Twenty epithelial-type UM tissue specimens and 16 spindle-type UM tissue specimens were retrieved from UM patients who underwent enucleation in Tianjin Medical University Eye Hospital from October 2015 to April 2018. All patients were confirmed by pathological examination after operation, who were classified according to the 1980 WHO classification of histological types. The age range was 24-67 years. The inclusion criteria were as follows (1): the pathological diagnosis was primary UM; (2) the submitted tissue sections were relatively complete; (3) the patient had no other types of cancer; and (4) the patient had not received radio- or chemotherapy. This study was approved by the Ethics Committee of Tianjin Medical University Eye Hospital (2017KY(L)-53). All patients provided written informed consent.

The patients’ eye tumor specimens were fixed in neutral formalin and embedded in paraffin, followed by sectioning, deparaffinization, antigen retrieval, blocking, antibody incubation, DAB (diaminobenzidine) staining, hematoxylin staining, back blue, and dehydration coverslips. Primary antibodies were anti-rabbit (Proteintech, China), and secondary antibodies were goat anti-rabbit IgG (H+L) (Jackson Immuno Research Inc., US). As previously described, the staining scoring was calculated according to the following formula: H-score = staining intensity * number of positive cells/total number of cells (20). Three random fields were chosen, and the staining intensity was manually evaluated: 0 point for no yellow precipitate, 1 point for light yellow precipitate, 2 points for brownish-yellow precipitate, and 3 points for dark brown-yellow precipitate; positive cell counts were manually counted using Image J software.

Human retinal epithelial cell line (APRE-19), human umbilical vein endothelial cells (HUVEC) as well as three human UM cell lines were acquired from the Chinese Academy of Sciences (Shanghai, China), including primary spindle-type OCM-1 cells, metastatic epithelial Mum-2B cells, and metastatic spindle-type Mum-2C cells. All cells were cultured in an incubator (Thermo Fisher Scientific, US) with 37°C, 5% CO₂. Subculture was carried out when the cells adhered and reached about 80% confluence. After discarding the medium, they were washed with Dulbecco’s Phosphate-Buffered Saline (DPBS; Gibco, US), and digested with 1 mL of 0.25% trypsin-EDTA (Gibco, US). Thereafter, Dulbecco’s modified eagle medium (DMEM) containing 100 U/mL penicillin, 100 μg/mL streptomycin, and 10% fetal bovine serum (FBS; Gibco, US) was used to terminate the digestion. The cells were centrifuged at 1000 rpm and resuspended in DMEM for future experiments.

UM cells were lysed by RIPA cell lysate, and protein concentrations were measured utilizing BCA Protein Assay Kit (Thermo Scientific, US). There were 20 μg proteins separated by 10% SDS-PAGE electrophoresis as well as transferred onto polyvinylidene fluoride membrane. The membrane was sealed with 5% nonfat dry milk and then incubated with primary antibodies against SOST (1:1000; 21933-1-AP; Proteintech, China), Wnt-1 (1:1000; 27935-1-AP), β-catenin (1:5000; 51067-2-AP), cyclin-D1 (1:2000; 26939-1-AP), MMP2 (1:500; 10373-2-AP), and MMP9 (1:500; 10375-2-AP) as well as β-Tubulin (1:1000; 10094-1-AP) at 4°C overnight. The next day, the membrane was incubated with goat anti-rabbit secondary antibody (Cell Signaling Technology, US) for 1 h at room temperature. Exposure was performed using Amersham ECL prime chemiluminescent fluid (GE Healthcare, US) in a dark room. The gray value was determined with ImageJ software.

The design and synthesis of the sequences of siRNA targeting human SOST (gene ID: 50964) and negative control siRNA (si-NC) were achieved by Shanghai Sangon Bioengineering Co., Ltd., (China). The sequence of SOST siRNA (si-SOST) was 5′-GGCGTTCAAGATGATGCCACGGAA-3′. OCM-1, Mum-2B, and Mum-2C cell lines were randomly divided into three groups: blank group (no transfection), si-NC group (transfection with si-NC plasmid), and si-SOST group (transfection with si-SOST plasmid). The full length SOST cDNA was synthesized and sub-cloned into the pcDNA3.1 vector to construct pcDNA3.1-SOST overexpression (OE-SOST) plasmid. OCM-1, Mum-2B, and Mum-2C cell lines were randomly divided into two groups: empty vector group and OE-SOST group. These cells were trypsinized at 0.25% as well as centrifuged at 1000 rpm lasting 5 min. Then, they were resuspended in medium and counted using an automated cell counter (Invitrogen, US), and 4×10⁵ cells/well were seeded onto a six-well culture plate. When the cell confluency reached 50-80%, the complete medium was replaced with serum- and penicillin-free medium. Four hours later, using Lipofectamine 2000 (Invitrogen, US), plasmids encoding target sequences or siRNAs were transfected into the UM cell line. Twenty-four hours after transfections, the complete medium was exchanged for subculture. During this period, cellular morphology was investigated under a microscope (Olympus, Tokyo, Japan).

Total RNAs of UM cells were extracted utilizing Trizol kit (Invitrogen, US). The extracted RNAs were utilized for reverse transcription. Cellular cDNA synthesis was conducted according to the instructions of the reverse transcription kit (Transgen, Beijing, China). The PCR amplification reaction was carried out according to the instructions of RT-qPCR kit (Transgen, Beijing, China). Table 1 lists the primer sequences that were synthesized by Shanghai Sangon Bioengineering Co., Ltd. (China). The reaction condition contained pre-denaturation at 94°C lasting 4 min, denaturation at 94°C lasting 30 sec, annealing at 56°C lasting 30 sec, and extending at 72°C lasting 30 sec, in total 45 cycles. The reaction system contained 10 μl Premix Ex Taq or SYBR Green Mix, 1 μl forward primer, 1 μl reverse primer, 1 μl cDNA template, and ddH₂O supplemented to 20 μl. Glyceraldehyde phosphate dehydrogenase (GAPDH) acted as an internal reference gene, and the Ct values ​​(threshold of amplification curves) were examined with the control group set to 1. The iQ5 Real-Time PCR Amplifier was purchased from the Bio-Rad company (US). Relative expression values ​​were calculated by the 2^-ΔΔCt method.

UM cells were collected and digested with trypsin, followed by centrifugation at 1000 rpm for 5 min. After rinsing the cells with sterile PBS, 1 ml of pre-cooled 70% ethanol was utilized to resuspend the cells as well as fixed overnight at 4°C. They were then rinsed with PBS to remove the ethanol solution, followed by incubation with 500 μl of PI staining solution at room temperature lasting 15 min in the dark. The cell cycle was detected by flow cytometry.

UM cells were trypsinized, followed by centrifugation at 1000 rpm lasting 5 min. After rinsing the cells utilizing pre-cooled sterile PBS at 4°C, the cellular concentrations were adjusted to 5×10⁴/mL. Then, 10 μl of Annexin V-FITC and 10 μl of PI were added to 200 μl of the cell suspension and incubated for 10 min at room temperature in the dark, followed by the addition of 500 μl PBS. Apoptotic levels were examined by flow cytometry (BD, US) as well as analyzed with FlowJo flow cytometry software.

Following adjustment of the cellular concentration to 1×10⁴/mL, the cells were seeded in 96-well plates, with six duplicate wells for each group. The edge wells were filled with 0.01 mol/L sterile PBS. After 24 h at constant temperature, the medium in the wells was aspirated and discarded, and 90 μL sterile PBS supplemented with 10 μL 0.5% MTT (Sigma-Aldrich, US) were added. After 4 h, the culture was terminated. Thereafter, 100 μl of dimethyl sulfoxide (DMSO; Sigma-Aldrich, US) was added to each well. The plate was shaken well for 10 min, and the absorbance (optical density, OD) value at 490 nm wavelength of each well was measured with a microplate reader (Promega, US).

UM cells were seeded in 24-well plates (1 mL/well) containing 1 cm × 1 cm sterile glass slides. The plates were pre-incubated for 24 h at 37°C in a 5% CO₂ incubator. After transfection, 20 μM EdU working solution pre-warmed at 37°C was added to a 24-well plate. The cells continued to incubate for 2 h. After removing the culture medium, the cells were fixed with 1 ml of 4% paraformaldehyde for 15 min at room temperature. After washing, the cells were incubated with 1 ml of 0.3% TritonX-100 in PBS per well lasting 10 min at room temperature, followed by 0.5 ml Click addictive reaction lasting 30 min in the dark. DAPI (1000×) was diluted 1:1000 in PBS. After removing the washing solution, they were incubated with 200 μl of 1× DAPI solution at room temperature lasting 5 min in the dark. Finally, fluorescence detection was performed.

For invasion assay, Matrigel was diluted 8:1 with serum-free DMEM. Then 100 μL diluted Matrigel was added to the upper chamber of the transwell, and the whole process was performed on ice to prevent it from solidifying. The residual liquid was aspirated from the culture plate. UM cells were trypsinized, resuspended, and concentration adjusted. The upper chamber of a 24-well plate transwell covered with Matrigel was evenly seeded 1×10⁵ cells/well as well as maintained by adding serum-free DMEM. Meanwhile, 600 μl of complete medium with 10% FBS was added to the lower chamber. Following 24 h, the chamber was removed, rinsed twice with PBS, and then gently wiped with a cotton swab to remove unpenetrated cells on the upper layer of the chamber membrane. The cells were fixed with 4% paraformaldehyde for 30 min in a 24-well plate, stained with crystal violet staining solution lasting 15 min. The cells invaded through Matrigel were observed under an inverted microscope. Three areas were randomly selected to count the number of invasive cells through Matrigel. The experimental procedure for the migration assay was the same as for the invasion assay, except that Matrigel was not added.

UM cells transfected with si-NC or si-SOST were co-cultured with HUVEC, and the tube formation ability was detected. Briefly, Matrigel was removed from -20°C and placed in a 4°C refrigerator overnight. The next day, 50 μl Matrigel was added to pre-cooled 96-well plates, and maintained in a 37°C incubator for 45 min. The co-cultured HUVEC was routinely digested, resuspended in cell medium containing 10% FBS. Thereafter, they were seeded into 96-well plates (1×10⁴ cells/well) as well as incubated at 37°C. The tube formation was investigated under an inverted microscope.

The primer of SOST was designed by querying the CDS sequence on the NCBI website as well as the Thermo fisher website, with a primer sequence of 5’- AAAAGCAGGCGTTCAAGAATGATGCTTGGATCCAAGCATCATTCTTGAACGCCTGC-3’. The primer was synthesized by Shanghai Sangon Bioengineering Co., Ltd. (China). With pENTRTM H1/TO as vector, shRNA SOST (sh-SOST) or negative control (sh-NC) plasmids were constructed by enzyme digestion, ligation, transformation, and other operations, and then the plasmid identification was carried out. There were 293 T cells used to package the target plasmid, and when the density of OCM-1 cells reached 40%-50%, a stable transfection cell line was constructed after screening.

Six-week-old BALB/c nude female mice (14 ± 2 g) were purchased from Vital River Company (Beijing, China). All experiments gained the approval of the Laboratory Animal Care and Use Committee of Tianjin Medical University Eye Hospital (TJYY2020122054). All animals were separated into four groups (3 mice/group), containing a control group, model group, model + sh-NC group, and model + sh-SOST group. Normal OCM-1 cells or OCM-1 cells stably transfected with sh-NC or sh-SOST were trypsinized, centrifuged as well as suspended in Hanks’ balanced salt at 1×10⁷/mL. The abdominal skin of the mice was disinfected with iodophor, and the mice were anesthetized by intraperitoneal injection of 0.5% pentobarbital sodium. The right eye of the mice was instilled with proparacaine for ocular surface anesthesia, and the left eye was coated with ophthalmic gel to prevent keratitis. Under the operating microscope, a 30G needle was pointed at the oblique angle to create a tunnel via the sclera, and to penetrate the choroid without retina. There were 34G blunt needles attached to 10 μL Hamilton syringes inserted into the subretinal space with the bevel toward the inner eye. The mice in the model group, model + sh-NC group, and model + sh-SOST group were separately injected by 5 μL (5×10⁴ cells) of the cell suspension of normal OCM-1 cells or OCM-1 cells stably transfected by sh-NC or sh-SOST. For the control group, the mice did not receive any treatment. After the injection, a white bump was found in the fundus of the mouse, indicating that the injection was successful. Tobramycin ophthalmic ointment was then applied to the eye to reduce the risk of infection. After successful injection, the state of the mice was observed and weighed every other week, and the time of tumor formation in the eyes of the mice was recorded. After six weeks, all mice were euthanized, and their eyeballs were enucleated. The length (L) as well as width (W) of the murine eyeballs were measured utilizing a vernier caliper. The tumor volume (V) was measured in accordance with the following formula: V = π×L×W²×1/6.

A TUNEL kit was purchased from the Roche Company (Switzerland). The tissue slides were soaked in xylene for 10 min to dewax. For hydration, the sections were soak in graded ethanol (100%, 95%, 85%, 70%, 50%) for 30 sec, respectively. After washing twice with 1×PBS for 5 min each, 20 μg/ml proteinase K was dropwise added. After 15 min at 37°C, the sections were washed 3 times with 1×PBS. The labeling solution was mixed with buffer evenly at a volume ratio of 1:9. Thereafter, 100 μL TUNEL reaction solution was dropwise added. The reaction area with a cover glass was covered, and reacted at 37°C in a humid, dark environment for 60 min. After washing three times with 1×PBS, 50 μL of DAPI buffer (0.1 mg/mL; Sigma, US) was added to each well, and incubated in the dark at room temperature for 5 min. After washing 3 times with 1×PBS, images were photographed under a fluorescence microscope.

Paraffin sections of murine in situ UM tumors in eyeball were deparaffinized, antigen retrieved, blocked with 5% BSA, incubated with primary and secondary antibodies, and finally stained with DAPI (Sigma, US) and mounted with resin. Primary antibodies included SOST, LRP5, and LRP6 antibodies (Affinipure, Jackson, US), and fluorescent-labeled secondary antibodies included Alex Fluor 488 and Alex Fluor 610 goat anti-rabbit antibodies (Affinipure, Jackson, US). The sections were then observed under a fluorescence microscope and three fields of view were randomly selected to be photographed. The nuclei stained with DAPI were blue under the excitation of ultraviolet, and the positive expression was fluorescein-labeled green or red.

All statistical analysis was carried out utilizing GraphPad Prism software (version 8.0.1, GraphPad Software, US). Student’s t-test was implemented for comparisons between two groups, and one-way analysis of variance (ANOVA) was implemented for comparisons between multiple groups. Pearson correlation test was used for correlation analysis. P<0.05 was considered statistically significant.

Our study analyzed the expressions of SOST and Wnt-1 as well as β-catenin in 20 cases of epithelial type UM tissues with a high degree of malignancy and 16 cases of spindle type UM with a low degree of malignancy via immunohistochemistry. The results showed that SOST expression was lower as well as Wnt1 and β-catenin expressions were higher in epithelial type in comparison to spindle type UM tissues ( Figures 1A, B ). This indicated that low SOST expression as well as activation of Wnt/β-catenin signaling were linked to malignant progression of UM. We also calculated the correlation between SOST and Wnt1 and β-catenin in UM. Our results showed that SOST was negatively correlated to Wnt1 (p=0.0003 and r=-0.9869) and β-catenin (p<0.0001 and r=-0.9953). The STRING online database also confirmed their protein-protein interactions ( Figure 1C ).

SOST expression was further examined in human retinal epithelial cell line (APRE-19) and three human UM cell lines containing primary spindle-type OCM-1 cells, metastatic epithelial Mum-2B cells, and metastatic spindle-type Mum-2C cells. As shown in Figures 2A, B , SOST expression was lower in three human UM cell lines in comparison to APRE-19 cells. Compared with OCM-1 cells, lower SOST expression was found in Mum-2B and Mum-2C cells. Meanwhile, SOST expression was lower in Mum-2C cells than Mum-2B cells. This indicated the down-regulation of SOST in metastatic cells. To silence the expression of SOST, siRNA against SOST was transfected into OCM-1, Mum-2B, and Mum-2C cells. In Figure 2C , 48 h after transfection, the expression of SOST mRNA was successfully decreased in three human UM cell lines. Further analysis showed that the expressions of WNT1, CTNNB1, CCND1, MMP2, and MMP9 mRNAs were decreased in si-SOST-transfected OCM-1, Mum-2B, and Mum-2C cells than controls ( Figures 2D–H ). As expected, SOST protein was reduced following si-SOST transfection than control in three human UM cell lines ( Figures 2I, J ). Additionally, Wnt-1, β-catenin, Cyclin-D1, MMP2, and MMP9 proteins presented higher levels when SOST expression was knocked out ( Figures 2K–O ). Collectively, silenced SOST could activate the Wnt/β-catenin pathway in UM cells.

Flow cytometry showed that after si-SOST transfection, the proportion of G1 phase was lower as well as the proportions of S phase and G2 phase were higher in three cell lines OCM-1, Mum-2B, and Mum-2C compared with that of the si-NC group, indicating that silencing SOST expression was capable of increasing the proportion of UM cells in interphase ( Figures 3A–F ). Additionally, the apoptotic levels of si-SOST-transfected OCM-1, Mum-2B, and Mum-2C cells were decreased in comparison to that of the si-NC group ( Figures 3G–L ). Altogether, SOST knockdown enabled us to accelerate cell cycle progression as well as decrease apoptosis in UM cells.

Through CCK-8 as well as EdU staining assays, this research evaluated the effects of SOST on proliferation of UM cells. As depicted in Figures 4A–C , the cell viability of si-SOST-transfected OCM-1, Mum-2B, and Mum-2C cells was elevated in comparison to that of the si-NC group. To overexpress SOST, UM cells were transfected with OE-SOST plasmid. RT-qPCR confirmed that SOST expression was significantly up-regulated in UM cells ( Figure 4D ). Lower cell viability was found in SOST-overexpressed OCM-1, Mum-2B, and Mum-2C cells ( Figure 4E ). Moreover, proliferation levels of three human UM cell lines were enhanced by si-SOST transfection ( Figures 4F, G ) as well as weakened by SOST overexpression ( Figures 4H, I ). Thus, reduced expression of SOST enabled us to strengthen proliferative capacities of UM cells.

Transwell experiment results showed that compared with the si-NC group, the number of OCM-1, Mum-2B, and Mum-2C cells that penetrated the upper chamber of transwell in the si-SOST group was elevated in the si-SOST group ( Figures 5A, B ). This indicated the increase in migration capacities of UM cells. In Figures 5C, D , the number of three human UM cell lines invaded through Matrigel was increased in the si-SOST group in comparison to the si-NC group, indicating the enhancement in invasion capacities of UM cells. Following co-culture with si-SOST-transfected OCM-1, Mum-2B, and Mum-2C cells, the tube formation number of HUVEC was increased ( Figures 5E, F ). Altogether, SOST knockdown could enhance migration and invasion as well as angiogenesis of UM cells.

To further verify the impact of SOST on UM, we constructed a stable SOST knockdown expression OCM-1 cells as well as ocular orthotopic tumor models in nude mice. Following six weeks, we removed the eyeballs and measured the tumor volume. In comparison to the control group, the tumor volume of the model group was increased. Meanwhile, higher tumor volume was found in the model + sh-SOST group than the model + sh-NC group ( Figures 6A, B ), indicating that SOST suppression accelerated tumor growth. TUNEL staining also demonstrated that proliferative capacity of tumor cells in the model + sh-SOST group was heightened in comparison to that of the model + sh-NC group ( Figure 6C ).

Immunohistochemistry demonstrated that SOST expression was lower in tumor cells of the model + sh-SOST group in comparison to that of the model + sh-NC group ( Figure 7 ). Additionally, Wnt-1, β-catenin, cyclin-D1, MMP2, and MMP9 as well as Ki-67 levels were higher in tumor cells of the model + sh-SOST group than that of the model + sh-NC group, demonstrating that SOST suppression was capable of activating Wnt/β-catenin signaling in ocular orthotopic tumor models.

Further analysis showed that LRP5 and LRP6 membrane receptors presented lower expressions in tumor cells of the model + sh-SOST group than that of the model + sh-NC group, demonstrating that SOST suppression could decrease the expressions of LRP5 and LRP6 ( Figures 8A, B ). Also, SOST co-localized with LRP5/6 in UM cells, indicating that SOST was interacted with LRP5/6.