These data support previous studies in human GBM demonstrating that SULF2, an extracellular endosulfatase that removes 6- em O /em -sulfate from trisulfate disaccharides and can promote PDGFRA signaling, is most highly expressed in a subset of GBM (21). demonstrate differences in HS disaccharide content and structure across four patient-derived Ercalcidiol tumorsphere lines (GBM1, 5, 6, 43) and between two murine tumorsphere lines derived from murine GBM with enrichment of mesenchymal and proneural gene expression (mMES and mPN, respectively) markers. In GBM, the heterogeneous HS content and structure across patient-derived tumorsphere lines suggested diverse functions in the GBM tumor microenvironment (TME). In GBM5 and mPN, elevated expression of sulfatase 2 (SULF2), an extracellular enzyme that alters ligand binding to HS, was associated with low trisulfated HS disaccharides, a substrate of SULF2. In contrast, other primary tumorsphere lines had elevated expression of the HS-modifying enzyme heparanase (HPSE). Using gene editing strategies to inhibit HPSE a role for HPSE in promoting tumor cell adhesion and invasion was identified. These studies characterize the heterogeneity in HS glycosaminoglycan content and structure across GBM and uncover their role in tumor cell invasion. sulfate from trisulfate disaccharides (12,43). GBM5 had the lowest trisulfate abundance relative to the other GBM (Physique 1F (inset), p 0.05). GBM5 also had the highest expression of mRNA across the four lines (Physique 1G, p 0.0001). In contrast, was expressed at similar levels (Physique 1H). Open in a separate window Physique 1 Analysis of HS disaccharide structure and amount from human GBM tumorspheres(ACD) Extracted ion chromatography profile of HS disaccharides isolated from human GBM tumorspheres demonstrating intensity profiles (counts per second) over time (min) across tumors. Based on the analysis of disaccharide standards (Supplemental Figures 2C3), each peak could be assigned a disaccharide with different sulfate patterns. (E) Absolute HS abundance normalized to total protein extracted (ng/mg) across tumorspheres. (F) Relative HS disaccharide abundance as a percentage (%) of HS disaccharide composition. Inset highlights the comparison of the relative percentage of trisulfate content. (GCH) Relative SULF1 and SULF2 gene expression levels in tumorspheres, normalized to GBM43 (n=3). Representative data (ACD) from biologic replicates. Symbols denote: NAc, N-acetylglucosamine; NS, N-sulfated glycosamine; 6S, 6-O-sulfated glucosamine; 2S, 2-O-sulfated iduronic acid; and trisulfate, NS2S6S. Mean SEM. *, p 0.05; **, p 0.01; ***, p 0.001; ****, p 0.0001. Heparanase (was most highly expressed in the mesenchymal GBM subtype (Physique 2A; p 0.0001), a subtype proposed to be the most therapy resistant (47). To examine potential functions for HPSE in GBM43, CRISPR-Cas9 gene editing was used to introduce deletions in exon 3 of HPSE. Clones with heterozygous deletions in and reduced HPSE expression relative to control clones were isolated (Physique 2B p 0.001). HPSE reduced clones did not exhibit a growth defect relative to control clones in vitro (Supplemental Physique 4B). To assess cell invasion in the context of a three-dimensional matrix, tumorspheres were embedded in matrigel and tumor cell invasion away from the sphere was imaged and quantified over time. Reduced HPSE conferred a marked decrease in tumor cell invasion (Physique 2CCD, p 0.0001). Cell invasion in a three-dimensional matrix requires coordinated regulation of cell attachment, migration, and ECM breakdown. Given Ercalcidiol the reported functions for HPSE and HS in cell invasion and adhesion, we asked whether decreased invasion in HPSE-reduced cells was associated with reduced cell adhesion. reduced GBM43 had markedly reduced cell adhesion to laminin relative to control cells (Physique 2ECF, p 0.01). Addition of exogenous HPSE promoted invasion of HPSE-reduced GBM43 cells relative to control treated cells (Physique 2GCH, p 0.05). Open in a separate window Physique 2 HPSE expression is increased in the mesenchymal GBM subtype and can promote tumor cell invasion(A) Increased expression of HPSE in human GBM of the mesenchymal (MES) transcriptional subtype. Bars denote the mean SEM Z-score for tumors in the specified subtype (1) with mesenchymal (red) (n=56) and non-mesenchymal (white) (n=139); The Cancer Genome Atlas (TCGA) Data Portal (63). (B) Focal deletion in HPSE confers decreased HPSE mRNA expression in.N.S., not significant (n=6). GBM with enrichment of mesenchymal and proneural gene expression (mMES and mPN, respectively) markers. In GBM, the heterogeneous HS content and structure across patient-derived tumorsphere lines suggested diverse functions in the GBM tumor microenvironment (TME). In GBM5 and mPN, elevated expression of sulfatase 2 (SULF2), an extracellular enzyme that alters ligand binding to HS, was associated with low trisulfated HS disaccharides, a substrate of SULF2. In contrast, other primary tumorsphere lines had elevated expression of the HS-modifying enzyme heparanase (HPSE). Using gene editing strategies to inhibit HPSE a role for HPSE in promoting tumor cell adhesion and invasion was identified. These studies characterize the heterogeneity in HS glycosaminoglycan content and structure across GBM and uncover their role in tumor cell invasion. sulfate from trisulfate disaccharides (12,43). GBM5 had the lowest trisulfate abundance relative to the other GBM (Physique 1F (inset), p 0.05). GBM5 also had the highest expression of mRNA across the four lines (Physique 1G, p 0.0001). In contrast, was expressed at similar levels (Physique 1H). Open in a separate window Physique 1 Analysis of HS disaccharide structure and amount from human GBM tumorspheres(ACD) Extracted ion chromatography profile of HS disaccharides isolated from human GBM tumorspheres demonstrating intensity profiles (counts per second) over time (min) across tumors. Based on the analysis of disaccharide standards (Supplemental Figures 2C3), each peak could be assigned a disaccharide with different sulfate patterns. (E) Absolute HS abundance normalized to total protein extracted (ng/mg) across tumorspheres. (F) Relative HS disaccharide abundance as a percentage (%) of HS disaccharide composition. Inset highlights the comparison of the relative percentage of trisulfate content. (GCH) Relative SULF1 and SULF2 gene expression levels in tumorspheres, normalized to GBM43 (n=3). Representative data (ACD) from biologic replicates. Symbols denote: NAc, N-acetylglucosamine; NS, N-sulfated glycosamine; 6S, 6-O-sulfated glucosamine; 2S, 2-O-sulfated iduronic acid; and trisulfate, NS2S6S. Mean SEM. *, p 0.05; **, p 0.01; ***, p 0.001; ****, p 0.0001. Heparanase (was most highly expressed in the mesenchymal GBM subtype (Physique 2A; p 0.0001), a subtype proposed to be the most therapy resistant (47). To examine potential functions for HPSE in Ercalcidiol GBM43, CRISPR-Cas9 gene PP2Bgamma editing was used to introduce deletions in exon 3 of HPSE. Clones with heterozygous deletions in and reduced HPSE expression relative to control clones were isolated (Physique 2B p 0.001). HPSE reduced clones did not exhibit a growth defect relative to control clones in vitro (Supplemental Physique 4B). To assess cell invasion in the context of a three-dimensional matrix, tumorspheres were embedded in matrigel and tumor cell invasion away from the sphere was imaged and quantified over time. Reduced HPSE Ercalcidiol conferred a marked decrease in tumor cell invasion (Physique 2CCD, p 0.0001). Cell invasion in a three-dimensional matrix requires coordinated regulation of cell attachment, migration, and ECM breakdown. Given the reported functions for HPSE and HS in cell invasion and adhesion, we asked whether decreased invasion in HPSE-reduced cells was associated with reduced cell adhesion. reduced GBM43 had markedly reduced cell adhesion to laminin relative to control cells (Physique 2ECF, p 0.01). Addition of exogenous HPSE promoted invasion of HPSE-reduced GBM43 cells relative to control treated cells (Physique 2GCH, p 0.05). Open in a separate window Physique 2 HPSE expression is increased in the mesenchymal GBM subtype and can promote tumor cell invasion(A) Increased expression of HPSE in human GBM of the mesenchymal (MES) transcriptional subtype. Bars denote the mean SEM Z-score for tumors in the specified subtype (1) with mesenchymal (red) (n=56) and non-mesenchymal (white) (n=139); The Cancer Genome Atlas (TCGA) Data Portal (63). (B) Focal deletion in HPSE confers decreased HPSE mRNA expression in human GBM43 (HPSE del., n=7, and Ctrl, n=3). (CCD) Decreased 3D invasion of Ercalcidiol GBM43 with HPSE deletion as compared to Ctrl. Representative images and quantification (biologic triplicate) at 24 h post plating. (ECF) Decreased laminin adhesion of GBM43 with HPSE deletion relative to Ctrl. (GCH). Increased invasion of HPSE deletion GBM43 cells treated with conditioned media from HPSE overexpressing cells as compared to control vector (Ctrl) overexpressing cells (HPSE, n=5, and Ctrl, n=5). Mean SEM. Representative images and data are from biologic triplicate experiments; *, p 0.05; **, p 0.01; ***, p 0.001; ****, p 0.0001. (C, F, H) Scale bars denote 100m and (E) scale bar denotes 20m. Heparin, a highly sulfated HS, can bind to and inhibit HS-interacting factors including enzymes such as HPSE (48,49). Due to the anticoagulant properties of heparin, however, heparan sulfate mimetics have been developed with reduced anti-coagulant properties. These mimetics have been used for their ability to inhibit HPSE and the function of other HS-interacting factors in therapeutic trials for cancer, including necuparanib (formerly.