A total of 20 normal salivary glands (SGs) and 46 SACC tumor tissues, which were surgically resected between January 2001 and December 2013 at Dalian Medical University (Dalian, China), were included in the present study. All patient samples were collected under the approved guidelines provided by the Institutional Review Boards at Dalian Medical University. All collected tissues were clinically examined by oral pathologists and informed consent forms were obtained from all patients. The demographic and clinical pathological information for each patient is listed in Table I.

The human SACC cell line SACC83 was generously provided by Dr Shenglin Li (Department of Oral and Maxillofacial Surgery, Peking University School and Hospital of Stomatology, Beijing, China). Cells were subcultured in RPMI-1640 medium (Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA), supplemented with 10% fetal bovine serum (FBS; Gibco; Thermo Fisher Scientific, Inc.) and 1X penicillin/streptomycin (Gibco; Thermo Fisher Scientific, Inc.). All cell cultures were maintained at 37°C in a humidified incubator containing 5% CO₂. To inhibit the PI3K pathway, SACC cells were treated with 1 µM GDC-0941 (LC Laboratories, Woburn, MA, USA) or vehicle (dimethylsulfoxide) control for 4 h. To knock down the expression of WDR66, SACC83 cells were transfected with the WDR66 short interfering RNA (siRNA) or control siRNA using Lipofectamine™ 3000 Transfection Reagent (Invitrogen; Thermo Fisher Scientific, Inc.), according to the manufacturer's protocol. Three siRNAs against WDR66 were tested. The sequence of these three siRNAs were: #1918, 5′-GCGGUGUACCACUUAACAATT-3′; #1566, 5′-GCUGGUUGUCUGGGACAUATT-3′; and #969, 5′-GCAGAGAGUUCUUCUGUAUTT-3′. The sequence for the negative control siRNA was 5′-UUCUCCGAACGUGUCACGUTT-3′. PTEN knockdown was achieved by transfecting with PTEN siRNA (5′-GAUCUUGACAAAGCAAAUATT-3′). At between 48 and 72 h post-transfection, SACC83 cells were harvested and the knockdown efficiencies were determined using the reverse transcription-quantitative polymerase chain reaction (RT-qPCR) and western blotting.

Formalin-fixed paraffin-embedded (FFPE) slides of SACC biopsies (4-mm thick) from patients and normal HSG tissue sections were used for IHC staining according to a protocol described previously (25). Antibody binding was visualized using a Vectastain^® Avidin-Biotin Complex kit (cat. no. PK-4001 [rabbit immunoglobulin G (IgG)] or PK-4002 (mouse IgG); Vector Laboratories, Inc., Burlingame, CA, USA) and a Diaminobenzidine kit (cat. no. DAB-1031, Fuzhou Maixin Biotech Co., Ltd., Fuzhou, China), according to the manufacturer's protocol, and counterstained with hematoxylin. Images of the slides were captured using a BX43 light microscope (Olympus Corporation, Tokyo, Japan) controlled with CellSens software (version 1.12; Olympus Corporation). The primary antibodies used for IHC were mouse anti-WDR66 (1:200; cat. no. ab83694; Abcam, Cambridge, UK) and mouse anti-PTEN (1:200; cat. no. 18-0652; Zymed; Thermo Fisher Scientific, Inc.). The quantitative analysis of IHC staining followed the criteria we have described previously (24). The staining index was determined as follows: 0–2, low (−); 3–6, medium (+); 7–9, high (++) expression.

Double IF was performed as described previously (26) using mouse anti-WDR66 (as aforementioned) and rabbit anti-PTEN (1:100; cat. no. 60300; ProteinTech Group, Inc., Chicago, IL, USA) primary antibodies, and Alexa Fluor 488-conjugated goat anti-rabbit IgG [heavy and light chains (H+L)] (1:500; cat. no. ab181448; Abcam, UK) and Alexa Fluor 594-conjugated goat anti-mouse IgG (H+L) (1:500; cat. no. ab150120; Abcam, UK) secondary antibodies. Nuclei were labeled with DAPI (1:200; cat. no. C1005; Beyotime Institute of Biotechnology, Haimen, China).

Total RNA was isolated with TRIzol^® reagent (Invitrogen; Thermo Fisher Scientific, Inc.), according to the manufacturer's protocol. Total RNA samples were reverse-transcribed into cDNA using the PrimeScript™ RT reagent kit with gDNA Eraser (cat. no. DRR047A; Takara Bio, Inc., Otsu, Japan). qPCR was performed using a Thermal Cycler Dice Real-Time system (TP800; Takara Bio, Inc.) using SYBR Premix Ex Taq II kit (cat. no. DRR820A; Takara Bio, Inc.), according to the manufacturer's protocol. GAPDH served as an internal control gene. All primers were synthesized by Takara Bio, Inc., and the sequences are presented in Table II.

SACC cells were harvested and lysed in radioimmunoprecipitation buffer containing protease/phosphatase inhibitor cocktail (cat. no. KGP250, Nanjing KeyGen Biotech Co., Ltd., Nanjing, China). Total proteins (40 µg) were separated by SDS/PAGE (10% gels) prior to transfer onto nitrocellulose membranes, which were blocked with fat-free milk powder (5%) diluted in Tris-buffered saline containing 0.1% Tween-20 at room temperature for 1 h then incubated with primary antibody at 4°C overnight followed by horseradish peroxidase-conjugated secondary antibodies [HRP-conjugated goat anti-rabbit IgG (H+L); 1:2,000; cat. no. ABC-AS014; and HRP-conjugated goat anti-mouse IgG (H+L); 1:2,000, cat. no. ABC-AS003; both from ABclonal Biotech Co., Ltd., Woburn, MA, USA] for 2 h at room temperature. A Prime Western Blotting Detection Reagent enhanced chemiluminescence kit (cat. no. K-12045-D10; Advansta Inc., San Jose, CA, USA) was used to detect the signals.

Primary antibodies used for the western blotting were: Mouse Anti-WDR66 (1:500; as aforementioned), rabbit anti-epithelial (E-)cadherin (1:2,000; cat. no. 20874; ProteinTech Group, Inc.), rabbit anti-vimentin (1:2,000; cat. no. 10366; ProteinTech Group, Inc.), rabbit anti-aldehyde dehydrogenase 1 (ALDH1; 1:400; cat. no. 15910; ProteinTech Group, Inc.), rabbit anti-Oct3/4 (1:500; cat. no. 11263; ProteinTech Group, Inc.), rabbit anti-Nanog (1:2,000; cat. no. 14295; ProteinTech Group, Inc.), rabbit anti-sex-determining region Y box 2 (Sox2; 1:1,000; cat. no. 3579; Cell Signaling Technology, Inc., Danvers, MA, USA), mouse anti-vinculin (1:10,000; cat. no. ab73412; Abcam) and mouse anti-GAPDH (1:1,000; cat. no. AG019; Beyotime Institute of Biotechnology).

SACC cells were harvested and were lysed with ice-cold IP buffer. The cell lysates were incubated with Protein G-conjugated sepharose beads at 4°C for 2 h and centrifuged at 1,000 × g for 5 min at 4°C. The extract was incubated with pre-coupled antibodies bound to Protein G-conjugated beads at 4°C overnight. Samples were resolved by SDS-PAGE (10% gels) and transferred onto polyvinylidene fluoride membranes. Primary antibodies against PTEN (1:100; as aforementioned) and WDR66 (1:100; as aforementioned) were used for IP.

Cell proliferation was evaluated using a Cell Counting Kit-8 (CCK-8; Dojindo Molecular Technologies, Inc., Kumamoto, Japan). Briefly, cells were seeded in a 96-well plate at a density of 0.5×10⁴ cells/well in RPMI-1640 with 10% FBS. Cells were harvested from day 1 to day 6. CCK-8 reagent was added to each well and incubated at 37°C for 4 h before each time point of harvest. The absorbance values were determined at a wavelength of 450 nm.

Cells at 6×10⁴ cells/well in serum-free RPMI-1640 medium were added to the upper chamber of a 6.5-mm Transwell with 8.0-µm pore polyester membrane inserts (Corning Incorporated, Corning, NY, USA) and the inserts were placed in 24-well plates with RPMI-1640 containing 10% FBS in the lower chamber. For invasion assays, the Transwell membranes were pre-coated with 50 µg/ml Matrigel solution (BD Biosciences, San Jose, CA, USA), whereas Matrigel was not used for cell migration assays. After 48 h of incubation at 37°C, cells that had invaded to the reverse side of the Transwell membrane were fixed with methanol and stained with 0.1% crystal violet at room temperature for 20 min. For each filter, five randomly selected fields were analyzed under an inverted light microscope at ×20 magnification.

SACC cells (~1×10⁶) were seeded onto 6-well plates at 37°C for 24 h. The plated cells were then scratched with a 200 µl sterile pipette tip. Images of the wounded areas were captured under a microscope at ×10 magnification after 0, 12 and 24 h of incubation at 37°C, and the distances between the wound edges were determined at each time point.

To investigate the gene status of WDR66 in human cancer, a cross-cancer genomic analysis of WDR66 was performed using The Cancer Genome Atlas (TCGA) database (www.cbioportal.org/public-portal) using WDR66 as a key word.

All statistical analyses were performed using GraphPad Prism (version 6; GraphPad Software, Inc., La Jolla, CA, USA) and SPSS software (version 19.0; IBM Corp., Armonk, NY, USA). All data were acquired from at least three independent experiments and are presented as the mean ± standard error of the mean. For two group-designed experiments, comparisons were performed using unpaired Student's t-test; for multiple comparisons, analysis of variance (ANOVA) with two-way ANOVA followed by Dunn's post hoc test or two-way repeated-measures ANOVA followed by Bonferroni's post hoc test was performed. Additionally, the correlations of PTEN and WDR66 were assessed using Spearman rank correlation coefficients. Survival curves were estimated using the Kaplan-Meier method and differences in survival distributions were evaluated by the log-rank test. P<0.05 was considered to indicate a statistically significant difference.

We previously reported a gene expression profile by microarray analysis using human SACC83 cells with ectopic knockdown of PTEN (24). Initial review of the microarray data revealed a total of 244 mRNAs to be differentially expressed (fold change >3) between SACC83 cells and SACC83 cells with PTEN knockdown (24). Following reanalysis of these microarray data for the present study, EMT genes were identified to be particularly enriched in SACC83 cells with PTEN knockdown (Fig. 1A). Selected genes were validated by RT-qPCR. Decreases in CDH1 (E-cadherin) and increase in VIM (vimentin) RNA expression levels were determined in SACC83 cells with PTEN knockdown (Fig. 1B), suggesting that loss of PTEN may promote EMT in SACC cells. In addition, the transcription factor of EMT, zinc finger E-box-binding homeobox 1 (27), and a confirmed EMT promoter, transforming growth factor β1 (28), were also validated to reveal increased expression levels in SACC83 cells with PTEN knockdown (Fig. 1B). Among the validated genes, WDR66 (Fig. 1A and B) was selected for further functional investigation as it was identified to be the least-studied EMT gene candidate in the context of cancer development and maintenance.

To investigate whether the inverse association between PTEN and WDR66 is due to PTEN's regulation of the PI3K signaling pathway (29), the SACC83 cells were treated with a PI3K inhibitor, GDC-0941 (1 µM), to inhibit PI3K activity. However, PI3K inhibitor treatment failed to cause the expected decrease in WDR66 expression (Fig. 1C), suggesting that the inverse association between PTEN and WDR66 expression levels may not depend entirely on the function of PTEN in regulating the PI3K signaling pathway. WDR66 belongs to the WD-repeat containing family of proteins, which collectively functions in the formation of protein-protein complexes in a variety of signaling pathways (30). Thus, a co-IP experiment between WDR66 and PTEN was performed, and the results suggested that there may be an interaction between PTEN and WDR66 in SACC83 cells (Fig. 1D).

To further determine the effect of the PTEN-WDR66 interaction on SACC pathogenesis, double IF staining was performed to examine the intracellular distribution of WDR66 in the control and under PTEN-deficient conditions. As presented in Fig. 1E, WDR66 expression (green) was inversely associated with PTEN expression (red).

To facilitate the functional studies of WDR66, WDR66 mRNA was knocked down using three siRNAs, and siRNA #1918 was selected on the basis of its optimal knockdown efficacy as determined using RT-qPCR (Fig. 2A). Knockdown efficacy of siRNA #1918 was validated by western blotting (Fig. 2B). siRNA #1918 was used to determine the function of WDR66 in mediating the EMT phenotype induced by PTEN deficiency. Downregulation of CDH1 and upregulation of VIM mRNAs were confirmed in the SACC83 cells with PTEN knockdown (Fig. 2C). However, simultaneous knockdown of WDR66 mRNA with PTEN mRNA knockdown led to a complete reversal of the effect of PTEN siRNA on decreasing the CDH1 expression level (Fig. 2C). Similar observations were made at the protein expression level on the basis of western blot results for E-cadherin and vimentin (Fig. 2D). Collectively, these results suggest that WDR66 may be essential for the maintenance of the EMT phenotype induced by PTEN deficiency in SACCs.

Since EMT has been suggested to confer CSC-like properties on neoplastic cells (31), the function of WDR66 in regulating the expression of genes associated with cancer cell stemness in the context of PTEN deficiency was also investigated. Similar to the gene changes in EMT, the embryonic stem cell genes NANOG and OCT4, and the putative CSC genes ALDH1 and SOX2 in SG cancer (32) were all increased following PTEN knockdown in SACC cells. The expression of NANOG and OCT4 genes was decreased following WDR66 knockdown with or without PTEN knockdown (Fig. 3A). Similar observations were made at the protein expression level on the basis of western blot results for NANOG and OCT4 (Fig. 3B). Collectively, these results suggest that WDR66 may be required to maintain the expression of genes associated with CSCs in the context of PTEN deficiency.

To determine the cellular effects of WDR66 in the context of PTEN deficiency, cellular proliferation assays were performed in SACC83 cells carrying WDR66 knockdown with or without PTEN knockdown. As expected, knockdown of PTEN promoted cellular proliferation rates, but this effect was fully nullified with WDR66 knockdown (Fig. 4A). Similarly, knockdown of PTEN accelerated wound closure, but simultaneous knockdown of WDR66 led to the attenuation of this effect to the point that there was no significant difference between the siPten + siWDR66 group and the siControl group (Fig. 4B). Finally, the migratory and invasive abilities that were enhanced by PTEN knockdown were also decreased with WDR66 knockdown (Fig. 4C).

To search for genetic alterations of WDR66 in human malignancies, a cross-cancer genomic analysis of WDR66 was performed using The Cancer Genome Atlas (TCGA) database. Somatic point mutations and gene amplification were the most common genetic alterations in WDR66 across various types of cancer. Notably, amplification of WDR66 was the only genetic alteration noted in patients with SACC (Fig. 5). Next, WDR66 expression levels were determined in patient samples and associated with PTEN expression in SACCs via IHC. A total of 46 SACC and 20 normal SGs were included. WDR66 was weakly expressed in normal SGs, and the majority of the tubular and cribriform SACCs, but it was markedly expressed in tumor cells of solid-pattern SACCs (Fig. 6A). This staining pattern of WDR66 was the opposite to the staining pattern of PTEN, where it was markedly expressed in normal SGs, tubular and cribriform SACCs, but was lost from the solid pattern SACCs (Fig. 6A). This inverse association of WDR and PTEN expression is exemplified in their opposite staining pattern in the same fields of tumor cells, as presented in Fig. 6A. The IHC staining results were summarized in Fig. 6B, and revealed a statistically significant inverse correlation between WDR66 and PTEN expression in patient tissues (P=0.044). Finally, the significance of WDR66 on prognosis of the 46 patients with SACC was investigated. Kaplan-Meier analysis demonstrated that the patients with positive WDR66 expression had significantly decreased overall survival times compared with patients with negative WDR66 expression (70.1 months vs. 121.8 months, P=1.30×10⁻⁴; log-rank test; Fig. 6C), which revealed that WDR66 expression was negatively associated with the overall survival periods of patients with SACC.