPL mRNA level was also decreased in the MB cell lines compared with the NHA cells (Fig. 1B). A total of four pairs of MB samples and normal human cerebellum were detected by western blotting to explore whether they exhibited similar tendencies. The present study demonstrated that SPOPL was differentially expressed in all four human MB samples compared with the normal human cerebellum, as presented in Fig. 1C. This result is similar to the results obtained for the mRNA level. As demonstrated in Fig. 1D, the tumor/normal ratio of SPOPL signals exhibited an 0.25–0.7-fold difference in the four tissue pairs. These results indicate decreased SPOPL expression in cancer lesions.

On the basis of the aforementioned results, whether SPOPL exhibits decreased expression levels in a larger cohort of clinical samples was assessed. In total, 58 archived MB tissues were examined. SPOPL was detected in 56/58 (96.7%) cases. According to immunohistochemical staining evaluation, the 56 cases were separated into two groups: 42 cases with decreased expression (75%) and 14 cases with increased expression (25%). In contrast, more intense SPOPL staining was observed in the normal human cerebellum (Fig. 2). Statistical analyses were performed to explore the association between SPOPL and the clinicopathological characteristics of MB. As summarized in Table II, SPOPL expression was markedly associated with the tumor differentiation level (P=0.011), whereas no association was identified with the patient age, WHO histological subtype or tumor risk with MB.

SPOPL expression in patients with MB was significantly associated with the survival time of patients (P<0.05) at a coefficient of 0.187, which manifested in increased SPOPL expression and survival time in patients with MB (Table III). The effects of SPOPL and classic clinicopathological characteristics (including age, sex, WHO histological subtype, residual tumor size, metastatic status and differentiation level) on the survival rates by Kaplan-Meier estimator analysis and the log-rank test were calculated. The results indicated that the survival time significantly differed between the low and high SPOPL expression groups (P<0.05). As presented in Fig. 3, the cumulative 5-year survival rate for the low SPOPL expression group was only 35.1% [95% confidence interval (CI) 0.439–0.720], whereas it was 83.6% in the high SPOPL expression group (95% CI 0.668–0.940). Furthermore, multivariate survival analysis was performed to test the SPOPL expression level, age, sex, WHO histological subtype, metastatic status and differentiation level to determine whether SPOPL was an independent prognostic factor of the outcome of patients with MB. The residual tumor size and SPOPL expression were independent prognostic factors (Table IV). Thus, it was concluded that the SPOPL gene is associated with MB prognosis.