Journal: Oncotarget
Article Title: Chondroitin sulfatases differentially regulate Wnt signaling in prostate stem cells through effects on SHP2, phospho-ERK1/2, and Dickkopf Wnt signaling pathway inhibitor (DKK3)
doi: 10.18632/oncotarget.22152
Figure Lengend Snippet: (A) Phospho-ERK1/2 was determined by sandwich ELISA, in which total ERK1/2 was first captured in the wells of an ELISA plate. A second antibody was used to detect phospho-ERK1/2. In the prostate stem cells, GALNS overexpression and ARSB silencing increased the phospho-ERK1/2 (p<0.001, n=3). In contrast GALNS silencing and ARSB OE reduced the phospho-ERK1/2 (p<0.001, n=3). The ERK activity inhibitor peptide I was effective in reversing the effect of the GALNS OE and ARSB silencing. (B) In human prostate tissue, phospho-ERK1/2 was significantly increased in the malignant tissue, compared to normal (p<0.001, n=6; unpaired t-test, two-tailed). (C) Decline in SHP2 activity, due to transfection with a dominant negative SHP2 DNA construct, led to significant increase in phospho-ERK1/2 in the prostate stem cells (p<0.001, n=3). In contrast, the constitutively active SHP2 construct reduced the phospho-ERK1/2 (p<0.001, n=3). (D) SHP2 activity was determined by measurement of phosphate released from a synthetic phosphopeptide, following isolation of SHP2 by anti-SHP2 antibody conjugated to agarose beads. ARSB silencing and GALNS overexpression reduced the SHP2 activity in the prostate stem cells (p<0.001, n=3). In contrast, GALNS silencing and ARSB overexpression increased the SHP2 activity (p<0.001, n=3). These effects are attributed to increased binding of SHP2 to C4S when ARSB was silenced or GALNS was overexpressed. (E) In the prostate stem cells, ARSB silencing significantly reduced the SHP2 activity. The dominant negative (DN) SHP2 DNA construct further reduced the SHP2 activity (p<0.001, n=3). The effect of ARSB silencing was inhibited by the constitutively active (CA) SHP2 DNA construct (p<0.001, n=3). (F) In the malignant human prostate tissue, the SHP2 activity was reduced ∼50% (p<0.001, n=6; unpaired t-test, two-tailed), attributable to the previously determined increase in C4S in the malignant tissue. (G) Both the dominant negative SHP2 DNA construct and PHPS1 (30 μM x 24 h), a chemical SHP2 inhibitor, blocked DKK3 mRNA expression. In contrast, constitutively active SHP2 increased the mRNA DKK3 expression (p<0.001, n=6). These results indicate the involvement of SHP2 in the expression of DKK3. [ARSB=arylsulfatase B=N-acetylgalactosamine-4-sulfatase; CA=constitutively active; DKK=Dickkopf Wnt inhibitory factor; DN=dominant negative; DNMT=DNA methyltransferase; ERK=extracellular-signal regulated kinase; GALNS=galactosamine-(N-acetyl)-6-sulfatase; N-acetylgalactosamine-6-sulfatase; galactose-6-sulfate sulfatase; OE=overexpression; SHP2=non-receptor tyrosine phosphatase; si=siRNA].
Article Snippet: The human prostate stem cell line was obtained from ATCC (CRL-2887; Manassas, VA) and grown in Keratinocyte Serum Free Medium (K-SFM) with 0.05 mg/ml bovine pituitary extract (BPE) and 5 ng/ml epidermal growth factor (EGF), and maintained at 37°C in a humidified, 5% CO 2 environment with replenishment of media every third day, as recommended.
Techniques: Sandwich ELISA, Enzyme-linked Immunosorbent Assay, Over Expression, Activity Assay, Two Tailed Test, Transfection, Dominant Negative Mutation, Construct, Phospho-proteomics, Isolation, Binding Assay, Expressing
ATCC is a verified supplier 
ATCC manufactures this product 
![(A) ARSB activity was measured in <t>prostate</t> <t>stem</t> cells using the exogenous substrate 4-methylumbelliferyl sulfate. ARSB activity was significantly reduced by ARSB knockdown by specific siRNA and increased by ARSB overexpression using ARSB plasmid in a pCMV6-XL4 vector in the prostate stem cells (p<0.001, n=3). GALNS silencing or overexpression did not affect the ARSB activity. (B) GALNS activity was measured using the exogenous substrate 4-methylumbelliferyl-β-D-galactoside-6-sulfateNH 4 . GALNS activity was significantly reduced by GALNS siRNA and increased by GALNS overexpression using GALNS plasmid in a pCMV6-XL4 vector (p<0.001, n=3). ARSB silencing or overexpression did not affect the GALNS activity. (C) In malignant prostate tissue, the ARSB activity was significantly lower than in the normal <t>human</t> prostate tissue (p<0.0001, n=6, unpaired t-test, two-tailed). (D) In contrast, the GALNS activity was significantly higher in the malignant tissue (p<0.0001, n=6, unpaired t-test, two-tailed). [ARSB = arylsulfatase B = N-acetylgalactosamine-4-sulfatase; GALNS = galactosamine-(N-acetyl)-6-sulfatase; N-acetylgalactosamine-6-sulfatase; galactose-6-sulfate sulfatase; OE = overexpression; si = siRNA].](https://pub-med-central-images-cdn.bioz.com/pub_med_central_ids_ending_with_5016/pmc05725016/pmc05725016__oncotarget-08-100242-g001.jpg)
![(A) Total sulfated glycosaminoglycans (GAGs) were measured using the Blyscan™ assay which detects sulfated GAGs by binding to 1,9-dimethylmethylene blue. In the prostate stem cells, total sulfated glycosaminoglycans (GAGs) were increased following silencing of ARSB or of GALNS (p<0.001, n=3). In contrast, overexpression of ARSB or of GALNS decreased the total sulfated GAGs (p<0.001, n=3). (B) Chondroitin-4-sulfate (C4S) was measured by the Blyscan™ assay, following immunoprecipitation by antibody specific for C4S. C4S was significantly increased following ARSB silencing and reduced when ARSB was overexpressed (p<0.001, n=3). Changes in GALNS expression did not affect the level of C4S. (C) Chondroitin 6-sulfate was detected by the Blyscan™ assay, following immunoprecipitation with an antibody specific for C6S. When GALNS was silenced, chondroitin 6-sulfate (C6S) increased significantly, and declined when GALNS was overexpressed (p<0.001, n=3). Changes in ARSB expression did not affect the C6S level. (D) The C4S/C6S ratio was calculated and shown to be increased when ARSB was silenced or GALNS was overexpressed (p<0.001, n=3). The ratio was reduced when ARSB was overexpressed or GALNS was silenced. (E) In the human prostate tissues, C6S and C4S were measured by the Blyscan™ assay. C4S was increased and C6S was reduced in the malignant tissue (p<0.001, n=6; unpaired t-test, two-tailed), consistent with decrease in ARSB activity and increase in GALNS activity. Overall, total sulfated GAGs were significantly increased in the malignant tissue, compared to the normal tissue (p<0.01, n=6). (F) The C4S/C6S ratio was calculated and was increased in the malignant tissue, compared to the normal tissue (p<0.001, n=6; unpaired t-test, two-tailed). [ARSB=arylsulfatase B=N-acetylgalactosamine-4-sulfatase; C4S=chondroitin 4-sulfate; C6S=chondroitin 6-sulfate; GALNS=galactosamine-(N-acetyl)-6-sulfatase; N-acetylgalactosamine-6-sulfatase; galactose-6-sulfate sulfatase; GAG=glycosaminoglycan; OE=overexpressed; si=siRNA].](https://pub-med-central-images-cdn.bioz.com/pub_med_central_ids_ending_with_5016/pmc05725016/pmc05725016__oncotarget-08-100242-g002.jpg)
![(A) Nuclear β-catenin was measured by ELISA in nuclear extracts of the prostate stem cells following ARSB and GALNS silencing and overexpression. Nuclear ß-catenin increased significantly following ARSB silencing or GALNS overexpression (p<0.001, n=3). Inversely, GALNS silencing or ARSB overexpression reduced the nuclear ß-catenin (p<0.001, n=3). (B) Nuclear ß-catenin was measured in nuclear extracts from normal and malignant human prostate tissue. Nuclear ß-catenin was significantly increased in the malignant tissue (p<0.01, unpaired t-test, two-tailed, n=6). (C) Nuclear DNA-bound TCF/LEF was determined by a transcription factor reporter assay in the prostate stem cells. A biotin-labeled TCF/LEF DNA binding sequence probe which detected TCF/LEF bound to DNA was mixed with nuclear extracts to form TCF/LEF-DNA complexes. A filter plate was used to retain the bound DNA probe and remove free probe. The bound prelabeled DNA probe was eluted from the filter and collected for quantitative determination. The bound TCF/LEF increased following either ARSB silencing or GALNS overexpression (p<0.001, n=3). In contrast, ARSB overexpression and GALNS silencing inhibited the increase (p<0.001, n=3). (D) Further demonstration of the impact of the chondroitin sulfatases was shown by effects on the mRNA expression of Wnt/ß-catenin dependent genes. QPCR showed increased expression of c-Myc and GATA-3 following ARSB silencing or GALNS overexpression. In contrast, overexpression of ARSB or silencing of GALNS reduced the mRNA expression of c-Myc and GATA-3 (p<0.001, n=6). [ARSB=arylsulfatase B=N-acetylgalactosamine-4-sulfatase; C4S=chondroitin 4-sulfate; C6S=chondroitin 6-sulfate; GALNS=galactosamine-(N-acetyl)-6-sulfatase; N-acetylgalactosamine-6-sulfatase; galactose-6-sulfate sulfatase; GAG=glycosaminoglycan; OE=overexpressed; si=siRNA; TCF/LEF=T-cell factor/lymphoid enhancer-binding factor].](https://pub-med-central-images-cdn.bioz.com/pub_med_central_ids_ending_with_5016/pmc05725016/pmc05725016__oncotarget-08-100242-g003.jpg)
![(A) When the prostate stem cells were treated with the DNA hypomethylating agent 5-azacytidine (10 μM x 24 h), the ARSB silencing- or GALNS overexpression- induced increases in nuclear β-catenin were inhibited (p<0.001, n=3). This indicated that a transcriptional mechanism was required for the effects of ARSB siRNA and GALNS overexpression on nuclear ß-catenin. (B) Similarly, the effects of ARSB silencing or GALNS overexpression on TCF/LEF binding to nuclear DNA were inhibited by treatment with the DNA hypomethylating agent, 5-azacytidine, (p<0.001, n=3). This indicated that a transcriptional mechanism was required for the activation of Wnt/β-catenin signaling, as manifested by effects of ARSB siRNA and GALNS overexpression on TCF/LEF nuclear-DNA binding. (C) QPCR was performed using standard quantitative methods and established primers. The increased mRNA expression of c-Myc and of GATA-3 following either ARSB silencing or GALNS overexpression was inhibited by 5-azacytidine (p<0.001, n=6). These effects are consistent with dependence on DNA methylation for the observed increases in manifestations of Wnt/ß-catenin signaling following changes in activity of chondroitin sulfatases ARSB and GALNS. (D) Treatment with JW67 (4 mg/ml x 24 h), an inhibitor of the Wnt/ß-catenin signaling pathway, also blocked the ARSB silencing-induced increases in mRNA expression of c-Myc and GATA-3 (p<0.001, n=6). This finding indicates that the increased activation of Wnt/β-catenin signaling was also required to increase the expression of these Wnt target genes. (E) The effect of GALNS overexpression on mRNA expression of c-Myc and GATA-3 was also inhibited by JW67 (p<0.001, n=6). This finding indicated that the effects of ARSB silencing and GALNS overexpression on Wnt target genes were both mediated by activation of Wnt/β-catenin signaling. [ARSB=arylsulfatase B=N-acetylgalactosamine-4-sulfatase; 5-AZA=5-azacytidine; GALNS=galactosamine-(N-acetyl)-6-sulfatase; N-acetylgalactosamine-6-sulfatase; galactose-6-sulfate sulfatase; OE=overexpression; si=siRNA; TCF/LEF=T-cell factor/lymphoid enhancer-binding factor].](https://pub-med-central-images-cdn.bioz.com/pub_med_central_ids_ending_with_5016/pmc05725016/pmc05725016__oncotarget-08-100242-g004.jpg)
![(A) Whole genomic DNA from prostate stem cells in which ARSB and GALNS had been silenced or overexpressed and from control samples was obtained and fractionated. Methylated DNA was isolated by binding to the methyl-CpG binding domain of human MBD2 protein, which was coupled to paramagnetic Dynabeads R M-280 Streptavidin via a biotin linker. The methylated fragments were then eluted and subjected to QPCR with specific primers to the DKK3 promoter. DKK3 promoter methylation was increased when ARSB was silenced or GALNS overexpressed (p<0.001, n=6), and reduced when GALNS was silenced or ARSB overexpressed (p<0.001, n=6). (B) By methylation specific PCR using primers specific for both the methylated and unmethylated DKK3 promoter, the expression of the methylated DKK3 promoter was demonstrated on a 2% agarose gel. Band density was increased following GALNS overexpression and ARSB silencing (p<0.001, n=3). (C) Genomic DNA was isolated from normal and malignant prostate tissue and was fractionated. The methylated dsDNA was isolated by binding to MBD2 which was coupled to Dynabeads, as above. QPCR was performed to quantify the DKK3 promoter methylation. In the malignant prostate tissue, DKK3 promoter methylation was increased (p<0.0001, n=6; unpaired t-test, two-tailed), thereby inhibiting DKK3 expression and permitting increased Wnt signaling. [ARSB=arylsulfatase B=N-acetylgalactosamine-4-sulfatase; DKK=Dickkopf inhibitor of Wnt signaling pathway; GALNS=galactosamine-(N-acetyl)-6-sulfatase; N-acetylgalactosamine-6-sulfatase; galactose-6-sulfate sulfatase; OE=overexpression; si=siRNA].](https://pub-med-central-images-cdn.bioz.com/pub_med_central_ids_ending_with_5016/pmc05725016/pmc05725016__oncotarget-08-100242-g006.jpg)
![(A) Phospho-ERK1/2 was determined by sandwich ELISA, in which total ERK1/2 was first captured in the wells of an ELISA plate. A second antibody was used to detect phospho-ERK1/2. In the prostate stem cells, GALNS overexpression and ARSB silencing increased the phospho-ERK1/2 (p<0.001, n=3). In contrast GALNS silencing and ARSB OE reduced the phospho-ERK1/2 (p<0.001, n=3). The ERK activity inhibitor peptide I was effective in reversing the effect of the GALNS OE and ARSB silencing. (B) In human prostate tissue, phospho-ERK1/2 was significantly increased in the malignant tissue, compared to normal (p<0.001, n=6; unpaired t-test, two-tailed). (C) Decline in SHP2 activity, due to transfection with a dominant negative SHP2 DNA construct, led to significant increase in phospho-ERK1/2 in the prostate stem cells (p<0.001, n=3). In contrast, the constitutively active SHP2 construct reduced the phospho-ERK1/2 (p<0.001, n=3). (D) SHP2 activity was determined by measurement of phosphate released from a synthetic phosphopeptide, following isolation of SHP2 by anti-SHP2 antibody conjugated to agarose beads. ARSB silencing and GALNS overexpression reduced the SHP2 activity in the prostate stem cells (p<0.001, n=3). In contrast, GALNS silencing and ARSB overexpression increased the SHP2 activity (p<0.001, n=3). These effects are attributed to increased binding of SHP2 to C4S when ARSB was silenced or GALNS was overexpressed. (E) In the prostate stem cells, ARSB silencing significantly reduced the SHP2 activity. The dominant negative (DN) SHP2 DNA construct further reduced the SHP2 activity (p<0.001, n=3). The effect of ARSB silencing was inhibited by the constitutively active (CA) SHP2 DNA construct (p<0.001, n=3). (F) In the malignant human prostate tissue, the SHP2 activity was reduced ∼50% (p<0.001, n=6; unpaired t-test, two-tailed), attributable to the previously determined increase in C4S in the malignant tissue. (G) Both the dominant negative SHP2 DNA construct and PHPS1 (30 μM x 24 h), a chemical SHP2 inhibitor, blocked DKK3 mRNA expression. In contrast, constitutively active SHP2 increased the mRNA DKK3 expression (p<0.001, n=6). These results indicate the involvement of SHP2 in the expression of DKK3. [ARSB=arylsulfatase B=N-acetylgalactosamine-4-sulfatase; CA=constitutively active; DKK=Dickkopf Wnt inhibitory factor; DN=dominant negative; DNMT=DNA methyltransferase; ERK=extracellular-signal regulated kinase; GALNS=galactosamine-(N-acetyl)-6-sulfatase; N-acetylgalactosamine-6-sulfatase; galactose-6-sulfate sulfatase; OE=overexpression; SHP2=non-receptor tyrosine phosphatase; si=siRNA].](https://pub-med-central-images-cdn.bioz.com/pub_med_central_ids_ending_with_5016/pmc05725016/pmc05725016__oncotarget-08-100242-g008.jpg)