In this study, the role of the PrP C -HOP complex in the regulation associated with GSC biology was evaluated. First, we demonstrated which our model of neurospheres expressed several markers of stemness. The particular expression of CD15, CD133, Oct4, Musashi-1, and Sox2 was increased in neurospheres compared to a non-stem problem (monolayer), validating our experimental model (Fig.   1 ). PrP C expression was higher in neurospheres than in monolayer cultures (Fig.   2a ) and, since the enrichment of GBM cultures with originate cells leads to more malignant tumors in vivo [ 27 ], these data support previous results from our team which show that PrP C expression is correlated with tumor aggressiveness [ 6 , 15 ].

PrP C expression seems to be associated with stem-like properties, since its silencing led to a differentiated expression of stem cell guns. Compared to parental neurospheres, CD133 expression decreased and Sox2 expression became undetectable in PrP C -depleted populations (Fig.   2e and f ). Previous studies have shown that CD133 and Sox2 are usually exclusively expressed at perinecrotic and perivascular regions related to stem-like cell pools, and that nestin and Musashi-1 are usually homogeneously expressed across the tumor, identifying precursors [ 28 ]. PrP2 neurospheres expressed Musashi-1 in the cytoplasm in contrast to parent neurospheres, where it segregated preferentially to cell nuclei (Fig.   2f ). Nuclear expression of Musashi-1 has been associated with activation of the Level pathway in gliomas [ 29 , 30 ] which, subsequently, may lead to increased tumor malignancy via induction of expansion and therapy resistance [ 31 ]. The expression associated with nestin was similar in parental and PrP2 neurospheres (Fig.   2f ). Appearance of differentiation markers is more evident in PrP2 neurospheres compared to the parental population after serum stimuli (Fig.   2g ). Together, our outcomes suggest that parental neurospheres have a stem-like phenotype with some precursor cells, and that the PrP2 population has precursors plus cells committed to a specific phenotype. Indeed, the function associated with PrP C in stem cellular biology has been broadly studied. PrP C has been described as an important molecule for neural dedication and for the proliferation of precursors [ 32 ]. Within tumor stem-like cells, PrP C promotes proliferation and migration [ 11 ]. PrP D interacts with the cell surface protein CD44, a marker for several types of cancer stem-like cells [ 33 ] that associates with tumor-initiating and metastatic capacities and promotes epithelial-mesenchymal transition (EMT) and growth after resection [ 34 ].

A possible interaction between PrP C and CD133 was also observed given their co-expression and localization on the cell surface of parental neurospheres (Fig.   2b plus c ). PrP C and CD133 were previously proven to localize to the same membrane domains (lipid rafts), modulating differentiation and stemness, respectively [ 35 ]. Additionally , a decrease in cell surface expression of CD133 associated with PrP C was observed after copper incitement, suggesting PrP C as a service provider for CD133 internalization (Fig.   2c and d ). PrP C is constitutively endocytosed via clatrin-coated pits [ 36 ] plus copper ions reversibly stimulate this endocytic pathway [ 37 ]. CD133 has been shown to affect the clathrin-endocytosis process [ 38 ] and trafficking down the endosomal and lysosomal pathway for degradation [ 39 ]. Remarkably, the cytoplasmatic domain of CD133 binds β -catenin and the downregulation of CD133 increases β -catenin degradation and affects tumor growth in vitro and in vivo [ 39 ]. Indeed, the β -catenin localization was perturbed within PrP C -depleted cells (Fig.   5f ), suggesting the main role of PrP C within the stabilization of the signaling module on the cell surface.

PrP C -depleted populations formed less secondary neurospheres (Fig.   5a ), suggesting that withdrawing PrP C impairs self-renewal ability. This particular supports the hypothesis that PrP C can act as an important player in stemness servicing and that its dowregulation induces cell line commitment, proving recent evidence that showed a less oncogenic phenotype in PrP C -depleted GSCs [ 40 ]. Thus, it is possible that PrP C acts as an essential molecule for GSC biology and it is capable of maintaining an undifferentiated state in this GBM subpopulation and, since its expression may indicate tumor richness with stem-like cells, it may be used as a tumor development marker. Interesting, the effect of temozolomide, a common chemotherapeutic broker for brain tumors, is enhanced in PrP C -depleted glioma cells, supporting PrP C as an effective target for GBM [ 41 ].

As a quantity of studies show that PrP C provides a scaffold protein, assembling signaling platforms on the plasma membrane layer to elicit several biological processes including in originate cells (reviewed in [ 8 , 42 ]), we appeared for alterations in the expression of cell adhesion healthy proteins on the cell surface of GSCs. We observed the decrease in E-cadherin and integrin α 6 expression over the cell surface of PrP C -silenced populations (Fig.   5c , d and i ), and detected E-cadherin within the cytoplasm (Fig.   5e ). The expression of these proteins was also associated with PrP C , as E-cadherin- or integrin α 6-positive cells were almost exclusively positive for PrP C (Fig.   5d and i ). Integrin β 1 expression was detected both in parental and PrP2 neurospheres, and co-localized with PrP C on the cell surface (Fig.   5j and k ). In addition , GSC immigration on laminin was impaired in PrP D -depleted cells (Fig.   5g and h ). Indeed, it has been demonstrated that PrP D participates in E-cadherin recruitment to the cellular surface [ 43 ] with no significant differences validated in transcripts levels [ 43 , 44 ]. Interestingly, it was proven that PrP C -null mice existing increased paracellular permeability, with lower levels of E-cadherin, desmoplakin, occluding, and other proteins related to cell-cell junctions in digestive tract tissues [ 45 ]. Furthermore, it was reported that PrP C is able to regulate β one integrin adhesiveness modulating ligand-induced changes in integrin service [ 46 ]; however , its depletion had no impact on total β 1 integrin expression levels [ 47 ]. These data suggest that PrP C may be capable of recruiting cell adhesion molecules to the cellular surface of GSCs, raising the hypothesis of PrP C modulating invasion-related processes.

We also reported that GSCs have high expression of HOP and PrP C (Fig.   3b ) and described the HOP-PrP D engagement on their plasma membrane (Fig.   3a ). HOP expression plus secretion was similar in parental, PrP1, and PrP2 populations (Fig.   3c f ), as demonstrated formerly by Santos and collaborators in murine wild-type plus PrP C -null neurospheres [ 6 ]. We demonstrated that HOP is able to modulate GSC expansion (Fig.   4d ) plus self-renewal, depending on its interaction with PrP D (Figs.   3g , 4d , and 5a ); inhibiting the PrP C -HOP interaction with a HOP peptide, which usually mimics the PrP C holding site, abrogates the effects of recombinant HOP (Figs.   3g and 5a ). Similar effects were also noticed in another glioblastoma cell line (U251) confirming the importance of these types of complexes in GSC biology (Fig.   3j ). Another HOP peptide (anti-TPR), which usually inhibits the HOP-Hsp90 interaction, has been described to cause cell death in several cancer cell lines [ 48 ] and produce a cytotoxic effect in glioblastomas [ 49 ], highlighting HOP as a potential target for GBM therapy. Previous reports [ 15 ] as well as information from this study (Fig.   3e and h ) also described that PrP C and secreted HOP interact on the cell surface associated with human glioblastoma cell lines and modulate GBM development by promoting proliferation through activation of the Erk1/2 path [ 16 ]. Remarkably, a higher basal phosphorylation of Erk was found in PrP C -depleted tissue when compared to their counterparts (Fig.   3h and i ), confirming our prior data in retinae [ 24 ] and hippocampal neurons [ 25 ], along with those from other authors in adult brain and cerebellum extracts [ 26 ]. This feature suggests that high process of Erk could be a compensatory effect for prion protein amputation [ 42 ]. Indeed, the targeting of stress reaction proteins has remarkable potential for the development of molecular treatments. Simultaneous inhibition of expression of both HOP ligands Hsp70 and Hsp90 reduced proliferation and promoted apoptosis within GBM cell lines in vitro [ 50 ]. Additionally, HOP performs an important function in compensatory mechanisms associated with stress responses in tumor cells [ 51 ], additional supporting its relevance in GBM maintenance.

Finally, we observed that PrP C and/or HOP knockdown affects the particular proliferative and tumorigenic capacity of GSCs (Fig.   4e and f ) in vivo, and facilitates PrP C and HOP substances as potential new targets for developing more efficient restorative strategies.