DT ^270‐326, a Truncated Diphtheria Toxin, Increases Blood–Tumor Barrier Permeability by Upregulating the Expression of Caveolin‐1

Summary

Aim

The nontoxic mutant of diphtheria toxin (DT) has been demonstrated to act as a receptor‐specific carrier protein to delivery drug into brain. Recent research showed that the truncated “receptorless” DT was still capable of being internalized into cells. This study investigated the effects and potential mechanisms of DT ^270‐326, a truncated “receptorless” DT, on the permeability of the blood–tumor barrier (BTB).

Methods

BTB and GECs were subjected to DT ^270‐326 treatment. HRP flux assays, immunofluorescent, co‐immunoprecipitation, Western blot, CCK‐8, and Flow cytometry analysis were used to evaluate the effects of DT ^270‐326 administration.

Results

Our results revealed that 5 μM of DT ^270‐326 significantly increased the permeability of BTB in vitro, which reached its peak at 6 h. The permeability was reduced by pretreatment with filipinIII. DT ^270‐326 co‐localized and interacted with caveolin‐1 via its caveolin‐binding motif. The mRNA and protein expression levels of caveolin‐1 were identical with the changes of BTB permeability. The upregulated expression of caveolin‐1 was associated with Src kinase‐dependent tyrosine phosphorylation of caveolin‐1, which subsequently induced phosphorylation and inactivation of the transcription factor Egr‐1. The combination of DT ^270‐326 with doxorubicin significantly enhanced the loss of cell viability and apoptosis of U87 glioma cells in contrast to doxorubicin alone. Conclusions: DT ^270‐326 might provide a novel strategy to increase the delivery of macromolecular therapeutic agents across the BTB.

Introduction

Glioma is a highly malignant central nervous system neoplasm with a poor prognosis. Adjuvant chemotherapy is of great importance in the treatment of glioma. However, macromolecular antiglioma drugs cannot penetrate into the tumor tissue due to the existence of the blood–tumor barrier (BTB) formed by capillary endothelial cells and glioma cells, which seriously limits the efficacy of chemotherapy 1. Therefore, efficient opening of the BTB and effective delivery of antitumor drugs into the glioma tissue are the keys to improving the survival of glioma patients.

Diphtheria toxin (DT) is a bacterial exotoxin synthesized by Corynebacterium diphtheriae 2. It has been well studied as a carrier protein for polysaccharide vaccines against various pathogens 3. Diphtheria toxin has three structural domains: N‐terminal catalytic domain that is involved in cytotoxicity (C‐domain), domain that helps to transport the catalytic domain structure through cell membrane (T‐domain) and C terminal receptor binding region (R‐domain) 4. T, C, and R structural domains of diphtheria toxin are independent of each other and can function separately even after in vitro deletion and transformation of other domains 5. Recombinant R‐ and T‐domains of DT promote the drug transport across the cell membrane 6. The R‐domain is activated by binding to specific receptors. However, whether the T‐domain is implicated in the drug transport remains unclear 7, 8. Our previous work reported that CRM197 increased the BTB permeability via transcellular pathway 9. Sequence analysis reveals that the T‐domain of DT possesses a caveolin‐binding motif (273FAGANYAAW281, the underlined abbreviation represents aromatic amino acids) 10. These results prompted us to investigate whether the caveolin‐binding motif of T‐domain is the functional domain in caveolae‐dependent BTB permeability increase induced by CRM197.

Caveolae is a specialized cytoplasmic membrane protein that transports macromolecular substances into and out of cells 11. In endothelial cells, caveolin‐1 is the main structural protein necessary for the caveolae formation 12, 13. Phosphorylation of caveolin‐1 at Y14 is prerequisite for caveolae formation and is possibly an intermediate step in the process of caveolae signal cascade 14, 15. It has been proved that caveolin‐1 phosphorylation is closely related to the BBB opening in subarachnoid hemorrhage 16. It has further been established that caveolin‐1‐mediated endocytosis and Src kinase signaling pathway are involved in the regulation of BBB permeability 17, 18.

Zinc finger transcription factor Egr‐1 (early growth responsive‐1, Egr‐1) is one of the most important members of the immediate early response gene family 19, 20. Egr‐1 is a nuclear protein that has transcriptional regulatory function 21. Gene expression of Egr‐1 might be regulated by various factors such as cell oxidative stress and mechanical damage 22, 23. Egr‐1 is the substrate of multiple kinases (such as PKC) as well 24.The transcriptional regulation of Egr‐1 is negatively regulated by its phosphorylation 25. The activated nonphosphorylated Egr‐1 regulates the expression of target genes by binding to the GC promoter binding element. Egr‐1 is deactivated by phosphorylation and cannot bind to the promoter region of target genes, leading to inhibited gene transcription 26. Joshi et al. 27 confirmed that tyrosine phosphorylation of caveolin‐1 triggered the phosphorylation and inactivation of Egr‐1, resulting in increased caveolin‐1 expression in MDA‐435 breast cancer cells.

This study reveals that the recombinant truncated T‐domain of DT (DT^270‐326 containing caveolin‐binding motif) increases the BTB permeability mainly through caveolae‐dependent pathway. Our work further suggests that DT^270‐326 activates Src kinase and phosphorylates caveolin‐1. Phosphorylated caveolin‐1 then upregulates total caveolin‐1 expression via the phosphorylation and inactivation of transcriptional repressor Egr‐1. DT^270‐326 combined with doxorubicin enhances doxorubicin‐induced inhibition of cell viability and apoptosis of glioma cells in the in vitro BTB model.

Materials and Methods

Cell Lines and Cell Cultures

The immortalized human cerebral microvascular endothelial cells line hCMEC/D3 (ECs) was obtained from Dr. Couraud (Institute Cochin, Paris, France). The cells were cultured on culture flasks coated with 150 μg/mL of cultrex rat collagen I (R&D Systems, Minneapolis, MN, USA). The culture medium contained endothelial basal medium (EBM‐2; Lonza, Walkersville, MD, USA), supplemented with 5% fetal bovine serum (FBS; PAA Laboratories GmbH, Pasching, Austria), 1% Penicillin–Streptomycin (Life Technologies Corporation, Paisley, UK), 1.4 μM hydrocortisone (Sigma‐Aldrich, St Louis, MO, USA), 1% chemically defined lipid concentrate (Life Technologies Corporation), 5 μg/mL ascorbic acid (Sigma‐Aldrich), 10 mM HEPES (PAA Laboratories GmbH), and 1 ng/mL human basic fibroblast growth factor (bFGF; Sigma‐Aldrich). ECs used for the study were between passages 30 and 40.

Human glioblastoma cell lines U87 were purchased from the Shanghai Institutes for Biological Sciences Cell Resource Center and maintained in Dulbecco's modified Eagle's medium (DMEM) of high glucose with 10% fetal bovine serum, 100 U/mL penicillin, and 100 μg/mL streptomycin (Life Technologies Corporation). Cells were maintained in a humidified incubator at 37°C and 5% CO₂, and medium was refreshed every 2 days. Cells used for the study were between passages 30 and 40. The cell proliferation was determined by counting the viable cells with trypan blue exclusion. Culture plastic wares were purchased from Corning (Corning, NY, USA). The cells were used within 50 passages.

Generation of Stably Transfected GECs

ECs stably transfected with the mammalian expression vector pCMV‐myc (Clontech, Mountain View, CA, USA) containing full‐length caveolin‐1 (Cav1wt), dominant‐negative Y14F caveolin‐1 (Cav1Y14F), and phosphomimetic Y14D caveolin‐1 (Cav1Y14D) were generated as follows: the cells were performed using Lipofectamine LTX and Plus Reagents (Life Technologies Corporation), according to the manufacturer's instruction. Approximately 2 weeks after transfection, G418‐resistant cells were selected and expanded in EBM‐2 medium supplemented with 5% fetal bovine serum and 400 μg/mL of G418. The empty pCMV‐myc vector was also transfected as a negative control (NC). Then, the stably transfected ECs' cell lines were cocultured with U87 glioma cells to acquire the stably transfected GECs.

Establishment of the BTB Model In Vitro

In vitro BTB model was generated in a well‐established transwell system using ECs and U87 glioma cells as described by Hurst and Fritz 28. For coculture chambers, U87 glioma cells were seeded at 2 × 10⁴ per well in 6‐well or 24‐well plates with suitable culture medium. When 80% of U87 glioma cells were confluent, ECs were seeded at 2 × 10⁵ per well in the upper chamber of the transwell filters (0.4 μm pore size; Corning, NY, USA) coated freshly with 150 μg/mL of cultrex rat collagen I. The cells became confluent after 4–5 days of coculture, and the BTB model in vitro was successfully established 29. Through the establishment of BTB model in vitro, the ECs were cocultured with glioma cells to obtain the glioma endothelial cells (GECs).

Synthesis of Truncated Diphtheria Toxin and its Mutants

The truncated DT (DT^270‐326) and three kinds of point mutants were synthesized by Sangon Biotech Company (Shanghai, China). Mutation sites are designed within the caveolin‐binding motif (273FAGANYAAW281) in T‐domain of DT: DT^F273A, DT^Y278A, and DT^W281A, respectively, represent Phe²⁷³→Ala²⁷³, Tyr²⁷⁸→Ala²⁷⁸, and Trp²⁸¹→Ala²⁸¹. The truncated DT^270‐326 and point mutants of DT^270‐326 all contain N‐terminal hexahistidine (6 × his)‐tags.

HRP Flux Measurement

To examine transport of HRP protein across the BTB models in vitro, horseradish peroxidase (HRP; 10 μg/mL; Sigma‐Aldrich) alone or together with DT^270‐326 with 0, 1.25, 2.5, 5, 10, and 20 μM was added to the upper chamber of the transwell system. After administration for 6 h, the media from the lower chamber were collected, and 5 μL of the media were incubated with 200 μL of 3, 3′, 5, 5′‐Tetramethylbenzidine liquid substrate in 96‐well plate at 37°C for 30 min. The absorption was read at 370 nm by a microplate reader (Molecular Devices, Sunnyvale, CA, USA). The HRP was serially diluted at concentration from 0.625 μg/mL to 20 μg/mL to obtain a standard curve. Further, the time‐dependent effect of DT^270‐236 on the permeability of BTB in vitro was detected as described above.

In order to determine the endocytotic processes involved in the transport of HRP, GECs in the upper chamber of the transwell system were pretreated with 5 μg/mL of filipin III for 30 min at 37°C, then cells were washed and incubated in the same medium containing 10 μg/mL of HRP and 5 μM of DT^270‐326. HRP content of the samples was assayed as described above. HRP flux was expressed as picomoles passed per square centimeter surface area per hour.

Transmission Electron Microscopy

The GECs in the upper chamber of the transwell were divided into three groups: control group; HRP group (10 μg/mL of HRP for 6 h); and DT^270‐326+HRP group (5 μM of DT^270‐326 combined with 10 μg/mL of HRP for 6 h). After incubation with DMEM containing HRP alone or in combination with DT^270‐326 for 6 h at 37°C, confluent GECs monolayers of BTB model in vitro were washed and fixed with 2.5% glutaraldehyde for 2 h at 4°C. HRP was used as a macromolecular model because its reaction product can be visualized in cells by electron microscopy 30. After that, according to the standard procedures, semithin and ultrathin sections were stained with uranyl acetate and lead citrate, and then the ultrastructure of pinocytotic vesicles was observed by transmission electron microscopy (JEM‐1200; JEOL Ltd., Tokyo, Japan). A double‐blind approach was used to measure the ultrastructural changes.

Dual Immunofluorescence Assays

To examine the codistribution of DT^270‐326 and caveolin‐1 in the GECs, dual immunofluorescent staining was carried out. The GECs in the upper chamber of the transwell were fixed with 4% paraformaldehyde for 20 min and permeabilized with 0.2% Triton X‐100 for 10 min at room temperature. The cells were incubated in 5% BSA blocking buffer for 2 h at room temperature. Then, the cells were incubated in the mixture of two primary antibodies which are raised in different species at the recommended concentration in 1% BSA in PBS in a humidified chamber overnight at 4°C. Thereafter, the cells were incubated in the mixture of two secondary antibodies with two different fluorochromes in 1% BSA for 2 h at room temperature in dark. The cells were then washed, and DNA was stained with 4, 6‐diamidino‐2‐phenylindole (DAPI; Beyotime Institute of Biotechnology, Jiangsu, China; 0.5 μg/mL) for 5 min. The coverslips were mounted with antifading medium. The stainings were observed using immunofluorescence microscopy (Olympus, Tokyo, Japan) and merged by Chemi Imager 5500 V2.03 software. The antibodies used were the followings: mouse anti‐6 × his monoclonal antibody (diluted 1:200; Abcam, MA, USA), rabbit anti‐caveolin‐1 polyclonal antibody (diluted 1:50; Santa Cruz Biotechnology, Santa Cruz, CA, USA), Cy3‐labeled goat anti‐mouse IgG, and FITC‐labeled anti‐rabbit IgG (diluted 1:100; Beyotime Institute of Biotechnology, Jiangsu, China).

Co‐immunoprecipitation

The GECs were treated with DT^270‐326 and three kinds of mutations (DT^F273A, DT^Y278A, and DT^W281A) at 5 μM for 6 h. And then, cells were washed three times in PBS, harvested by scraping, and centrifuged for 5 min at 206 g. The pelleted cells were homogenized at 4°C with cell lysis buffer (Beyotime, Jiangsu, China), and the homogenates were centrifuged for 30 min at 26494 g at 4°C. The supernatants including 40 μg of protein of each were combined with 1 μg of either mouse IgG (IgG group) or anti‐6 × his primary antibody (anti‐6 × his group) and incubated on a rocker overnight at 4°C. Then, 20 μL of resuspended volume of Protein A/G Plus‐Agarose was added into the supernatants, which were incubated at 4°C on a rocker platform for 6 h. Then immunoprecipitates were collected by centrifugation at 573 g for 5 min and washed with lysis buffer. The samples were resuspended in 20 μL of 2 × SDS loading buffer (Sigma) and analyzed by SDS‐PAGE using anti‐caveolin‐1 antibody. The immunoblots were visualized by enhanced chemiluminescence (ECL kit, Santa Cruz Biotechnology) and scanned using Chemi Imager 5500 V2.03 software.

RNA Isolation and Reverse‐Transcription PCR

Total cellular RNA was extracted using Trizol reagent (Life Technologies Corporation) according to the manufacturer's protocol. The RNA concentration and quality were determined for each sample by the 260/280 nm ratio using a Nanodrop Spectrophotometer (ND‐100). Reverse transcription was performed using the High Capacity cDNA Reverse Transcription Kits (Takara Biotechnology, Dalian, China) according to the manufacturer's protocol. The PCR primers along with their Tm and amplicon sizes for CAV‐1 and GAPDH were as follows: CAV‐1 5′‐GACTTTGAAGATGTGATTGC‐3′ (forward) and 5′ –AGATGGAATAGACACGGCTG‐3′ (reverse) (Tm = 56°C, 253 bps), GAPDH 5′‐ CGCTGAGTACGTCGTGGAGT‐3′ (forward) and 5′‐CGTCAAAGGTGGAGGAGTGG‐3′ (reverse) (Tm = 56°C, 617 bps). The PCR products were then resolved on agarose gels and photographed using the Chemi Imager 5500 gel image analysis instrument. Integrated density values (IDV) of PCR product bands were measured by Chemi Imager 5500 software (Alpha Innotech, CA, USA).

Western blot Analysis

The cells were lysed with ice‐cold RIPA buffer (50 mM Tris‐HCl, pH 8.0, 150 mM NaCl, 0.1% SDS, 1% NP‐40, 0.5% sodium deoxycholate, and 1 mM EDTA) supplemented with phosphatase inhibitors or protease inhibitors (10 mg/mL aprotinin, 10 mg/mL phenyl‐methylsulfonyl chloride, and 50 mM sodium orthovanadate) and incubated on ice for 20 min. The lysates were centrifuged at 26494 g for 30 min at 4°C. The supernatant extracts were determined with the BCA protein assay kit (Beyotime Institute of Biotechnology). Equal amounts of total protein (50 μg) were separated by SDS‐PAGE and then gels were transferred to a PVDF membrane, blocked with 5% nonfat dry milk in TBST for 2 h, and subsequently incubated with primary rabbit anti‐Src polyclonal antibody (diluted 1:1000; Abcam Inc.), rabbit anti‐Src (phospho Y418) antibody (diluted 1:1000; Abcam Inc.), rabbit anti‐p‐caveolin‐1 polyclonal antibody (diluted 1:1000; CST Inc., Danvers, MA, USA), rabbit anti‐Egr‐1 polyclonal antibody (diluted 1:200; Santa Cruz Biotechnology), mouse anti‐caveolin‐1 monoclonal antibody (diluted 1:1000; Abcam Inc.), and mouse anti‐myc‐caveolin‐1 monoclonal antibody (diluted 1:1000; Abcam Inc.) separately over night at 4°C. The expression levels were normalized to GAPDH housekeeping gene (diluted 1:1000; Santa Cruz Biotechnology). The membrane was washed and incubated with respective HRP‐conjugated secondary antibody. The immunoblots were visualized by enhanced chemiluminescence (ECL kit; Santa Cruz Biotechnology) and scanned using Chemi Imager 5500 V2.03 software, and integrated light density values (IDVs) were calculated by Fluor Chen 2.0 software and normalized with that of GAPDH.

Cell Viability Assay (CCK‐8 Assay)

The cell viability was detected by Cell Counting Kit‐8 (CCK‐8) (Dojindo Laboratories, Kumamoto, Japan) according to the manufacturer's instructions. U87 cells were seeded at a density of 40,000 cells per well in 24‐well plates and cocultured with the ECs for 4 days using transwell inserts to establish the BTB model in vitro. Then, 200 μL of culture media was added in the upper insert separately containing DMSO (1 μM, as control group), doxorubicin (1 μM) or DT^270‐326 (5 μM) alone or in combination. After 12‐h incubation, the media were replaced and 20 μL CCK‐8 solution was added into each well. After incubating for 1 h in the incubator, the cell viability was determined by measuring the absorbance at 450 nm using microplate reader. Five replicate wells were prepared for each group, and repeated three independent experiments were performed.

Analysis of Apoptosis by Flow Cytometry

After the GECs growing on the transwell were confluent to monolayer, they were treated with DMSO (1 μM, as control group), doxorubicin (1 μM) or DT^270‐326 (5 μM) alone or in combination for 12 h. After treatment, 1 × 10⁵ U87 cells in the lower chamber were resuspended in Annexin V binding buffer and stained with Annexin V‐FITC and Propidium Iodide (PI) (1 μg/mL) using Annexin V‐FITC Apoptosis Detection Kit (Beijing Biosea Biotechnology Co. Ltd., China) following manufacturer's instruction. After incubation at room temperature, the apoptotic cells were quantified by flow cytometry (BD Bioscience, Franklin Lakes, NJ, USA). Early apoptosis is defined by Annexin V⁺/PI⁻ staining (lower right quadrant, LR), and late apoptosis is defined by Annexin V⁺/PI⁺ staining (upper right quadrant, UR).

Statistical Analysis

The SPSS 18.0 statistical analysis software was used for statistical analysis. Data were expressed as the mean ± standard deviation (SD). One‐way analysis of variance (ANOVA) followed by Bonferroni's post hoc test was used to determine the significant differences among multiple groups, and considered significant when P < 0.05.

Results

DT^270‐326 Increased the Transcytosis of HRP Across the BTB In Vitro

To detect the effects of DT^270‐326 on the transcytosis of macromolecular substance across the BTB in vitro, the transcytosis assay was performed. The leakage rates of HRP were increased in a dose‐dependent manner from various concentration levels of DT^270‐326 (0, 1.25, 2.5, 5 μM). Compared with the 0 μM of DT^270‐326 group, the leakages rates of HRP in the 2.5, 5, 10, and 20 μM group were significantly increased (P < 0.01). The leakage rates in the 5, 10, and 20 μM group were significantly higher than the 2.5 μM group (P < 0.05). There was no significant difference among the 5, 10, and 20 μM groups (P > 0.05, Figure 1A). About 5 μM of DT^270‐326 is the first dose to reach the platform phase of HRP leakage rates; therefore, the dose of 5 μM was chosen for the subsequent experiments. After the treatment with 5 μM of DT^270‐326, HRP leakage rates were increased in a time‐dependent manner. HRP leakage rate reached the peak at 6 h and then decreased gradually. After the treatment of DT^270‐326 for 24 h, HRP leakage rate showed no significant difference compared with the control group at the same time point. After pre‐incubating with filipin III, a kind of caveolae inhibitors, for 30 min, the GECs were treated with HRP in combination with DT^270‐326 for 24 h. At various time points (3, 6, 12 and 18 h), the HRP leakage rates of filipin III+DT^270‐326 group were significantly lower than those of DT^270‐326 group at the same time points(P < 0.05). The HRP leakage rates of filipin III+DT^270‐326 group at 6 and 12 h were significantly higher than those of control group at the same time points(P < 0.05,Figure 1B). Pretreatment with filipin III inhibited the transendothelial transport of HRP significantly, which suggested that a caveolae‐mediated pathway was involved in this transport.

Figure 1.

DT ^270‐326 increased the endocytosis of HRP across the BTB in vitro. (A) HRP flux was increased significantly in a dose‐dependent manner. Data represent means ± SD (n = 4, each). *P < 0.05, **P < 0.01 versus DT ^270‐326 0 μM group; ^# P < 0.05 versus DT ^270‐326 2.5 μM group. (B) In combination with DT ^270‐326 (5 μM), HRP flux increased in a time‐dependent manner and peaked at 6 h. At the same time, there was a significant increase in HRP flux compared with control group (*P < 0.05 and **P < 0.01). However, filipin III pretreatment significantly reduced the rates of the HRP flux compared with DT ^270‐326 group (^# P < 0.05 and ^## P < 0.01). The HRP leakage rates of filipin III+DT ^270‐326 group at 6 and 12 h were significantly higher than those of control group at the same time points(*P < 0.05). Data represent means ± SD (n = 4, each). (C) Treatment of DT ^270‐326 resulted in a significant ultrastructural change in GECs. Untreated GECs showed few pinocytotic vesicles (a and b). Endocytic vesicles in GECs had no obvious increase in the number when HRP was given alone (c and d). However, the number of HRP‐containing endocytic vesicles increased obviously after the treatment of HRP in combination with DT ^270‐326 (e and f). b, d, and f are the magnifications of a, c, and e respectively. Arrows show the endocytic vesicles. Scale bars: 200 nm (a,c,e); 30 nm (b,d, and f).

Ultrastructural Changes of GECs In Vitro BTB Model Induced by DT^270‐326

The results obtained using transmission electron microscopy showed that the GECs in the control group had few endocytic vesicles (Figure 1Ca,b). When treated with HRP for 6 h, the number of vesicles in GECs had no obvious increase compared with the control group. No HRP was observed in the vesicles (Figure 1Cc,d). However, after treatment with HRP in combination with DT^270‐326 for 6 h, the high electron density particles were found in the vesicles, which is a result of HRP uptake 31. Meanwhile, the number of HRP‐containing endocytic vesicles was increased in GECs (Figure 1Ce,f). These results support the potential role of DT^270‐326 to transport macromolecular substance toward the brain.

Expression and Localization of DT^270‐326 and Caveolin‐1 in GECs

Immunofluorescent staining results showed that the expression of caveolin‐1 in the cytoplasm of GECs was increased in a time‐dependent manner after the treatment of DT^270‐326. The expression levels of caveolin‐1 reached the maximum at 6 h and then decreased gradually. The expression and distribution of DT^270‐326 showed a similar tendency as caveolin‐1. The staining results mentioned above demonstrated that there was a co‐localization between caveolin‐1 and DT^270‐326 proteins in the cytoplasm of GECs (Figure 2A Merge).

Figure 2.

Subcellular localization of DT ^270‐326 and caveolin‐1 and the interaction between DT ^270‐326 (or DT ^270‐326 mutants DT^{F 273A}, DT^{Y 278A}, and DT^{W 281A}) and caveolin‐1 in GECs. (A) The GECs were treated with DT ^270‐326 for different times and then were stained by dual immunofluorescence and visualized under Cy3 excitation (DT ^270‐326), FITC excitation (caveolin‐1), and simultaneous excitation of both fluorochromes (merge). Nuclei were labeled with DAPI. Scale bars represent 20 μm. The results are representative of at least three independent experiments. (B) Schematic representation of DT ^270‐326 and three kinds of mutant DT ^270‐326 proteins. (C) Compared with DT ^270‐326 group, Western blot results showed a lower expression of caveolin‐1 in the input, as well as in the 6 × his‐DT ^270‐326 mutants groups. (D) Quantification of caveolin‐1 expression in total proteins and co‐IPed proteins, **P < 0.01 compared with DT ^270‐326 group (ratio of input to control); ^## P < 0.01 compared with the DT ^270‐326 group (ratio of co‐IPed proteins by 6 × his antibody to input). All experiments were run in parallel. Data are given as mean ± SD (n = 4, each).

DT^270‐326 Binds with Caveolin‐1 via Caveolin‐Binding Domain

Figure 2B represents a schematic diagram of the single aromatic amino acid mutation sites (DT^F273A, DT^Y278A, DT^W281A) in the caveolin‐binding motif of DT^270‐326. As shown in Figure 2C,D, after treatment with DT^270‐326 and three mutants for 6 h, co‐immunoprecipitated (co‐IPed) caveolin‐1 and DT^270‐326 were apparent in either total proteins or 6 × His‐DT^270‐326 immunoprecipitated complex. Compared with 6 × His‐DT^270‐326 group, the co‐IPed caveolin‐1 in three mutant groups showed a significantly decreased binding levels, among which 6 × his‐DT^F273A > 6 × his‐DT^Y278A > 6 × his‐DT^W281A group (**P < 0.01, ^## P < 0.01).

DT^270‐326 Upregulated the mRNA and Protein Expression Levels of Caveolin‐1 in GECs

After the treatment of DT^270‐326, the mRNA and protein expression levels of caveolin‐1 in GECs showed the same time‐dependent changing tendency. Both the mRNA and protein expressions of caveolin‐1 reached the maximum at 6 h and 12 h, decreased gradually, and then return to the normal level before treatment at 24 h (Figure 3A,B).

Figure 3.

DT ^270‐326 (5 μM) upregulated the mRNA and protein expression of caveolin‐1 in GECs. (A) Representative RT‐PCR on an agarose gel showed bands corresponding to caveolin‐1 mRNA. Results of IDVs analysis of caveolin‐1 are shown. Data are given as mean ± SD (n = 4, each). *P < 0.05 and **P < 0.01 versus DT ^270‐326 0 h group. (B) The protein expression levels of caveolin‐1 were assessed by Western blot. Values represent the means ± SD (n = 3, each group). **P < 0.01 versus control group.

DT^270‐326 Increased the Phosphorylation of Src, Caveolin‐1, and Egr‐1 in GECs

As shown in Figure 4A, DT^270‐326 induced a significant increase in phosphorylated Src and caveolin‐1 at 1 h, decreased gradually, and then recovered to the normal level at 6 h. Meanwhile, the phosphorylation statue of Egr‐1 induced by DT^270‐326 increased at 1 h, peaked at 6 h, and decreased gradually to the normal level at 18 h. The total protein expressions of caveolin‐1 and Egr‐1 induced by DT^270‐326 also showed a similar time‐dependent changes, which peaked at 6 h and returned to the normal levels at 24 h.

Figure 4.

DT ^270‐326 upregulated protein levels of p‐Src, p‐caveolin‐1, Egr‐1, and caveolin‐1 in GECs, and the increased levels could be inhibited by PP2 pretreatment. (A) Representative Western blot bands showing the expression of p‐Src, p‐caveolin‐1, p‐Egr‐1 in GECs. (B) Representative Western blot bands showing the phosphorylated levels of Src, caveolin‐1, Egr‐1, and the protein expression level of caveolin‐1 in GEC pretreated with or without PP2. (C) Results of IDVs analysis of p‐Src, p‐caveolin‐1, p‐Egr‐1 (normalized to total Src, caveolin‐1, and Egr‐1, respectively), and Src, caveolin‐1, and Egr‐1 (normalized to GAPDH) were shown. The Egr‐1 is the sum of phosphorylated and nonphosphorylated Egr‐1. Data are given as mean±SD of n = 4 individual experiments. *P < 0.05 and **P < 0.01 versus DT ^270‐326 0 h group. (D) Quantitative analysis of the bands of Western blot. The p‐Src, p‐caveolin‐1, and p‐Egr‐1 were normalized to total Src, caveolin‐1, and Egr‐1, respectively. Src, caveolin‐1, and Egr‐1 were normalized to the corresponding GAPDH. The Egr‐1 is the sum of phosphorylated and nonphosphorylated Egr‐1. Data are given as mean±SD of n = 4 individual experiments. ^▲ P < 0.05 and ^▲▲ P < 0.01 versus control group. *P < 0.05 and **P < 0.01 representative PP2 + DT ^270‐326 1 h group versus DT ^270‐326 1 h group. ^# P < 0.05 and ^## P < 0.01 representative PP2 + DT ^270‐326 6 h group versus DT ^270‐326 6 h group.

DT^270‐326 Increased the Phosphorylation of Caveolin‐1 and Egr‐1 via a Src Kinase Pathway

As shown in Figure 4B, after treatment of DT^270‐326 for 1 h, the phosphorylation level of Src was increased significantly. Then pretreatment with Src phosphatase inhibitor 1‐tert‐Butyl‐3‐(4‐chlorophenyl) ‐1H‐pyrazolo [3, 4‐d] pyrimidin‐4‐amine (PP2) for 30 min at 37°C, the phosphorylated Src was decreased significantly compared with DT^270‐326 1 h group (P < 0.01). Compared with the control group, the phosphorylation level of Src in DT^270‐326 6 h group was not significantly changed (P > 0.05). There was no statistical difference of the total Src protein levels among all groups (P > 0.05). Compared with the control group, the total and phosphorylated levels of caveolin‐1 were significantly increased in DT^270‐326 1 h group; and PP2 pretreatment significantly inhibited the total and phosphorylated caveolin‐1 induced by DT^270‐326. In DT^270‐326 6 h group, the total caveolin‐1 protein level increased significantly (P < 0.01), whereas p‐caveolin‐1 showed no significant change (P > 0.05). After pretreatment of PP2, the total protein expression level as well as the p‐caveolin‐1 was significantly decreased compared with DT^270‐326 6 h group (P < 0.01). Compared with the control group, the total and phosphorylated levels of Egr‐1 were significantly increased in DT^270‐326 1 h and 6 h groups as well (P < 0.01). The pretreatment of PP2 significantly inhibited the changes of total and phosphorylated Egr‐1 mentioned above (P < 0.05), compared with DT^270‐326 1 h and 6 h groups. But these total and phosphorylated Egr‐1 expressing levels were higher than their respective control groups.

Caveolin‐1 Y14 Phosphorylation Upregulated Caveolin‐1 Expression by Inactivating Egr1

In order to explore the role of DT^270‐326 in upregulating the caveolin‐1 expression in GECs, Egr‐1 phosphorylation level and caveolin‐1 protein level were screened by Western blot in stable Cav1wt‐, Cav1Y14F (caveolin‐1 phosphorylation significantly negative mutant)‐ and Cav1Y14D (caveolin‐1 phosphorylation mimics)‐transfected GECs after the treatment of DT^270‐326 for 6 h. As shown in Figure 5, compared with the Cav1wt group, Egr‐1 phosphorylation level and caveolin‐1 protein level were significantly decreased in Cav1Y14F group (P < 0.01), whereas significantly increased in Cav1Y14D group (P < 0.01, P < 0.05). By detecting the expression of myc‐caveolin‐1 in three kinds of stable transfected cell lines, the expression levels of exogenous caveolin‐1 were verified to be not statistically different (P > 0.05).

Figure 5.

Cav1 Y14 phosphomimetic induces expression of p‐Egr‐1 and caveolin‐1. (A) Representative Western blot bands showing the expression of phosphorylated, nonphosphorylated Egr‐1, caveolin‐1, and myc‐caveolin‐1 in the Cav1Y14D, Cav1Y14F, and NC group. (B) Densitometric quantification of p‐Egr‐1 relative to Egr‐1 (the sum of phosphorylated and nonphosphorylated Egr‐1), caveolin‐1, and myc‐caveolin‐1 relative to GAPDH. Data represent means ± SD (n = 4, each). *P < 0.05, **P < 0.01 versus NC group values; ^# P < 0.05, ^## P < 0.01 versus Cav1wt group values.

DT^270‐326 Increased the Antitumor Potency of Doxorubicin by Enhancing Transport across the BTB In Vitro

CCK‐8 assay was performed to measure the U87 cell viability in the lower chamber. Compared with control group, the U87 cell viability was not markedly changed in DT^270‐326 group (P > 0.05), whereas was significantly decreased in Dox group (P < 0.05). Compared with Dox group, the U87 cell viability was significantly decreased in DT^270‐326+Dox group (Figure 6A).

Figure 6.

The changes of viability and apoptosis of U87 cells induced by DT ^270‐326. (A) U87 cell viability was significantly reduced by doxorubicin combined with DT ^270‐326 group. Data are given as mean ± SD of n = 6 individual experiments. *P < 0.05 and **P < 0.01 versus control group and ^## P < 0.01 versus Dox group. (B) Apoptosis of U87 cells changed by doxorubicin and DT ^270‐326 alone or in combination for 6 h. Dot plots indicating the data of a representative experiment. (C) Percentages of apoptotic cells in different treatment groups are shown. Data are given as mean ± SD of n = 4 individual experiments. *P < 0.05, **P < 0.01 versus control group and ^## P < 0.01 versus Dox group.

Doxorubicin‐induced apoptosis of U87 cells was determined by double staining with Annexin V‐FITC and PI. As shown in Figure 6C, compared with control group, the apoptosis rate of U87 cells showed no significant difference in DT^270‐326 group, whereas the apoptosis rate of U87 cells treated with doxorubicin alone or in combination with DT^270‐326 was significantly increased (P < 0.05). Compared with Dox group, the apoptosis rate was significantly increased in U87 cells treated with doxorubicin in combination with DT^270‐326 (P < 0.05).

Discussion

Studies have demonstrated that DT³⁸⁸ (a truncated DT composing of C‐ and T‐domains) coupling with IL‐3 showed enhanced binding to the IL‐3 receptor and greater cytotoxicity to human leukemia cells 32. And DT³⁸⁵, known as the “receptorless” truncated DT, could be transported into human U87 and U251 glioma cells to generate cytotoxic effect. It is possible that DT³⁸⁵ could enter tumor cells via other endocytic pathways rather than the receptor‐mediated endocytosis. However, the detailed mechanism is unclear 33. By analyzing the structure of DT, the caveolin‐binding motif (²⁷³ FAGANYAAW ²⁸¹) was identified in the T‐domain. Wang P et al. also confirmed that the nontoxic mutant of DT could promote the transportation of macromolecular substances across BTB via caveolae‐dependent transcytosis 9, 34. We speculated that the T‐domain of DT might be associated with the nonreceptor‐mediated endocytosis, for example, the caveolae‐dependent transcellular pathway. We proved that the permeability of BTB was significantly increased in a time‐ and dose‐dependent manner in vitro by administration of DT^270‐326, a truncated T‐domain of DT containing caveolin‐binding motif. After treatment of DT^270‐326 at a dose of 5 μM for 6 h, the permeability of BTB reached a peak. Meanwhile, DT^270‐326 significantly increased the quantity of pinocytotic vesicles containing macromolecular substance‐HRP (40KD) in GECs. After pretreated with caveolae inhibitor filipin III, the HRP leakage rates were significantly decreased in the in vitro BTB model. It is reasonable that DT^270‐326 increases the permeability of BTB mainly via caveolae‐mediated transcytosis 35.

Caveolin‐l is the major structural protein that is essential for the formation of caveolae, and plays a key role in the endocytotic process of tumor vascular endothelial cells 36, 37. It can also interact with proteins containing the caveolin‐binding motif (ΦXΦXXXXΦ or ΦXXXXΦXXΦ or ΦXΦXXXXΦXXΦ, Φ represents aromatic amino acids, and X represents any amino acid) 38, 39 through its 82‐101 amino acid fragment known as the caveolin scaffolding domain (CSD) 37, 40. The present study also found that DT^270‐326 colocalized with caveolin‐1 in GECs. Their interaction was further demonstrated by co‐IP. To verify whether DT^270‐326 binds to caveolin‐1 through its caveolin‐binding motif, individual point mutations of aromatics within this motif (DT^F273A, DT^Y278A, DT^W281A) were constructed, respectively. Compared with the 6 × his‐DT^270‐326 group, the expression of caveolin‐1 co‐IPed with three mutants was decreased significantly. The binding levels were shown as 6 × his‐DT^W281A > 6 × his‐DT^Y278A > 6 × his‐DT^F273A, indicating that caveolin‐binding motif was crucial for the interaction between caveolin‐1 and DT^270‐326, and phenylalanine (F) played a bigger role than any other aromatic amino acids during this process. This study also verified that the co‐IPed caveolin‐1 was upregulated by the treatment of DT^270‐326. Compared with the 6 × his‐DT^270‐326 group, the co‐IPed caveolin‐1 was significantly decreased in three mutant groups. The reduced levels of co‐IPed caveolin‐1 were in accordance with the weakened effect of the interaction between DT^270‐326 and caveolin‐1. These results suggest that DT^270‐326 upregulates the expression of caveolin‐1 by interacting with caveolin‐1 through the caveolin‐binding motif.

Caveolin‐1 is the major tyrosine‐phosphorylated substrate of Src kinase 17. Phosphorylation of caveolin‐1 is the crucial step in the process of signal cascade triggered in the caveolae‐mediated endocytosis, and it further affects the permeability of BTB 41, 42. In subarachnoid hemorrhage, the Src phosphatase inhibitor PP2 decreased BBB permeability by inhibiting the phosphorylation of caveolin‐1 16. To further explore the potential mechanism in the regulation of caveolin‐1 expression by DT^270‐326, the phosphorylation of caveolin‐1 and the potential upstream and downstream signaling molecules were detected in GECs. Results demonstrated that the phosphorylated and total caveolin‐1 were significantly increased, accompanied by the activation of Src kinase in GECs after treated with DT^270‐326 for 1 h. Pretreatment of PP2 significantly reversed the DT^270‐326‐induced upregulation of phosphorylated and total caveolin‐1 compared with control group at the same time points. These results suggest that DT^270‐326 promotes the phosphorylation of caveolin‐1 by activating Src kinase.

The transcriptional activity of Egr‐1 is closely related to its phosphorylation level. The phosphorylated Egr‐1 attenuates its binding to the target genes, whereas the nonphosphorylated Egr‐1 enhances the transcription of target genes by promoting the binding to them 26. Joshi et al. 27 confirmed that tyrosine‐phosphorylated caveolin‐1 inhibited the negative regulatory effect of Egr‐1 on caveolin‐1 gene expression by increasing the phosphorylation level of Egr‐1, resulting in upregulated expression of total caveolin‐1 in MDA‐465 breast cancer cells. In the present study, DT^270‐326 upregulated phosphorylated Egr‐1 (top band of Egr‐1 land in Figure 4A), and the expression pattern was consistent with the upregulated caveolin‐1. The phosphorylated Egr‐1 and total caveolin‐1 expression levels reached peak at 6 h, while the phosphorylated Src kinase and caveolin‐1 arrived at their highest levels at 1 h. The phosphorylated Egr‐1 and total caveolin‐1 expression levels in DT^270‐326 6 h groups were still much higher than the control group even after PP2 pretreatment, meaning that besides Src kinase, expression of other activators involved in the Egr‐1 phosphorylation might be induced by DT^270‐326 at 6 h. Chu et al. 43 showed that both ERK1/2 and P38 MAPKs are the upstream kinases of Egr‐1. We would like to investigate them in our future researches. To further investigate the relationship between the expression of phosphorylated caveolin‐1, phosphorylated Egr‐1, and total caveolin‐1, Cav1Y14F (caveolin‐1 phosphorylation dominant‐negative mutant)‐ and Cav1Y14D (caveolin‐1 phosphorylation mimics)‐stably transfected GEC cell lines were established. Results showed that DT^270‐326 induced a significant elevation of the phosphorylated (top band) Egr‐1 accompanied by upregulated total caveolin‐1 in Cav1Y14D‐transfected GECs. On the contrary, the phosphorylated Egr‐1 and total caveolin‐1 were significantly reduced in Cav1Y14F‐transfected GECs. The above results suggest that phosphorylated caveolin‐1 might inhibit the binding of Egr‐1 to caveolin‐1 gene by enhancing the phosphorylation level of Egr‐1, leading to attenuated negative regulation as well as upregulated expression of total caveolin‐1 27. After the treatment of DT^270‐326, caveolin‐1 mRNA and protein expression levels reached peaks at 6 h and returned to the basal levels until 24 h. The pattern of caveolin‐1 expression was consistent with the change of BTB permeability. Authors 44, 45 also suggested that the upregulation of caveolin‐1 could increase the number of pinocytotic vesicles in endothelial cells and could enhance the permeability of BTB via transcellular pathway, which was consistent with the results of this research.

Doxorubicin is clinically used as a therapeutic agent against several tumors 46, 47, 48. However, it is seldom used to treat brain glioma due to its low penetration across BTB 49. Currently, various methods have been developed to promote the delivery of doxorubicin across BTB, including solid lipid nanoparticle carriers; folic acid and transferrin double‐targeted adriamycin liposome; focused ultrasound 50, 51, 52. To further verify the effect of DT^270‐326 on the transport of doxorubicin through BTB, the antitumor effects of doxorubicin alone or combined with DT^270‐326 were analyzed. The results showed that compared with the doxorubicin group, the combination of DT^270‐326 and doxorubicin significantly increased the apoptotic rate of U87 glioma cells and reduced the viability of U87 cells in the in vitro BTB model, suggesting that DT^270‐326 might enhance the antitumor effects of doxorubicin by promoting the transport across BTB in vitro.

The present research demonstrates for the first time that DT^270‐326 increases the permeability of BTB via a caveolae‐dependent transcellular pathway. DT^270‐326 interacts with caveolin‐1 through its caveolin‐binding motif, which might activate Src kinase and phosphorylate caveolin‐1; the phosphorylated caveolin‐1 induces phosphorylation and inactivation of Egr‐1 to upregulate the expression of caveolin‐1. DT^270‐326 could enhance the antitumor effects of doxorubicin by promoting BTB penetration. Therefore, coadministration of DT^270‐326 with macromolecular anticancer drugs might become a new strategy for the treatment of brain glioma.

Disclosure

PW and YXX conceived and designed the experiments. YL, PW, and XLS performed the experiments. YL, PW, YHL, and YXX analyzed the data. YL, PW, YHL, and YXX contributed reagents/materials/analysis tools. YL, PW, and YXX wrote the manuscript. All authors read and approved the final manuscript. The authors declare no competing financial interests.