Glioblastomas (GBMs) are resistant to current therapy protocols and identification of molecules that target these tumors is crucial. Interaction of secreted heat-shock protein 70 (Hsp70)–Hsp90-organizing protein (HOP) with cellular prion protein (PrPC) triggers a large number of trophic effects in the nervous system. We found that both PrPC and HOP are highly expressed in human GBM samples relative to non-tumoral tissue or astrocytoma grades I–III. High levels of PrPC and HOP were associated with greater GBM proliferation and lower patient survival. HOP–PrPC binding increased GBM proliferation in vitro via phosphatidylinositide 3-kinase and extracellular-signal-regulated kinase pathways, and a HOP peptide mimicking the PrPC binding site (HOP230–245) abrogates this effect. PrPC knockdown impaired tumor growth and increased survival of mice with tumors. In mice, intratumor delivery of HOP230–245 peptide impaired proliferation and promoted apoptosis of GBM cells. In addition, treatment with HOP230–245 peptide inhibited tumor growth, maintained cognitive performance and improved survival. Thus, together, the present results indicate that interfering with PrPC–HOP engagement is a promising approach for GBM therapy.
Astrocytomas, the most common type of primary brain tumor, are classified by the World Health Organization into four malignancy grades: I (pilocytic astrocytoma), II (diffuse astrocytoma), III (anaplastic astrocytoma) and IV [glioblastoma (GBM)] that is the most aggressive subtype of adult human brain tumors.1 GBM usually presents in the sixth or seventh decade of life and carries an average prognosis of <12 months.2 It exhibits a relentless malignant progression characterized by uncontrolled cellular proliferation, diffuse infiltration, propensity for necrosis, robust neovascularization and resistance to traditional and newer targeted therapeutic approaches.3 This aggressive phenotype is the result of a variety of genetic and epigenetic alterations that lead to the deregulation of intracellular signaling pathways.4
We showed previously that the interaction of heat-shock protein 70 (Hsp70)–Hsp90-organizing protein (HOP, aka stress-inducible phosphoprotein 1, STI1 or STIP1) with cellular prion protein (PrPC) leads to the activation of signaling pathways that promote neuron/glial cell survival and differentiation.5,6 Some of these pathways also modulate proliferation of GBM cells in vitro.7 HOP is a 543-residue (66-kDa) co-chaperone that binds Hsp70 and Hsp90, modulating their folding activities.8, 9, 10 Despite its cytoplasmic localization, HOP is constitutively secreted mainly in extracellular vesicles.11 Once in the extracellular milieu, HOP (at amino acids (aa) 230–245) interacts with membrane-bound PrPC (at aa 113–128).12,13
PrPC and HOP have been implicated in cancer progression. The expression of an altered form of PrPC, pro-PrPC, that interacts with filamin A is associated with poorer clinical outcome and survival of patients with pancreatic ductal adenocarcinoma.14 Pro-PrPC is also expressed in melanoma cells, although its role in melanomagenesis is unclear.15 In colorectal tumors, PrPC may regulate glycolytic rates by modulating GLUT1 levels.16 PrPC expression is higher in metastatic than in non-metastatic gastric cancer where it is thought to promote adhesion, migration and invasion.17
Two independent studies demonstrated elevated expression of HOP, with Hsp70 or Hsp90, in colon tumors.18,19 High levels of these proteins have a negative impact on patient survival.18,19 Elevated HOP expression has been observed in pancreatic and ovarian tumors and in hepatocellular carcinoma.20, 21, 22 Secretion of HOP by GBM cells and other cancer cells is thought to influence cell proliferation, invasion and angiogenesis.7,22, 23, 24 Indeed, HOP has been shown to be secreted together with Hsp90 by breast cancer, pancreatic tumor and fibrosarcoma cells, making Hsp90-bound HOP available to form extracellular active complexes with molecules such as matrix metalloproteinase-2, a collagenase related to tumor invasiveness.23, 24, 25 HOP secreted by ovarian cancer tissues can be detected in peripheral blood, making it a biomarker that, when found in association with CA125, allows for early detection of these tumors.22 When secreted, HOP also associates with activin A receptor, type II-like kinase-2, activating SMAD (Sma- and Mad-related proteins) signaling, and thereby inducing proliferation in ovarian tumor cells.26
We found that in human GBM cells in vitro, HOP binds specifically to PrPC and modulates proliferation through extracellular-signal-regulated kinase (ERK1/2) and phosphatidylinositide 3-kinase (PI3K) pathways,7 suggesting that HOP–PrPC binding could be an important therapeutic target in GBMs. It is noteworthy that we demonstrated that mice bearing orthotopic xenograft GBM exhibit early cognitive decline,27 consistent with the brain dysfunction and cognitive decline described for human patients with brain tumors.28
Herein, the expression of HOP and PrPC was evaluated in human glioma specimens, grades I–IV, and correlated with tumor proliferation and patient survival. The role of PrPC–HOP engagement and the impact of diminishing it on tumor growth, cognitive performance and survival were evaluated in vivo using orthotopic xenograft mouse models.
Overexpression of HOP and PrPC in GBMs
Evaluation of HOP gene expression with quantitative PCR (qPCR) experiments demonstrated that GBM (grade IV) samples had higher HOP mRNA levels than non-neoplastic tissues and grade II astrocytoma samples (Figure 1a). According to q–PCR findings, a meta-analysis using the database from The Cancer Genome Atlas (TCGA) for GBM samples also showed increased levels of mRNA of both HOP (Agilent: 2.0-fold; Affymetrix: 1.5-fold) and PrPC (Agilent: 7.01-fold; Affymetrix: 6.96-fold) when compared with non-neoplastic tissues (Table 1). Moreover, the protein–protein interaction network reveals that HOP (k=21) and PRNP (k=88) present a high connectivity degree (Supplementary Figure 1).
Complementary immunohistochemistry experiments performed in a different set of astrocytoma specimens organized in tissue microarrays demonstrated significantly higher expression of both HOP (Figures 1b and c) and PrPC (Figures 1d and e) in GBM samples versus non-neoplastic and grade I–III astrocytoma samples.
GBM samples expressing high levels (i.e., above the median) of both HOP and PrPC presented a higher proliferation ratio (indexed by Ki-67 labeling) than those with high levels of HOP and low levels of PrPC (Figure 1f), indicating that the two proteins may act in cooperation to promote proliferation. Furthermore, Kaplan–Meier survival analysis indicated that patients with high levels of HOP and PrPC had shorter survival than those with high levels of HOP and low levels of PrPC (Figure 1g). No difference in survival was found in patients with lower HOP levels (Figure 1g). These findings indicate that HOP and PrPC are upregulated in GBM and that elevated expression of both proteins within tumors is correlated with increased tumor proliferation and decreased patient survival.
HOP mediates tumor growth in a PrPC-dependent manner
Our previous results showed that HOP secreted by a non-tumorigenic glioma cell line (A172) induces proliferation in a PrPC-dependent manner.7 Here, the proliferative effect mediated by secreted HOP through membrane-attached PrPC was evaluated in a tumorigenic human glioblastoma cell line (U87). Flow cytometry analysis in U87 cells confirmed the high expression of PrPC at the cell membrane (Figure 2a). Immunofluorescence analysis of intracranial U87-xenografted tumors showed that PrPC and HOP were highly and uniformly expressed and colocalized, indicating that both proteins may interact in tumors in vivo (Figure 2b), as previously demonstrated in normal brain.13 As shown in Figure 2c, immunoblots of conditioned medium (CM) from cultured xenograft tumors confirmed the presence of secreted HOP (66 kDa) in the absence of CDK4 (confirming the absence of cell lysis). ELISA quantification showed that the concentration of the secreted HOP increased with increasing tumor size (Figure 2d).
The expression of PrPC in U87 cells was knocked down by infection with lentiviral particles carrying anti-PrPC short hairpin RNA (shRNA) sequence 1 (U87 shRNA 1) direct against mouse mRNA, used as a control, and sequence 2 (U87 shRNA 2) against human mRNA. Flow cytometry analysis (Figure 2e) showed that shRNA 1-infected cells present a small inhibition (17.4±2.7%) of PrPC levels owing to its partial homology with the human PrPC sequence, whereas shRNA 1-infected cells show 84.6±2.9% less PrPC than non-infected cells. Treatment with recombinant HOP induced proliferation in shRNA 1-infected cells and non-infected cells, but had no effect on the shRNA 2-infected cells. In addition, treatment with HOP lacking the PrPC binding site (HOPΔ) had no effect on proliferation (Figure 2f).
Control U87, shRNA 1-infected and shRNA 2-infected cells were orthotopically xenografted into nude mice and osmotic pumps (OPs) were implanted 9 days later to deliver saline, recombinant HOP, or HOPΔ for 14 days. Tumors derived from shRNA 2 cells were smaller than those derived from non-infected or shRNA 1 cells. HOP treatment caused substantial increases in tumor size in mice bearing the non-infected and shRNA 1 cells, but tumors derived from shRNA 2 cells were unaffected. HOPΔ treatment did not affect tumor growth in any of the groups (Figure 2g). Mice implanted with shRNA 2-infected cells (84.6% suppressed PrPC) survived significantly longer than mice implanted with shRNA 1-infected cells (17.4% suppressed PrPC) or non-infected cells (Figure 2h). These results are in agreement with findings in human GBMs (Figures 1f and g) in that with high levels of HOP in U87 cells (as observed in Figures 2f and g), PrPC levels correlate directly with proliferation and correlate inversely with survival, indicating that U87 cells can recapitulate the properties observed here in human tumors.
HOP230–245 peptide abrogates GBM proliferation mediated by HOP
We hypothesized that the HOP230–245 peptide that mimics the PrPC binding site12 could be used to disrupt HOP–PrPC interaction in GBMs, and, consequently, control tumor growth (Figure 3a). When U87 cells were pretreated with HOP peptides (HOP230–245 or control peptide HOP61–76) and then incubated with recombinant HOP, we found that HOP230–245 treatment inhibited proliferation, whereas the control peptide (HOP61–76) had no effect (Figure 3b). Similar results were obtained with A172 cells (Figure 3e), which were originally used to describe the role of HOP in GBM proliferation,7 and also with two additional human GBM cell lines, U118 and U251 (Figures 3c and d, respectively). To address the activation of intracellular signaling triggered by HOP, cells were treated with recombinant HOP and/or peptides and then submitted to immunoblotting to evaluate ERK1/2 phosphorylation. As described previously,7 HOP induced ERK1/2 phosphorylation in GBM cell lines (Figures 3i–k) and HOP230–245 peptide abolished this effect, whereas control peptide did not alter HOP-induced ERK1/2 phosphorylation. Proliferation assays performed in the presence of ERK1/2 and PI3K inhibitors (U0126 and LY29004, respectively), as previously described in A172 cells,7 demonstrated that both pathways are involved in HOP-mediated proliferation (Figures 3f–h) in GBM cells. Taken together, these data support the notion that proliferation mediated by HOP–PrPC association via ERK1/2 and PI3K signaling pathways is impaired by HOP230–245 peptide in GBM cells.
HOP230–245 reduces tumor growth of human GBM xenografts and increases survival.
To evaluate the capacity of HOP230–245 to inhibit tumor growth in vivo, mice were inoculated with U87 cells and treated with saline or peptides using a guide-screw system (GS) (Figures 4a–c and e) or OPs (Figures 4d and e). At 14 days postinoculation (p.i.) (summary time line in Figure 4a), the mean tumor volume in mice treated with HOP230–245 was reduced significantly and in a dose-dependent manner compared with control mice given saline. Tumor size in HOP61–76-treated animals was similar to that in saline controls, demonstrating specificity of the HOP230–245 peptide on tumor growth inhibition (Figure 4a).
Histological sections of resected tumors showed that HOP230–245 treatment caused a fivefold reduction in cell proliferation (Ki-67 labeling) and a 2.2-fold increase in apoptosis (cleaved-caspase-3 labeling) compared with controls treated with HOP61–76 (Figures 4b and c). Intratumoral microvessel density (CD31 labeling) was not significantly affected by HOP230–245 treatment (Figures 4b and c).
A tumor growth curve pointed to day 9 p.i. as the appropriate day to initiate HOP230–245 treatment of pre-established tumors (Figure 4d), since that was when tumor volume could be measured easily (volume range, 6.6–11.2 mm3) and animals did not present neurologic effects for the subsequent 2 weeks. In animals implanted with OP for intratumor delivery of HOP230–245 or HOP61–76, mean tumor volume was 8.0±3.4 mm3 at day 9 p.i. (pretreatment). Tumor growth was reduced in animals that received HOP230–245 (26.6±5.0 mm3) for 14 days (day 23 p.i.) relative to that observed in animals that received the control peptide HOP61–76 (50.5±8.7 mm3 for the same time (Figure 4d).
U87 tumor-bearing mice treated with HOP230–245 (delivered via GS every 2 days or via an OP continuously, both starting from day 9 p.i.) showed grade 4 neurologic signs (see Materials and methods) later than HOP61–76-treated mice. Moreover, intratumoral administration of HOP230–245 using either GS or OP increased animal survival compared with the HOP61–76 group (Figure 4e), suggesting that HOP230–245 peptide can decrease GBM growth in vivo and extend animal survival. The effect of HOP230–245 in reducing tumor growth was also evaluated in animals bearing U251 tumors (Supplementary Figure 2). The results suggest that HOP230–245 treatment decreases tumor size when compared with HOP61–76 peptide treatment (Supplementary Figure 2c), similar to what was described for U87 tumors. Taken together, these data indicate that HOP230–245 peptide can decrease GBM growth in vivo and extend animal survival.
HOP230–245 preserves cognition in mice bearing orthotopic U87 xenografts
The same groups of peptide-treated mice used in the above survival experiments (Figure 4e) were subjected to OR cognitive testing. At day 9 p.i. (pretreatment), mice in the GS groups, despite having substantial tumors (~8 mm3), were able to recognize new objects in a 90-min short-term memory (STM) test (memory evidenced by exploration time; Figure 5a). Mice treated with HOP230–245 continued to recognize new objects until day 25 p.i., (Figure 5b), whereas HOP61–76-treated mice exhibited cognitive deficits at day 12 p.i. (Figure 5c), almost 2 weeks before the onset of neurologic signs. A GS-implanted control mice (without tumor) exhibited normal behavior (Figure 5d).
Mice in the OP groups showed no evidence of cognitive deficits in 90-min STM and 24-h long-term memory (LTM) tests of object recognition (OR) on days 8/9 p.i. (Figure 5e). Mice treated with HOP230–245 continued to show normal cognitive performance at days 13/14, 17/18 and 21/22 p.i. Conversely, mice treated with HOP61–76 had impaired STM but normal LTM on days 13/14, and then exhibited severe STM and LTM impairments on days 17/18 and 21/21 p.i. (Figure 5f). Thus, local treatment of tumors with HOP230–245 peptide decreased the cognitive deficits that otherwise occurred with tumor growth.
In this study, we showed that human GBMs present higher levels of HOP and PrPC than non-tumor brain tissue or astrocytoma grades I–III. Elevated levels of both HOP and PrPC were associated with augmented proliferation and reduced survival, suggesting a functional role of these proteins in tumor progression. These findings fit with previous reports of altered PrPC processing in pancreatic ductal adenocarcinoma and of elevated expression of PrPC in colorectal tumors, melanoma and metastatic gastric cancer correlating with tumor relapse, poor clinical outcome, and survival.14, 15, 16, 17,29, 30, 31
Our findings also extend prior evidence showing that pancreatic,14 hepatocellular21 and ovarian22 carcinomas express high levels of HOP and that HOP depletion can reduce invasion and proliferation of highly aggressive cell lines.20,21,24 Furthermore, it has also been shown that specific inhibitors32 or a competitor peptide for the Hsp90-HOP33 interaction inhibit tumor development, pointing to the Hsp90–HOP complex as a potentially selective anticancer therapeutic target.33,34 Mechanistically, secreted HOP35 carried by exosome-like extracellular vesicles11 may promote PrPC-dependent proliferation via activation of PI3K and ERK1/27 or by binding to activin A receptor, type II-like kinase-2 and activating the SMAD-ID signaling.26
Remarkably, our results demonstrate that PrPC is expressed in U87 cells and that its knockdown decreases proliferation in xenografted tumors and increases animal survival. These findings are consistent with reports showing that PrPC inhibition decreases growth and induces programmed cell death in glioma36 and colorectal tumor cells,16 and sensitizes tumor cells to cytotoxic drugs.37,38 It has been suggested that HOP knockdown may reduce invasion of pancreatic cancer cells by decreasing expression of its downstream target gene matrix metalloproteinase-224 and by reducing pseudopodia formation and migration in breast cancer.39 We did not test HOP knockdown in this study, since this approach would not allow distinguishing effects on HOP co-chaperone activity from effects on the HOP–PrPC interaction. To isolate the role of the PrPC–HOP interaction, we conducted in vitro and in vivo experiments using GBM cells with knocked down PrPC or mutant HOP (PrPC binding site deleted) and found that the availability of the PrPC–HOP complex was associated with tumor proliferation and reduced survival. Importantly, these experimental data are in agreement with our patient data.
We used the HOP230–245 peptide, which competes with full-length HOP for PrPC binding,12 to test whether blockage of HOP–PrPC association could interfere with tumor growth. We found that HOP230–245 inhibited proliferation of four GBM cell lines in vitro. Moreover, intratumor delivery (a successfull approach used for other drugs in phase I and II clinical trials)3,40,41 of HOP230–245 peptide also inhibited tumor growth in vivo and improved survival of the animals implanted with GBM cells. Thus, confirming the role of HOP230–245 peptide in the inhibition of GBM growth. It is important to note that these findings will require further validation using primary GBM cultures, since many GBM histopathologic features are not fully recapitulated in cell lines.42
It is important to consider that PrPC–HOP interaction itself has been implicated in self-renewal of neural stem cells.43 On the other hand, temozolomide treatment commonly given to patients with GBM induces PrPC overexpression44 and also selects a subset of tumor stem cells,45 which might inadvertently contribute to GBM recurrence. It remains to be determined whether the PrPC–HOP complex also favors GBM stem cell self-renewal; if so, HOP230–245 could potentially block this effect.
Finally, it is noteworthy that we found that HOP230–245 appeared to prevent tumor-induced cognitive deficits in tumor-bearing mice. This beneficial effect was probably due to inhibition of tumor growth as demonstrated previously.27 This is particularly important since cognitive performance is increasingly valued as secondary end point in clinical trials.46 On the other hand, a few years ago, we demonstrated that HOP230–245 peptide had trophic properties13 and was able to improve memory formation when infused directly into the hippocampus.47 However, it is unlikely that the present results can be attributed to peptide diffusion to other brain regions such as the hippocampus. It is possible, although, that HOP230–245 peptide could protect neurons against apoptosis or degeneration caused by toxic antitumor drugs and radiotherapy.48
Owing to intrinsic resistance of aggressive malignant astrocytomas to established therapies, our ability to improve the survival of these patients depends largely on strategies to target tumor cell resistance and the identification of novel antitumor agents. The advancement of new approaches to treat patients with GBM tumors may depend on identifying tumor-expressed molecules that are associated with tumor growth and developing new approaches to inhibit them. The present results show that the PrPC–HOP complex is a key element in GBM proliferation. These findings further suggest that PrPC–HOP complex formation may be blocked by HOP230–245 peptide or peptide-mimetic drugs, and thus point to a novel approach for treating patients with GBM, which with currently available treatments have a very poor prognosis.
Materials and methods
Fresh surgical astrocytoma and non-neoplastic brain tissue samples were from University of São Paulo (approved by the Ethics committee—691/05). RNA was extracted with RNeasy Mini Kits (Qiagen, Venlo, Netherlands) and subjected to reverse transcription. Relative expression levels were determined by qPCR (ABI Prism 7500; Applied Biosystems, Foster City, CA, USA). A geometric mean of the expression of reference genes (BCRP, HPRT and GUSB) was used as a quantitative standard for relative expression analysis. Primer sequences are listed in Supplementary Table 1. The 1.73−ΔΔCt equation was applied to calculate relative HOP expression, where ΔΔCt=ΔCt tumor−mean ΔCt non-neoplastic brain tissues and ΔCt=Ct HOP−Ct geometric mean of reference genes.
As validation purposes, we used two types of publicly available data sets from TCGA (The Cancer Genome Atlas): gene expression measurements using Agilent G4502A-07 and Affymetrix U133. For transcript-level analysis, we use TCGA's normalization and gene summaries. It was considered the probe to gene mapping provided by TCGA in the Array Definition Files. The protein–protein interaction network was evaluated using the NAViGaTOR software (Toronto, ON, Canada).49
Patients and tissue microarray
Two tissue microarrays were prepared as in Kononen et al.50 from 185 formalin-fixed, paraffin-embedded astrocytomas (from A. C. Camargo Cancer Center, São Paulo, Brazil) and 14 non-neoplastic samples (from surgically remediable patients with intractable mesial temporal lobe epilepsy). Samples were prepared after approval from A. C. Camargo Cancer Center Research and Ethics Committee (Process 1613/11). The clinicopathologic characteristics of patients are available in the Supplementary Information.
For immunohistochemistry, tissue microarrays were incubated overnight with the primary antibody (Supplementary Table 2) at 4 °C, after antigen retrieval. Reactions were developed as described previously.51 Virtual slides were created with the ScanScope System (Aperio Technologies, Vista, CA, USA). Positive pixels were quantitated and staining intensity was classified as 0 (negative), 1 (weakly positive), 2 (positive) or 3 (strong positive). HSCOREs were as Σ(i × Pi), where Pi is % positive pixels and i is the staining intensity classification (0–3), yielding an HSCORE range of 0–300. The percentage of positive Ki-67 nuclei was calculated according to the internal algorithm of with ImageScope (IHC Nuclear v.1, Vista, CA, USA).
Proliferation and survival
HSCOREs for PrPC and HOP expression in samples from 91 patients with GBM were compared with the % of Ki-67 positive nuclei. Samples were divided into high- and low-expression groups, and split at the HSCORE median. Associations between Ki-67 and PrPC and HOP expression groupings were performed. PrPC and HOP HSCOREs were used to construct Kaplan–Meier postsurgery survival curves.
Proteins and peptides
Mouse recombinant HOP or HOPΔ (deletion mutant for amino acids 230–245) were purified12 and stripped of lipopolysaccharides with detoxi-gel endotoxin removing gel (Thermo Scientific Pierce, Waltham, MA, USA). Human HOP230–245 peptides (ELGNDAYKKKDFDTAL) and amino-terminal peptide HOP61–76 (GCKTVDLKPDWGKGYS) were synthesized by GenScript (Piscataway, NJ, USA).
Human U87MG, A172 and U118 cell lines were purchased from American Type Collection (ATCC, Manassas, VA, USA). U215 cell were kindly provided by Drs Sueli Oba Shinjo and Suely Marie. Cells were cultured in Dulbecco's modified Eagle's medium (Gibco, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (FBS). Aseptically resected tumor specimens were cultured for 48 h in Dulbecco's modified Eagle's medium F12 supplemented with B27 and gentamicin (40 μg/ml). The CM was centrifuged (2000 g) and filtered.
Cell proliferation assay
Cells were cultured in Dulbecco's modified Eagle's medium plus 10% FBS at 37 °C for 12 h. Cells were washed and maintained in Dulbecco's modified Eagle's medium F12 for 48 h, and then treated for 24 h with 0.2 μM recombinant HOP and HOPΔ, 0.2 or 8 μM HOP or 10% FBS or pharmacologic inhibitors of ERK1/2 and PI3K (U0126 and LY29004, respectively; Calbiochem, San Diego, CA, USA). Cells were exposed to bromodeoxyuridine (BrdU) (35 μM) 2 h before fixation and were processed for immunofluorescence. The percentage of BrdU-positive nuclei was calculated. Proliferation assays for U118 and U251 cells were evaluated using BrdU Cell Proliferation Assay Kit (Cell Signaling, Danvers, MA, USA) according to the manufacturer's instructions.
Deficient lentiviral particles were produced in HEK293FT cells using the ViraPower Lentiviral Expressing System (Invitrogen, Carlsbad, CA, USA). The pLenti constructs for the shRNA sequence were as follows: shRNA 1 for a PrP mouse sequence: 5′-IndexTermCACCAGAACAACTTCGTGCACGACTCGAAAAGTCGTGCACGAAGTTGTTC-3′; and shRNA 2 for the human sequence 5′-IndexTermCACCGCGTCAATATCACAATCAAGCCGAAGCTTGATTGTGATATTGACGC-3′. U87 cells were infected (multiplicity of infection=25) with lentiviral vectors carrying two constructs targeting PrPC (shRNA 1 and 2). After infection, cells were selected with Zeocin (600 μg/ml) for 10 days.
Orthotopic glioma model
Female or male 8–10-week-old Balb/C nude mice (nu/nu; Taconics, Hudson, NY, USA) were used. Institutional guidelines for animal welfare and experimental conduct were followed and the study approved by Animal Ethics Committee of A. C. Camargo Cancer Center (process 025/08). A guide-screw (Plastic One, Roanoke, VA, USA) was used for intracerebral tumor cell engraftment and intratumoral injections.52 Three days after screw implantation, 0.5 × 106 U87 cells were inoculated through the screw into the right striatum. Drugs were injected through the screw over 6 min. For dose–response curves, treatment was initiated on the same day as inoculation and continued every 2–3 days over 2 weeks. For tumor assays, treatment was initiated at day 9 p.i. and continued every 2 days over 14 days or until animals presented neurologic symptoms. For treatment using OPs, U87 (0.5 × 106) or U251 (1.5 × 106) cells were implanted into the right striatum using a Hamilton syringe, and at day 9 p.i. (for U87-implanted animals) or at day 34 (for U251-implanted animals), mice were anesthetized and OPs (Alzet, Cupertino, CA, USA) were implanted subcutaneously and peptides were delivered to the tumors. Neurologic symptoms were assessed by modified neurologic scores:53 grade 0, none; grade 1, tail weakness/paralysis; grade 2, hind leg para/hemiparesis; grade 3, hind leg para/hemiparalysis; grade 4, tetraplegia/moribund/death. Animals were euthanized when they presented grades 3 and 4. All mice (n=169) were euthanized by CO2 saturation.
The volume (mm3) of the tumors derived from U87 cells were calculated as (length × width2)/2 as described.27 The tumors were fixed in 4% paraformaldehyde, sectioned (10 μm) and processed for immunofluorescence. Animals bearing U251 tumors have their brains removed, fixed, paraffin-embedded and hematoxylin and eosin stained. Tissues were evaluated by a pathologist and tumor areas were used to estimate tumor extension in serial sectioning.
Fixed cells were treated with 2 N HCl for 30 min followed by borate buffer for 10 min. Cells were blocked, permeabilized, incubated with anti-BrdU biotinylated antibody (Millipore, Billerica, MA, USA) overnight and labeled with Strepta-Alexa 488 (Molecular Probes, Carlsbad, CA, USA) and 4',6-diamidino-2-phenylindole (DAPI) for 1 h. Sections were blocked for 1 h and incubated overnight with primary antibody (Supplementary Table 2). Secondary antibody (Supplementary Table 2) was applied for 1 h, followed by DAPI staining. Slides were mounted with Fluorsave (Calbiochem). Sections were viewed under an Olympus immunofluorescence microscope (Center Valley, PA, USA).
CM was prepared from 48-h cultures of xenograft tumors, mixed in Laemmli buffer, resolved in 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis and immunoblotted13 with actin (loading), IgG (negative) and CDK4 (lysis) controls (Supplementary Table 2). Cells were lysed with Laemmli sample buffer and extracts were used for sodium dodecyl sulfate–polyacrylamide gel electrophoresis (10% gels). Proteins were transferred to nitrocellulose membranes (GE Healthcare, Little Chalfont, UK) and immunoblotted to detect phosphorylated (p)-(T202/Y204)-ERK1/2 and total ERK1/2. Bands were imaged with UVITEC system and quantified using ImageJ software (NIH, Bethesda, MD, USA). Levels of the p-ERK1/2 were relativized to total ERK1/2. Antibodies were purchased from Cell Signaling.
U87 cells were incubated with primary and secondary antibodies (Supplementary Table 2) for 30 min each at 4 °C. FACS analyses were performed with FACSCalibur (BD Biosciences, Franklin Lakes, NJ, USA).
Binding plates (96-well) were coated overnight at 4 °C with CM from U87 cultures or tumors. Washed wells were then filled with phosphate-buffered saline (with 5% milk) and incubated with primary (2 h) and secondary (1 h) antibodies at 37 °C. The assay was developed in orthophenilenediamine solution (0.33 mg/ml 0.5 M citrate buffer, pH 5.2, 0.4% H2O2) for 5 min at room temperature. The reaction was stopped with 4 M H2SO4 and the absorbance (490 nm) measured using a microplate reader (Bio-Rad, Hercules, CA, USA).
Animals inoculated with HOP230–245 or HOP61–76 and implanted with GS or OP were evaluated for OR.27 They were habituated to an open-field polyvinyl chloride plastic arena (30 × 25 × 20 cm3) with 20 min per day free exploration for 4 days. Then, they were placed in the arena with two identical objects on day 9, 12, 17 or 25 p.i. and left to explore for 5 min. STM was tested in a 5-min trial 90 min later, wherein one of the identical objects was replaced by a novel one. A second group of animals were inoculated with tumor cells and 9 days p.i. trained as described previously, but STM was tested at 3 h and LTM was tested at 24 h. At the test, new objects replaced one of the identical ones used in training. The exploratory behavior of the mice toward familiar and novel objects was quantified. After the first OR session at day 9, mice were implanted with OP to deliver peptides HOP230–245 or HOP61–76 and further OR sessions were performed at days 13/14, 17/18 and 21/22 p.i. for STM (3 h) and LTM (24 h) no-tumor, GS-implanted mice served as controls.
Mean values represent at least three independent data sets; error bars represent standard errors of the mean. One-way analysis of variance (ANOVA) followed by Tukey's post hoc test was used for multiple comparisons. Unpaired Student’s t-test was used to compare Ki-67-, CD31- and cleaved-caspase3-immunostained groups. Nonparametric ANOVA Kruskal–Wallis test followed by Dunn’s multiple comparison test were used to analyze immunochemistry and qPCR data. Significance was accepted at P<0.05. Differences between Kaplan–Meier survival curves were evaluated with the log-rank test.
Furnari FB, Fenton T, Bachoo RM, Mukasa A, Stommel JM, Stegh A et al. Malignant astrocytic glioma: genetics, biology, and paths to treatment. Genes Dev 2007; 21: 2683–2710.
Ohgaki H, Kleihues P . The definition of primary and secondary glioblastoma. Clin Cancer Res 2013; 19: 764–772.
Westphal M, Lamszus K . The neurobiology of gliomas: from cell biology to the development of therapeutic approaches. Nat Rev Neurosci 2011; 12: 495–508.
Jones TS, Holland EC . Standard of care therapy for malignant glioma and its effect on tumor and stromal cells. Oncogene 2012; 31: 1995–2006.
Linden R, Martins VR, Prado MAM, Cammarota M, Izquierdo I, Brentani RR . Physiology of the prion protein. Physiol Rev 2008; 88: 673–728.
Martins VR, Beraldo FH, Hajj GN, Lopes MH, Lee KS, Prado MAM et al. Prion protein: orchestrating neurotrophic activities. Curr Issues Mol Biol 2010; 12: 63–86.
Erlich RB, Kahn SA, Lima FRS, Muras AG, Martins RAP, Linden R, et al. STI1 promotes glioma proliferation through MAPK and PI3K pathways. Glia 2007; 55: 1690–1698.
Maciejewski A, Prado MA, Choy W-Y . (1)H, (15)N and (13)C backbone resonance assignments of the TPR1 and TPR2A domains of mouse STI1. Biomol NMR Assign 2013; 7: 305–310.
Lee C-T, Graf C, Mayer FJ, Richter SM, Mayer MP . Dynamics of the regulation of Hsp90 by the co-chaperone Sti1. EMBO J 2012; 31: 1518–1528.
Muller P, Ruckova E, Halada P, Coates PJ, Hrstka R, Lane DP et al. C-terminal phosphorylation of Hsp70 and Hsp90 regulates alternate binding to co-chaperones CHIP and HOP to determine cellular protein folding/degradation balances. Oncogene 2013; 32: 3101–3110.
Hajj GNM, Arantes CP, Dias MVS, Roffé M, Costa-Silva B, Lopes MH et al. The unconventional secretion of stress-inducible protein 1 by a heterogeneous population of extracellular vesicles. Cell Mol Life Sci 2013; 70: 3211–3227.
Zanata SM, Lopes MH, Mercadante AF, Hajj GNM, Chiarini LB, Nomizo R et al. Stress-inducible protein 1 is a cell surface ligand for cellular prion that triggers neuroprotection. EMBO J 2002; 21: 3307–3316.
Lopes MH, Hajj GNM, Muras AG, Mancini GL, Castro RMPS, Ribeiro KCB et al. Interaction of cellular prion and stress-inducible protein 1 promotes neuritogenesis and neuroprotection by distinct signaling pathways. J Neurosci 2005; 25: 11330–11339.
Li C, Yu S, Nakamura F, Yin S, Xu J, Petrolla AA et al. Binding of pro-prion to filamin A disrupts cytoskeleton and correlates with poor prognosis in pancreatic cancer. J Clin Invest 2009; 119: 2725–2736.
Li C, Yu S, Nakamura F, Pentikainen OT, Singh N, Yin S et al. Pro-prion binds filamin A, facilitating its interaction with integrin 1, and contributes to melanomagenesis. J Biol Chem 2010; 285: 30328–30339.
Li Q-Q, Sun Y-P, Ruan C-P, Xu X-Y, Ge J-H, He J et al. Cellular prion protein promotes glucose uptake through the Fyn-HIF-2α-Glut1 pathway to support colorectal cancer cell survival. Cancer Sci 2011; 102: 400–406.
Pan Y . Cellular prion protein promotes invasion and metastasis of gastric cancer. FASEB J 2006; 20: 1886–1888.
Kubota H, Yamamoto S, Itoh E, Abe Y, Nakamura A, Izumi Y et al. Increased expression of co-chaperone HOP with HSP90 and HSC70 and complex formation in human colonic carcinoma. Cell Stress Chaperones 2010; 15: 1003–1011.
Ruckova E, Muller P, Nenutil R, Vojtesek B . Alterations of the Hsp70/Hsp90 chaperone and the HOP/CHIP co-chaperone system in cancer. Cell Mol Biol Lett 2012; 17: 446–458.
Walsh N, O’Donovan N, Kennedy S, Henry M, Meleady P, Clynes M et al. Identification of pancreatic cancer invasion-related proteins by proteomic analysis. Proteome Sci 2009; 7: 3.
Sun W, Xing B, Sun Y, Du X, Lu M, Hao C et al. Proteome analysis of hepatocellular carcinoma by two-dimensional difference gel electrophoresis: novel protein markers in hepatocellular carcinoma tissues. Mol Cell Proteom 2007; 6: 1798–1808.
Wang T-H, Chao AA-S, Tsai C-L, Chang C-L, Chen S-H, Lee Y-S et al. Stress-induced phosphoprotein 1 as a secreted biomarker for human ovarian cancer promotes cancer cell proliferation. Mol Cell Proteom 2010; 9: 1873–1884.
Sims JD, McCready J, Jay DG . Extracellular heat shock protein (Hsp)70 and Hsp90α assist in matrix metalloproteinase-2 activation and breast cancer cell migration and invasion. PLoS One 2011; 6: e18848.
Walsh N, Larkin A, Swan N, Conlon K, Dowling P, McDermott R et al. RNAi knockdown of Hop (Hsp70/Hsp90 organising protein) decreases invasion via MMP-2 down regulation. Cancer Lett 2011; 306: 180–189.
Eustace BK, Jay DG . Extracellular roles for the molecular chaperone, hsp90. Cell Cycle. 2004; 3: 1098–1100.
Tsai C-L, Tsai C-N, Lin C-Y, Chen H-W, Lee Y-S, Chao A et al. Secreted stress-induced phosphoprotein 1 activates the ALK2-SMAD signaling pathways and promotes cell proliferation of ovarian cancer cells. Cell Rep 2012; 2: 283–293.
Wasilewska-Sampaio AP, Santos TG, Lopes MH, Cammarota M, Martins VR . The growth of glioblastoma orthotopic xenografts in nude mice is directly correlated with impaired object recognition memory. Physiol Behav 2013; 123C: 55–61.
Caine C, Mehta MP, Laack NN, Gondi V . Cognitive function testing in adult brain tumor trials: lessons from a comprehensive review. Expert Rev Anticancer Ther 2012; 12: 655–667.
Sy M-S, Altekruse SF, Li C, Lynch CF, Goodman MT, Hernandez BY et al. Association of prion protein expression with pancreatic adenocarcinoma survival in the SEER residual tissue repository. Cancer Biomark 2012; 10: 251–258.
Antonacopoulou AG, Palli M, Marousi S, Dimitrakopoulos F-I, Kyriakopoulou U, Tsamandas AC et al. Prion protein expression and the M129V polymorphism of the PRNP gene in patients with colorectal cancer. Mol Carcinogen 2010; 49: 693–699.
Antonacopoulou AG, Grivas PD, Skarlas L, Kalofonos M, Scopa CD, Kalofonos HP . POLR2F, ATP6V0A1 and PRNP expression in colorectal cancer: new molecules with prognostic significance? Anticancer Res 2008; 28: 1221–1227.
Pimienta G, Herbert KM, Regan L . A compound that inhibits the HOP–Hsp90 complex formation and has unique killing effects in breast cancer cell lines. Mol Pharm 2011; 8: 2252–2261.
Horibe T, Kohno M, Haramoto M, Ohara K, Kawakami K . Designed hybrid TPR peptide targeting Hsp90 as a novel anticancer agent. J Transl Med 2011; 9: 8.
Horibe T, Torisawa A, Kohno M, Kawakami K . Molecular mechanism of cytotoxicity induced by Hsp90-targeted Antp-TPR hybrid peptide in glioblastoma cells. Mol Cancer 2012; 11: 59.
Lima FRS, Arantes CP, Muras AG, Nomizo R, Brentani RR, Martins VR . Cellular prion protein expression in astrocytes modulates neuronal survival and differentiation. J Neurochem 2007; 103: 2164–2176.
Barbieri G, Palumbo S, Gabrusiewicz K, Azzalin A, Marchesi N, Spedito A et al. Silencing of cellular prion protein (PrPC) expression by DNA-antisense oligonucleotides induces autophagy-dependent cell death in glioma cells. Autophagy 2011; 7: 840–853.
Yu G, Jiang L, Xu Y, Guo H, Liu H, Zhang Y et al. Silencing prion protein in MDA-MB-435 breast cancer cells leads to pleiotropic cellular responses to cytotoxic stimuli. PLoS One 2012; 7: e48146.
Meslin F, Conforti R, Mazouni C, Morel N, Tomasic G, Drusch F et al. Efficacy of adjuvant chemotherapy according to prion protein expression in patients with estrogen receptor-negative breast cancer. Ann Oncol 2007; 18: 1793–1798.
Willmer T, Contu L, Blatch GL, Edkins AL . Knockdown of Hop downregulates RhoC expression, and decreases pseudopodia formation and migration in cancer cell lines. Cancer Lett 2013; 328: 252–260.
Allhenn D, Boushehri MAS, Lamprecht A . Drug delivery strategies for the treatment of malignant gliomas. Int J Pharm 2012; 436: 299–310.
Quant EC, Wen PY . Novel medical therapeutics in glioblastomas, including targeted molecular therapies, current and future clinical trials. Neuroimaging Clin N Am 2010; 20: 425–448.
Joo KM, Kim J, Jin J, Kim M, Seol HJ, Muradov J et al. Patient-specific orthotopic glioblastoma xenograft models recapitulate the histopathology and biology of human glioblastomas in situ. Cell Rep 2013; 3: 260–273.
Santos TG, Silva IR, Costa-Silva B, Lepique AP, Martins VR, Lopes MH . Enhanced neural progenitor/stem cells self-renewal via the interaction of stress-inducible protein 1 with the prion protein. Stem Cells 2011; 29: 1126–1136.
Zhuang D, Liu Y, Mao Y, Gao L, Zhang H, Luan S et al. TMZ-induced PrPc/par-4 interaction promotes the survival of human glioma cells. Int J Cancer 2012; 130: 309–318.
Chen J, Li Y, Yu T-S, McKay RM, Burns DK, Kernie SG et al. A restricted cell population propagates glioblastoma growth after chemotherapy. Nature 2012; 488: 522–526.
Butowski N, Chang SM . Endpoints for clinical trials and revised assessment in neuro-oncology. Curr Opin Neurol 2012; 25: 780–785.
Coitinho AS, Lopes MH, Hajj GN, Rossato JI, Freitas AR, Castro CC et al. Short-term memory formation and long-term memory consolidation are enhanced by cellular prion association to stress-inducible protein 1. Neurobiol Dis 2007; 26: 282–290.
Gehring K, Sitskoorn MM, Aaronson NK, Taphoorn MJB . Interventions for cognitive deficits in adults with brain tumours. Lancet Neurol 2008; 7: 548–560.
Brown KR, Otasek D, Ali M, McGuffin MJ, Xie W, Devani B et al. NAViGaTOR: Network Analysis, Visualization and Graphing Toronto. Bioinformatics 2009; 25: 3327–3329.
Kononen J, Bubendorf L, Kallioniemi A, Bärlund M, Schraml P, Leighton S et al. Tissue microarrays for high-throughput molecular profiling of tumor specimens. Nat Med 1998; 4: 844–847.
Alvarenga AW, Coutinho-Camillo CM, Rodrigues BR, Rocha RM, Torres LFB, Martins VR et al. A comparison between manual and automated evaluations of tissue microarray patterns of protein expression. J Histochem Cytochem 2013; 61: 272–282.
Lal S, Lacroix M, Tofilon P, Fuller GN, Sawaya R, Lang FF . An implantable guide-screw system for brain tumor studies in small animals. J Neurosurg 2000; 92: 326–333.
Weissert R, Wallström E, Storch MK, Stefferl A, Lorentzen J, Lassmann H et al. MHC haplotype-dependent regulation of MOG-induced EAE in rats. J Clin Invest 1998; 102: 1265–1273.
This work was supported by Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP 09/14027-2, 07/08410-2 and 04/12133-6) and Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq). Fellowships from FAPESP (to TGS) (2009/51653-9), NGQ (2009/51751-0), BC-S (2008/55381-0), BRR (2010-13654-0, 2012/19019-0), APW-S (2010/20796-6), and from Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) (to APW-S) are gratefully acknowledged. We are thankful for Maria Del Mar Inda and Severino da Silva Ferreira for technical assistance. Drs Maria Dirlei Begnami, Victor Piana de Andrade and Martín Roffe contributed with helpful discussions. We thank the AC Camargo Biobank for providing the astrocytoma samples used in this study.
The authors declare no conflict of interest.
Supplementary Information accompanies this paper on the Oncogene website
About this article
Cite this article
Lopes, M., Santos, T., Rodrigues, B. et al. Disruption of prion protein–HOP engagement impairs glioblastoma growth and cognitive decline and improves overall survival. Oncogene 34, 3305–3314 (2015). https://doi.org/10.1038/onc.2014.261
Journal of Molecular Medicine (2021)
Cytokine & Growth Factor Reviews (2021)
Journal of Oral Pathology & Medicine (2021)