Article pubs.acs.org/bc
Glioma-Targeted Drug Delivery Enabled by a Multifunctional Peptide Mingfei Zhang,†,∥ Xishan Chen,†,∥ Man Ying,† Jie Gao,† Changyou Zhan,‡ and Weiyue Lu*,†,§ †
Department of Pharmaceutics, School of Pharmacy, Fudan University and Key Laboratory of Smart Drug Delivery, Ministry of Education, 826 Zhangheng Road, Shanghai 201203, China ‡ Department of Pharmacology, School of Basic Medical Sciences and §State Key Laboratory of Medical Neurobiology, The Collaborative Innovation Center for Brain Science, Fudan University,Shanghai 200032, China ABSTRACT: The rapid proliferation of glioma relies on vigorous angiogenesis for the supply of essential nutrients; thus, a radical method of antiglioma therapy should include blocking tumor neovasculature formation. A phage display selected heptapeptide, the glioma-initiating cell peptide GICP, was previously reported as a ligand of VAV3 protein (a Rho GTPase guanine nucleotide exchange factor), which is overexpressed on glioma cells and tumor neovasculature. Therefore, GICP holds potential for the multifunctional targeting of glioma (tumor cells and neovasculature). We developed GICP-modified micelle-based paclitaxel delivery systems for antiglioma therapy in vitro and in vivo. GICP and GICP-modified PEG−PLA micelles (GICP−PEG−PLA) could be significantly taken up by U87MG cells, a human cell line derived from malignant gliomas and human umbilical vein endothelial cells (HUVECs). Furthermore, GICP−PEG−PLA micelles demonstrated enhanced penetration in a tumor spheroid model in vitro in comparison to unmodified micelles. In vivo, DiR-loaded GICP−PEG−PLA micelles exhibited superior accumulation in the tumor region by targeting neovasculature and glioma cells in nude mice bearing subcutaneous glioma. When loaded with paclitaxel, GICP−PEG−PLA micelles could more effectively suppress tumor growth and neovasculature formation than unmodified micelles in vivo. Our results indicated that GICP could serve as a promising multifunctional ligand for glioma targeting.
■
INTRODUCTION Glioma is the most common and aggressive type of brain tumor, accounting for 45% of malignant primary brain and central nervous system (CNS) tumors.1 The conventional therapeutic modalities include surgical resection, radiotherapy, and chemotherapy. According to previous reports,2,3 the 5 year survival rate of glioma patients receiving standard postoperative radiotherapy was only 3%, whereas that of patients treated with temozolomide-combined radiotherapy was 11%, indicating that chemotherapy is effective to some extent for glioma treatment. However, chemotherapy of glioma still remains challenging in clinic. There are many factors affecting the clinic outcome of chemotherapy. Conventional chemotherapeutics distribute without selectivity after systemic administration, which may cause damage to normal tissues.4,5 Angiogenesis always plays pivotal roles in tumor growth, evasion, and metastasis;6,7 thus, it would be promising for cancer therapy by blocking angiogenesis. Actively targeted drug delivery employs targets (such as overexpressed antigens and receptors) in tumor or tumorrelated tissues to achieve higher efficacy and lower side effects in comparison to the conventional chemotherapy.8−10 In the previous reports,11−14 VAV3 overexpression has been verified to be tightly associated with tumor growth, apoptosis, invasion, proliferation, metastasis, and poor prognosis. Liu et al.15 reported that VAV3 was up-regulated in glioma, especially in © XXXX American Chemical Society
glioma-initiating cells (GICs), suggesting that VAV3 would be a potential target for glioma targeting. A heptapeptide gliomainitiating cell peptide, GICP (SSQPFWS), was identified by phage display as a ligand of VAV3 receptors with high affinity.15 In our preliminary study, GICP also exhibited high binding affinity to HUVECs, suggesting that this peptide may also be able to target neovasculature of glioma. In the present work, paclitaxel (PTX), which can effectively inhibit the growth of glioma cells,16 was adopted as the model drug. A biocompatible and biodegradable diblock polymer poly(ethylene glycol)-block-poly(lactic acid) (PEG−PLA) was employed to encapsulate PTX by forming micelles.16−18 GICP was modified on the surface of PEG−PLA micelles (GICP− PEG−PLA), and the unmodified mPEG−PLA micelles were prepared as the control. Cellular uptake efficiency, tumor spheroid penetration, and cytotoxicity of micelles were evaluated in vitro. Biodistribution of near-infrared dye (1,1′dioctadecyl-3,3,3′,3′-tetramethyl indotricarbocyanine iodide, DiR) loaded micelles was conducted in nude mice bearing Special Issue: Peptide Conjugates for Biological Applications Received: October 25, 2016 Revised: November 27, 2016
A
DOI: 10.1021/acs.bioconjchem.6b00617 Bioconjugate Chem. XXXX, XXX, XXX−XXX
Article
Bioconjugate Chemistry
Tumor Spheroid Penetration. mPEG−PLA/C6 and GICP−PEG−PLA/C6 micelles containing 160 ng/mL C6 were incubated with U87MG tumor spheroid for 1 h. The fluorescence was captured under the same condition by Z-stack mode of laser scanning confocal microscope with an interval of 5 μm (Figure 3); the penetration depth of mPEG−PLA/C6 was 43.59 μm ± 4.15 μm, and the penetration depth of GICP− PEG−PLA/C6 was 56.87 ± 5.34 μm. It was indicated that GICP could enhance the penetration of micelles into tumor spheroid. Biodistribution. To evaluate the biodistribution of GICP− PEG−PLA and mPEG−PLA micelles in nude mice bearing subcutaneous U87MG tumor, six nude mice were randomly assigned into two groups. A total of 200 μL of DiR-loaded micelles with a concentration of 80 μg/mL was injected into the tail vein, and the mice were imaged by an IVIS system at different time points (Figure 4A). It appeared that both populations of micelles accumulated in tumor region at all tested time points. However, GICP-modified micelles exhibited much more accumulation in tumor region than did mPEG− PLA micelles (Figure 4B). A total of 24 h after injection, mice were sacrificed, and all organs were harvested to measure the fluorescence intensity. The results were consistent with that of ex vivo imaging study, indicating that GICP could significantly facilitate tumor targeting of micelles (Figure 4C). Co-localization with Neovasculature. To investigate the localization of micelles in vivo, C6 was loaded in both GICP− PEG-PLA micelles and mPEG−PLA micelles with the same concentration of 30 μg/mL. The immunofluorescence images showed that GICP-modified micelle could completely colocalize with tumor neovasculature, but unmodified micelles could just partially co-localize. (Figure 5). Cytotoxicity. The in vitro cytotoxicity of Taxol, GICP− PEG−PLA/PTX micelles, and mPEG−PLA/PTX micelles were evaluated in both U87MG cells and HUVECs (Figure 6). The IC50 value of Taxol, GICP−PEG−PLA/PTX micelles, and mPEG−PLA/PTX micelles in U87MG cells were, respectively, 13, 20, and 23 nM. The IC50 value of Taxol, GICP−PEG−PLA/PTX micelles, and mPEG−PLA/PTX micelles in HUVECs were 9, 17, and 18 nM, respectively. In comparison to micelle formulations, the slightly better efficacy of Taxol might be due to the incomplete release of paclitaxel from the aforementioned micelles. In Vivo Antiglioblastoma and Antiangiogenesis Effect. Saline, Taxol, mPEG−PLA/PTX, or GICP−PEG− PLA/PTX were injected via the tail vein at the 1st, 3rd, 5th, 7th, 9th, and 11th day with a total dosage of 36 mg/kg paclitaxel. A total of 24 h after the last administration, one nude mouse from each group was sacrificed, and tumors were harvested for TUNEL and CD31 immunohistochemical examination (Figure 7). The proportion of TUNEL-positive cells of normal saline, Taxol, mPEG−PLA/PTX, and GICP−PEG−PLA/PTX groups were, respectively, 5.46% ± 1.19%, 14.8% ± 1.20%, 22.7% ± 1.48%, and 40.6% ± 4.53%. The angiogenesis inhibition rate of Taxol, mPEG−PLA/PTX and GICP−PEG−PLA/PTX groups were 13.7% ± 4.69%, 33.5 ± 5.68%, and 78.3% ± 6.00%. Tumor volume and body weight were measured every 2 days. All nude mice were sacrificed on the 16th day. The final tumor inhibition rate of Taxol, mPEG−PLA/PTX, and GICP−PEG− PLA/PTX groups were, respectively, 60.2% ± 8.66%, 64.2% ± 7.10%, and 75.7% ± 6.30% (Figure 7D). The data implied that GICP−PEG−PLA/PTX could significantly improve tumor inhibition efficacy in comparison to mPEG−PLA/PTX (p