Nanoplatform Assembled from a CD44-Targeted ... - ACS Publications

Jan 19, 2018 - theless, it is not cytotoxic and therefore cannot kill tumor cells, ..... compared to that on the first day of treatment (Figure 6A). I...
1 downloads 0 Views 3MB Size
ACS Nano 2018.12:1519-1536. Downloaded from pubs.acs.org by UNITED ARAB EMIRATES UNIV on 01/01/19. For personal use only.

Nanoplatform Assembled from a CD44Targeted Prodrug and Smart Liposomes for Dual Targeting of Tumor Microenvironment and Cancer Cells Yaqi Lv, Chaoran Xu, Xiangmei Zhao, Chenshi Lin, Xin Yang, Xiaofei Xin, Li Zhang, Chao Qin, Xiaopeng Han, Lei Yang, Wei He,* and Lifang Yin* Department of Pharmaceutics, School of Pharmacy, China Pharmaceutical University, Nanjing 210009, P.R. China S Supporting Information *

ABSTRACT: The tumor microenvironment (TME) plays a critical role in tumor initiation, progression, invasion, and metastasis. Therefore, a therapy that combines chemotherapeutic drugs with a TME modulator could be a promising route for cancer treatment. This paper reports a nanoplatform self-assembled from a hyaluronic acid (HA)-paclitaxel (PTX) (HA-PTX) prodrug and marimastat (MATT)-loaded thermosensitive liposomes (LTSLs) (MATT-LTSLs) for the dual targeting of the TME and cancer cells. Interestingly, the prodrug HA-PTX can selfassemble on both positively and negatively charged liposomes, forming hybrid nanoparticles (HNPs, 100 nm). Triggered by mild hyperthermia, HA-PTX/MATT-LTSLs HNPs rapidly release their payloads into the extracellular environment, and the released HA-PTX quickly enters 4T1 cells through a CD44-HA affinity. The HNPs possess promoted tumor accumulation (1.6-fold), exhibit deep tumor penetration, and significantly inhibit the tumor growth (10-fold), metastasis (100%), and angiogenesis (10-fold). Importantly, by targeting the TME and maintaining its integrity via inhibiting the expression and activity of matrix metalloproteinases (>5-fold), blocking the fibroblast activation by downregulating the TGF-β1 expression (5-fold) and suppressing the degradation of extracellular matrix, the HNPs allow for significant metastasis inhibition. Overall, these findings indicate that a prodrug of an HA−hydrophobic-active compound and liposomes can be self-assembled into a smart nanoplatform for the dual targeting of the TME and tumor cells and efficient combined treatment; additionally, the co-delivery of MATT and HA-PTX with the HNPs is a promising approach for the treatment of metastatic cancer. This study creates opportunities for fabricating multifunctional nanodevices and offers an efficient strategy for disease therapy. KEYWORDS: thermosensitive liposomes, prodrug, matrix metalloproteinases, tumor microenvironment, extracellular matrix, metastatic breast cancer

T

promotion of cancer progression and tumor metastasis, especially due to their ability to alter the microenvironment.6 The high expression of MMPs, especially MMP-2 and MMP-9, has been found in a variety of common tumors, such as colorectal carcinomas, breast cancer, brain cancer, and non-small-cell lung cancer.7−11 Previous reports have also suggested that interfering with MMP function or expression blocks the metastatic cascade and tumor growth.12 Accordingly, it is widely assumed that regulating the expression of MMPs in the TME can suppress tumor growth and inhibit metastatic spread.3,13,14

he tumor microenvironment (TME) plays a critical role in tumor initiation, progression, metastasis, and resistance to therapy. It is composed of noncancer cells including endothelial cells, immune cells and fibroblasts, an altered extracellular matrix (ECM), and a vascular and lymphatic network.1 Among these elements, the ECM contains a large number of matrix macromolecules and consequently provides physical scaffolding for embedded cells, and it is also a predominant factor influencing cancer cell migration and drive invasion.2 As a result, changes in the ECM regulate the TME and thus affect cellular processes, including invasion and migration.3,4 Matrix metalloproteinases (MMPs), which are extracellular proteinases, can degrade ECM components, such as collagens, gelatins, and fibronectin and cleave relevant proteins.5 Therefore, MMPs play a vital role in the © 2018 American Chemical Society

Received: November 14, 2017 Accepted: January 19, 2018 Published: January 19, 2018 1519

DOI: 10.1021/acsnano.7b08051 ACS Nano 2018, 12, 1519−1536

Article

www.acsnano.org

Cite This: ACS Nano 2018, 12, 1519−1536

Article

ACS Nano

Figure 1. Schematic illustration of HNP preparation and intended mechanism for metastatic breast cancer treatment. The components of HNPs are MATT-LTSLs and the prodrug HA-PTX. HA-PTX was synthesized and self-assembled onto the surface of MATT-LTSLs to form spherical hybrid drug-loaded nanoparticles. When HNPs pass through the tumor tissue, the release of HA-PTX and MATT encapsulated in LTSLs can be triggered by local heating of the tumor site. The released MATT will bind with the MMPs in the TME to prevent the TME against destruction and block the metastasis of tumor cells, whereas the HA-PTX prodrug, via targeting the CD44 receptor of tumor cells, will undergo CD44-mediated endocytosis and thus kill the cancer cells.

In this study, hybrid nanoparticles (HNPs) directly selfassembled from an HA-PTX prodrug and MATT-loaded LTSLs (MATT-LTSLs) were developed to deliver MATT and HA-PTX to the TME and cancer cells, respectively, for the treatment of metastatic breast cancer. The formation of this hybrid nanosystem (HA-PTX/MATT-LTSL HNPs) and the proposed mechanism of action after intravenous administration are illustrated in Figure 1. Unlike traditional targeted delivery routes, this nanosystem was fabricated by assembling a targeting prodrug and another drug payload liposome that simultaneously accomplishes combined therapy for the disease and activetargeted delivery. Importantly, the versatile HNPs are able to simultaneously target the disease microenvironment and diseased cells, representing an efficient approach of disease therapy. This report indicates that a prodrug composed of insoluble drugs and polymers and liposomes can be assembled directly into a nanoscaffold. This study provides a strategy for preparing smart nanoparticles for drug delivery. To study the formation of this nanosystem and explore its potential application for disease treatment, comprehensive experiments were performed.

Marimastat (MATT), a broad-spectrum synthetic enzyme inhibitor, can achieve greater than 50% enzyme inhibition of collagenases, gelatinases, and MMPs, even at a nanomolar concentration; it mimics MMP substrates and thus displays competitive and potent, but reversible, suppression.12,15,16 MATT inhibits tumor progression by inhibiting metastasis. Nevertheless, it is not cytotoxic and therefore cannot kill tumor cells, which dramatically compromises its antitumor efficacy. Consequently, delivering MATT alone to target MMPs in the TME is insufficient to eliminate tumor cells. Thus, a potent approach for cancer therapy based on MATT is highly desirable. Generally, researchers have assumed that overexpressed receptors in cancer cells provide a well-defined strategy for selectively killing tumor cells.17 CD44 is a transmembrane cellsurface protein that is highly expressed on various types of cancer cells, including breast, pancreas, colon, prostate, and stomach cancer,18,19 and CD44 determines the tumorigenic and metastatic capacities of many cancer cells.20 Hyaluronic acid (HA), a natural acidic polysaccharide that is nontoxic and biodegradable, possesses a strong affinity for CD44 receptors and thus can be utilized as a ligand for tumor targeting.21 Paclitaxel (PTX) is one of the most effective cytotoxic drugs for treating diverse solid tumors, especially metastatic breast cancer.22,23 Treatment with HA-conjugated PTX improves cancer therapy due to the specific binding of HA to CD44.24 Therefore, it is expected that the combination therapy with MATT and an HA-PTX prodrug will have a synergistic effect on the inhibition of tumor growth. However, it has been a great challenge to simultaneously load the prodrug and MATT in a single vehicle for their subsequent delivery and release to target sites. Lysolipid-containing thermosensitive liposomes (LTSLs) are smart nanocarriers that are stable at body temperature (37 °C) but disintegrate during a phase transition that occurs during mild hyperthermia (HT) at 42 °C to promote the rapid release of encapsulated drugs in a local region.25,26 LTSLs loaded with doxorubicin (ThermoDox) have entered human clinical trials,27−29 and accordingly, LTSLs are a promising drug delivery system.

RESULTS Preparation and Characterization. HA-PTX/MATTLTSL HNPs consisted of the prodrug HA-PTX and MATTLTSLs. HA-PTX was synthesized according to a previous report.30 In brief, HA-PTX was synthesized by adipic dihydrazide (ADH)modified HA and PTX 2′-OH via a succinate ester bond and was characterized by infrared (IR), ultraviolet and visible spectra (UV/vis) spectra, and 1H NMR spectroscopy (Supplementary Figure S1). The PTX loading rate (%) in the prodrug was approximately 15%, as calculated by the characteristic peak of 1H NMR. MATT-LTSLs were prepared by film hydration followed by the probe supersonic method. The encapsulation percentage of MATT in LTSLs based on high-performance liquid chromatography (HPLC) assay was 56.36%. Additionally, cationic temperature-sensitive liposomes (LTDSLs) and long-circulating liposomes (SSLs) were included in the study to study the mechanism of formation for the self-assembled hybrid nanoparticles. 1520

DOI: 10.1021/acsnano.7b08051 ACS Nano 2018, 12, 1519−1536

Article

ACS Nano

FITC-HA-PTX and Rho-LTSLs from HNPs through a mild hyperthermia process that disintegrates LTSLs (Figure 2C). Figure 2E shows the confocal laser scanning microscope (CLSM) images of 4T1 cells incubated with FITC-HA-PTX/ Rho-LTSL HNPs at different ratios. Upon increasing the concentration of Rho-LTSLs (acceptor), red fluorescence intensified while the green fluorescence of FITC-HA-PTX (donor) gradually weakened. Next, we assumed that the membrane fluidity of liposomes would be affected if the HA-PTX assembly was true. Therefore, the measurement of fluorescence anisotropy was performed. 1,6-Diphenyl-1,3,5-hexatriene (DPH), a membrane-bound probe that reflects the interference in the hydrophobic region of lipid membrane,33,34 was loaded in the lipid bilayers of LTSLs in HNPs. As shown in Figure 2F, the anisotropy values obtained from HNPs with different HA/lipid ratios at 25 or 37 °C were increased by 15−27% compared with that from LTSLs, demonstrating that the DPH mobility was reduced after the assembly of HA-PTX. The decreased membrane fluidity of LTSLs, in turn, verified the strong interaction between HA-PTX and LTSLs and confirmed the assembly of HA-PTX on liposomes. Noticeably, at 42 °C, close to the phase transition temperature of LTSLs, the anisotropy values decreased significantly in comparison with that at 25 or 37 °C. Furthermore, the values from HNPs at 42 °C, despite the HA-PTX/LTSL ratios, were similar to that from LTSLs, indicating the maintenance of thermosensitivity. These data demonstrated that the coating of HA-PTX reduced the membrane fluidity of LTSLs and that the HNPs possessed similar thermosensitivity with LTSLs. Collectively, the measurements of FRET and fluorescence anisotropy demonstrated that the prodrug HA-PTX can be assembled onto the LTSLs, LTDSLs, as well as SSLs. Serum Stability and Temperature-Triggered Release. The stability was studied by FRET. After being stored in 10% fetal bovine serum (FBS) medium for 48 h at 37 °C, the FRET ratio in FITC-HA-PTX/Rho-LTSL HNPs was not altered (Figure 2G) and, accordingly, indicated that HNPs were stable against dissociation. To assess the thermosensitivity of prepared LTSLs and determine whether HA-PTX assembled on LTSLs would affect their responses to temperature, an examination of temperaturetriggered drug release was conducted for both LTSLs and HNPs. A fluorescence probe, 5(6)-carbocyfluorescein (CF), was loaded into the vehicles to study the release profiles at different temperatures. The assembly of HA-PTX onto the surface of LTSLs had little effect on the release of CF at 37 °C, however, and resulted in a retarded release at 42 °C 60 min later (Figure 2H). Nonetheless, at 42 °C, both LTSLs and HNPs released their payloads in a dramatically faster pattern than that at 37 °C (normal body temperature) due to the presence of lysolipids in the lipid bilayer of LTSLs, which leads to a less ordered phospholipid arrangement,31,35 demonstrating that HNPs had a similar thermoresponsive feature, as well. This result demonstrated that after an injection into the circulation system, most of the drug would be well encapsulated in HNPs, and rapid drug exposure in a local disease site could be achieved with HT treatment. CD44-Targeted Intracellular Uptake of HA-PTX and Trafficking. To study the expression of CD44 receptors on 4T1 cells, CLSM observation and quantification by flow cytometry were performed after staining with a FITC-labeled CD44 antibody. As depicted in Supplementary Figure S3A, intense green fluorescence on the cell surface was observed, which

Finally, the HNPs were prepared by self-assembling the prodrug HA-PTX with these liposomes at a HA/lipid mass ratio of 1:4. Table 1 summarizes the main physical−chemical properties of these key nanoparticles. Dynamic light scattering (DLS) Table 1. Physical and Chemical Properties of Key Nanoparticlesa nanoparticles HA-PTX MATT-LTSLs HA-PTX/ MATT-LTSL HNPs LTDSLs HA-PTX/ LTDSL HNPs SSLs HA-PTX/SSL HNPs

particle size (nm) ± SEM

PDI ± SEM

surface charge (mV) ± SEM

90.20 ± 0.70 95.60 ± 1.10

0.23 ± 0.01 0.24 ± 0.01

−5.73 ± 2.92 3.09 ± 2.30 −1.62 ± 1.25

102.10 ± 1.20 109.60 ± 1.90

0.24 ± 0.01 0.28 ± 0.01

10.41 ± 2.02 7.13 ± 1.86

82.40 ± 0.90 90.30 ± 1.40

0.19 ± 0.01 0.23 ± 0.01

−2.21 ± 1.35 −4.37 ± 1.72

a

The HA/lipid and PTX/MATT mass ratios in these HNPs was 1:4 and 1:1, respectively.

showed the MATT-LTSLs and HNPs were 90−100 nm in diameter with a polydispersity index (PDI) less than 0.3. The average particle size and PDI of HA-PTX/MATT-LTSL HNPs with a PTX/MATT mass ratio of 1:1 were a little greater than those of MATT-LTSLs, predominately due to the coating of HA-PTX (Figure 2A). Similar results were obtained with HAPTX/LTDSL HNPs and HA-PTX/SSL HNPs. In contrast to MATT-LTSLs with a positive surface charge, a slight negative surface charge (−1.62 mV) was detected in HA-PTX/MATTLTSL HNPs. Therefore, to some extent, HA-PTX was attached to the surface of MATT-LTSLs. The spherical morphology of MATT-LTSLs and HNPs was confirmed by transmission electron microscopy (TEM) (Figure 2D). Confirmation of Coassembly and Membrane Fluidity Measurement. To ascertain the assembly of HA-PTX on liposomes, measurements of fluorescence resonance energy transfer (FRET) and membrane fluidity were performed. First, FRET was utilized to confirm that HA-PTX can be assembled on the liposomes. FRET, a well-defined energy transfer process between a fluorescent donor and an acceptor within 1−10 nm, is an efficient tool to detect the dedicated interplay between nanoparticles and external or internal stimuli by monitoring the changes of fluorescence from both donor and acceptor.31,32 With this approach, fluorescein isothiocyanate (FITC), a donor, was conjugated with HA-PTX, and rhodamine B (Rho), an acceptor, was encapsulated in LTSLs, LTDSLs, and SSLs, with surface charges of approximately +3, +10 and −2 mV, respectively. Strong fluorescence from FITC-HA-PTX (FITC, 515 nm) and low emission from Rho-LTSLs (Rho, 575 nm) were displayed (Figure 2B), which indicated a profound FRET effect from the donor and acceptor. To study the energy transfer, HNPs with various FITC-HA-PTX/Rho-LTSL ratios were prepared, and their fluorescence emission spectra were monitored. Importantly, fluorescence intensity from the acceptor (Rho, 575 nm) increased with an increase of Rho-LTSLs, whereas the donor (FITC) fluorescence at 515 nm gradually declined (Figure 2B). Similar results were obtained with FITC-HAPTX/Rho-LTDSL HNPs and FITC-HA-PTX/Rho-SSL HNPs (Supplementary Figure S2), which exhibited gradually increased fluorescence quenching of FITC as the Rho in HNPs increased. Inversely, a decreasing trend in the acceptor fluorescence and an increase in donor fluorescence are shown after separating 1521

DOI: 10.1021/acsnano.7b08051 ACS Nano 2018, 12, 1519−1536

Article

ACS Nano

Figure 2. Characterization. (A) Size distributions of LTSLs and HNPs. (B) Fluorescence emission spectra of FITC-HA-PTX/Rho-LTSL HNPs with FITC/Rho mass ratios of 1:0, 2:1, 1:1, 1:2, and 0:1. Excitation wavelength: 450 nm. (C) Fluorescence emission spectra of FITC-HAPTX/Rho-LTSLs with (red) and without (black) HT treatment at 42 °C. (D) TEM images. (E) Images of 4T1 cells after being treated with FITC-HA-PTX/Rho-LTSL HNPs at ratios of 2:1, 1:1, and 1:2 (green, FITC; red, Rho) at 37 °C. The images were acquired by CLSM. The scale bar is 5 μm. (F) Fluorescence anisotropy of DPH in HNPs with HA/lipid mass ratios of 1:8, 1:4, and 1:2 (mean ± SEM, n = 3, *P < 0.05). The control was LTSLs. The determination was performed at 25, 37, and 42 °C. An increase in anisotropy value indicates reduction in membrane fluidity of liposomes. (G) Serum stability of FITC-HA-PTX/Rho-LTSL HNPs in 10% fetal bovine serum medium was studied by FRET at 37 °C within a 48 h period (mean ± SEM, n = 3). (H) In vitro release profile at 42 or 37 °C. Data were presented as mean ± SEM (n = 3,*P < 0.05, **P < 0.01).

by assembling Rho-HA-PTX onto CF-LTSLs. As shown in Supplementary Movie S1 and Figure 3A, after HT treatment, the distribution of green fluorescence from CF in the extracellular medium and profound red fluorescence (Rho) in the cytoplasm were observed and became stronger over time. This result revealed that after delivery to the local disease site and HT treatment, the HNPs were not ready for cellular uptake but released the payloads of CF and HA-PTX into the extracellular environment, and the HA-PTX achieved rapid cellular entry. Next, we investigated the role of the CD44 receptor in the cellular uptake of HA-PTX in 4T1 cells using CLSM and flow cytometry. In contrast with the control (free Rho, Supplementary Figure S4), after the cells were cultured, significant fluorescence (red) was visualized in perinuclear regions (Supplementary S5A), and the fluorescence became stronger with increasing Rho

indicated high expression of CD44 receptors on 4T1 cells. Additional quantitative analysis demonstrated 93.07 ± 1.29% CD44positive 4T1 cells (Supplementary Figure S3B). For the other intended target, collagen, especially collagen I, and gelatin, which are the two dominant components of the ECM,2 are rich in the 4T1 cell-induced TME and play vital roles in the migration and invasion of tumor cells. Therefore, 4T1 cells were chosen for our cell experiments, and 4T1 tumor-bearing mice were used to study in vivo performance. We first examined whether HNPs would promote a desired pattern using a real-time confocal technique. After reaching the local disease site and undergoing HT treatment at 42 °C, HNPs were disrupted and subsequently released the LTSL-encapsulated payload and HA-PTX, which could rapidly enter cancer cells via CD44 receptor association. To test this approach, hybrid dual-labeled Rho-HA-PTX/CF-LTSL HNPs were prepared 1522

DOI: 10.1021/acsnano.7b08051 ACS Nano 2018, 12, 1519−1536

Article

ACS Nano

Figure 3. Cellular uptake of Rho-HA-PTX via CD44 mediation. (A) Cellular uptake of Rho-HA-PTX/CF-LTSL HNPs with a Rho/CF mass ratio of 1:1 in 4T1 cells with or without HT treatment. Extracellular release of HA-PTX was triggered by HT treatment at 42 °C (green, CF; red, Rho). Images were acquired by real-time CLSM. The scale bar is 10 μm. (B) Concentration (Rho)-dependent uptake in 4T1 cells was measured by flow cytometry. Cellular uptake in 4T1 cells with or without preincubation with 10 mg/mL HA: (C) flow cytometry analysis, (D) CLSM observation, and (E) concentration (Rho)-related uptake (mean ± SEM, n = 3, *P < 0.05). Rho-HA-PTX was incubated with the cells for 4 h at 37 °C at a Rho concentration of 500 ng/mL. In CLSM examination, the nuclei were stained by DAPI (red, Rho; blue, DAPI). The scale bar is 5 μm. (F) CLSM images of 4T1 cells after 1, 2, and 4 h incubations at 37 °C with Rho-HA-PTX (red, Rho, 500 ng/mL). The late endosomes and lysosomes were stained by Lyso-tracker green. The scale bar is 5 μm.

concentration and over time (Figure 3B and Supplementary Figure S5B). We then incubated Rho-HA-PTX with 4T1 cells at different temperatures. The fluorescence intensity of Rho-HAPTX was significantly higher at 37 °C than at 4 °C (Supplementary Figure S5C,D). These results suggested that HA-PTX was effectively delivered into 4T1 cells in a time-dependent and energyrelated manner. To illustrate the contribution of CD44 receptors in cellular uptake, Rho-HA-PTX was incubated with 4T1 cells with CD44 receptors saturated by pretreatment with an excessive amount of free HA. After incubation with pretreated cells for 4 h with a fixed concentration of Rho-HA-PTX, measurement by flow cytometry showed that the fluorescence intensity was significantly reduced compared to that of untreated cells (Figure 3C). Visually, CLSM observation also revealed that the red fluorescence (Rho-HA-PTX) distributed in the perinuclear areas of cells pretreated with HA was markedly weaker than that in untreated cells (Figure 3D). Additional quantification indicated that at high concentrations of Rho-HA-PTX, the cellular uptake of Rho-HA-PTX was approximately only 50% of the uptake in untreated cells (Figure 3E). These results confirmed

that a CD44-mediated endocytosis pathway driven by CD44-HA affinity played a critical role in the uptake of Rho-HA-PTX. To study the intracellular trafficking of HA-PTX, lysosomes and late endosomes were marked with Lyso-tracker green. Yellow spots in the merged images were visualized after the colocalization of red fluorescence (Rho-HA-PTX) and green fluorescence (endo/lysosomes) (Figure 3F), which demonstrated that Rho-HA-PTX was detained by lysosomes before accumulating in cytoplasm. As a result, Rho-HA-PTX obtained cellular entry through an endocytic pathway.36 Moreover, the cellular uptake was significantly reduced by chlorpromazine (Cpz), cytochalasin-D (Cyto-D), and NaN3 with deoxyglucose (NaN3+DG) (Supplementary Figure S6), implying that this uptake was an energy-dependent process and mediated by micropinocytosis and a clathrin pathway. In Vitro Cytotoxicity and Synergistic Effect. Cell viability in the presence of the drug-free carriers, HA-ADH and LTSLs, was approximately 100%, even at high concentrations (Supplementary Figure S7A,B). Therefore, these carriers are safe. HA-PTX exhibited more profound cytotoxicity against 1523

DOI: 10.1021/acsnano.7b08051 ACS Nano 2018, 12, 1519−1536

Article

ACS Nano

Figure 4. Cytotoxicity and apoptosis in tumor cells and inhibition of fibroblast activation. Cell viability was assessed by an MTT assay at 37 °C. HNPs with a PTX/MATT ratio of 1:1 were used in these experiments. Cytotoxicity of (A) PTX and HA-PTX, (B) MATT and MATTLTSLs, MATT-LTSLs+HT, (C) HA-PTX/MATT-LTSL HNPs, and HA-PTX/MATT-LTSL HNPs+HT in 4T1 cells after a 48 h incubation at different concentrations (mean ± SEM, n = 5, *P < 0.05, **P < 0.01). (D) Fluorescence images were taken with fluorescence microscopy, and the nuclei were stained with DAPI. The scale bar is 20 μm. Normal cells displayed homogeneous chromatin and showed spherical bright fluorescence without segmentation and fragmentation. In contrast, apoptotic cells exhibited nuclear condensation and chromatin pyknosis. (E) Apoptosis rate of 4T1 cells was determined by Annexin V-FITC/PI staining. The lower-left, lower-right, upper-right, and upper left quadrants represented the viable, early apoptotic, late apoptotic and dead cells, respectively. The incubation was performed at a PTX concentration of 10 μg/mL or a MATT concentration of 5 μg/mL for 48 h at 37 °C. (F) CLSM images of NIH/3T3 cells cocultured with 4T1 cells in a transwell device after a 24 h incubation with different formulations at a PTX concentration of 5 μg/mL or a MATT concentration of 2.5 μg/mL at 37 °C. The α-SMA was stained red with α-SMA antibody, and the nuclei were stained blue with DAPI. The scale bar is 20 μm. (G) Western blot analysis of TGF-β1 and TNC expression. GADPH was used as a loading control. Dark bands indicate the protein expression. Quantitative analysis of the expression of (H) TGF-β1 and (I) TNC (mean ± SEM, n = 3, *P < 0.05, **P < 0.01).

appeared to have no cytotoxicity at any tested concentration, whereas MATT-LTSLs displayed significant cytotoxicity at high MATT concentrations independent of HT treatment (Figure 4B).

4T1 cells than free PTX, as the PTX concentration ranged from 1 to 10 μg/mL, mainly due to the improved uptake of HA-PTX via CD44 mediation (Figure 4A). Surprisingly, free MATT 1524

DOI: 10.1021/acsnano.7b08051 ACS Nano 2018, 12, 1519−1536

Article

ACS Nano

Figure 5. Tumor targeting in vivo. (A) In vivo fluorescence imaging of the 4T1 tumor-bearing BALB/c mice after intravenous injection with free DiR, DiR-labeled LTSLs (DiR-LTSLs) and HA-PTX/DiR-LTSL HNPs at different time points. The dose of DiR was 0.5 mg/kg based on animal’s weight. (B) Ex vivo fluorescence images of important tissues collected from the mice 24 h after administration and (C) fluorescence intensity of the DiR signal (mean ± SEM, n = 3, **P < 0.01). Colocalization of (D) Rho-HA-PTX and CF-LTSLs in tumor sections using (E) free Rho/CF as control. Nuclei were stained with DAPI (green, CF; red, Rho; blue, DAPI). Yellow fluorescence indicates intact HNPs. Tumor sections were obtained from isolated tumor tissues harvested from mice 2 h after injection with Rho-HA-PTX/CF-LTSL HNPs. The scale bar is 20 μm (enlarged view, 10 μm). The intratumoral distribution of (F) CF-LTSLs, (G) FITC-HA-PTX (green), (H) FITC-HAPTX/LTSL HNPs, and (I) HA-PTX/CF-LTSL HNPs was observed by CLSM. The dose of the dyes was 0.5 mg/kg according to animal’s weight. The tumor tissues were isolated at 2 h after injection. Microvessels were stained red with Cy7-labeled CD31 antibody, and nuclei were stained blue with DAPI. The microvessels inside tumors exhibited red fluorescence, whereas fluorescent probe-labeled nanoparticles exhibit green fluorescence. Yellow fluorescence indicates the colocalization of nanoparticles with microvessels. The scale bar is 20 μm (enlarged view, 10 μm).

To examine the cytotoxicity of HNPs, we first evaluated the synergistic effects between HA-PTX and MATT by calculating the combination index (CI). As depicted in Supplementary Figure S8A,B, at high inhibition rate (Fa), the CI values from the HNPs with PTX/MATT ratios of 3:1, 2:1, 1:1, and 1:2 were less than 1 and therefore demonstrated the synergism between HA-PTX and MATT. However, at MATT concentrations ≥5 μg/mL, the ratios of 2:1, 1:1, and 1:2 exhibited higher toxicity against 4T1 cells than the 3:1 ratio (Supplementary Figure S8A), demonstrating promoted in vitro synergistic effect. Indeed, HNPs with PTX/MATT ratio of 1:1 showed greater toxicity to 4T1 cells than MATT-LTSLs or HA-PTX when the concentration of PTX was higher than 1 μg/mL (Figure 4C), which demonstrated synergistic cancer cell killing. Additionally, for HNPs, there was little impact of HT treatment on cytotoxicity at low drug concentrations (