Engagement of cellular prion protein with the co-chaperone Hsp70/90 organizing protein regulates the proliferation of glioblastoma stem-like cells

Abstract

Background: Glioblastoma (GBM), a highly aggressive brain tumor, contains a subpopulation of glioblastoma stem-like cells (GSCs) that play roles in tumor maintenance, invasion, and therapeutic resistance. GSCs are therefore a promising target for GBM treatment. Our group identified the cellular prion protein (PrP^C) and its partner, the co-chaperone Hsp70/90 organizing protein (HOP), as potential target candidates due to their role in GBM tumorigenesis and in neural stem cell maintenance.

Methods: GSCs expressing different levels of PrP^C were cultured as neurospheres with growth factors, and characterized with stem cells markers and adhesion molecules markers through immunofluorescence and flow cytometry. We than evaluated GSC self-renewal and proliferation by clonal density assays and BrdU incorporation, respectively, in front of recombinant HOP treatment, combined or not with a HOP peptide which mimics the PrP^C binding site. Stable silencing of HOP was also performed in parental and/or PrP^C-depleted cell populations, and proliferation in vitro and tumor growth in vivo were evaluated. Migration assays were performed on laminin-1 pre-coated glass.

Results: We observed that, when GBM cells are cultured as neurospheres, they express specific stemness markers such as CD133, CD15, Oct4, and SOX2; PrP^C is upregulated compared to monolayer culture and co-localizes with CD133. PrP^C silencing downregulates the expression of molecules associated with cancer stem cells, upregulates markers of cell differentiation and affects GSC self-renewal, pointing to a pivotal role for PrP^C in the maintenance of GSCs. Exogenous HOP treatment increases proliferation and self-renewal of GSCs in a PrP^C-dependent manner while HOP knockdown disturbs the proliferation process. In vivo, PrP^C and/or HOP knockdown potently inhibits the growth of subcutaneously implanted glioblastoma cells. In addition, disruption of the PrP^C-HOP complex by a HOP peptide, which mimics the PrP^C binding site, affects GSC self-renewal and proliferation indicating that the HOP-PrP^C complex is required for GSC stemness. Furthermore, PrP^C-depleted GSCs downregulate cell adhesion-related proteins and impair cell migration indicating a putative role for PrP^C in the cell surface stability of cell adhesion molecules and GBM cell invasiveness, respectively.

Conclusions: In conclusion, our results show that the modulation of HOP-PrP^C engagement or the decrease of PrP^C and HOP expression may represent a potential therapeutic intervention in GBM, regulating glioblastoma stem-like cell self-renewal, proliferation, and migration.

Keywords: Cellular prion protein; Glioblastoma; Hsp70/90 organizing protein; Proliferation; Stem cells.

Fig. 1

Characterization of glioblastoma U87 and U251 neurospheres. a Immunofluorescence for CD133 (green) in U87 cultured as monolayer plus serum (left) or neurospheres (right). Nuclei staining (TO-PRO) shown in red. b Dot plot for CD133 expression in monolayer cultured with serum (left) and neurospheres (right). CD133⁺ cells shown in red and CD133^– cells in black . c Immunostaining for the stem cells markers Oct4, Musashi-1, Sox2, and Nestin (green) in monolayer (upper) and neurospheres (lower), with nuclei staining (TO-PRO) shown in red. d Dot plot graph for CD15 expression in monolayer cultured with serum (left) and as neurospheres (right). CD15⁺ shown in red and CD15^– in black. e Cellular prion protein (PrP ^C) expression assessed by flow cytometry in parental (orange), shRNA-PrP1 (green), or shRNA-PrP2 (red) populations. Negative control shown in blue (only secondary antibody staining). f Immunofluorescence for PrP^C (green), in parental (left), shRNA-PrP1 (middle), or shRNA-PrP2 (right) populations. Arrow indicates staining on the cell surface and arrowhead in the perinuclear region. Nuclei staining (TO-PRO) shown in red. g Immunoblot for Hsp70/90 organizing protein (HOP) (top) and PrP^C (bottom) expression in U251 knockout (PrP ^KO) clones (1 and 2) compared to the parental (Ptl) population. GAPDH was used as the loading control. Note that the smear for PrP^C immunostaining is due to the different glycosylated isoforms. h Flow cytometry for PrP^C expression in the U251 populations parental (left) and U251 PrP-knockout clone 2 (PrP^KO) (middle). IgG isotype (right) was used as the negative control

Fig. 2

Stem cells marker expression in cellular prion protein (PrP ^C)-depleted neurospheres. a PrP^C expression assessed by flow cytometry in parental monolayer (red) and neurosphere (green) cultures. Negative control shown in blue (only secondary antibody staining). b Immunofluorescence for PrP^C (red) and CD133 (green) in parental neurospheres shows co-localization on the cell surface. c Dot plot of CD133 and PrP^C expression in parental neurospheres in the absence (–Cu²⁺) or presence (+Cu²⁺) of CuSO₄ 250 μM. d Histogram for PrP^C and CD133 in the absence (–Cu²⁺) and presence (+Cu²⁺) of CuSO₄ 250 μM. Negative control shown in blue (only secondary antibody staining). e Dot plot of CD133 expression in parental (left) and shRNA-PrP2 (right) neurospheres. CD133⁺ shown in red and CD133^– shown in black. f Immunofluorescence for the stem cells markers musashi-1, nestin, Sox2, and CD133 (green) in parental (upper) and shRNA-PrP2 (lower) neurospheres. Nuclei staining (TO-PRO) shown in red. g Immunofluorescence for the cell differentiation markers GFAP and βIII-tubulin (green) in parental (left) and shRNA-PrP2 (right) neurospheres after 5 days of serum treatment. Nuclei staining (TO-PRO) shown in red

Fig. 3

Hsp70/90 organizing protein (HOP) promotes proliferation of neurospheres dependent on cellular prion protein (PrP ^C) by activating the Erk1/2 signaling pathway. Immunofluorescence for PrP^C (green) and HOP (red) in parental neurospheres shows co-localization on the cell surface in non-permeabilized cells (a) and expression of both proteins in the cytoplasm in permeabilized cells (b). Nuclei staining (TO-PRO) in shown in blue. A higher magnification field is shown in the inset. c Immunoblot for HOP in parental (Ptl), shRNA-PrP1 (PrP1), and shRNA-PrP2 (PrP2) neurospheres. GAPDH was used as the loading control. d Densitometry analysis of immunoblot; values of three independent experiments are expressed relative to control (parental). e Immunoblot of conditioned medium of parental (Ptl), shRNA-PrP1 (PrP1), and shRNA-PrP2 (PrP2) PrP^C-depleted populations. Total protein extract (Ext). GAPDH immunodetection was used as cell lysis control. f Densitometry analysis of immunoblot; values of three independent experiments are expressed relative to control (parental). g U87 neurospheres treated with HOP and/or peptides pepHOP_230–245 or pepHOP_422–437 (1 μM), combined or alone, compared to untreated control. Percentage of BrdU-positive cells in all conditions in relation to total number of cells (n = 6, *p < 0.005). h Immunoblot of phosphorylated Erk1/2 and total Erk1/2 in parental and shRNA-PrP2 neurospheres for basal levels (Ctrl), fetal bovine serum (FBS), or recombinant HOP treatments (HOP) for 20 min. GAPDH was used as the loading control. i Densitometry of the relative values of ERK1/2 activation are represented by the ratio of p-ERK and total ERK1/2 in parental and shRNA-PrP2 neurospheres after treatment with recombinant HOP or FBS. j U251 neurospheres of parental and U251 knockout (PrP ^KO) populations treated with HOP and/or peptides pepHOP_230–245 or pepHOP_422–437 (1 μM), combined or alone, compared to untreated control. BrdU-positive cells detected by spectrophotometry (n = 4, p < 0.05, ANOVA followed by Tukey post-hoc test)

Fig. 4

Cellular prion protein (PrP ^C) and/or Hsp70/90 organizing protein (HOP) knockdown suppresses cell proliferation and tumor growth in vivo. a Immunoblot for HOP expression in U87 non-target (NT), shRNA-HOP (HOP ^KD), or shRNA-PrP2/HOP (PrP2/HOP ^KD) populations. Actin was used as the protein loading control. b Immunoblot densitometry analysis; values of three independent experiments are expressed relative to control (parental). c Dot plot for PrP^C expression of U87 parental and HOP^KD populations. PrP^C+ cells shown in red and PrP^C– cells shown in black. d. Colorimetric BrdU incorporation assay in U87 parental, HOP^KD, PrP2, or PrP2/HOP^KD populations. Values of four independent experiments are expressed relative to control (parental). *p < 0.05, ANOVA followed by Tukey post-hoc test. e U87 neurospheres cells from parental, PrP2, HOP^KD, or PrP2/HOP^KD populations (1 × 10⁶ cells) were implanted into the flank of nude mice and the tumor growth was monitored daily. Data represent tumor volume on day 10 after tumor detection (n = 4, *p < 0.05, ANOVA followed by Tukey post-hoc test). f Tumors were resected, fixed, paraffin embedded, and immunostained for Ki67. Representative images of Ki67 labeling (red) and DAPI (nuclei, blue). g Values represent the percentage of Ki67-positive cells relative to total number of cells (nuclei, DAPI staining)

Fig. 5

Cellular prion protein (PrP ^C) promotes GSC self-renewal binding HOP and modulates cell surface adhesion molecule stability. Neurosphere number (a) or size (b) after 1 week treatment every 48 h with Hsp70/90 organizing protein (HOP) and peptides pepHOP_230–245 and pepHOP_422–437 (1 μM), combined or alone (500nM), compared to control (n = 6, *p < 0.05, ANOVA followed by Tukey post-hoc test). c Dot plot of E-cadherin expression in parental and shRNA-PrP2 neurospheres. E-cad⁺ cells shown in red and E-cad^– cells shown in black. d Dot plot of E-cadherin and PrP^C expression in parental and shRNA-PrP2 neurospheres. e Immunofluorescence for E-cadherin (green) in parental and shRNA-PrP2 neurospheres, showing expression on the cell surface (parental) and in the perinuclear region (shRNA-PrP2). Nuclei (TO-Pro) stain shown in red. f PrP^C (green) and β-catenin (red) expression and co-localization (yellow) of parental and shRNA-PrP2 neurospheres. g Migration assay, ratio between cell migration distance (halo), and neurosphere size for parental and shRNA-PrP2 neurospheres 24 h after plating on laminin-1 (n = 3, *p < 0.05). h Cell scratch assay; images of three experimental replicates were acquired and the distance of each scratch closure after 24 h was measured by comparing with the images at time 0 h for parental and shRNA-PrP2 neurospheres plated on laminin-1 (n = 4, *p < 0.05). i Dot plot of α6 integrin and PrP^C expression in parental and shRNA-PrP2 neurospheres. j Immunofluorescence for β1 integrin (green) of parental and shRNA-PrP2 neurospheres. Nuclei (TO-PRO) stain shown in red. k. PrP^C (green) and β1 integrin (red) expression and co-localization (yellow) of parental and shRNA-PrP2 neurospheres. Nuclei (TO-PRO) stain shown in blue; a higher magnification is shown in the inset

Fig. 6

Model of prion protein (PrP) as a scaffold protein modulating GSC biology. Scheme illustrating how PrP^C may act as a scaffold protein, regulating stemness (dowregulation of CD133 and Sox2 and Msi1 sub-localization altered), recruiting cell adhesion molecules (E-cadherin and integrin α6β1), binding CD133-β-catenin to the plasma membrane, and modulating GSC proliferation and self-renewal through its interaction with Hsp70/90 organizing protein (HOP). Blockage of PrP^C-HOP interaction with HOP peptide impairs the binding of HOP and PrP^C and, consequently, Erk1/2 activation, affecting proliferation