Endoplasmic reticulum chaperone prolyl 4-hydroxylase, beta polypeptide (P4HB) promotes malignant phenotypes in glioma via MAPK signaling

Endoplasmic reticulum (ER) chaperone Prolyl 4-hydroxylase, beta polypeptide (P4HB) has previously been identified as a novel target for chemoresistance in glioblastoma multiforme (GBM). Yet its functional roles in glioma carcinogenesis remain elusive. In clinical analysis using human glioma specimens and Gene Expression Omnibus (GEO) profiles, we found that aberrant expression of P4HB was correlated with high-grade malignancy and an angiogenic phenotype in glioma. Furthermore, P4HB upregulation conferred malignant characteristics including proliferation, invasion, migration and angiogenesis in vitro, and increased tumor growth in vivo via the mitogen-activated protein kinase (MAPK) signaling pathway. Pathway analysis suggested genetic and pharmacologic inhibition of P4HB suppressed MAPK expression and its downstream targets were involved in angiogenesis and invasion. This is the first study that demonstrates the oncogenic roles of P4HB and its underlying mechanism in glioma. Since tumor invasion and Vascularisation are typical hallmarks in malignant glioma, our findings uncover a promising anti-glioma mechanism through P4HB-mediated retardation of MAPK signal transduction.


INTRODUCTION
Glioblastoma multiforme (GBM) is characterised by its highly infiltrative and proliferative invasion of surrounding tissues, which makes it one of the most lethal types of primary brain tumours [1,2]. Despite multimodal treatments involving surgical resection, chemotherapy and radiotherapy, patients with GBM have life expectancies of only around 15 months after diagnosis [3]. One major reason for treatment failure is GBM's aggressive behaviour that invade surrounding tissues, which precludes complete tumour excision [4,5]. Exploiting the signaling mechanisms that drive tumour progression and invasion could be important for the development of therapeutics for this deadly disease.
Endoplasmic reticulum (ER) is the major site for protein synthesis, and the ER chaperone functions to ensure proper folding of newly synthesized proteins. With the increment of protein synthesis in proliferating cancer cells, the activities of ER chaperones experience high demand that might lead to the activation of unfolded protein response (UPR) in cancer cells [6,7]. While this class of proteins was traditionally thought to be primarily responsible for protein folding and post-translational modification within the ER, they are now also found to be critical in regulating proliferation, apoptosis and immunity Research Paper www.impactjournals.com/oncotarget [6,8]. Moreover, due to the high interconnectivity of the ER with other cellular compartments, the functions of ER chaperones may also extend beyond the ER by regulating pro-oncogenic and pro-survival signals [9,10]. Therefore, perturbation of ER homeostasis could have critical roles in Tumourigenesis, and targeting ER chaperones represents a novel direction for developing anti-tumor therapy.
Prolyl 4-hydroxylase, beta polypeptide (P4HB) is one of the major chaperone proteins [11]. Like other ER chaperones, P4HB, is also located on the cell surface, with activities distinct from those in the ER [12,13]. These findings expanded the paradigm of P4HB functions, and implied a high versatility of this protein in different biological processes. Elevated levels of P4HB may confer tolerance against extracellular stresses such as hypoxia and ischemia [14]. Its expression following serial transplantations of clinical glioblastoma xenografts in the brains of immunodeficient rats also showed correlations with invasive properties [15].
Recently, we showed that upregulation of P4HB was associated with temozolomide (TMZ) resistance in malignant glioma, and that its inhibition may sensitize chemoresistant glioma to TMZ treatment [16]. This study aimed at delineating the oncogenic role of P4HB in malignant glioma. Our work provided novel findings that aberrant expression of P4HB promotes tumour invasion, angiogenesis and growth via the MAPK signaling pathways. These multifunctional roles implicate its clinical significance in cancer. Targeting P4HB with consequential blockage of these pathways may provide an alternative treatment approach for GBM.

Upregulated P4HB expression is associated with high-grade human glioma
We first assessed cytoplasmic P4HB expression in 64 human glioma specimens based on the staining intensities and percentage of positively stained cells. Levels of intensity varied among different cancer grades, in that P4HB was detectable in 30/48 (62.5%) high-grade, but only 5/16 (31.2%) in low-grade gliomas. Only 12.6% of low-grade glioma cases demonstrated more than 25% of P4HB-stained cells, which was around 30% relative to high-grade glioma (47.9% cases), suggesting that P4HB expression was correlated with glioma malignancy grades ( Figure 1A). High P4HB expression in high grade gliomas was further validated by western blot and end-point PCR analysis at the protein and mRNA levels ( Figure 1B).

P4HB expression is correlated with increased angiogenesis
In addition to clinical grading (p = 0.013), P4HB expression in malignant glioma was also found to be significantly correlated with angiogenesis as indicated by CD31 expression (p = 0.009), and a positive association with VEGF (p = 0.07) ( Table 1). Microvessels (arterioles, venules, and capillaries) densities and VEGF staining intensities were significantly stronger in high-grade (high P4HB expression) when compared to the low-grade glioma (low P4HB expression) (Figure 2A). Distinctive set of 64 cases. Scoring system was calculated based on intensity of staining (none, weak, moderate and strong; top) and percentage of positively stained cells (< 10%, 10%-25% 26%-50%, 51%-75% and > 75%; bottom) in low-grade glioma (LGG) and high-grade glioma (HGG) specimens. (B) Representative images demonstrated P4HB protein (upper panel) and gene expression (lower panel) in low-grade and high-grade glioma specimens as demonstrated by western blotting and RT-PCR, respectively. β-Actin was used as an internal loading control. vessel staining patterns, revealed by CD31 staining, were observed in a grade-dependent manner with differential P4HB expressions. In low-grade glioma with low P4HB expression, microvessels were predominantly pericyteliked with a capillary phenotype, whereas in highgrade glioma there was hypervascularity with enlarged, branched and disorganized vessel structures. Microvessel density (MVD) correlated positively with P4HB in both low-grade (p = 0.003) and high-grade gliomas (p = 0.0001) ( Figure 2B).

Differential gene expression and biological processes associated with P4HB gene expression
To interrogate the potential mechanisms underlying P4HB's action in the present context, we studied the gene expression profiles of 73 patient-derived tissues obtained from a publicly available database GSE16011. Hierarchical clustering graphically displayed distinct differential gene expression patterns between lowand high-P4HB groups ( Figure 3A). A total of 2227 differentially expressed genes were identified, of which 756 were showed two-fold or more differences.
Gene ontology (GO) enrichment analysis confirmed that most up-regulated genes were featured with aspect to blood vessel, vasculature and tissue development. Others were involved in cell adhesion, extracellular matrix organization and wound healing ( Figure 3B). The findings suggested that P4HB may regulate a wide range of gene functions but predominantly in angiogenesis and Tumour regulation. Using GeneSpring ™ , we have identified several canonical signaling pathways that are significantly correlated with high P4HB expression (p < 0.001) (Supplementary Table 1). Of interest is the involvement of MAPK signaling pathway together with focal adhesion and angiogenesis, which based on the degrees of correlation, our previous findings and literature reviews, were subject to further investigations in this study.

Involvement of MAPK signaling in P4HBdependent oncogenic effects
To study the role of MAPK, D54 and U87 cells were treated with increasing concentrations (0, 250 and 500 μM) of bacitracin (BAC), a P4HB inhibitor. A dose-dependent inhibition of MAPK phosphorylation was observed after 24 hour of treatment ( Figure 4A). Treatment with quercetin-3-rutinoside (Q3R), another potent selective inhibitor of P4HB, at increasing concentrations (0, 25, 50, 100μΜ) for 48 hours similarly suppressed MAPK phosphorylation in D54, U87 and U251 cells ( Figure 4B). The findings suggested that P4HB-dependent oncogenic activities are associated with MAPK signaling.

Over-expression of P4HB promotes cell proliferation, migration, invasion and tube formation ability in vitro
Having found that P4HB expression was associated with glioma malignant phenotypes, we then performed gain-of-function in vitro assays. Figure 5A illustrates that GBM cells with P4HB over-expression were successfully established in D54, U87 and U251 cells (D54-P4HB, U87-P4HB and U251-P4HB). When compared with the vectors controls (D54-Vec, U87-Vec and U251-Vec), cells with P4HB over-expression exhibited significantly (p < 0.05) higher proliferative rates on MTT assay at 120 hour ( Figure 5B). GBM cells with P4HB over-expression also exhibited greater migration abilities at 24 hour than vector controls ( Figure 5C). A similar trend was observed on matrigel invasion assay, with U87-P4HB and U251-P4HB cells showing greater invasion ability at 24 hours when compared to their respective vector controls (p < 0.001) ( Figure 5D). Angiogenic ability was also measured on tube formation assay, parental cells showed no branching and tube network formation while GBM cells with P4HB over-expression aligned to form branchlike and net like ( Figure 5E).
To further examine the involvement of MAPK activities in P4HB-dependent oncogenic effects, MAPK phosphorylation was shown to be enhanced in P4HB over-expressing U87 and U251 cells (U87 P4HB-1, U87 P4HB-2, U251 P4HB-1 and U251 P4HB-2) when compared with the vector controls (U87 Vec-Ctrl and U251 Vec-Ctrl) ( Figure 5F). Suppression of MAPK activities by using U0126, at 0, 1, 5, 10 μM for 2 hours in P4HB over-expressing U87 and U251 cells (U87 P4HB and U251 P4HB) was found to reduce VEGF expression without affecting P4HB level ( Figure 5G). The invasive and angiogenic abilities of U87 P4HB and U251 P4HB cells were affected phenotypically after the reduction of MAPK activity ( Figure 5H and 5I). These findings lend further support to the suggestion that the oncogenic activities of P4HB are, at least partially, mediated by downstream MAPK signaling.

DISCUSSION
Neoplastic progression is a multistep process of genetic mutations and adaptive responses that can override growth arrest or senescence controls [17]. Adaptive endoplasmic reticulum stress response (ERSR) due to chronic ER stress (e.g., caused by hypoxia and glucose deprivation) is a common feature in rapidly proliferating cancer cells [18], and may protect tumor cells against further exogenous insults such as chemotherapeutics [19,20]. ERSR is carried out by chaperones whose main functions are to facilitate protein folding and the eradication of malformed proteins [21], but recent discoveries indicate that chaperones may also affect tumor cell growth and signaling [22].
Glioma cells are in a constant state of low grade ERSR that possibly contributes to their resistance to chemoirradiation [21]. The most abundant and well-characterised ER chaperone, GRP78 has been identified to be critical for tumourigenesis and therapeutic resistance [6,23], exhibiting elevated expression in many cancers, including glioma [24]. Cancer cells with constitutively upregulated GRP78 could tolerate chemotherapeutic agents through the suppression of pro-apoptotic pathways [25]. For oncogenesis, GRP78 could induce the activation of AKT signaling during leukemogenesis and prostate tumorigenesis [26][27][28]. A conditional heterozygous deletion of GRP78 may reduce tumour angiogenesis and metastatic growth in tumor endothelial cells but not in normal tissue [29]. Other chaperones such as GRP 94, calreticulin (CRT) and protein disulfide isomerase (PDI) are similarly implicated [30][31][32].  Our group has previously identified another ER chaperone protein, P4HB, to be associated with TMZ resistance in malignant glioma [33]. Down-regulation of P4HB enhanced chemosensitivity via the ERSR signaling pathway both in vitro and in vivo, implying that it is a potential target for chemoresistant GBM [16]. The P4HB family also has neuroprotective actions in various neurodegenerative diseases and cerebral ischemia [34][35][36]. As such, it is conceivable, though unproven, that high P4HB expression in brain tumours may also exert impact on disease progression and clinical outcome. Goplen et al. were first to identify the role of P4HB in cancer invasion but no functional studies or clinical validation were reported [15]. Other investigators have also described its protective role in post-injury response and anti-tumor immunity, its role in glioma tumourigenesis has been elusive [37,38]. The present study is the first to describe the oncogenic role of P4HB in GBM and the mechanisms underlying its contribution to glioma progression.
MAPK signaling is one of the main EGFR downstream pathways that are known to be critical for glioma tumourigenesis [39,40]. MAPK signaling are evolutionarily important linkages to the machinery that controls fundamental cellular processes such as proliferation, migration and apoptosis [41]. Abnormalities in MAPK signaling play a critical role in the development and progression of cancer [42]. Recently, STL1, the co-chaperone stress-in-double protein 1, is reported to promote glioma proliferation through MAPK pathways, suggests a possible link between MAPK and chaperone activities in the development in glioma [43].
Interestingly, we found that up-regulation of P4HB was significantly associated with MAPK activation signaling and the critical downstream processes of angiogenesis and invasion. We also delineated the influences of P4HB on the regulations of VEGF. By adversely perturbing P4HB in vitro, MAPK and VEGF expressions decreased; contrawise, P4HB over-expression  had the opposite effects. To further verify our findings, we suppressed P4HB expression pharmaceutically. BAC is widely used as a P4HB inhibitor of uncertain specificity of action [44], whereas Q3R has more selective mode of action [13]. We observed that both inhibitors could suppress MAPK phosphorylation, suggesting that P4HB may regulate MAPK signaling. Conversely, inhibiting MAPK pathway with U0126 abrogated the induction of VEGF, tumor invasion and angiogenesis, without alterations in P4HB protein levels. The overall findings suggested that MAPK signaling was downstream to P4HB.

Human glioma specimens and cell lines
A total of 64 specimens (mean age: 49.48 ± 13.80 years) obtained from 2002 to 2010 were retrieved from our institution's tissue bank and comprised 16 lowgrade gliomas (WHO grade II), 12 anaplastic gliomas (WHO grade III) and 36 GBM (WHO grade IV) [2]. The diagnosis of malignant glioma was made radiologically via magnetic resonance imaging (MRI) before surgery, and confirmed histologically by a certified pathologist. The study was approved by the Institutional Review Board of our institution, and all tissues were collected with signed informed consent from patients.
Microvessel density (MVD) was evaluated after CD31 staining [46]. Only stained endothelial cells or endothelial cell clusters clearly separated from adjacent microvessels, tumour cells, and connective tissue elements were considered as single, countable microvessel. All slides were initially scanned at low magnification (40× or 100×) for the identification of most prominent vascular 'hot spots' followed by MVD counts on five random 200× microscopic fields. The mean value of five counts was recorded and the total score was calculated based on the average of individual counts made by 2 independent observers. Immunoblot 25µg of total protein lysates were separated on 12% polyacrylamide SDS gels and electroblotted on nitrocellulose membranes as previously described [47]. Briefly, after blocking with 5% non-fat milk in TBS-T (20 mM Tris, 137 mM NaCl, 0.1% Tween-20, pH 7.6), the membrane was probed with one of the following primary antibodies (at 1:1000 dilution) at 4 o C overnight: rabbit monoclonal antibodies against P4HB, total p44/42 MAP kinase (Erk 1/2) (total MAPK), phosphor-p44/42 MAP kinase (Erk1/2) (Thr202/Tyr204) (pMAPK), and VEGF (all from Cell Signaling Technology Inc.). The membranes were washed three times with TBS/T followed by incubation with 1:10,000 diluted HRP-conjugated secondary antibodies (Invitrogen-Zymed Laboratories) at 4 o C for 1 hour. Signal on blots was developed using ECL detection system (GE Biosciences, Buckinghamshire, England).

Semi-quantitative RT-PCR
1 μg of RNA extracts was reverse-transcribed into cDNA using the TaqMan reverse transcription reagents (Applied Biosystems, CA, USA). PCR amplification was performed in a 20 μLreaction mixture containing 3 μLof cDNA template and 200nM of each primer [P4HB or beta actin (ACTB)]. ACTB was used as an internal control for PCR quality. The primer sequences for P4HB were 5′-GCTGATGACATCGTGAACTGG-3′ (sense) and 5′-TTGGAGAACACGTCACTGTTG-3′ (antisense) and ACTB were 5′-CCAACCGCGAGAAGATGA-3′ (sense) and 5′-CCAGAGGCGTACAGGGATAG-3′ (antisense). The RT-PCR was programmed and started at initial incubation at 94°C for 5 min to activate the Taq DNA polymerase followed by 35 cycles at 94°C for 45 s, 58°C for 45 s, 72°C for 40 s and final extension at 72°C for 10 min. The RT-PCR assay was repeated twice and PCR products were separated by electrophoresis on 2% agarose gel and visualized under UV light after ethidium bromide staining.

Affymetrix microarray dataset
To interrogate the molecular mechanism underlying P4HB's actions, we studied the gene expression profiles of 73 human specimens from Genomic Spatial Event (GSE) 16011 at the National Center for Biotechnology Information (NCBI) Gene Expression Omnibus (GEO) database (http://www.ncbi.nlm.nih.gov/geo/). These included 7 normal controls, 17 World Health Organization (WHO) grade II, 29 grade III, and 20 grade IV gliomas. The raw CEL files generated from Affymetrix GeneChip Human Genome U133 Plus 2.0 Array were Robust Multi-Array Average (RMA) preprocessed and normalized using GeneSpring software version 12.5 (Agilent Technologies, Santa Clara, CA, USA). The differential gene expression profiles of tissues with low-or high-P4HB expression (using the median expression level as cut-off) were identified. GO enrichment analysis was performed to delineate the predominant functions of up-regulated genes (p ≤ 0.00001) under the GO category 'biological processes'.

Intracranial xenograft model
Evaluation of tumorigenicity on parental GBM cells and GBM cells with P4HB over-expression was performed using orthotopic tumor model based on guidelines approved by the Committee on the Use of Live Animal for Teaching and Research (CULATR). Immunocompromised [athymic nude (nu/nu)] mice at about six weeks of age were anesthetized by intra-peritoneal injection of 50 mg/kg ketamine/xylazine. Mice were fitted onto a stereotaxic device for intracerebral injections (RWD Life Science Co., Ltd., ShenZhen, GuangDong Province, China). Briefly, a small skin incision was made to expose the bregma suture followed by the creation of a small burr hole at anteriorposterior (AP) = +1 and medial-lateral (ML) = −2.5 from bregma by a micromotor drill (Hager & Meisinger GmbH, Neuss, Germany). 1 × 10 6 GBM cells in 5µL of PBS (U87-Vec-Ctrl, P4HB-1 and P4HB-2; U251-Vec-Ctrl, P4HB) were slowly deposited (1µL/min) in the right striatum at a depth of −3.5 mm from dura with a 10-µL syringe (26-gauge needle; Hamilton Co., Reno, NV, USA). The needle was left in place for 1 min and slowly withdrawn in 3 min followed by closure of the skin incision with sutures.

Statistical analysis
All statistical analyses were performed using PASW Statistics 18.0 (SPSS Inc., Chicago, IL, USA). Differentially expressed genes with a fold change ≥ 2 and p value ≤ 0.001 were included for Hierarchical Clustering Analysis (HCA) [48] and Gene Ontologies (GO) analysis for the three categories:-biological process (BP), molecular function (MF), and cellular component (CC) [49]. Pathway analysis in GeneSpring GX software was used to identify canonical pathways associated with the differentially expressed genes with p value ≤ 0.001. Correlation of P4HB expression with different pathological grades were analyzed by Pearson's chi-squared (χ 2 ) test. Samples belonging to the P4HB High and P4HB Low groups were defined based on the median value of the sample population. Continuous data were presented as mean ± SD. Student's t-test was used to determine whether a difference existed between two groups. One-way ANOVA analysis was used for comparison of more than 2 groups. P values of less than 0.05 were considered statistically significant.

CONCLUSIONS
This study is the first to describe the oncogenic effects of P4HB in glioma and the mechanistic linking of P4HB-mediated MAPK activation in glioma progression, invasion and angiogenesis. It provides new knowledge on the significance of chaperone proteins dysregulation in malignant glioma both clinically and at a molecular level. These findings supplement our previous finding that inhibition of P4HB may resensitise chemoresistant GBM to TMZ. P4HB may be further exploited as a potential predictive marker for GBM prognosis and an alternative therapeutic approach for GBM and possibility in other cancers. Future studies will help to delineation the whole picture of P4HB dependent MAPK signaling together with their molecular interactions in glioma malignancy.