The human chondrosarcoma JJ012 cells were obtained from the laboratory of Dr Joel Block, Rush University, Chicago, IL, USA. The cells were maintained in complete growth medium containing the following: Dulbecco's modified Eagle's medium (DMEM) + GlutaMAX™, MEM (minimum essential medium) supplemented with F-12 + GlutaMAX™ Nutrient mixture (Ham), 10% fetal bovine serum (FBS), 1% penicillin/streptomycin (Thermo Fisher Scientific, Inc., Waltham, MA, USA); 25 µg/ml ascorbic acid, 100 ng/ml insulin and 100 nM hydrocortisone (Sigma-Aldrich; Merck KGaA). The cells were incubated at 37°C in a humidified atmosphere of 5% CO₂ and periodically checked for mycoplasma.

With the MycoAlert^® Mycoplasma Detection kit (Lonza, Inc., Rockland, ME, USA; cat. no. LT07-118; lot. no. 0000678423) 2 ml of cell culture or culture supernatant was transferred into a centrifuge tube and centrifuged at 1,500 rpm (200 × g) for 5 min. Cleared supernatant (100 µl) was transferred into a luminometer well. MycoAlert™ reagent (100 µl) was added to each sample and incubated at room temperature (20°C) for 5 min. The plate was placed in the luminometer and the program was initiated for Reading A. MycoAlert™ substrate (100 µl) was added to each sample and incubated at room temperature for 10 min. The plate was placed in the luminometer for Reading B. The ratio was calculated as Reading B/Reading A. If contamination was detected, cells were seeded in full growth media and treated with Plasmocin™ (InvivoGen, Inc., San Diego, CA, USA; cat. no. ant-mpt; lot. no. MPT-38-03A) with a concentration of 25 µg/ml for 14 days, removing and replacing with fresh Plasmocin™ treatment containing medium every 3–4 days.

A subset of cultured human JJ012 chondrosarcoma cells were treated with 10 µg/ml of PRP-1 and incubated at 37°C in a humidified atmosphere of 5% CO₂ for a period of 24 h prior to the Aldefluor^® assay. A subset of cultured human JJ012 chondrosarcoma cells were treated with incrementally increasing doses of 1, 5, 10 and 20 µg/ml PRP-1 and incubated at 37°C in a humidified atmosphere of 5% CO₂ for a period of 24 h prior to cytoplasmic and nuclear fractionation. PRP-1 was initially isolated from the brain hypothalamus (4).

To measure cells with ALDH activity, the Aldefluor^® assay was carried out as described according to manufacturer's protocol (Aldefluor™ kit, cat. no. 01700; StemCell Technologies, Vancouver, BC, Canada). Briefly, cells were harvested and resuspended in Aldefluor™ assay buffer at a concentration of 1×10⁶/ml. To activate the Aldefluor™ reagent, first 25 µl of DMSO was added and incubated for 15 min with 25 µl of 2N HCl, then 360 µl of assay buffer was added to the vial. The cells were then incubated with the activated Aldefluor™ reagent for 45 min at 37°C. Diethylaminobenzaldehyde (DEAB), a specific ALDH inhibitor, was added as a negative control. Following incubation, all tubes were centrifuged for 5 min at 250 × g, and the supernatant was removed, and resuspended in Aldefluor™ assay buffer. The cells were then transferred and strained onto Falcon 5 ml polystyrene round bottom tube with a cell strainer cap (cat. no. 352235). After labeling, the samples were sorted on a BD Biosciences (San Jose, CA, USA) Special Order Research Product (SORP) FACSAria II, using BD FACSDiva software (version 6.1.3) into ALDH^low and ALDH^high cells with and without PRP-1 treatment. Data analysis was performed using FlowJo software (FlowJo LLC, Ashland, OR, USA) (version 10).

The human chondrosarcoma JJ012 cells were sorted into cryovials and flash frozen in liquid nitrogen for shipment. Each group was repeated in triplicate. The RT² profiler PCR array was carried out by Qiagen, where RNA isolation and quality control were completed (WNT Signaling Pathway RT² Profiler PCR Array, cat. no. PAHS-043Z; Qiagen, Inc., Valencia, CA, USA).

Fold-change was calculated using the ΔΔCt method. Fold-change (2^−ΔΔCt) is the normalized gene expression in the test sample divided by the normalized gene expression in the control sample. Fold-regulation represents fold-change results in a biologically meaningful way. Fold-change values greater than one indicated a positive or an upregulation, and the fold-regulation is equal to the fold-change. Fold-change values less than one indicated a negative or downregulation, and the fold-regulation was the negative inverse of the fold-change. The P-values were calculated based on a Student's t-test of the replicate 2^−ΔΔCt values for each gene in the control group and treatment groups.

JJ012 chondrosarcoma cells were cultured and incubated to confluency. A subset of cells was treated with 10 µg/ml of PRP-1 for 24 h. Cells were collected using trypsin and then seeded directly onto coverslips (5×10⁵ cells/coverslip) placed into 6-well clusters. Cells were cultured overnight at 37°C in a 5% CO₂ incubator. After 24 h, the medium was removed, and samples were fixed using 1 ml of 4% formaldehyde solution (F8775; Sigma-Aldrich; Merck KGaA) in phosphate-buffered saline (PBS), pH 7.4 1X (Gibco; Thermo Fisher Scientific) (10010–023) for 15 min in the incubator at 37°C. Samples were then washed with PBS twice. Permeabilization of samples was completed with PBS/Triton X-100 1% (T9284; Sigma-Aldrich; Merck KGaA) for 5 min at room temperature. The detergent was removed, and non-specific sites were blocked using PBS containing 2% bovine serum albumin (BSA, A2153; Sigma-Aldrich; Merck KGaA) at room temperature for 30 min. Further incubation of samples was completed adding primary antibodies: Alexa Fluor Conjugate 594 WGA (Thermo Fisher Scientific, Inc.; cat. no. W11262) at a dilution of 1:200 incubated for 10 min at room temperature in the dark; E-cadherin (Abcam; AB1416) at a dilution of 1:100 incubated for 60 min at room temperature in the dark; and β-catenin (Abcam; cat. no. AB223075) at a dilution of 1:100 incubated for 24 h in a cold room (4°C) on a rocker. The following morning, two consecutive washes and incubation were completed with PBS solution. Secondary antibodies anti-rabbit DyLight 488 (Abcam; cat. no. AB96899) at a dilution of 1:500 and anti-mouse DyLight 550 (Abcam; cat. no. AB96872) at a dilution of 1:500 were added and samples were incubated for 60 min at room temperature in the dark. The second fixation step using formaldehyde for 15 min at room temperature was performed followed by two washing steps. 4′,6-Diamino-2-phenylindole dihydrochloride (3 µM) (DAPI; D1306; Thermo Fisher Scientific, Inc.) was used for nuclear staining for 10 min at room temperature. This was followed by two PBS washes. Antifade mounting medium was used to mount the samples on coverslips. ProLong Gold Antifade reagent (p10144; Life Technologies) was applied directly to fluorescently labeled cells on microscope slides to be used as a liquid mount and to protect fading of fluorescent dyes during microscopy.

Image acquisition was performed by the Analytical Imaging Core Facility at DRI/SCCC, University of Miami (FL, USA). Zeiss 200M, ApoTome fluorescent microscope, DAPI 49, GFP 38HE, Cy3 43, Cy5 50 filter cubes (Carl Zeiss Miscroscopy), heated stage, Orca II ERG Hamamatsu b/w 14-bit camera and AxioVision acquisition software were used. The coverslips were placed in regular 35-mm Petri dishes and the cells were grown on them, covered with medium. Once the cells were grown, the coverslips were taken out, and the cells were fixed, stained and mounted on glass slides. For imaging controls secondary antibodies were used without the primaries.

JJ012 chondrosarcoma cells were cultured and incubated to confluency. Cells were collected using trypsin and then seeded into Petri dishes at a concentration of 1×10⁶ cells/ml. The cells were incubated for 24 h at 37°C in a 5% CO₂ incubator. The next day, an ice-cold phosphate-buffered saline wash was performed, and protease inhibitor was added to the cell lysis buffer (C2978; Sigma-Aldrich; Merck KGaA) in a 1:100 ratio. After the collection of cells with a rubber scraper and lysis of cell membranes with an 18-gauge needle, the cells were centrifuged at 15,000 × g at 4°C. The supernatant was then collected, and protein content was measured using NanoDrop^® spectrophotometer (Thermo Fisher Scientific, Inc.). The supernatant was frozen at −80°C until loading onto the gels (20 µg/lane). Polyacrylamide gel electrophoresis and western blotting reagents were supplied by Lonza, Inc. (Allendale, NJ, USA) and related procedures were followed in accordance with the company's protocol. The catalog numbers for the reagents and suppliers are listed below. Pager Gold Precast Gels (59502; 10% Tris-glycine; Lonza, Inc.); ECL reagent (RPN2109; GE Healthcare, Little Chalfont, UK); Western Blocker solution (W0138; Sigma-Aldrich; Merck KGaA); ProSieve Quad Color Protein marker (4.6–300 kDa, 00193837; Lonza, Inc.); 20X reducing agent for ProSieve ProTrack Dual Color Loading buffer (00193861; Lonza, Inc.); ProTrack loading buffer (00193861; Lonza, Inc.); ProSieve ProTrack Dual Color Loading buffer EX running buffer (00200307; Lonza, Inc.); ProSieve EX Western Blot Transfer buffer (00200309; Lonza, Inc.); Immobilon^®-P polyvinylidene difluoride membranes (P4188; Sigma-Aldrich; Merck KGaA).

Cytoplasmic and nuclear fractions of cells were isolated following the manufacturer's instructions (cat. no. 40010, lot. no. 22118072; Active Motif) using phosphatase inhibitors, protease cocktail inhibitor and sonicator.

CGP57380 is a cell-permeable selective inhibitor of β-catenin nuclear translocation which was added in incrementally increasing doses of 1, 5, 10 and 20 µM for a period of 24 h prior to the assay (C0993, Sigma-Aldrich; Merck KGaA).

DEAB, a specific ALDH inhibitor, from the Aldefluor™ kit was added as a negative control (5 µl/ml) for a period of 24 h prior to lysate generation.

Rabbit polyclonal antibody to β-catenin was applied as a primary antibody (ab2365, Abcam) at a dilution of 1:1,000 and goat anti-rabbit IgG peroxidase conjugate as a secondary antibody (A0545; Sigma-Aldrich; Merck KGaA) at a dilution of 1:5,000. As housekeeping proteins, mouse anti-tubulin antibody was applied for cytoplasmic fractions (T5168; Sigma-Aldrich; Merck KGaA) at a dilution of 1:2,000 and anti-mouse IgG (A4416; Sigma-Aldrich; Merck KGaA) at a dilution of 1:5,000 was applied as a secondary antibody. Mouse anti-TBP was used for nuclear fractions (T1827, Sigma-Aldrich; Merck KGaA) at a dilution of 1:1,000 and anti-mouse IgG (A4416; Sigma-Aldrich; Merck KGaA) at a dilution of 1:5,000 was applied as a secondary antibody. Incubations for all primary antibodies were carried out in a cold room while rocking for a period of 24 h, while secondary antibodies were incubated for 2 h under the same conditions.

Quantitative analysis and densitometry were obtained using integrated density analysis on ImageJ 1.52e (NIH; National Institutes of Health, Bethesda, MD, USA) to calculate relative optical density (OD) of β-catenin to the housekeeping protein. Bulk JJ012 was used as the control.

Statistical analyses were performed using individual unpaired t-tests for flow cytometry experiments, which were repeated 10 times. Statistical analyses of relative ODs were completed using one-way analysis of variance (ANOVA) with a post hoc Dunnett's multiple comparisons test of all samples to control bulk JJ012 cells expressed as 95% confidence intervals of mean difference. All western blot experiments were repeated 2 times. All statistical analyses were completed using GraphPad Prism 8.0.2 (GraphPad Software, Inc., San Diego, CA, USA). A P-value <0.05 was considered significant. Error bars represent standard error of the mean (SEM) in all graphs with *P<0.05, **P<0.01, ***P<0.001 (as indicated in the figures and figure legends).

Considering the demonstrated cytostatic, antiproliferative and tumor-suppressor properties of PRP-1 and the established ALDH^high CSC populations in chondrosarcoma cell lines, we sought to determine whether PRP-1 plays a role in the expression of stem cell characteristics in bulk JJ012 human chondrosarcoma cells. To test this, we cultured a group of bulk JJ012 with PRP-1 and without PRP-1. These cells underwent Aldefluor^® assay and ALDH expression was detected using FACS. ALDH inhibitor, DEAB, served as the negative control for both treated and untreated cells to ensure the accuracy of the analysis. Unstained untreated and PRP-1-treated cells showed no background fluorescence. A set of untreated cells stained with Aldefluor^® in the presence of DEAB showed 0.5% ALDH^high cells (Fig. 1A) and PRP-1-treated cells stained with Aldefluor^® in the presence of DEAB showed 0.1% ALDH^high cells (Fig. 1C).

A set of PRP-1-treated cells in the absence of DEAB showed 3.1% ALDH^high cells (Fig. 1D) compared to a set of untreated cells in the absence of DEAB which showed 75.3% ALDH^high cells (Fig. 1B). A statistically significant difference was found between the ALDH^low populations in the untreated [mean (M)=51.10%, SEM=12.00%, n=10] and PRP-1-treated JJ012 cells (M=88.39%, SEM=2.14%, n=10) stained with Aldefluor^®; t(18)=3.059, P=0.0068 (Fig. 1E). There was also a statistically significant difference between the ALDH^high populations in the untreated (M=42.49%, SEM=11.82%, n=10) and PRP-1-treated JJ012 cells (M=6.93%, SEM=1.35%, n=10) stained with Aldefluor^®; t(18)=2.990, P=0.0079 (Fig. 1E). These results demonstrated that PRP-1 treatment significantly decreased the ALDH^high CSC population from a bulk JJ012 cell population. Subpopulations were collected using flow cytometry and consisted of: ALDH^high cells sorted from untreated JJ012 cells (ALDH^{high-untreated}), ALDH^low cells sorted from untreated JJ012 cells (ALDH^{low-untreated}) and ALDH^low cells sorted from PRP-1-treated JJ012 cells (ALDH^low-PRP-1). ALDH^high cells were not able to be collected from the PRP-1-treated JJ012 cells due to the exceedingly low levels of these CSC populations after PRP-1 treatment.

After demonstrating the ability of PRP-1 to eliminate CSCs in JJ012 cells, we sought to understand the pathways involved in this population shift. Studies connecting ALDH expression to Wnt/β-catenin signaling pathway activation in many tumor types, including prostate cancer, liver cancer and breast cancer, drove us to explore this pathway in chondrosarcoma cells (19–21). We used the RT² profiler PCR arrays performed by Qiagen to explore the Wnt/β-catenin signaling pathway in PRP-1-treated and untreated cells.

Our first comparison was between ALDH^{high-untreated} cells and bulk JJ012 cells. Wnt signaling genes were profiled on three samples with technical triplicates for each group. Notably, ALDH^{high-untreated} cells demonstrated significant downregulation of 6 Wnt signaling genes, including TCF7L1, WNT3, FZD7, FOSL1, WNT7B and FZD6 and upregulation of only one Wnt signaling gene, PORCN, with putative transcription factors involved listed (Table I). TCF7L1 had the highest fold downregulation (6.33-fold) (P=0.000543). WNT3, FZD7, FOSL1, WNT7B and FZD6 were downregulated in a range from 2.48- to 2.76-fold. PORCN showed 2.43-fold upregulation (P=0.006511). miRNAs that regulate these downregulated Wnt signaling genes (FZD7, FZD6, FOSL1, TCF7L1 and WNT7B) in the ALDH^{high-untreated} cells were identified (Table II).

In the present study, we compared ALDH^low-PRP-1 cells to bulk JJ012 cells to determine whether addition of PRP-1 results in a differential expression of Wnt genes compared to an untreated and unsorted JJ012 population using a RT² profiler PCR array. We identified significant upregulation of 5 Wnt signaling genes in the PRP-1-treated ALDH^low cells, including BCL9, PORCN, RHOU, FZD2 and RPLP0, with putative transcription factors involved listed (Table III). MMP7 was 3.50-fold down-regulated (P=0.018004) in the ALDH^low-PRP-1 cells compared to the bulk JJ012 cells. Notably, PORCN was 2.43-fold upregulated (P=0.006511) in the ALDH^{high-untreated} cells (Table I) while PORCN was 3.54-fold upregulated (P=0.030693) in the ALDH^low-PRP-1 cells (Table III) compared to the bulk JJ012 cells. miRNAs that regulate the upregulated Wnt signaling genes RHOU and BCL9 in ALDH^low-PRP-1 cells were identified (Table IV).

After showing PRP-1 treatment eliminates CSCs in JJ012 cells, we aimed to determine the differences in Wnt signaling genes between ALDH^{high-untreated} cells and ALDH^low-PRP-1 cells using a RT² profiler PCR array.

The arrays identified two significantly downregulated cancer genes from the Wnt pathway in the ALDH^low-PRP-1 cells, including the CCND2 gene, encoding G1/S specific cyclin D2 and the MMP7 gene, encoding a matrix metalloproteinase, with putative transcription factors involved listed (Table V). CCND2 was downregulated 4.51-fold (P=0.004252) and MMP7 was downregulated 3.25-fold (P=0.000044). miRNAs that regulate the downregulated Wnt signaling gene CCND2 in the ALDH^low-PRP-1 cells were identified (Table VI). Numerous studies have demonstrated the importance of miRNA regulators, their downregulation, and their role in overexpression of CCND2 in association with high-grade osteosarcomas and resistance to chemotherapy (22–25). In fact, CCND2 was found to be upregulated in metastatic osteosarcoma compared to the primary tumor (26). Together, these experimental results demonstrate the important role that CCND2 plays in the development of the progression of cancer, chemoresistance and metastasis that may also be present in chondrosarcoma.

Immunofluorescent stained cell images demonstrated that β-catenin was present in the nuclei in the PRP-1-treated JJ012 cells vs. in the cytoplasm of untreated JJ012 cells. Single-cell images are depicted (Fig. 2A-D) and a wider field of multiple cells is shown (Fig. 2E and F).

To confirm the findings from the RT² profiler PCR arrays and immunocytochemistry, western blot analysis was carried out. One-way ANOVA found significant differences in the mean relative optical density (OD) of β-catenin between bulk JJ012 cells, ALDH^{low-untreated} and ALDH^low-PRP-1 cells [F(2,3)=70.60, P=0.0030]. Post hoc Dunnett's tests determined no significance difference between ALDH^{low-untreated} and bulk JJ012 cells, but a significant increase in relative OD of β-catenin was found between ALDH^low-PRP-1 and bulk JJ012 cells (1.834±0.611, P=0.0023) (Fig. 2G and H). This supports the finding that PRP-1 regulates β-catenin protein expression.

To determine whether PRP-1 has an effect on subcellular regulation of the Wnt/β-catenin pathway, western blot analysis was performed on cytoplasmic and nuclear fractions using bulk JJ012 cells as a control compared to various combinations of PRP-1, CGP57380 (CGP; a β-catenin nuclear translocation inhibitor) and DEAB (a direct ALDH inhibitor). One-way ANOVA of the western blot experiments with incrementally increasing PRP-1 concentrations found a statistically significantly difference in relative OD of β-catenin between the groups [F(4,5)=67.93, P=0.0002]. Post hoc Dunnett's test demonstrated a significant decrease in the mean relative OD in cells treated with 1 µg PRP-1 (−0.1867±0.0762, P=0.0011), 5 µg PRP-1 (−0.1923±0.0763, P=0.0010), 10 µg PRP-1 (−0.2555±0.0762, P=0.0003) and 20 µg PRP-1 (−0.3485±0.0762, P<0.0001) compared to the bulk JJ012 control cells (Fig. 3A and F). One-way ANOVA of experiments with the addition of β-catenin nuclear translocation inhibitor CGP57380 (27,28) in increasing concentrations demonstrated no significant differences in mean relative OD of cytoplasmic β-catenin protein between the groups [F(4,5)=2.649, P=0.1570] (Fig. 3B and G). When adding CGP57380 (CGP) with increasing PRP-1 concentrations, one-way ANOVA demonstrated significant differences in mean relative OD of cytoplasmic β-catenin protein between the groups [F(4,5)=16.71, P=0.0043]. Post hoc analysis using Dunnett's test demonstrated a significant increase in relative OD of cytoplasmic β-catenin with 5 µg PRP-1 + 5 µM CGP (2.076±1.1198, P=0.0040), 10 µg PRP-1 + 10 µM CGP (1.881±1.0504, P=0.0062) and 20 µg PRP-1 +20 µM CGP (1.754±1.1199, P=0.0083) compared to the bulk JJ012 control cells (Fig. 3C and H). One-way ANOVA analysis of JJ012 cells treated with DEAB found significant differences in the mean relative OD of cytoplasmic β-catenin between groups [F(3,4)=62.33, P=0.0008]. Post hoc Dunnett's test found a significant decrease in relative OD of cytoplasmic β-catenin following treatment with DEAB alone (−0.2969±0.2876, P=0.0452), 10 µg PRP-1 alone (−0.7082±0.2876, P=0.0020) and 10 µg PRP-1 + DEAB (−1.006±0.2873, P=0.0005) compared to the bulk JJ012 control cells (Fig. 3D and I). When examining the relative OD of nuclear β-catenin in the JJ012 cells treated with PRP-1 and CGP, one-way ANOVA found significant differences in means between groups [F(3,4)=1182, P<0.0001]. Post hoc analysis using Dunnett's test demonstrated a significantly increased intensity of nuclear β-catenin in the high-dose (20 µg) PRP-1 (0.2612±0.0797, P=0.0007) and significantly decreased intensity of nuclear β-catenin following 20 µg PRP-1 + 20 µM CGP (−0.9474±0.0797, P<0.0001) compared to the bulk JJ012 control cells (Fig. 3E and J).