Designing Liposomes to Suppress Extracellular ... - ACS Publications

quantified using Image J pro analysis software with the same threshold; results were expressed as percentage staining per visual field. Five fields we...
0 downloads 10 Views 2MB Size
Designing Liposomes To Suppress Extracellular Matrix Expression To Enhance Drug Penetration and Pancreatic Tumor Therapy Tianjiao Ji,‡,†,∇,# Jiayan Lang,†,§,∇,# Jing Wang,†,∇,# Rong Cai,†,∇ Yinlong Zhang,†,⊥,∇ Feifei Qi,†,∇ Lijing Zhang,‡ Xiao Zhao,†,∇ Wenjing Wu,†,⊥,∇ Jihui Hao,∥ Zhihai Qin,*,‡ Ying Zhao,*,†,∇ and Guangjun Nie*,†,∇ ‡

The First Affiliated Hospital, Zhengzhou University, Zhengzhou 450052, China CAS Key Laboratory for Biomedical Effects of Nanomaterials and Nanosafety, CAS Center for Excellence in Nanoscience, National Center for Nanoscience and Technology, Beijing 100190, China § Sino-Danish Center for Education and Research, Sino-Danish College of UCAS, Beijing 100190, China ∥ Department of Pancreatic Carcinoma Tianjin Medical University Cancer Institute and Hospital, National Clinical Research Center of Cancer, Key Laboratory of Cancer Prevention and Therapy, Tianjin 300060, China ⊥ College of Pharmaceutical Science, Jilin University, Changchun 130021, China ∇ University of Chinese Academy of Sciences, Beijing 100049, China †

S Supporting Information *

ABSTRACT: During pancreatic tumor development, pancreatic stellate cells (PSCs) proliferate exuberantly to secrete extracellular matrix (ECM) in the tumor stroma, which presents major barriers for drug delivery and penetration in tumor tissue. Thus, down-regulating ECM levels via regulation of the PSCs may allow enhanced penetration of therapeutic drugs and thereby enhancing their therapeutic efficacy. To regulate the PSCs, a matrix metalloproteinase-2 (MMP-2) responsive peptide-hybrid liposome (MRPL) was constructed via coassembly of a tailor-designed MMP-2 responsive amphiphilic peptide and phospholipids. By utilizing the MMP-2-rich pathological environment, the pirfenidone (PFD) loaded MRPL (MRPL-PFD) can specifically release PFD at the pancreatic tumor site and down-regulate the multiple components of ECM expressed by the PSCs. This resulted in a significant increase in the penetration of gemcitabine into the tumor tissue and enhanced the efficacy of gemcitabine for pancreatic tumor. Our design tailored for antifibrosis of pancreatic cancer may provide a practical approach to build functional liposomes through supramolecular assembly, and regulation of ECM may be a promising adjuvant therapeutic strategy for pancreatic and other ECM-rich tumors. KEYWORDS: peptide-hybrid liposome, MMP-2 responsive, pancreatic stellate cells, down-regulation of ECM, enhanced pancreatic tumor therapy

P

which is the fourth leading cause of cancer-related death in Europe and North America.5,6 Therefore, barrier penetration strategies are urgently needed to enhance drug perfusion and to improve therapeutic efficiency for pancreatic cancer.7−9

ancreatic tumors contain the highest amount of stroma in almost all solid tumors.1,2 The stroma contains multiple extracellular matrix (ECM) components and forms a physical barrier for the delivery of cytotoxic chemotherapeutics, molecular targeted biologics, or nanomedicines to the tumoral milieu, thus diminishing their tumor-targeting and penetrating performance and antitumor efficacy.3,4 This is one of the major reasons leading to the treatment failure of pancreatic cancer, © 2017 American Chemical Society

Received: February 14, 2017 Accepted: August 14, 2017 Published: August 14, 2017 8668

DOI: 10.1021/acsnano.7b01026 ACS Nano 2017, 11, 8668−8678

Article

www.acsnano.org

Article

ACS Nano

Scheme 1. Proposed Mechanism of MMP-2 Responsive Peptide Hybrid Liposome (MRPL) for down-Regulation of ECM in Pancreatic Tumorsa

a Pancreatic tumors possess an ECM-enriched microenvironment (left), and the MRPL can deliver the antifibrosis agent pirfenidone (PFD) to the tumor site and achieve a tumor-specific release via MMP-2 cleavage in tumor stroma. The released PFD down-regulates multiple components of ECM expressed by PSCs (middle) and increases the perfusion of small molecules (probes and chemotherapeutics) in the tumor tissue (right).

light,33,34 heat,35−37 and ultrasound,38−40 which is not benefit to mass production or clinical application. In addition, considering the rigid desmoplasia of pancreatic tumor tissues, it is extremely difficult for nanocarriers to penetrate the tumor stroma to target PTCs. Therefore, pathological and/or environmental stimulus triggered responsive release and regulation of pancreatic tumor microenvironment in the tumor stroma may be a suitable strategy for improved therapeutic efficacy. In this work, in contrast to functionalizing liposomes with chemical conjugation, we constructed a matrix metalloproteinase-2 (MMP-2) responsive peptide-hybrid liposome (MRPL) by coassembly of a tailor-designed MMP-2 responsive amphiphilic peptide (MRP) with a phospholipid (L-αphosphatidylcholine) with environmental responsive feature (Scheme 1). MRPL delivers the soluble antifibrosis drug pirfenidone (PFD) and specifically releases it at the tumor site, responsive to the extensive expression of MMP-2 by both PSCs and tumor cells (but not by normal tissues).24,41 The released PFD down-regulated the expression of multiple ECM components in tumors, with subsequent decrease in the desmoplasia, and enhanced the perfusion of therapeutics (including chemotherapy drugs) into the deep tumor tissue. Given its easy operation and tumor-specific drug delivery capacity, this coassembly strategy may provide opportunities to build up functional liposomes for regulating the stromal ECM.

In pancreatic cancer, pancreatic stellate cells (PSCs) are the major stromal cell type presenting in tumor tissues and the principal source of ECM production, responsible for stromal components up to 70% in some pancreatic tumors.10,11 The components of ECM interact closely with pancreatic tumor cells (PTCs) to create a tumor promoting microenvironment that facilitates and stimulates both local tumor growth and distant metastasis.12,13 Among PSC-specific ECM components, collagen I and fibronectin are the key components of stroma in both primary and metastatic sites, helping to shape the emerging neoorgan and promote the survival of tumor cells;14−17 tenascin C can enhance pancreatic cancer cell growth and motility through the activation of integrin intracellular signaling pathway,18,19 and versican, a large proteoglycan in extracellular peri-tumoral, facilitates tumor invasion and metastasis by decreasing cell-ECM adhesion.20,21 In addition, these ECM components form physical barriers to inhibit the penetration and distribution of agents for tumor imaging and therapy.22 Consequently, regulating the activities of PSCs in pancreatic cancer may be a specific and efficient method to inhibit the secretion of ECM, which may provide an efficient adjuvant strategy for chemotherapy enhancement. Reviewing the ECM construction, both PTCs and PSCs secrete multiple matrix metalloproteinases (MMPs),23,24 which in turn provides opportunities to design responsive materials for pancreatic tumor imaging and drug delivery. Liposome-drug complex is one of the most popular formulations for cancer therapy, and several liposomal formulations have been approved by US Food and Drug Administration (FDA), such as liposomal doxorubicin (Doxil, Janssen Biotech, Inc.) and liposomal irinotecan (Onivyde, Merrimack) and others under active clinical trials.25−27 Thus, developing improved functionality of liposomes may hold great potential for clinical applications. Some well-modified liposomalchemotherapeutic drug formulations with targeting or penetrating functionalities have been developed, achieving improved therapy efficacy.28−30 However, most functional liposomes need chemical conjugation of lipids with functional molecules,31,32 and some responsive liposomes need additional triggers such as

RESULTS AND DISCUSSION MRP Design and MMP-2 Responsiveness. The MRP (SDK(C18)SGPLG-IAGQSK(C18)DS) is a key element of the MRPL. The hydrophilic domain of the amphiphilic peptide contains an MMP-2-specific cleavable sequence (GPLGIAGQ),31,42 and two other hydrophilic amino acids aspartic acid (D) and serine (S), containing carboxyl and hydroxyl groups, were included to increase the hydrophilicity. In addition, two octadecanoic acid chains were linked to the sidechains of the lysine residues to serve as the hydrophobic domain, which can interact with the tails of the phospholipid through hydrophobic force (Figure S1A). The purity of MRP was over 8669

DOI: 10.1021/acsnano.7b01026 ACS Nano 2017, 11, 8668−8678

Article

ACS Nano

Figure 1. The morphology of MPR self-assembly, liposomes, and MRPL with different peptide and lipid ratios, characterized by TEM. Scale bar, 200 nm.

Figure 2. The fine-tuned drug release profiles of the MRPL and MNPL. (A) The drug release curves of MRPL-PFD with different P:L ratios. The P:L = 1:3 group spontaneously leaked drug at a rapid rate, indicating instability of the nanocarrier. (B) The drug release profiles of MRPL-PFD (with different P:L ratios) incubated with MMP-2. A favorable responsive release profile was exhibited by the P:L = 1:5 group of MRPL. (C) TEM images of MRPL-PFD and MNPL-PFD (with P:L = 1:5) before and after incubated with MMP-2. The red boxed region is shown magnified in the inset. The scale bar in the top and middle rows, 200 nm; in the bottom row, 50 nm.

phospholipid at peptide/lipid (P:L) ratios (mass/mass) of 1:10, 1:8, or 1:5, the coassembly products transformed into stable hybrid liposomal nanospheres with uniform size distributions (Figure 1 and Table S1). If the ratio of peptide/lipid was further changed to 1:3, fiber-like structures emerged in the solution, supposedly resulting from the self-assembly of the excess MRP. This result indicated that the insertion of MRP into hybrid liposomes reached saturation at the ratio of 1:3 (P:L), and continuous increase in MRP may decrease the stability and uniformity of the hybrid liposomes. To assess the stability of MRPLs and MNPLs (MMP-2 nonresponsive peptide-hybrid liposomes) with different hybrid ratios (1:10, 1:8, 1:5, or 1:3), we investigated the PFD release profiles at 37 °C in PBS. The MRPL showed a MMP-2 responsive fast drug release (Figure 2A), while MNPL with P:L = 1:3 spontaneously released the drug due to instability (Figure

95%, as shown in Figure S1B. We also synthesized an MMP-2nonresponsive peptide (MNP, DSK(C18)DGSALGQIGPDSK(C18)DS) with a structure and sequence similar to that of MRP as a negative control (Figure S2). The responsiveness of MRP and MNP toward MMP-2 was assessed by matrix-assisted laser desorption/ionization time-offlight mass spectrometry (MALDI-TOF MS). After incubation with MMP-2 for 3 h, the peak of MRP (M = 2079.8) (Figure S3A) disappeared with the emergence of two fragment peaks (M1 = 1025.9 and M2 = 1071.0), whose sizes corresponded to the two predicted digested segments (Figure S3B). By contrast, no responsiveness of MNP (M = 2309.1) toward MMP-2 was observed (Figure S3C, D). MRP Self-Assembly and Hybrid-Liposome Construction. The MRP tends to self-assemble to micrometer-long nanofibers (Figure 1). When MRP was mixed with the 8670

DOI: 10.1021/acsnano.7b01026 ACS Nano 2017, 11, 8668−8678

Article

ACS Nano

Figure 3. The effects of MRPL-PFD on the key compotents of ECM in vitro and in vivo. (A) Expression levels of multiple ECM components (collagen I, fibronectin, versican, and tenascin C) after incubation with various PFD formulations (PFD concentration: 0.3 mg/mL) detected by immunocytochemistry measurement for in vitro samples (upper panel, 48 h treatment) and immunohistochemistry measurment for in vivo tissue samples (lower panel, 2 week treatment, PFD dose: 30 mg/kg). MRPL-PFD significantly down-regulated the levels of four key components of ECM in both in vitro and in vivo samples. The scale bar, 100 μm. (B) Quantificative analysis of positive staining areas treated by the various formulations based on immunocytochemistry measurement (left panel) and immunohistochemistry measurement (right panel). Five random fields were observed for each group. The brown stained areas were taken into account. Data are presented as mean ± SD **p < 0.01, *p < 0.05. The detailed data were presented in Tables S3 and S4.

MRP needed to be increased. For the hybrid liposomes with low MRP proportion (P:L = 1:10 and 1:8), they exhibited limited MMP-2 responsive behavior, and their PFD release efficacy was not significantly increased compared with the pure liposomes and MNPL (Figures S5 and S6). When the P:L ratio was

S4), while the other three groups did not exhibit a significantly different drug release rate compared to the liposomes without peptide component. Consequently, the ratios of P:L below 1:3 were considered as relatively stable formulations. However, to achieve a high MMP-2 responsive profile, the component of 8671

DOI: 10.1021/acsnano.7b01026 ACS Nano 2017, 11, 8668−8678

Article

ACS Nano

Figure 4. Evaluation of Rhd penetration and distribution and therapeutic efficacy of gemcitabine in tumors. (A) Rhd penetration and distribution in pancreatic tumor (PSCs/Mia-paca-2 co-implanted) tissues after 2 weeks’ treatment of the different PFD formulations. Frozen tumor sections were stained with DAPI (blue) to label nuclei. Red: Rhd. Dotted lines show the border between edge and core of tumor tissue. Scale bar, 100 μm. (B) Quantification of the depth of Rhd penetration in tumors treated by the PFD formulations. Five random fields were observed for each group under confocal microscopy. Data are presented as mean ± SD **p < 0.01. (C) The growth curves of PSCs and Mia-paca-2 co-implanted pancreatic tumors in mice treated by the different PFD formulations. The dose of PFD was 30 mg/kg in all groups at days 3, 5, 7, 9, 11, 13, and 15 (hollow arrows). At days 17, 19, and 21 (solid arrows), the various PFD treatments were terminated, and gemcitabine was injected (gemcitabine dose: 20 mg/kg). Data are presented as mean ± SD (n = 8). **p < 0.01. (D) Histological analysis of tumor slices of each group after treated with gemcitabine. H&E staining, the top row; TUNEL immunohistochemical staining, the bottom row. Scale bar, 100 μm. The tumor tissue of MRPLPFD pretreated group exhibited the largest apoptosis areas and the highest TUNEL-positive cells among all groups.

increased to 1:5, over 80% of encapsulated PFD was released promptly and adequately within 180 min in the presence of MMP-2 (Figure 2B). Based on these results, we chose the P:L = 1:5 MRPL as the optimal carrier for subsequent experiments (Table S2). The transmission electron microscopy (TEM) morphology changes may also be an evidence of the MRPL’s MMP-2 responsiveness. In the presence of MMP-2, approximately 90%

liposomal vesicles were not smooth anymore, most likely due to the cleavage of MMP-2 responsive peptides (MRPs) (Figures 2C and S7). In contrast, the morphology of MNPL did not exhibit any observable changes (Figures 2C and S7). These results suggested that the MRPL can release the drug via a membrane disturbance when the MRPs were cleaved by MMP-2. This specific drug release behavior is supposed to increase the 8672

DOI: 10.1021/acsnano.7b01026 ACS Nano 2017, 11, 8668−8678

Article

ACS Nano

every 2 days (PFD dose: 30 mg/kg). Tumors were removed after 2 weeks for immunohistochemistry (IHC) analysis (Figure 3A, B). IHC images showed obvious down-regulation of the expression of all the four major ECM components in MRPLPFD treated tumors compared with other groups (Figure 3A), with collagen I down-regulated to 19.2 ± 1.0%, fibronectin to 15.3 ± 1.4%, versican to 36.3 ± 2.6%, and tenascin C to 38.5 ± 2.2%, compared to the control group. Although free PFD also down-regulated ECM components to some degree, its efficacy was much more limited (65.3 ± 4.8% of collagen, 60.2 ± 4.7% of fibronectin, 85.3 ± 4.8% of versican, and 77.4 ± 5.9% of tenascin C, respectively, compared to the control group) (Figure 3B, Table S4). This occurred probably due to the short circulation time of free PFD in vivo and the low tumor accumulation efficiency. MNPL-PFD showed slightly superior efficacy than free PFD, which was different from the in vitro results, and may be attributed to its prolonged circulation time compared to free PFD in vivo as well as its accumulation as whole particles at the tumor site via the enhanced access of tumor vasculature.45,46 However, without MMP-2 responsive property, even though the MNPL accumulated similar amount as of MRPL, MNPL nanoformulation cannot efficiently release the loaded cargos, which limited cargo perfusion in the tumor tissue (Figure S10). Based on these results, the MRPL-PFD-mediated PFD in vivo delivery and tumor release by responding to the MMP-2 at the tumor site and exhibited the highest efficacy of the ECM downregulation in vivo. Enhanced Penetration and Therapeutic Efficacy. The overexpression of ECM is a major barrier to drug delivery, therefore the inhibition of ECM expression may increase the perfusion of therapeutics in the tumor. To verify this hypothesis, free Rhd was chosen to visualize the perfusion efficiency of small molecules at the tumor site after different treatments of PFD formulations for 2 weeks. One hour after injection, in vivo images showed the strongest fluorescence signal in the MRPL-PFD treated group (Figure S11A), which was about 5.3 times stronger than the PBS and MRPL treated groups and 2.8 and 2.5 times stronger than the PFD and MNPL-PFD groups, respectively (Figure S11B). All of these results indicate the maximum tumor perfusion of Rhd in the MRPL-PFD treated mice. Confocal microscopy images showed a gradient fluorescence of Rhd from the tumor edge to the core. Importantly, the fluorescence intensity of Rhd was much wider and stronger in MRPL-PFDtreated tumors compared with MNPL-PFD, free PFD, blank MRPL, or control groups (Figure 4A). Further analysis of the average penetration depth of Rhd in each group showed that the penetration depth in MRPL-PFD treated group was 972.2 ± 28.3 μm, which was approximately 5.3, 3.0, and 9.2 times deeper than free PFD (182.3 ± 14.7 μm), MNPL-PFD (323.6 ± 21.5 μm), control (105.3 ± 12.1 μm), and MRPL (104.7 ± 11.2 μm) groups, respectively (Figure 4B). Taken together, these data demonstrated that we had successfully constructed a nanocarrier MRPL-PFD for down-regulating ECM and increasing the perfusion of small molecules in the pancreatic tumors. Since the down-regulation of ECM can increase the perfusion of small molecules, efficacy of chemotherapy drugs may also be enhanced. To explore the impact of ECM regulation on chemotherapeutic efficacy, PSCs/Mia-paca-2 pancreatic tumor bearing mice were administered (iv) PBS, PFD, MNPL-PFD, or MRPL-PFD (n = 8) when the tumor volume reached approximately 50 mm3. A dose of 30 mg/kg PFD in each formulation was given every 2 days for a total of seven injections (days 3, 5, 7, 9, 11, 13, and 15). The results showed that the

intratumoral drug concentration and therefore improve the efficacy. Extracellular Release Behavior of MRPL. Due to the high interstitial fluid pressure of pancreatic tumors,43,44 most nanomaterials are not able to penetrate the tumor stroma; therefore, the drugs need to be released when the nanocarriers reach the stroma, so that the released drug can be promptly taken up by PSCs. To confirm the specific expression of MMP-2 in tumor tissues, MMP-2 activity in PSCs and Mia-paca-2 (a pancreatic tumor cell line) and MMP-2 expression in the major organs in the PSC and Mia-paca-2 co-implanted tumor bearing mice were examined (Figure S8). The enzyme activity of MMP2 in PSCs was even higher than the tumor cells (Figure S8A), and the tumor tissue exhibited overexpression of MMP-2 compared to other organs (Figure S8B). This result allowed us to conclude that a sufficient amount MMP-2 is active in the tumor tissue to trigger the release of PFD from MRPL-PFD. Next, we used a small molecular dye, rhodamine (Rhd) as a model drug to examine the extracellular release of MRPL carrier and the small molecule uptake behavior of PSC. The Rhd uptake by PSCs in the Rhd-encapsulated MRPL (MRPL-Rhd) group showed a fluorescence intensity similar to the free Rhd group, measured by laser confocal microscopy and flow cytometry (Figure S9), indicating the sufficient release of Rhd by MRPLRhd. In contrast, the MNPL-Rhd cannot efficiently release its Rhd content due to its nonresponsiveness to MMP-2. Only a weak Rhd signal likely derived from a slight uptake of MNPLRhd by PSCs was measured. In PSCs treated with an MMP-2 inhibitor (ARP100) before the incubation with cells, the MRPLRhd was not cleaved, and the fluorescence signal was similar to the MNPL-Rhd group (Figure S9A, B). These results led us conclude that only when the MRPL nanocarriers encountered MMP-2, efficient cargo release was triggered, resulting in subsequent enhanced cellular uptake of loaded small molecules. These factors ensured the specificity and efficiency of drug release at the tumor site. ECM down-Regulation of MRPL-PFD in Vitro and in Vivo. Collagen I and fibronectin are the key stromal components secreted by PSCs,14−17 and tenascin C and versican are also important PSCs-specific ECM components,18−20 thereby we chose these four components to evaluate the ECM regulation efficacy of each PFD formulation in cell culture and tumor tissues. After incubation with different formulations (MRPL, free PFD, MNPL-PFD, or MRPL-PFD) (PFD concentration: 0.3 mg/mL) for 48 h, the PSCs were fixed, and each ECM component was detected by immunocytochemistry. ECM expression in the blank MRPL treated group was not significantly different compared with the control group (Figure 3A), indicating no ECM regulatory effect from the nanocarrier itself. However, all four ECM components were significantly downregulated by MRPL-PFD, with 28.1 ± 1.2% of Collagen I, 27.4 ± 1.5% of fibronectin, 51.7 ± 3.2% of versican, and 13.4 ± 1.7% of tenascin C, compared to the control group (100%), a similar efficacy to that of free PFD (Figure 3B, Table S3). However, MNPL-PFD down-regulated each ECM component to a very limited degree, probably due to ineffective PFD release and limited cell uptake as demonstrated previously. These results demonstrated that MRPL-PFD could efficiently release the loaded PFD in response to MMP-2, which enabled efficient PFD-mediated down-regulation of ECM expression. To further explore the ECM suppression efficacy of the nanoformulations in vivo, PBS, MRPL, free PFD, MNPL-PFD, or MRPL-PFD were intravenously injected into tumor bearing mice 8673

DOI: 10.1021/acsnano.7b01026 ACS Nano 2017, 11, 8668−8678

Article

ACS Nano

Self-Assembly of MRP and MNP. 0.05 mg MRP or MNP was dissolved in 5.0 μL dimethyl sulfoxide (DMSO) (Sigma-Aldrich, USA) and then added into 1.0 mL distilled water to be ultrasonicated (KQ2200E, China) at 100 W for 1 min. The resulted nanostructures were characterized by TEM (HT7700, HITACHI, Japan) using a negative staining method with phosphotungstic acid at different times of incubation. Preparation and Characterization of Liposomes, MNPL, and MRPL. Liposomes were prepared by a thin film hydration method followed by membrane extrusion. 35 Briefly, 1.0 mg of L -αphosphatidylcholine (soybean lecithin, J&K Scientific, China), 0.02 mg of 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[methoxy(polyethylene glycol)-2000] (DSPE-PEG2000) (Nanocs, USA), and 0.25 mg of cholesterol (J&K Scientific, China) were dissolved using 10.0 mL of methylene chloride in a round-bottom flask and incubated for 5 min. Then a phospholipid film was formed through a rotary evaporator under reduced pressure. 1.0 mL of PBS was added into the roundbottom flask with phospholipid film for hydration for 1 h. After hydration, closely graded unilamellar liposomes were obtained by 11 times extrusion through 200 and 100 nm membranes using an extruder (LiposoFast-basic. Avestin, CA). To prepare the MRPL, MRP was dissolved in methanol at first, and then lecithin, 0.02 mg of PEG-DSPE, and 0.25 mg of cholesterol were dissolved using 10.0 mL of methylene chloride. The total mass of the MRP and lecithin was 1.0 mg but with the different ratios of P:L (such as 1:10, 1:8, 1:5, 1:3). The methanol and methylene chloride solutions were mixed, and the phospholipid film was prepared. The morphology of liposome and MRPL was characterized by TEM (HT7700, HITACHI, Japan) using a negative staining method. The size, polydispersity index, and zeta-potential of them were measured by dynamic light scattering (DLS, Zetasizer Nano ZS90, Malvern, UK). The MNPLs were prepared using the same method as the MRPLs. Preparation and Characterization of PFD Loaded Liposomes, MNPL, and MRPL. After the phospholipid film was formed through a rotary evaporator under reduced pressure, 1.0 mL PFD-containing PBS was added into the round-bottom flask with phospholipid film for hydration for 1 h. After hydration, closely graded unilamellar liposomes were obtained by 11 times extrusion through 200 and 100 nm membranes using an extruder. Next, the liposome, MNPL, or MRPL was collected after centrifugation at 10,000 g for 10 min at room temperature. For evaluating PFD encapsulation efficiency, lyophilized PFD-loaded liposome MNPL or MRPL was dissolved in DMSO and measured by UV−vis spectrophotometer (LAMBDA650, PerkinElmer, USA) after making a standard curve with free PFD (the detected wavelength was 263 nm). The encapsulation efficiency (EE) was calculated according to the following formula:

average tumor volume in the PFD treated groups was slightly smaller, but not significant, than that of the control group (Figure 4C), which may be due to fast growth of the tumors regulated by some autocrine regulation pathways in this tumor model.47−49 A dose of 20 mg/kg gemcitabine was given at days 17, 19, and 21. Further monitoring of tumor growth showed that the tumors continued growing in the control group, indicating that gemcitabine did not effectively inhibit the fast-growing tumors, neither did the free PFD and MNPL-PFD groups (Figure 4C). However, the size of tumors in the MRPL-PFD group decreased significantly after initiation of gemcitabine treatment (**p < 0.01), compared to PBS, PFD, and MNPL-PFD groups. Mice were sacrificed 7 days after gemcitabine treatment because of the excess large tumor volume in the control group and the serious side effects of gemcitabine (Figure S12). H&E stained tumor tissues in the MRPL-PFD treated group exhibited larger areas of apoptosis compared to other groups (Figure 4D). The percentage of apoptotic cells in each group was further determined by TUNEL immunohistochemical staining, and the TUNEL-positive cells in MRPL-PFD group were 59.7 ± 5.7%, while other groups were much lower (control: 9.8 ± 2.4%, PFD: 17.3 ± 3.4%, MNPL-PFD: 28.4 ± 4.5%) (Figures 4D and S13). To directly measure drug perfusion and accumulation, the sum of the concentrations of gemcitabine and its metabolite, 2′,2′-difluorodeoxyuridine (dFdU)50 in tumor tissue detected by high-performance liquid chromatography (HPLC), was used to evaluate the drug perfusion efficiency. The tumor drug concentration in MRPL-PFD (6.2 ± 0.8 pmol/mg protein) was three times higher than control group (2.1 ± 0.4 pmol/mg protein), two times higher than PFD group (3.0 ± 0.3 pmol/mg protein), and 1.5 times higher than MNPL-PFD group (4.3 ± 0.8 pmol/mg protein) (Figure S14). These results indicated that MRPL-PFD significantly enhanced the therapeutic efficacy of gemcitabine by increasing the perfusion of gemcitabine.

CONCLUSION In conclusion, considering the pathological environment of pancreatic tumors via carefully engineering molecular selfassemblies, we have constructed an MMP-2 responsive liposome (MRPL) through coassembly of a functional peptide and phospholipid. The designed hybrid nanosystem achieved tumor-specific delivery and release of PFD at the PSCs-enriched pancreatic tumors. The released PFD down-regulates expression of multiple ECM components of PSCs, which increase the perfusion of small molecules in the tumor, thereby enhancing therapeutic efficacy of gemcitabine. Due to the ease of production and tumor-specific drug delivery, the peptide-hybrid liposomal nanoformulations may provide an attractive approach to build stroma responsive liposomes with the functionality of regulating the tumor stroma and enhancing drug penetration into deep tumor tissue. Regulation of ECM may become an efficient adjuvant strategy to improve the efficacy of pancreatic tumor therapy.

EE(%) = (mass of PFD encapsulated in nanoparticles /mass of PFD added) × 100% The in vitro PFD release kinetics of liposome-PFD, MNPL-PFD, and MRPL-PFD was measured via dialysis. Briefly, 1.0 mL of liposome-PFD, MNPL-PFD, or MRPL-PFD solution was injected into a dialysis cartridge with a molecular weight cutoff value of 2 kDa. The cartridge was dialyzed against 10.0 mL of PBS (10.0% FBS) and shaken at 37 °C at 100 rpm with activated charcoal to create an ideal sink condition. The concentration of PFD remaining in the dialysis cartridge at different time points was measured by UV−vis spectrophotometry. Preparation of Rhd Loaded MNPL and MRPL. After the phospholipid film was formed through a rotary evaporator under reduced pressure, 1.0 mL of Rhodamine 6G (Rhd) (Aladdin Industrial Inc., China) containing PBS was added into the round-bottom flask with phospholipid film for hydration for 1 h. After hydration, closely graded unilamellar liposomes were obtained by 11 times extrusion through 200 and 100 nm membranes using an extruder. Next, the MNPL-PFD or MRPL was collected after centrifugation at 10,000 g for 10 min at room temperature. Cell Culture and Animals. We separated human pancreatic stellate cells (PSCs) from pancreatic cancer surgical specimens using the outgrowth method described by Bachem and Apte,51,52 and all of the established PSCs were used between passage 3 and 8. The use of

EXPERIMENTAL SECTION Enzyme-Responsiveness of MRP and MNP. Recombinant human MMP-2 (rhMMP-2, R&D system, USA) was diluted to 0.2 μg/mL in the assay buffer (50.0 mM Tris, 10.0 mM CaCl2, 150.0 mM NaCl, 0.05% (w/v) Brij 35, pH 7.5). After 50.0 μL of rhMMP-2 (0.2 μg/ mL) was loaded into a plate, the reaction was started by adding 50.0 μL of 100.0 μM MRP (dissolved in assay buffer). After 3 h, the reaction product was analyzed by MALDI-TOF MS (Bruker, Germany). The MS detection range was 600−2200 Da for MRP and 600−2500 for MNP. 8674

DOI: 10.1021/acsnano.7b01026 ACS Nano 2017, 11, 8668−8678

Article

ACS Nano pancreatic cancer surgical specimens was approved by the department of pancreatic carcinoma Tianjin Medical University, Cancer Institute and Hospital National Clinical Research Center of Cancer, Key Laboratory of Cancer Prevention and Therapy (Tianjin, China) and was conducted according to the Ethical Guidelines for Human Genome/Gene Research enacted by the Chinese Government and the Helsinki Declaration. Miapaca-2 (pancreatic cancer cell) cell line was purchased from the National Platform of Experimental Cell Resources for Sci-Tech (China), and the cell line was passaged in the laboratory for