GLIPR1-ΔTM protein as a systemic therapeutic approach for prostate cancer

Abstract

GLIPR1 is a p53 target gene known to be downregulated in prostate cancer, and increased endogenous GLIPR1 expression has been associated with increased production of reactive oxygen species, increased apoptosis, decreased c-Myc protein levels, and increased cell cycle arrest. Recently, we found that upregulation of GLIPR1 in prostate cancer cells increases mitotic catastrophe through interaction with Hsc70 and downregulation of Aurora kinase A and TPX2. In the current study, we evaluated the mechanisms of recombinant GLIPR1 protein (GLIPR1-ΔTM) uptake by prostate cancer cells and the efficacy of systemic GLIPR1-ΔTM administration in a prostate cancer xenograft mouse model. GLIPR1-ΔTM was selectively internalized by prostate cancer cells, leading to increased apoptosis through reactive oxygen species production and to decreased c-Myc protein levels. Interestingly, GLIPR1-ΔTM was internalized through clathrin-mediated endocytosis in association with Hsc70. Systemic administration of GLIPR1-ΔTM significantly inhibited VCaP xenograft growth. GLIPR1-ΔTM left no evidence of toxicity after it was completely removed from the mouse models 8 hours after injection. Our results demonstrate that GLIPR1-ΔTM is selectively endocytosed by prostate cancer cells, leading to increased reactive oxygen species production and apoptosis, and that systemic GLIPR1-ΔTM significantly inhibits growth of VCaP xenografts without substantial toxicity.

Introduction

For men in Western societies, prostate cancer is the most frequently diagnosed cancer and the second-leading cause of cancer-related death. The standard approaches for patients with localized prostate cancer are surveillance, surgery, and/or radiation therapy, depending on prostate-specific antigen values and the stage and grade of the disease. Adjuvant therapy and neoadjuvant therapy for high-risk patients have not yet been fully validated.¹ Metastatic prostate cancer treated mainly with hormonal therapy is associated with a poor prognosis, and the only chemotherapy agents approved by the United States Food and Drug Administration for metastatic prostate cancer are docetaxel and cabazitaxel, which are considered palliative approaches.^{2, 3} New, effective, well-tolerated agents for systemic therapy of prostate cancer need to be identified.

Human glioma pathogenesis-related protein 1 (GLIPR1) is a p53 target gene that is downregulated in prostate cancer owing to methylation of its promoter.^{4, 5} GLIPR1 has been found to be downregulated in other human cancers as well, such as leukemia and Wilms tumor.^{6, 7} Glipr1^{−/ −} mice had significantly shorter tumor-free survival times than Glipr1^+/+ and Glipr1^+/− mice for both the p53^+/+ and the p53^+/− genotypes owing to the unique array of malignancies in the Glipr1^{−/ −} mice.⁸ Overall, these studies suggest that GLIPR1 is a tumor suppressor. The orthotopic injection of an adenoviral vector overexpressing GLIPR1 in a mouse model of metastatic prostate cancer led to decreased microvessel density, indicating that GLIPR1 has anti-angiogenic activity and increases infiltration of tumor-associated macrophages and cytotoxic T cells.⁹ Recently, intraprostatic injection of the GLIPR1-expressing adenoviral vector prior to radical prostatectomy has been shown to be safe for patients with intermediate- or high-risk localized disease, and the preliminary data suggested a stimulated immune response against the tumor and significant antitumor activity, with increased apoptosis and p27 expression in the prostatectomy specimens.¹⁰ Finally, a vector-mediated, GLIPR1-modified, tumor cell–based vaccine inhibited the development of prostate cancer in mice injected orthotopically with 178-2 BMA cells and increased the activity of natural killer cells and cytotoxic T cells.¹¹ These data suggest that GLIPR1 therapy is a rational approach for patients with localized or metastatic prostate cancer.

Multiple established pathways are related to the antitumor effect of GLIPR1 on prostate cancer. A study by Li et al. showed that GLIPR1 upregulation increases the production of reactive oxygen species (ROS), leading to p53-independent activation of the JNK/c-Jun pathway and inhibition of the anti-apoptotic molecule Bcl2.⁸ More recent data suggest that GLIPR1 upregulation can lead to redistribution of casein kinase 1α from the Golgi complex to the cytoplasm, where it can phosphorylate β-catenin and c-Myc, leading to their degradation. The decreased β-catenin signaling leads to decreased expression of the β-catenin target gene MYC and increased p21 expression and cell cycle arrest.¹² These studies show that increased apoptosis and cell cycle arrest contribute to the antitumor effects of GLIPR1.

We recently showed that GLIPR1 gene transfer led to direct interaction with heat shock cognate protein 70 (Hsc70), a member of the heat shock protein 70 family, leading to destabilization and degradation of Sp1 and c-Myb and to decreased mitotic spindle stabilization and mitotic catastrophe in prostate cancer cells through downregulation of Aurora kinase A and TPX2.¹³ Importantly, we showed that a recombinant, truncated form of GLIPR1—GLIPR1-ΔTM—can induce apoptosis and mitotic catastrophe in prostate cancer cells in vitro and suppress tumor growth in vivo after systemic injection. These effects were accompanied by suppression of c-Myb, Aurora kinase A, and TPX2.¹³

Although GLIPR1-ΔTM antitumor efficacy has been established previously,¹³ there are very limited data regarding dose- and time-dependent therapeutic response to GLIPR1-ΔTM, the mechanism of GLIPR1-ΔTM uptake in prostate cancer cells and in non-tumorigenic prostate epithelial cells, and the biodistribution, half-life, and toxicity of systemic GLIPR1-ΔTM.

In the current study, we aim to further analyze GLIPR1-ΔTM cytotoxic activities in prostate cancer cells in vitro and in vivo; determine the specificity, kinetics, and mechanism of GLIPR1-ΔTM uptake by prostate cancer cells; determine the biodistribution and protein half-life of GLIPR1-ΔTM in vivo; and determine the toxicity of systemic intravenous administration of GLIPR1-ΔTM in vivo.

Materials and Methods

Cell lines

PrEC (from Lonza, Walkersville, MD), was grown in PrEGM^™ Bullet Kit^™ complete medium and PWR-1E and RWPE-1 (both from ATCC) were grown in complete Keratinocyte Serum Free Medium (K-SFM). LNCaP, DU145, PC-3 and VCaP were grown as described previously.¹³ All the cell lines were validated by short tandem repeat DNA fingerprinting with the AmpFlSTR Identifiler Kit (Applied Biosystems, Inc., Foster City, CA) in the Characterized Cell Line Core Facility at The University of Texas MD Anderson Cancer Center.

Recombinant GLIPR1-ΔTM purification

The GLIPR1-ΔTM–with his and myc tag plasmid was constructed using a pSecTag2/Hygro expression vector. Recombinant GLIPR1-ΔTM–with his and myc tag (hereafter, GLIPR1-ΔTM) was purified as described previously.¹³

Apoptosis analysis

Apoptotic nuclear morphology was analyzed with fluorescence microscopy after 4′,6-diamidino-2-phenylindole (DAPI) staining (2 μg/mL). DNA fragmentation analysis was performed using Cell Death Detection ELISA (Roche Applied Science, Indianapolis, IN) according to the manufacturer’s instructions.

Cell viability assay

Cell viability was evaluated using an MTS CellTiter 96^® AQueous One Solution Cell Proliferation Assay (Promega, Madison, WI) according to the manufacturer’s instructions as previously described.¹⁴

ROS detection and measurement

ROS levels were measured using a TECAN GENios microplate reader (MTX Lab Systems, Vienna, VA) after incubation with CM-H₂DCFDA as previously described.⁸ The antioxidant butylated hydroxyanisole (BHA) was purchased from Sigma-Aldrich (St. Louis, MO).

Western blot analysis

PC-3, VCaP and DU145 cells were treated with 20, 40, or 80 μg/mL of GLIPR1-ΔTM for 6, 12, or 24 hr in RPMI-1640 medium containing 0.1% serum for PC-3 cells and in high glucose DMEM containing 0.5% serum for VCaP cells. The treatment was followed by 10% serum stimulation for 30 min and lysates were collected. Antibodies against GLIPR1-ΔTM (Myc-tag,) were purchased from Cell Signaling Technology (9B11; #2276, Danvers, MA), antibodies against c-Myc were purchased from Santa Cruz Biotechnology (9E10; #sc-40; Santa Cruz, CA), antibodies against p21 were purchased from BD Biosciences (610233; Franklin Lakes, NJ), and antibodies against β-actin were purchased from Sigma-Aldrich (A2228 Saint Louis, Mo).

Immunofluorescence assay

Alexa 488 labeling of recombinant GLIPR1-ΔTM was prepared with an Alexa Fluor 488 Protein Labeling Kit (Invitrogen) according to the manufacturer’s instructions. Alexa 594 labeling of cholera toxin B, transferrin, and dextran was prepared with an Alexa Fluor 594 Protein Labeling Kit (Invitrogen) according to the manufacturer’s instructions. Chlorpromazine (CPZ), methyl-β-cyclodextrin (MβCD), and wortmannin inhibitors were purchased from Sigma-Aldrich. Hsc70 antibody was purchased from Abcam (ab19136; Cambridge, MA). Prostate cancer cells or non-tumorigenic prostate epithelial cells were washed twice with PBS buffer and incubated with Alexa 488–labeled GLIPR1-ΔTM, Alexa 594–labeled cholera toxin B, Alexa 594-labeled transferrin, and Alexa 594-labeled dextran. After 6 hr or 24 hr, cells were rinsed with PBS buffer and fixed in 4% formaldehyde for 5 min at room temperature. For the inhibitors study, cells were treated with CPZ, MβCD, or wortmannin 1 hr before treatment with labeled GLIPR1, transferrin, cholera toxin B, or dextran. The immunofluorescence of the target molecules was evaluated using an Eclipse 90i image analysis system (Nikon Instruments, Tokyo, Japan) with NIS-Elements AR software (version 3.0).

Red fluorescent protein (RFP) constructs of cellular organelle markers were purchased from OriGene (Rockville, MD). Cells were treated with FuGENE HD Transfection Reagent (Roche Applied Science) and 2 μg/mL DNA according to the manufacturer’s protocol. After 24 hr, the cells were washed and then cultured with Alexa 488–labeled GLIPR1-ΔTM in serum-free medium for 24 hr. Transfected cells were detected by fluorescence, and the internalization of GLIPR1-ΔTM was evaluated by confocal microscopy analysis. For the Hsc70 colocalization study, cells were treated with 20 μg/mL Alexa 488–labeled GLIPR1-ΔTM for 24 hr and then incubated with Hsc70 antibody at 1:100 dilution, and anti-rat secondary antibody conjugated with Alexa 594 at 1:400 dilution was used.

Tumor induction in mice

VCaP cells were transduced with lentivirus stably expressing luciferase. Aliquots of 3 × 10⁶ VCaP-luciferase cells in 25 μL PBS were injected directly into the right lobe of the dorsolateral prostate in athymic nude male mice (Taconic Farm, Hudson, NY) to induce orthotopic tumors. The tumors were allowed to grow for 14 days before treatment.

GLIPR1-ΔTM efficacy study

Three times a week for 2 weeks, the experimental groups received intra-tail vein injections of escalating doses of GLIPR1-ΔTM (20, 40, 80, and 160 μg/injection), while the control group received intra-tail vein injections of PBS. Tumor size was monitored by measuring the luminescence signal using the IVIS 200 Series (PerkinElmer, Wellesley, MA). Five weeks after the initial injection, the tumor-bearing mice were sacrificed and the tumors were collected and weighted.

Bioclearance study

Twenty CD-1 mice (22–24 g) purchased from Charles River were included in the bioclearance study. GLIPR1-ΔTM protein was iodinated using the Thermo Fisher Scientific Pierce Iodination Reagent (Waltham, MA) according to the manufacturer’s instructions as previously described.¹⁵ The mice were injected with 80 μg ¹²⁵I–GLIPR1-ΔTM with tag/50 μL. Serum, brain, thyroid gland, heart, liver, lung, spleen, bladder, prostate (ventral (VP), dorsolateral (DLP), and anterior (AP)), and kidney samples were removed 10 and 30 min and 1, 2, 4, 8, and 24 hr after injection. A gamma counter was used to measure emissions.

Toxicity study

Twenty three CD-1 mice (Charles River) were included in the toxicity study. The control group received intra-tail vein injections of PBS while experimental groups received intra-tail vein injections of 640 μg or 2.4 mg GLIPR1-ΔTM protein three times per week for 2 weeks. Then peripheral blood was collected from the mice for analysis of the following serum chemistry: blood urea nitrogen (BUN) (mg/dl), serum aspartate transaminase (AST) (U/L), serum alanine transaminase (ALT) (U/L), alkaline phosphatase (ALK) (U/L), total bilirubin (mg/dl), total protein (g/dl), albumin (g/dl), and globulin (g/dl). In addition, hematology parameters including total leukocyte count (WBC) (10³/μl), red blood cell (RBC) (10⁶/μl), hemoglobin (Hb) (g/dl), hematocrit (Hct) (%), and platelets (Plt) (10⁵/μl) were measured.

Statistical analysis

Unpaired t-tests were used in the experiments in which probability was determined. The p-values obtained were two-tailed and determined to be significant at p < 0.05.

Results

GLIPR1-ΔTM selectively induces apoptosis in prostate cancer cells through ROS generation

The prostate cancer cell lines DU145, PC-3, LNCaP, and VCaP and non-tumorigenic prostate epithelial cell lines PrEC, PWR-1E, and RWPE-1 were treated with 20 μg/mL GLIPR1-ΔTM protein or PBS to evaluate apoptosis and survival responses. DAPI staining and DNA fragmentation analysis showed that cancer cells treated with GLIPR1-ΔTM for 24 or 48 hr had higher rates of apoptosis than those treated with PBS, but non-tumorigenic cells treated with GLIPR1-ΔTM and those treated with PBS did not have significantly different rates of apoptosis (Figs. 1a and 1b). Moreover, we examined apoptosis and survival rates after administration of increasing concentrations of GLIPR1-ΔTM (0, 10, 20, 40, and 80 μg/mL) for 48 hr in VCaP and PC-3 cells. DAPI staining (Figs. 1c and 1d) showed that the apoptosis rate was significantly higher in treated cells than in untreated cells, including those treated with 10 μg/mL GLIPR1-ΔTM (p < 0.0001 in VCaP cells and p = 0.0002 in PC-3 cells). DNA fragmentation (Figs. 1e and 1f) confirmed that 10 μg/mL GLIPR1-ΔTM significantly increased apoptosis rates compared with PBS (0 μg/mL) in both cell lines (p = 0.01 in VCaP cells and p = 0.02 in PC-3 cells), while concentrations of 20–80 μg/mL GLIPR1-ΔTM significantly increased apoptosis rates compared with 10 μg/mL only in VCaP cells (p = 0.002). We used the MTS assay to analyze the survival rates of these cells after 48 hr of GLIPR1-ΔTM treatment and found that survival was significantly decreased with GLIPR1-ΔTM treatment, even for those treated with 10 μg/mL GLIPR1-ΔTM (p = 0.0001 in VCaP cells and p = 0.03 in PC-3 cells; Figs. 1g and 1h).

Figure 1.

GLIPR1-ΔTM selectively induces apoptosis in prostate cancer cells through ROS generation. A. DAPI staining evaluating apoptosis of prostate normal cells and cancer cells treated with PBS or 20 μg/mL GLIPR1-ΔTM at the indicated time showed increased apoptosis only in prostate cancer cells. B. DNA fragmentation in the same cells under the same conditions confirmed that GLIPR1-ΔTM induced apoptosis only in cancer cells and not in normal prostate cells. C. DAPI staining evaluating apoptosis of VCaP cells treated with 0 (PBS), 10, 20, 40, or 80 μg/mL GLIPR1-ΔTM for 48 hr showed that the apoptotic rate significantly increased, even with 10 μg/mL GLIPR1-ΔTM (p < 0.0001). D. DAPI staining evaluating apoptosis of PC-3 cells in the same conditions showed increased apoptosis with 10 μg/mL GLIPR1-ΔTM (p = 0.0002). E. DNA fragmentation showed that 10 μg/mL GLIPR1-ΔTM increased apoptosis in VCaP cells (p = 0.01); 20 μg/mL led to increased apoptosis compared to 10 μg/mL in VCaP cells (p = 0.002). F. DNA fragmentation showed that 10 μg/mL GLIPR1-ΔTM increased apoptosis in PC-3 cells (p = 0.02), but there was no difference between 10 μg/mL and other doses with regard to apoptotic rate. G. MTS assay evaluating viability of VCaP cells under the same conditions showed that the number of viable cells significantly decreased, even with 10 μg/mL (p = 0.0001). H. MTS assay evaluating viability of PC-3 cells under the same conditions showed that the number of viable cells significantly decreased, even with 10 μg/mL (p = 0.03). K. Treatment with 20 μg/mL GLIPR1-ΔTM for 48 hr results in elevated ROS levels in prostate cancer cells but not in non-tumorigenic prostate cells. L. Antioxidant BHA effectively blocked production of cellular ROS in prostate cancer cells. M. BHA inhibited apoptosis based on DAPI staining in prostate cancer cells. (− represents absence of GLIPR1-ΔTM treatment and + represents treatment with 20μg/ml of GLIPR1-ΔTM).

According to a report by Li et al.,⁸ endogenous GLIPR1 upregulation is related to increased ROS production and apoptosis induction. To determine whether exogenous GLIPR1 increases ROS production, we analyzed ROS levels in non-tumorigenic prostate epithelial cells (PrEC, PWR-1E, and RWPE-1) and prostate cancer cells (LNCaP, DU145, PC-3, and VCaP) after 20 μg/mL GLIPR1-ΔTM or PBS administration for 48 hr. We found that ROS production was markedly higher in treated cancer cells than in untreated cancer cells and treated and untreated non-tumorigenic cells (Fig. 1k). Because it is known that ROS production can be a mechanism of apoptosis induction, we used the antioxidant BHA to evaluate the effect of decreased ROS production on apoptosis rate. The administration of escalating doses of BHA (0, 10, 50, 100, and 200 μmol/L) decreased ROS production and apoptosis rates in prostate cancer cells despite the GLIPR1-ΔTM treatment (Figs. 1l and 1m). These results indicate that GLIPR1-ΔTM administration induces apoptosis through ROS production in prostate cancer cells but not in non-tumorigenic prostate epithelial cells.

GLIPR1-ΔTM is selectively taken up by prostate cancer cells

The significantly increased efficacy of GLIPR1-ΔTM protein in cancer cells compared with non-tumorigenic prostate epithelial cells led us to hypothesize that prostate cancer cells selectively take up GLIPR1-ΔTM protein, whereas non-tumorigenic prostate epithelial cells do not. To test this hypothesis, we first determined an optimal concentration and treatment time by treating PC-3 cells with Alexa 488–labeled GLIPR1-ΔTM in various concentrations (0, 1, 2, 5, 10, and 20 μg/mL) for 24 hr and with 10 μg/mL Alexa 488–labeled GLIPR1-ΔTM over various amounts of time (0, 6, 12, 18, 24, and 48 hr). We found that GLIPR1-ΔTM was taken up by the cancer cells in a dose-dependent fashion, with significantly increased uptake at 1 μg/mL compared with 0 μg/mL (p = 0.02). GLIPR1-ΔTM (10 μg/mL) uptake was also time-dependent (p < 0.0001; Figs. 2a and 2b). On the basis of these data, we treated the non-tumorigenic prostate epithelial cells (all lines) and the prostate cancer cells (all lines) with 10 μg/mL Alexa 488–labeled GLIPR1-ΔTM for 24 hr, and evaluated its uptake by confocal microscopy. The cancer cells internalized substantial amounts of labeled GLIPR1-ΔTM as previously has been shown,¹³ whereas the non-tumorigenic cells took up minimal to undetectable amounts of GLIPR1-ΔTM (Fig. 2c). Most or all of the cancer cells internalized the GLIPR1-ΔTM, whereas the non-tumorigenic cells took up only 20–30% of the GLIPR1-ΔTM (p < 0.0001; Fig. 2d).

Figure 2.

Prostate cancer cells selectively internalize GLIPR1-ΔTM. A. Percentage of PC-3 cells taking up GLIPR1-ΔTM protein after treatment with 0, 1, 5, 10, or 20 μg/mL for 24 hr. B. Percentage of PC-3 cells internalizing GLIPR1-ΔTM protein after treatment with 10 μg/mL at various time-points. C. Representative micrographs showing internalization of labeled GLIPR1-ΔTM by prostate normal and cancer cells treated with 10 μg/mL of labeled GLIPR1-ΔTM for 24 hr visualized by confocal microscopy. D. Percentage of cells with internalized GLIPR1-ΔTM after treatment of non-tumorigenic prostate epithelial cells and prostate cancer cells with 10 μg/mL for 24 hr (p < 0.0001).

GLIPR1-ΔTM suppresses c-Myc and increases p21 in PC-3, VCaP and DU145 cells

We have previously shown that overexpression of GLIPR1 leads to increased S252 phosphorylation and degradation of c-Myc, resulting in cell cycle arrest.¹² Additional studies showed that GLIPR1 overexpression also led to increased p21 in multiple prostate cancer cell lines.¹² We evaluated the effect of GLIPR1-ΔTM administration on c-Myc and p21 levels in PC-3, VCaP and DU145 cells. Specifically, the cells were treated with 0, 20, 40, or 80 μg/mL of GLIPR1-ΔTM for 6, 12, or 24 hr followed by 10% serum stimulation for 30 min or no serum stimulation. Western blotting analysis showed that GLIPR1-ΔTM reduced c-Myc and increased p21 (Figs. 3a and 3b), suggesting that the uptake of exogenous GLIPR1 decreases c-Myc signaling, as previously shown for GLIPR1 overexpression.¹² Apart from c-Myc downregulation other mechanisms may contribute to the induction of p21 explaining the more prominent p21 induction compared to c-Myc reduction. In particular, negative correlation between c-Myb and p21 expression has been reported during colon cell differentiation and it’s conceivable that GLIPR1-mediated downregulation of c-Myb may play a role in this regard.^13,16 Finally, it should be noted that serum deprivation downregulates c-Myc which is shown by our results, while the presence of serum may interfere with the endocytosis of GLIPR1-ΔTM explaining its decreased uptake by cancer cells. These results suggest that GLIPR1-ΔTM increases c-Myc degradation as previously shown for endogenous GLIPR1 induction¹² leading to decreased c-Myc upregulation after serum supplementation. Finally, our results show that p21 upregulation by GLIPR1-ΔTM leads to a plateau especially in PC-3 cells early during treatment (Fig. 3a).

Figure 3.

GLIPR1-ΔTM protein suppresses c-Myc and increases p21 in PC-3, VCaP and DU145 cells. A. PC-3 cells treated with 20, 40, or 80 μg/mL of GLIPR1-ΔTM protein for 6, 12, or 24 hr, with (+SS) or without (−SS) 10% serum stimulation for 30 min. GLIPR1-ΔTM decreased c-Myc and increased p21 protein levels. B. Results from treatment of VCaP cells at the same conditions. C. Results from treatment of DU145 cells at the same conditions.

GLIPR1-ΔTM is endocytosed through the clathrin pathway, accumulates in endosomes and lysosomes, and is colocalized with Hsc70

It is known that the two main mechanisms of endocytosis are clathrin-dependent and clathrin-independent and that the two most frequent mechanisms of clathrin-independent endocytosis are caveolae-dependent endocytosis and macropinocytosis.¹⁷ We have previously reported GLIPR1-ΔTM is taken up through the clathrin pathway.¹³ The aim of this study was to establish a mechanism of GLIPR1-ΔTM uptake, examine the accumulation of GLIPR1-ΔTM in different intracellular compartments and to connect these findings with its established mechanisms of therapeutic activity. We treated PC-3 cells with specific pathway inhibitors for 6 hr and then administered labeled GLIPR1-ΔTM. Specifically, we used CPZ, which inhibits clathrin-dependent endocytosis; MβCD, which inhibits caveolae-dependent endocytosis; and wortmannin, which inhibits PI3K/Akt-mediated macropinocytosis. Then, cells were treated for another 6 hr with Alexa 488–labeled GLIPR1-ΔTM (20 μg/mL); Alexa 594–labeled transferrin, which is known to be endocytosed by the clathrin pathway; Alexa 594-labeled cholera toxin B, which is endocytosed by a caveolae-dependent mechanism; or Alexa 594-labeled dextran, which can be endocytosed by both clathrin-dependent endocytosis and macropinocytosis. We found that GLIPR1-ΔTM uptake and transferrin uptake were inhibited only by CPZ (p = 0.01 for GLIPR1-ΔTM uptake and p = 0.03 for transferrin; Fig. 4a). As expected, cholera toxin B uptake was inhibited by MβCD (p = 0.0008; Fig. 4a) while dextran uptake was inhibited by both CPZ (p = 0.008) and wortmannin (p = 0.002; Fig. 54a). These data clearly show that GLIPR1-ΔTM is endocytosed by a clathrin-dependent pathway, with no contribution from the caveolae-dependent or macropinocytosis pathways. To further support this conclusion, we treated PC-3 cells with 20 μg/mL Alexa 488–labeled GLIPR1-ΔTM in combination with 10 μg/mL Alexa 594–labeled transferrin for 6 hr. Our results demonstrated colocalization of GLIPR1-ΔTM and transferrin in PC-3 cells after 6 hr of treatment, further indicating that GLIPR1-ΔTM is taken up by a clathrin-mediated mechanism (Fig. 4b).

Figure 4.

GLIPR1-ΔTM is endocytosed through a clathrin-dependent pathway and accumulates in endosomes and lysosomes. A. PC-3 cells were treated with CPZ, a clathrin-dependent endocytosis inhibitor; MβCD, a caveolae-dependent endocytosis inhibitor; or wortmannin, a macropinocytosis inhibitor, for 6 hr before treatment with 20 μg/mL GLIPR1-ΔTM or Transferrin, which is endocytosed with clathrin-mediated endocytosis; Cholera toxin B, which is endocytosed by caveolae-dependent endocytosis; or Dextran, which is endocytosed by both clathrin-mediated endocytosis and macropinocytosis, for another 6 hr. GLIPR1-ΔTM uptake was inhibited only by CPZ, suggesting that this protein is taken up by clathrin-mediated endocytosis. Based on the percentage of cells taking up the particular molecules, CPZ inhibits GLIPR1-ΔTM and transferrin uptake (p = 0.01 and p = 0.03 respectively), MβCD effectively inhibits cholera toxin B uptake (p = 0.0008) and CPZ (p = 0.008) and wortmannin (p = 0.002) effectively inhibit dextran uptake. (cont: control and wort: wortmannin) B. PC-3 cells were treated with Alexa 594-labeled transferrin and 20 μg/mL Alexa 488-labeled GLIPR1-ΔTM for 6 hr and visualized by confocal microscopy. GLIPR1-ΔTM and transferrin were colocalized in cells’ cytoplasm. C. PC-3 cells incubated with various organelle markers were treated with 10 μg/mL of GLIPR1-ΔTM for 24 hr and visualized by confocal microscopy. GLIPR1-ΔTM was localized in early endosomes, late endosomes, and lysosomes. (ER: Endoplasmic reticulum) D. PC-3, DU145, and VCaP cancer cells and PWR-1E normal prostate cells were incubated with Rab5, an early endosome marker, and then treated with 10 μg/mL GLIPR1-ΔTM for 24 hr. GLIPR1-ΔTM was localized in early endosomes in all treated cancer cells but not in non-tumorigenic prostate epithelial cells. E. VCaP cells were treated with 20 μg/mL of labeled GLIPR1-ΔTM for 24 hr and then stained for Hsc-70. GLIPR1-ΔTM was conjugated with Alexa 488 (green fluorescence) while anti-rat secondary antibody for Hsc-70 detection was conjugated with Alexa 594 (red fluorescence). Hsc-70 was diffusely localized in the cytoplasm, but areas of colocalization with internalized GLIPR1-ΔTM were observed.

It is known that clathrin-dependent endocytosis leads to accumulation of endocytosed protein in early and late endosomes and finally in lysosomes.¹⁸ To examine the accumulation of GLIPR1-ΔTM in various intracellular organelles, we transfected PC-3 cells with RFP constructs expressing specific cellular organelle markers and then treated them with 10 μg/mL GLIPR1-ΔTM for 24 hr. Our results showed that GLIPR1-ΔTM accumulates in early endosomes, late endosomes, and lysosomes but not in mitochondria, the endoplasmic reticulum, or the Golgi complex (Fig. 4c). On the basis of these results, we transfected PC-3, DU145, and VCaP prostate cancer cells and PWR-1E non-tumorigenic prostate cells with an RFP construct for the early endosome marker Rab5 and then treated them with 10μg/ml Alexa 488-labeled GLIPR1-ΔTM for 24 hr. Interestingly, we found that PC-3, DU145, and VCaP cells demonstrated variable accumulation of GLIPR1-ΔTM in their early endosomes (Fig. 4d). As expected, we did not find any accumulation of GLIPR1-ΔTM in early endosomes of non-tumorigenic PWR-1E, supporting our previous findings that non-tumorigenic prostate epithelial cells take up limited GLIPR1-ΔTM (Fig. 4d).

According to a recent report by Li et al., overexpression of endogenous GLIPR1 can lead to its increased interaction with Hsc70 and subsequent destabilization and degradation of Sp1 and c-Myb.¹³ In addition, Hsc70 is implicated in various levels of clathrin-dependent endocytosis: uncoating of clathrin-coated pits,^19,20 endosomal trafficking, and targeting in lysosomes.^21,22 Thus, interaction between exogenous GLIPR1-ΔTM and Hsc70 would suggest that the mechanism of GLIPR1-ΔTM uptake is directly related to its antitumor activity. To evaluate this interaction, we treated VCaP cells with 20 μg/mL Alexa 488–labeled GLIPR1-ΔTM for 24 hr, and found that this protein was colocalized with the Hsc-70 in the cytoplasm (Fig. 4e). This result suggests that clathrin-dependent endocytosis promotes the interaction between exogenous GLIPR1 and Hsc70, and that Hsc70 is a critical mediator of the antitumor activity of GLIPR1-ΔTM.

Systemic GLIPR1-ΔTM treatment inhibits tumor growth in VCaP xenografts

Next, we evaluated the efficacy of systemic GLIPR1-ΔTM by dose response studies using VCaP-luc xenografts. VCaP-luc cells were injected in nude mice, the tumors were allowed to grow for 14 days, and then 21 mice were treated with PBS, 17 mice with 20 μg of GLIPR1-ΔTM, 16 mice with 40 μg of GLIPR1-ΔTM, 15 mice with 80 μg of GLIPR1-ΔTM, and 10 mice with 160 μg of GLIPR1-ΔTM intravenously three times a week for 2 weeks. Three weeks after the initiation of treatment, the tumors were resected and weighted. Luminescence photon signal was monitored in vivo by using IVIS 200 Series. We found that 80 μg and 160 μg of GLIPR1-ΔTM significantly decreased photon emission compared with PBS control (p = 0.00069 and p = 0.0044, respectively), whereas 20 μg and 40 μg of GLIPR1-ΔTM decreased it but not significantly (p = 0.31 and p = 0.058, respectively; Fig. 5a). However tumors collected from mice treated with 20 μg of GLIPR1-ΔTM weighted less than those of the PBS controls (p = 0.0038), and treatment with 40 μg, 80 μg, and 160 μg of GLIPR1-ΔTM led to even greater inhibition of tumor growth (p < 0.0001 for all doses) (Fig. 5b). Importantly, in mice treated with 80 μg and 160 μg, we detected no residual tumor cells at the site of injection upon gross and microscopic analysis in two (13%) of 15 and three (30%) of ten mice, respectively. These data suggest that 80 μg is the optimal therapeutic dose for systemic GLIPR1-ΔTM administration.

Figure 5.

GLIPR1-ΔTM systematic administration leads to inhibition of tumor growth in an orthotopic VCaP xenograft model. A. VCaP cells were orthotopically injected in nude mice given PBS or one of various doses of GLIPR1-ΔTM (20, 40, 80, and 160 μg) three times a week for 2 weeks. According to the luminescence signal 21 days after orthotopic injection, 80 and 160 μg of GLIPR1-ΔTM protein led to significantly decreased photon emission. B. Tumors were collected 1 week after the end of therapy and weighted. All the doses of GLIPR1-ΔTM led to development of significantly smaller tumor compared to controls. Gross examination and microscopic analysis failed to detect any tumor cells in 2 of 15, and 3 of 10, tumor inoculation sites from mice that received 80μg and 160μg GLIPR1-ΔTM, respectively.

Systemic GLIPR1-ΔTM demonstrates a unique pattern of clearance from the serum and of biodistribution

After establishing the efficacy of systemic GLIPR1-ΔTM, we evaluated the kinetics of GLIPR1-ΔTM clearance from the serum and the organs. First, we injected mice with 80 μg of ¹²⁵I–GLIPR1-ΔTM and collected samples of serum and organs such as the prostate, liver, kidneys, and lungs at various time-points. Gamma counter measurements showed that serum levels of GLIPR1-ΔTM decreased progressively during the first few hours after injection: the half-life was 60 min, and the protein was almost completely removed from the serum 8 hours after injection (Fig. 6a). Evaluation of GLIPR1-ΔTM biodistribution and clearance from various organs showed that despite its rapid absorption by the spleen, liver and kidneys GLIPR1-ΔTM was rapidly removed from the body during the first 2 hr after injection, probably through secretion from the liver and kidneys (Fig. 6b). Finally, it should be noted that GLIPR1-ΔTM demonstrated a biphasic and sustained pattern of absorption by prostate compared to other organs did, and it was not detected in brain (Fig. 6b), suggesting that GLIPR1-ΔTM does not penetrate the blood-brain barrier.

Figure 6.

Evaluation of GLIPR1-ΔTM clearance from serum, biodistribution and toxicity following systemic intravenous injection. A. Alteration of serum levels of GLIPR1-ΔTM protein by time (min). The protein levels decreased progressively during the first hours after intravenous injection. The half-life was 60 min, and the protein levels significantly decreased 120 min after injection. B. Evaluation of GLIPR1-ΔTM bioclearance from various organs showed that the protein was absorbed by the spleen, liver and kidneys for a short time but was rapidly removed from the body, probably through secretion from the liver and kidneys. The normal prostate had more stable absorbance of the protein, and the protein was not detected at all in the brain. C. Based on measurement of serum BUN, AST, ALT, ALK, Total bilirubin, total protein, total albumin, globulin, WBC, RBC, Hb, Ht and Plt there was no significant difference between controls and treated groups apart from an increase of ALT with administration of 2.4mg of GLIPR1-ΔTM protein (p = 0.043).

Systemic GLIPR1-ΔTM did not cause toxicity in a mouse model

At this point, we reasoned that systemic GLIPR1-ΔTM treatment is a rational approach for primary and metastatic prostate cancer. From this perspective, potential toxicities related to systematic GLIPR1-ΔTM must be evaluated. Three times a week for 2 weeks, we administered 640 μg of GLIPR1-ΔTM intravenously in eight mice, 2.4 mg of GLIPR1-ΔTM in three mice, and PBS in 12 mice. Then, peripheral blood was collected from the mice for analysis and comparison of the serum chemistry and hematology parameters. We conducted a toxicity analysis and did not find any significant differences between the treated groups and control group other than an increase in alanine transaminase (p = 0.043) in those treated with 2.4 mg of GLIPR1-ΔTM (Fig. 6c), and no mice died during the treatment period. Notably, 2.4 mg is 30 times more than 80 μg, a dose that was shown to significantly decrease tumor weight (optimal therapeutic dose) in the mouse prostate tumor model. We conclude that systemic GLIPR1-ΔTM under our experimental conditions is not related to significant toxicity.

Discussion

The major aims of this study were to compare the uptake of purified GLIPR1-ΔTM protein by prostate cancer cells with its uptake by non-tumorigenic prostate epithelial cells, evaluate the mechanism of its endocytosis, relate this mechanism to its known antitumor activities, and examine its kinetics and toxicity profiles in a mouse model. We found that GLIPR1-ΔTM is selectively internalized by prostate cancer cells but not by non-tumorigenic prostate epithelial cells; this selective uptake is one of the most important properties in generating targeted therapies. Interestingly, we found that uptake of exogenous GLIPR1-ΔTM is related to increased ROS production and subsequent apoptosis in prostate cancer cells but not in non-tumorigenic prostate epithelial cells. We showed that inhibition of ROS production by antioxidant BHA decreased apoptosis. It is known that oxidative stress and particularly H₂O₂ administration increase c-Myc recruitment to γ–glutamylcysteine synthetase promoters, leading to increased γ–glutamylcysteine synthetase expression and subsequent glutathione production, resulting in decreased ROS production.²³ Thus, our findings indicate that the pro-apoptotic effects of GLIPR1-ΔTM are related to ROS production. Moreover, a recent review by Watson highlighted the role of antioxidants (e.g., glutathione), which are induced by activated oncogenic signaling, including Ras and c-Myc mediated pathways, as inhibitors of ROS accumulation and apoptosis in rapidly dividing cells in metastatic cancers.²⁴ Of note, we showed that exogenous GLIPR1-ΔTM decreased the levels of c-Myc in PC-3, VCaP and DU145 cells (Fig. 3). Thus, we suggest that this GLIPR1-ΔTM-mediated activity can increase the sensitivity of cancer cells to oxidative stress and contribute to the accumulation of ROS and that these activities promote cancer cell–selective apoptosis.

It is known that oxidative stress can promote tumor growth by activating oncogenic signaling while androgen deprivation therapy (ADT) induces oxidative stress leading to growth under castrated conditions mediated by evoked Twist1 and YB-1 signaling and androgen receptor (AR) over expression.²⁵ Based on our results, we propose that induction of ROS by GLIPR1-ΔTM may increase the sensitivity of ADT-resistant prostate cancer cells to oxidative stress beyond that consistent with viability, rendering oxidative stress the “Achilles’ heel” for prostate cancer cells under ADT. Moreover, it has been shown that AR inhibition downregulates c-Myc but prostate cancer cells overexpressing it continue to grow under ADT.²⁶ Given the established downregulation of c-Myc by GLIPR1-ΔTM in both AR positive and AR negative cancer cells it can be suggested that the combination of this agent with ADT may inhibit the development of CRPC phenotype. Further studies are needed to examine the efficacy of ADT and GLIPR1-ΔTM combinational therapies for patients with advanced prostate cancer.

The observation that GLIPR1-ΔTM protein is taken up selectively by prostate cancer cells led us to the hypothesis that this uptake may be mediated by endocytosis. We used endocytosis inhibitors to show that GLIPR1-ΔTM was endocytosed by a clathrin-mediated mechanism, which is the uptake of material into the cell from the surface using clathrin-coated vesicles and involves the formation of a complex that includes adaptor protein 2 and dynamin.¹⁸

We also showed that GLIPR1-ΔTM is accumulated in early and late endosomes and finally in lysosomes, which are the final destination of proteins internalized by clathrin-mediated mechanisms.¹⁷ Interestingly, Hsc70 protein functioning as ATPase catalyzes the uncoating of clathrin-coated pits, allowing detached and uncoated vesicles to travel and fuse with endosomes.^{19, 20} In addition, Hsc70 has been implicated in the late stages of endocytosis such as endosomal trafficking and protein targeting in lysosomes.^{21, 22} Recently, Li et al. found that endogenous GLIPR1 interacts with Hsc70 to promote Sp1 and c-Myb degradation and finally apoptosis and mitotic catastrophe in prostate cancer cells.¹³ We showed that endocytosed GLIPR1-ΔTM interacts with Hsc70 in prostate cancer cells’ cytoplasm. Taken together the results of Li et al, and our findings that show clathrin-mediated uptake of GLIPR1-ΔTM suggest that GLIPR1-Hsc70 interactions mediate both the uptake and cytotoxic activities of GLIPR1.

Finally, we showed that systemic administration of GLIPR1-ΔTM purified protein leads to inhibition of tumor growth in a VCaP-luc xenograft model, even at low doses, and our bioclearance experiments in CD-1 mice showed that this protein is cleared from the serum with a half-life of 60 min and selectively accumulates in the prostate before rapid excretion. Finally, we did not observe any significant toxicity even after total 6 doses of 2.4 mg, which is 30 times higher than the optimal therapeutic dose. We conclude that GLIPR1-ΔTM is an effective and safe drug with appropriate kinetics. Our results establish a foundation for the development of GLIPR1-ΔTM as a systemic neoadjuvant therapy in primary prostate cancer and systemic therapy in metastatic disease. Furthermore, in our view, a protein therapeutic based on an endogenous tumor suppressor protein that selectively targets cancer cells represents an unprecedented conceptual advance in cancer therapeutics.

Novelty and impact.

The present study demonstrated that GLIPR1-ΔTM protein is selectively taken up by cancerous, but not by normal, prostate cells and acts by the same mechanisms that endogenous GLIPR1 does: reactive oxygen species–mediated apoptosis, downregulation of c-Myc, and interaction with Hsc70. Systemic GLIPR1-ΔTM inhibited tumor overall growth and had rapid clearance and limited toxicity in mouse models. We establish the efficacy of an “auto-targeting,” cancer cell–selective therapeutic protein, demonstrating an unprecedented conceptual advance.

Abbreviations used

GLIPR1-ΔTM

Glioma pathogenesis-related protein 1 – transmembrane domain deleted

ROS

Reactive oxygen species

Hsc70

Heat shock cognate protein 70

ALT

Alanine transaminase