Tumor microenvironment-responsive ultra-small ... - ACS Publications

USNPs (5 - 10 nm) from the initial parent NPs (140 nm) upon arriving at the mild ..... assembly process reported here shed light on rational design an...
0 downloads 0 Views 6MB Size
Article Cite This: J. Am. Chem. Soc. 2018, 140, 14980−14989

pubs.acs.org/JACS

Tumor Microenvironment-Responsive Ultrasmall Nanodrug Generators with Enhanced Tumor Delivery and Penetration Pengfei Zhang,† Junqing Wang,† Hu Chen,† Lei Zhao,† Binbin Chen,† Chengchao Chu,† Heng Liu,† Zainen Qin,† Jingyi Liu,† Yuanzhi Tan,‡ Xiaoyuan Chen,§ and Gang Liu*,†,‡

Downloaded via KAOHSIUNG MEDICAL UNIV on November 9, 2018 at 10:28:49 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.



State Key Laboratory of Molecular Vaccinology and Molecular Diagnostics & Center for Molecular Imaging and Translational Medicine, School of Public Health, Xiamen University, Xiamen 361102, China ‡ State Key Laboratory of Physical Chemistry of Solid Surfaces & The MOE Key Laboratory of Spectrochemical Analysis & Instrumentation, College of Chemistry and Chemical Engineering, Xiamen University, Xiamen 361005, China § Laboratory of Molecular Imaging and Nanomedicine, National Institute of Biomedical Imaging and Bioengineering (NIBIB), National Institutes of Health (NIH), Bethesda, Maryland 20892, United States S Supporting Information *

ABSTRACT: Tumor microenvironment-induced ultrasmall nanodrug generation (TMIUSNG) is an unprecedented approach to overcome the drug penetration barriers across complex biological systems, poor circulation stability and limited drug loading efficiency (DLE). Herein, a novel strategy was designed to synthesize metal−organic nanodrug complexes (MONCs) through supramolecular coassembly of photosensitizer sinoporphyrin sodium, chemotherapeutic drug doxorubicin and ferric ions. Compared with the free photosensitizer, MONCs produced 3-fold more reactive oxygen species (ROS) through the energy transfer-mediated fluorescence quenching. Remarkably, the self-delivering supramolecular MONCs with high DLE acted as a potent ultrasmall-nanodrug generator in response to the mild acidic tumor microenvironment to release ultrasmall nanodrugs (5− 10 nm in diameter) from larger parental nanoparticles (140 nm in diameter), which in turn enhanced the intratumor permeability and therapeutic efficacy. The key mechanism of MONC synthesis was proposed, and we, for the first time, validated the generation of supramolecular scaffold intermediates between MONCs’ assembly/disassembly states, as well as their involvement in multidrug ligands interactions. This proof-of-concept TMIUSNG strategy provides a foundation for the rational design of analogous carrier-free nanotheranostics through the combination of multiple therapeutic agents and metal ions with imaging functions.



INTRODUCTION

diameter, despite having a long circulation half-life, generally penetrate only into a few cellular layers in the vicinity of tumor blood vessels.11 The limited penetration of NPs is largely due to the dense extracellular matrix and high interstitial fluid pressure of tumor tissues, possibly resulting in modest clinical therapeutic outcomes.12 On the contrary, small NPs (4−20 nm) have deeper tumor infiltration capabilities and perfuse more homogeneously within tumor tissue, but they more likely suffer from rapid clearance and inadequate drug retention.12 To overcome these biological barriers, size-switchable nanovehicles13−17 have been developed with proper initial particulate diameters in systemic circulation for EPR, and then generate and release small-sized NPs that have been encapsulated into or conjugated onto initial parental NPs upon arrival at solid tumors. Whereas the reported microsized

Various drug delivery platforms based on nanoparticles (NPs) have emerged as promising tools for cancer therapies.1 However, systemic delivery of most therapeutic NPs fails to penetrate efficiently through the tumor interstitium, even by taking advantage of enhanced permeability and retention (EPR) or active targeting mechanisms. Together, drug leakage from nanovehicles in the systemic circulation and limited drug loading efficiency (DLE)2 have overall hampered the clinical success of nanomedicine.3,4 While simply increasing systemic dosages is not a rational solution due to intensified severe side effects, nanovehicles must overcome multiple biological barriers to achieve deep tumor infiltration and effective drug concentration at tumor sites.5 The failure of nanovehicles to infiltrate entirely into tumor tissue is the major cause of tumor metastasis or recurrence.6−8 It is well-known that the size and surface charges of nanomaterials play important roles in the transport of NPs in tumors.9,10 NPs that are 100−200 nm in © 2018 American Chemical Society

Received: August 30, 2018 Published: October 16, 2018 14980

DOI: 10.1021/jacs.8b09396 J. Am. Chem. Soc. 2018, 140, 14980−14989

Article

Journal of the American Chemical Society

Scheme 1. Self-Assembly and Redelivery of MONCs Served as Ultrasmall Nanodrug Generators in Response to TMEa

a

(a) Photosensitizer porphyrin DVDMS, ferric ions and DOX coassembled into supramolecular MONCs at a molar ratio of 1:6:10. (b) Proton concentration reduction in MONC reaction system allowed the MONCs to grow into large-sized cylindrical particles (200−400 nm) from optimal spherical morphology (a, 140 nm). (c) On arrival at TME through EPR effects, MONCs with high drug loading contents partly disassembled into ultrasmall NPs (7−10 nm) in response to TME due to its intrinsic acid sensitiveness. (d) The generation of USNPs triggered by TME would give rise to deep tumor infiltration, thus achieving significant tumor inhibition. Besides, fluorescence quenching-mediated energy transfer in MONCs was demonstrated to remarkably enhance ROS production.

NP generators18 that could disassemble into 30−80 nm sized particles have shown some promise, several challenges such as the sophisticated design, multistep synthesis, and stimuliresponsive linkers are required to achieve the size-switchable feature of these nanocarriers.19 In this regard, a new generation of ultrasmall NP (USNP) generators is highly desirable with high DLE, satisfactory circulation stability, facile production, self-delivering characteristics, and biocompatibility. To achieve high DLE, a promising strategy has emerged to exploit the drug itself as a major building unit to generate carrier-free nanostructures.20 Various drug nanostructures have been reported, such as nanodrugs that are formed by selfassembly21 of free small molecules via a dilution approach, amphiphilic macromolecules containing polymer-drug conjugates/complexes22−24 and supramolecular NPs composed of amphiphilic drug−drug conjugates.25 The use of extra synthetic carriers can be circumvented by this manner, thereby resulting in enhanced drug loading.26 Yet, the alternative approach using covalent conjugation of hydrophobic drug molecules to hydrophilic segments/polymers might compromise the release of drug payloads and increase production cost, thus possibly impeding clinical translation of these NPs. Therefore, it is of great promise to explore the key nanostructure as well as mechanisms for the design of carrier-free nanosystems that integrate multiple drug components via noncovalent interactions.27,28 In this study, to overcome the aforementioned issues, we successfully synthesized metal−organic nanodrug complexes (MONCs) as a carrier-free multidrug system, which possesses self-delivering characteristics with high DLE, intrinsic acid sensitivity, and favorable biocompatibility. The MONCs were constructed by coassembly of photosensitizer DVDMS29

(Figure S1), chemotherapeutic drug doxorubicin (DOX) and ferric ions in an aqueous solution (Scheme 1). With the aid of ferric ions, two types of drug molecules coassembled into uniform NPs with an extremely high DLE of 78%. These strong interactions among multiple drug ligands and metal ions could maintain MONC circulation stability and improve therapeutic outcomes via EPR effects. Furthermore, we found that the MONCs could be harnessed as tumor microenvironment (TME)-triggered USNP (5−10 nm) generators, which decomposed into USNPs (5−10 nm) from the initial parent NPs (140 nm) upon arriving at the mild acidic TME. Due to its initial large hydrodynamic diameter (140 nm), MONCs were able to take the advantage of EPR effects.30 This acid-sensitive decomposition of MONCs led to whole and deep tumor infiltration, thereby achieving potent tumor inhibition through combinational chemo- and phototherapy. Interestingly, when compared with equivalent free photosensitizer DVDMS, MONCs produced approximately 3fold more reactive oxygen species (ROS) in vitro under laser irradiation, mainly because of fluorescence quenching-mediated energy transfer in MONCs. Importantly, unlike the selfassembly of amphiphilic macromolecules, we, for the first time, demonstrated that MONC formation was mediated through metal−organic supramolecular scaffold intermediate (SSI) between the assembly and disassembly states of MONC, and it was triggered by solubility changes of DOX due to proton shifts. The proposed mechanism involved in the supramolecular coassembly of multiple drugs has great potential for the rational design of multifunctional carrier-free nanodrugs on the basis of clinically used chemotherapeutics. 14981

DOI: 10.1021/jacs.8b09396 J. Am. Chem. Soc. 2018, 140, 14980−14989

Article

Journal of the American Chemical Society

Figure 1. (a−d) TEM images of MONCs at different ratios of DVDMS/Fe3+/DOX. Scale bar: 200 nm. (e) Schematic depicting of MONCs containing different proton contents in the same reaction system (DVDMS/Fe3+/DOX mixture ratio of 1:6:10) by adding exogenous hydrochloric acid or sodium hydroxide. (f−h) To modulate the proton concentration in MONC formulation, exogenous hydrochloric acid or sodium hydroxide was added to raw MONCs that had been prepared without purification at a DVDMS/Fe3+/DOX molar ratio of 1:6:10. After isolation of these NPs, it was found that reduced proton contents in reaction system allowed the optimal MONCs (f) to grow into large-sized cylindrical particles (g, 200− 400 nm). On the other hand, primary MONCs (f) partly decomposed into supramolecular scaffold intermediate (h, SSI) along with increased proton concentration, which was further validated by ICP/HPLC analysis. Scale bar: 100 nm. (i) Mechanisms of MONC formation, mediated by deposition of deprotonated DOX onto SSI. (1) Hydrophilic DOX·HCl (pink) first approached the chain-like structures (yellow pane on g and h) that consist of DVDMS (gray)/ Fe3+ (yellow dot) complexes. (2) Then, proton transfer from DOX·HCl to carboxylate group of DVDMS (gray) would result into generation of hydrophobic DOX (pink) from hydrophilic molecules. (3) Solubility changes of DOX due to proton shift triggered DOX deposition on the chain-like structures from SSI, which further contributed to the formation of spherical MONCs, driven by hydrophobic cores on long chains (yellow pane on g and h).



RESULTS AND DISCUSSION Preparation and Characterization of MONCs. To prepare carrier-free MONCs, negatively charged photosensitizer DVDMS was mixed with positively charged ferric ions in an aqueous solution to form a supramolecular chain or plateletlike structures (Figure S2),31 followed by adding DOX·HCl to the solution under vigorous stirring. Subsequently, after rest for 10 h and ultracentrifugation, the resulting MONCs were facilely obtained. In the absence of ferric ions, the mixing of protonated DOX with DVDMS in deionized water gave rise to a massive precipitation. This precipitate was not reduced after further addition of exogenous ferric ions to the DVDMS/DOX deposits, which thereby validates the importance in different adding order of reagents for MONC preparation. Furthermore, the ratios of these building ligands were shown to remarkably affect the morphology, size, and uniformity of MONCs. Through numerous attempts to test a series of ratios of formulation, we finally prepared optimal MONCs by coassembly of DVDMS, Fe3+, and DOX at a molar ratio of

1:6:10, which resulted in extremely high drug loading efficiency (DLE) (78%), low polydispersity (0.08−0.14), and hydrodynamic size ranging from 80 to 210 nm in diameter (Figure S3). Moreover, high-pressure liquid chromatography (HPLC) analysis proved the components of DVDMS/DOX in the purified MONCs with a molar ratio of 1:9.3, while almost all of the Fe3+ (99.2 ± 0.7%) were incorporated into MONCs as shown by inductively coupled plasma mass spectrometry (ICP-MS) assay. The purified MONC products in various ratios of formulations were analyzed by HPLC, ICP-MS, energy dispersive spectrometer (EDS) and X-ray photoelectron spectroscopy (XPS) (Figure S4, S5) to confirm the constituents of MONCs. Interestingly, by gradually increasing the DOX ratio in the metal-DVDMS mixtures, the size distribution of these supramolecules dramatically increased (Figure 1) and MONCs were transformed into dispersed spherical NPs (Figure 1c,d) from aggregated strip-like structures (Figure 1a,b) accordingly. These results showed that DOX incorporation into NPs indeed played a vital role in 14982

DOI: 10.1021/jacs.8b09396 J. Am. Chem. Soc. 2018, 140, 14980−14989

Article

Journal of the American Chemical Society

Figure 2. (a) Fluorescence quenching of DVDMS by Fe3+ or DOX, indicating the complex interactions among these building drug ligands. (b) In vitro 1O2 detection of different equivalent formulations under 630 nm laser irradiation by measuring SOSG FL intensity. (c) Intracellular ROS detection by DCFH-DA probe. Cumulative DOX release of MONCs in different pH conditions via dialysis with 100 kDa (d) of molecular weight cutoff (MWCO) or 10 kDa (e) of MWCO. (f) USNPs were generated by disassembly of nanodrug complex MONCs, and then passed through the dialysis membranes with 100 kDa MWCO rather than ones with 10 kDa of MWCO. In PBS at pH = 4.5, which mimicked acid environments in lysosomes, MONCs could release free DOX as well as DVDMS that were leaked through 10 kDa MWCO of dialysis membranes. (g) TEM images and (h) DLS analysis of MONCs after incubation in PBS at pH = 6.5. (i) MONCs stability in PBS or PBS plus 10% FBS by DLS analysis.

MONC formation, suggesting that intermolecular interactions between DOX and metal−DVDMS complexes effectively improved the drug loading capability, dispersity, and integrity of the MONCs. Critical Mechanism for MONC Formation. To explore the critical mechanism for MONC formation, we investigated the impact of proton concentrations on the morphology of MONCs by adding hydrochloric acid or sodium hydroxide (0.2 μM) to raw MONCs that had been prepared without further purification (DVDMS/Fe3+/DOX at molar ratio of 1:6:10) (Figure 1e−h). The resulting products containing different levels of protons were separated and observed with TEM imaging. As expected, reducing the proton content in the reaction systems drove the optimal MONCs (Figure 1f) to grow into large-sized cylindrical particles (Figure 1g, 200−400 nm), which might be ascribed to the further deposition of deprotonated DOX into the metal-DVDMS complexes. In contrast, in response to increased proton concentration, primarily spherical MONCs (Figure 1f) disassembled into interlaced net frameworks (Figure 1h), namely metal−organic supramolecular scaffold intermediates (SSIs). ICP-MS/HPLC demonstrated that the separated SSI structures were comprised of massive metal-DVDMS complexes and a small amount of DOX (DVDMS/Fe3+/DOX at molar ratio of 1:1.8−2.3:0.26− 0.41). Therefore, SSI formation was observed in a partial disintegration of MONCs with protonated DOX released under acidic condition. To further establish the presence of SSI

nanostructure in MONCs’ formation process, SSIs were successfully observed in the mixture of DVDMS, Fe3+ ions and DOX with a molar ratio of 1:2:0.3, which suggested that the relatively low ratio of building ligands were sufficient to generate SSI at initial growth stage (Figure S6). Notably, when exogenous HCl in the reactive system was neutralized by equivalent NaOH, the SSI nanostructures (Figure S7a), which were formed because of the addition of HCl to MONCs, were reversely transformed into large-sized parent MONCs (Figure S7b), indicating the critical role of proton content in SSImediated NP formation. Also, we found that both the decomposition of MONCs into SSI and MONCs’ growth into large-sized cylindrical particles were mediated by the long chain structures from SSI (yellow panel, Figure 1g,h). The SSImediated mechanism (Figure 1i) for MONC formation was further validated, based on distinct structural morphologies caused by different levels of proton content in the MONCs. Taken together, the SSI nanostructure indeed plays a framework-like role in the coassembly of multidrug ligands. On the basis of the experimental results in this work, we proposed and validated the key mechanisms of MONC formation, wherein negatively charged photosensitizer DVDMS, a symmetric biporphyrin structure, was initially associated with positively charged ferric ions via electrostatic attraction to form a supramolecular chain or platelet-like structure, similar to our previous report.31 Then, proton shift from DOX·HCl to carboxylate groups of DVDMS occurred 14983

DOI: 10.1021/jacs.8b09396 J. Am. Chem. Soc. 2018, 140, 14980−14989

Article

Journal of the American Chemical Society

Figure 3. (a) In vitro penetration of MONCs in MCF-7 multicellular spheroids (MCSs) 6 h postincubation in PBS at different pH using CLSM imaging. Scale bar: 50 μm. (b) Cellular uptake of MONCs at 3 h after incubation with MCF-7 cells. (c) The magnified images from merged b, showing that large-sized MONCs adsorbed to cellular plasma membrane and only a small fraction could infiltrate into cellular interior under physiology condition. (d) The mean DOX fluorescence intensity of treated cells with MONCs in pH 6.5 or pH 7.4 condition. (e) DOX availability via MONC delivery under different pH conditions, verified by cell viability detection at 48 h after 5 h incubation of MCF-7 cells with MONCs. (f) MTT assay was performed after incubation of MCF-7 cells with MONCs (630 laser irradiation, 100 mW/cm2 for 5 min) in cell culture medium.

and simultaneously hydrophilic protonated DOX was transformed into hydrophobic molecules, leading to strong hydrophobic interplay among the multiring structures of DOX and DVDMS. In this way, deposition of deprotonated DOX onto the SSI would facilitate the formation of hydrophobic sites that were scattered on the SSI. Ultimately, the metal−organic SSI, which mainly consisted of Fe-DVDMS complexes, was closely compacted into spherical NPs, driven by the DOX hydrophobic sites on the long chains, which in turn afforded the MONCs ultrahigh drug loading capacity.33 Notably, for the first time, the concept of metal−organic SSI between the assembly and disassembly states of MONC was introduced, based on TEM and DLS results. The metal− organic SSI structure indeed played a framework-like role in the formation of the self-delivering nanodrugs that combined multiple drug building ligands. On the other hand, solubility changes of DOX due to proton shifts were also considered to be a vital trigger for DOX deposition onto the SSI. It is believed that MONC coassembly is mediated by complex interactions among DVDMS, DOX, and ferric ions. The electrostatic interaction between the positively charged DOX and the negatively charged carboxylic acid, as well as the π−π stacking between the aromatic groups could be the major driving force for the MONC formation. These strong interactions are likely the key contributors to the intrinsic

acid sensitivity, favorable colloidal stability, and extremely high drug loading efficiency (DLE 78%) of MONCs. In Vitro ROS Generation and Nanodrug Release from MONC. The interplay among metal-drug building ligands was further evaluated by fluorescence measurements. Figure 2a shows that mixing DVDMS with Fe3+ or DOX caused a remarkable quenching, largely owing to strong intermolecular interactions. MONC coassembly was mediated by complex interactions among DVDMS, DOX, and Fe3+. These strong interactions, including electrostatic attraction and π−π stacking between DVDMS and DOX, played important roles in MONC stability in blood circulation. It is worth noting that we have demonstrated that DVDMS is an effective photosensitizer to produce ROS under 630 nm laser irradiation for tumor therapy.31,32 Next, we used Singlet Oxygen Sensor Green (SOSG) to assess the 1O2 generation of MONCs in deionized water under laser irradiation (Figure 2b, 100 mW/cm2 for 5 min). Intracellular ROS production of MONCs was detected by measuring the fluorescence intensity of DCFH-DA probes (Figure 2c). Interestingly, in contrast to the equivalent photosensitizer DVDMS, MONCs could generate approximately 3- and 2-fold-higher levels of reactive oxygen species (ROS), in vitro and in vivo respectively, which may be the result of fluorescent quenching-triggered energy transfer. It was conceivable that both the increase in the MONCs’ ROS levels 14984

DOI: 10.1021/jacs.8b09396 J. Am. Chem. Soc. 2018, 140, 14980−14989

Article

Journal of the American Chemical Society

Figure 4. (a) T1-weighted MR images and (b) MR signal intensities of tumors (white circles) in mice injected with 6 mg/kg of MONCs via tail veins. (c) FL imaging and (d) average FL radiant efficiency of DOX in vital mouse organs and tumors at 24 or 48 h following MONC treatment (K, Lu, S, Li, H and T are representatives for kidney, lung, spleen, liver, heart and tumor tissues, respectively). (e,f) 48 h following i.v. injection, the excised tumors (about volume of 90 mm3) were developed into frost slices, and then the DOX fluorescence distribution in entire tumors was revealed by CLSM observation (e, scale bar: 100 μm) and fluorescence microscope (f, scale bar: 200 μm), respectively. (g,h) HPLC quantitative analysis of DOX concentrations in (g) whole MCF-7 tumors or in (h) central interstitial regions (100 mm3) of tumors in mice following a single i.v. injection of DOXIL, free DOX and MONCs. (i) Penetration indexes (P-index) of MONCs were acquired through dividing DOX concentration (μg/g tissue) in central interstitial regions (100 mm3) of tumors by DOX concentration (μg/g tissue) in whole tumors. (j) Tumor growth profiles, (k) body weight changes and (l) representative hematoxylin and eosin (H&E) staining images of tumors in MCF-7 tumor xenograft mouse model. (1) PBS, (2) DOX + DVDMS, (3) DOX + DVDMS + laser, (4) DOXIL, (5) MONCs, (6) MONCs + laser.

and effective energy transfer indeed resulted from its unique nanostructure. Next, to evaluate the ability of MONCs to release drugs in an acid-responsive manner, the DOX release in different pH of phosphate buffered saline (PBS) was analyzed by HPLC. The release ratios of DOX via dialysis with a 100 kDa cutoff were much higher at pH 4.5 and pH 6.5 than at pH 7.4 (Figure 2d). Moreover, the DOX release ratio at pH 7.4 reached 13.6% in 120 h, indicating the favorable integrity of MONCs in PBS. However, the DOX release of MONCs via dialysis with 10 kDa cutoff was negligible at pH 6.5 (Figure 2e). The distinct DOX release profiles were found by dialysis at different molecular weight cutoffs (MWCO), so we postulated that MONCs could partially disassemble into ultrasmall NPs (5−10 nm) in PBS at pH 6.5, while free drugs, DOX and DVDMS, were completely

released in PBS at pH 4.5 and thereby passed through the dialysis bags with 10 kDa MWCO (Figure 2f). To validate this hypothesis, TEM images (Figure 2g) directly proved the generation process of USNPs on MONC surfaces in mildly acidic conditions. Also, as shown by DLS analysis in Figure 2h and Figure S8, USNPs (5−10 nm) were demonstrated to be produced due to the disassociation of large-sized parental NPs in the pH 6.5 solution. These USNPs were separated by size exclusion chromatography (SEC) and then analyzed using HPLC. Finally, it was established that USNPs were constituted from DVDMS/DOX/Fe3+ with a molar ratio of 1:17.4:2.1. Next, we assessed the DOX retention (Figure S9) in NPs by ultrafiltration assays with a 100 kDa cutoff (pH 7.4, 10% FBS). As shown in Figure S9, MONCs exhibited about 90% DOX preservation after 20 days, indicating the effective prevention 14985

DOI: 10.1021/jacs.8b09396 J. Am. Chem. Soc. 2018, 140, 14980−14989

Article

Journal of the American Chemical Society

5 μg/mL). As a control, MCF-7 cells treated with pH 6.5 PBS for 5 h did not exhibit observable cytotoxicity. To examine the combined therapy efficacy (Figure 3f) of MONCs, we added different formulations to cell culture wells, incubated them for 8 h, and then exposed the MCF-7 cells to 630 nm laser irradiation for 5 min (100 mW/cm2). At 32 h after laser irradiation, the cellular survival percent was measured using MTT assays. In contrast to the group treated with free DOX/ DVDMS plus laser, treatment of MCF-7 cells with MONCs plus laser caused more cell death at low drug concentration (1 μg/mL DOX), which was likely due to the energy-transfermediated enhancement of ROS production. In Vivo MR and Fluorescence Imaging of MONC. Given its biological safety and T1-weighted magnetic resonance (MR) imaging capabilities, ferric ions were chosen to be incorporated into the supramolecular drug nanostructures. In vitro MR imaging abilities of MONCs were determined and its r1 proton relaxivity reached 2.309 in the 1.5 T magnetic field (Figure S10). Although the r1 value of the MONCs was smaller than that of commercially available MR imaging contrast agents, the MONCs showed their potential as theranostic platforms for optical/MR imaging-guided chemo-photo combination therapy. We also monitored in vivo MR imaging effects of MONCs using animal MR imaging equipment with 9.4 T of magnetic field strength. The MCF-7-tumor-bearing mice were injected with MONC nanoparticles (6 mg/kg) through their tail veins. T1-weighted MR images (Figure 4a) were acquired at different time points postadministration. Signal intensity of T1-weighted imaging (T1-WI) gradually increased within 24 h and then plateaued (Figure 4b), suggesting effective accumulation of MONCs in tumor regions through EPR effects. The treated mice were sacrificed at 24 or 48 h postinjection, and their major organs were used to investigate the biodistribution of MONCs by FL imaging (Figure 4c). The average DOX fluorescence intensity of major organs as well as tumors was quantified (Figure 4d). In this manner, we verified that MONCs were accumulated largely in mouse tumors via passive targeting delivery, which is consistent with previous MR results. In Vivo Infiltration of MONC into Tumors. To further reveal the accurate distribution of MONCs and DOXIL within tumors, another batch of tumor-bearing mouse model was established. When tumors grew up to about 90 mm3 in volume, various formulations were injected into mice via tail veins. Subsequently, at 48 h postinjection, the tumors were developed into frost slices to monitor the DOX fluorescence distribution in entire tumors by CLSM (Figure 4e) and fluorescence (Figure 4f) imaging. Figure 4e,f shows the distinct location of DOXIL and MONCs in the entire tumor section. Abundant DOX signals in the MONC-treated group were evenly distributed throughout the tumors, whereas there were a few bright DOXIL spots sporadically located in the tumor tissues in the DOXIL-treated group. Next, to quantitatively compare the abilities of MONCs with DOXIL to penetrate the whole tumors, when the tumor volumes reached about 700 mm3, the tumor-bearing mice were i.v. injected with DOX, DOXIL, or MONCs containing 6 mg/ kg equivalent dose of DOX, then sacrificed at 24, 48, or 72 h after a single i.v. injection. Finally, either the centrally interstitial regions (100 mm3) of tumors or whole tumors in treated mice was dissected and isolated for DOX quantification by HPLC.34 Quantitative analysis of DOX concentrations in whole MCF-7 tumors (Figure 4g) or in centrally interstitial

of drug leakage during blood circulation. Furthermore, according to size measurements by DLS (Figure 2i), we demonstrated that MONCs maintained good polydispersity and hydrodynamic sizes over a period of 14 days in PBS and FBS-containing PBS. These data suggested that MONCs possessed favorable stabilities in the presence of serum, thus ensuring its structural integrity during blood circulation. This is a prerequisite for high drug accumulation in tumors through EPR effect. As some strong interplays, including electrostatic attraction in SSIs, metal−organic coordination and π−π stacking between the aromatic groups in USNP, played critical roles in MONC integrities, we suppose that the proton shift may disrupt metal−organic coordination as well as electrostatic interaction in formed SSIs, thus resulting in partial SSI dissociation and release of USNPs. DOX has been known to have an efficient coordination with ferric ions (Fe3+). Given the extremely high proportion of DOX in USNP, the USNP formation is likely due to π−π stacking among DOX and metal−organic coordination. In addition, we speculate that the proton shift from mild acidic environment to MONC is just a trigger to partially change the MONC structure. Phosphate anions or other anions may further facilitate the complete release of USNP, acting as competitors for binding Fe3+ ions. Next, we utilized a multicellular spheroids (MCSs) model to evaluate the in vitro penetration abilities of MONCs. The nanomaterials were incubated with MCSs for 6 h in different pH of PBS to monitor its infiltration depth using confocal laser scanning microscope (CLSM) imaging (Figure 3a). DOXIL, a PEGylated liposomal DOX, has already been extensively applied to clinical cancer treatment as a typical nanomedicine. After DOXIL treatment, strong DOX fluorescence was found only in the periphery of the MCS. In contrast, the red fluorescence signal was detected in the internal area of MCSs in our MONC-treated groups at pH 6.5. We next investigated the cellular uptake and intracellular localization of MONCs to evaluate its ability to penetrate cancer cells under different pH conditions (Figure 3b,c). Cellular uptake of MONCs by MCF7 cells was verified by probing the fluorescence (FL) signals from DOX. Following incubation with MONCs in PBS at different pH for 3 h, the MCF-7 cells were imaged using a CLSM. It is of note that most of the large-sized MONCs adsorbed to cellular surfaces and only a small fraction entered into tumor cells under physiological condition (Figure 3c), whereas strong DOX fluorescence was observed in the cytoplasm and nuclei in the mildly acidic PBS-treated group (Figure 3c). A second release of free DOX from USNP in endosomes (pH 4−5) may contribute to the effective drug infiltration into cellular nuclei (Figure 3c). The mean fluorescence intensity of the treated cells in the pH 6.5 medium was 2.3-fold higher than that in the pH 7.4 solution (Figure 3d), indicating that MONCs tend to accumulate into cellular interiors through acid-responsive USNP release. Subsequently, to further investigate the DOX availability (Figure 3e) from MONC delivery under different pH conditions, we added different contents of MONCs (1, 3, or 5 μg/mL) to 96-well plates, incubated them with MCF-7 cells for 5 h in PBS at pH 6.5 or pH 7.4, then replaced the PBS with standard, fresh growth medium and finally conducted an MTT assay 48 h after PBS replacement. Figure 3e shows that cell viabilities in the pH 6.5 PBS-treated groups was significantly lower than that in the pH 7.4 PBS-treated groups, which suggested the efficient delivery of DOX by MONCs under mildly acid conditions (at the MONC concentrations of 3 and 14986

DOI: 10.1021/jacs.8b09396 J. Am. Chem. Soc. 2018, 140, 14980−14989

Journal of the American Chemical Society

Article



regions (Figure 4h) (100 mm3) of tumors enabled us to introduce a new parameter, penetration index, to objectively evaluate the ability of NPs to infiltrate into tumor tissues. Penetration indexes (Figure 4i, P-index) of DOXIL, free DOX, and MONCs were calculated by dividing DOX concentration (μg/g tissue) in the centrally interstitial regions (Figure 4h, 100 mm3) of the tumors by the DOX concentration (μg/g tissue) in the whole tumors (Figure 4g). Combinational Therapy In Vivo. Encouraged by the results of the MR/FL imaging, we investigated the in vivo anticancer efficacy of MONCs, PBS, free DOX plus DVDMS and DOXIL. Various formulations containing an equivalent dose of DOX (3 mg/kg) were intravenously administrated into female MCF-7-xenografted mice on days 0, 3, 6, 9, and 12. At 24 h postadministration, the treated mice were subjected to 630 nm laser irradiation for 10 min (200 mW/cm2). Figure 4j showed that both DOXIL and DOX exhibited similarly suboptimal tumor inhibition efficacy. Importantly, the therapeutic efficacy of the MONC-treated group was comparable to that of the mice treated with free DOX and DVDMS plus laser, while the treatment of MONCs plus laser achieved optimal tumor inhibition outcomes (Figure 4j). In addition, we observed that free DVDMS plus DOX administration caused a sharp body weight reduction, compared with the other groups (Figure 4k). On day 20 following the first treatment, the tumors and major organs of the treated mice were collected for hematoxylin and eosin (H&E) staining (Figure 4l). Furthermore, the DOX contents in the plasma were also analyzed using HPLC (Figure S11). As shown by Figure S12, no obvious kidney toxicities were observed in MONC-treated groups relative to the control group. Although MONC could lead to slight liver injury and inhibition effects against lymphocytes, the liver functions as well as total lymphocyte amounts were rapidly recovered once the drug administration ended. All data demonstrated that MONCs have a longer half-life in circulating blood, enhanced passive accumulation at tumor sites, and effective whole tumor penetration, thereby resulting in satisfying therapeutic outcomes as well as reduced adverse toxicity.5 The TME has been recognized as desirable targets for the design of theranostics against cancer. In this study, we demonstrated the acid-responsive MONCs that caused USNPs to be released in the TME, possess powerful abilities to penetrate into whole tumor regions. It also represented a promising strategy since the TME-responsive nanodrugs can be tailored and endowed with specific theranostic functionalities through incorporation of multiple drug components into one carrier-free NP. Our study also established a theoretical foundation to rationally design analogous multidrug nanostructures through diverse metal-drug ligands. Importantly, many of the clinical anticancer drug molecules (e.g., camptothecin) are composed of two moieties, one containing amino, carbonyl, or hydroxyl groups that enable its coordination to metal ions, and the other segment consisting of multiple ring structures that contribute to their hydrophobic interactions. Given these structural characteristics, this strategy is readily available by utilizing multiple drug ligands and metal ions as building units for the construction of carrier-free nanodrugs.35 It also avoids the use of nonclinically proven materials, allowing facile large-scale production of nanodrugs and thus facilitating their clinical translation.

CONCLUSION In conclusion, we reported a self-delivering supramolecular MONC fabricated by the coassembly of multiple metal-drug ligands acting as an USNP generator to produce ultrasmall nanodrugs (5−10 nm) in response to the mildly acidic TME. Due to the relatively long half-life in the systemic circulation, supramolecular MONCs with proper initial size accumulate at tumor site through EPR effect. Subsequently, MONCs decompose into numerous ultrasmall-sized NPs with enhanced penetration to reach deep-seated tumor cells. The multidrug coassembly process reported here shed light on rational design and controlled fabrication of analogous carrier-free nanoplatforms, showing great potential for the development of TMEresponsive nanodrugs.



EXPERIMENTAL SECTION

Materials. Sinoporphyrin sodium, referred to as DVDMS (Molecular weight: 1230), were synthesized and provided by Qinglong Hi-tech Co, Ltd. (China). Doxorubicin hydrochloride (DOX·HCl) was purchased from TCI Corporation. 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), and 2′,7′dichlorofluorescein diacetate (DCFH-DA) were purchased from Sigma Corporation and Beyotime (China), respectively. Preparation and Characterization of MONCs. Ten mM of DVDMS was first mixed with 60 mM of ferric ions in 100 mL of deionized water for 1 h. Then, 100 mM of DOX·HCl was quickly added into the above mixture under vigorous stirring at 28 °C. Next, after quiet placement for 10−15 h, the MONC NPs were formed from coassembly of DVDMS, ferric ions and DOX·HCl. Finally, the resulting MONCs were separated by centrifugation at 21 000g for 20 min and stored at −20 °C after freeze-drying. Drug loading efficiency (DLE) was measured according to the following formula: DLE (%) = (weight of loaded DOX/weight of MONCs) × 100%. The weight of loaded DOX in NPs were determined by HPLC analysis of DOX content in suspension following centrifugation, while the weight of MONCs were also examined after lyophilization. The morphology and size distribution of MONCs were monitored with TEM and DLS instrument (Malvern). In vitro fluorescent imaging capabilities and ultraviolet−visible spectra (UV−vis) of MONCs were respectively examined with animal optical imaging system (IVIS Lumina II) and microplate reader (Thermo Scientific). Drug Release from NPs in Vitro. The MONCs (1.5 mL, 10 mg/ mL) in dialysis bags with different cutoff (MWCO: 100 000 Da or MWCO: 10 000 Da) were immersed in 80 mL of PBS at pH 7.4, 6.5 or 4.5, respectively. The solution outside the dialysis bags was acquired at predetermined times. Released DOX from MONCs was added into dilute ammonium hydroxide, then extracted with organic solvent and quantified by HPLC analysis. The ferric ion content in MONCs was measured by ICP. Singlet Oxygen (1O2) Detection of MONCs in Vitro. The capabilities of MONCs to produce singlet oxygen in deionized water under laser irradiation were assessed using 1O2 sensor SOSG probes. Briefly, 1 μL of SOSG (1 mM) was added into 2 mL of predetermined formulations containing equal contents of DVDMS (10 μg/mL). Subsequently, fluorescent emitting spectra of different formulations were monitored after a 630 nm laser irradiation (100 mW/cm2) for 5 min. The increase of SOSG FL intensity at 520 nm of emitting wavelength corresponded to the production of singlet oxygen. ROS Measurement of MONCs at Cellular Levels. ROS production in MCF-7 cells was measured via DCFH-DA probes (Beyotime). 5 × 105 of MCF-7 cells in dishes were incubated with 20 μg/mL of MONCs or equivalent free DVDMS for 12 h, followed by loading of DCFH-DA into the treated cells. Subsequently, the MCF-7 cells were washed twice and irradiated with 630 nm laser for 5 min (0.1 W/cm2), then collected following trypsin digestion, and finally its fluorescent spectra was detected by fluorescent spectrophotometer (Thermo Scientific). 14987

DOI: 10.1021/jacs.8b09396 J. Am. Chem. Soc. 2018, 140, 14980−14989

Article

Journal of the American Chemical Society Culture of MCF-7 Multicellular Spheroids (MCSs). Briefly, a 6well plate was covered by 0.5 mL of melted agarose liquid (1.5 w/v%) and cooled at room temperature to produce a layer of agaropectin on its surface. 3 × 105 of MCF-7 cells were seeded in the treated culture wells containing complete DMEM growth medium and cultured for 12 days to grow into cellular spheroids. During MCS culture period, this suspension was gently mixed using pipettors with specification of 1000 μL every 3 days. Infiltration of MONCs into MCSs Using CLSM. To examine the abilities of MONCs to penetrate into MCSs, MCSs were incubated with MONCs or Doxil containing equal dose of DOX (3 μg/mL) in the glass-bottom culture dished (NEST) for predetermined times and then observed with a confocal laser scanning microscope after centrifugation (CLSM, Zeiss, LSM 780). DOX Availability at Cellular Levels. We detected DOX availability via MONC delivery under different pH conditions by MTT assays. Briefly, we added 10 μg/mL of MONCs to cell culture wells, incubated them with MCF-7 cells in PBS at pH 6.5 or 7.4 for 5 h, then replaced the PBS with normal fresh growth medium and finally performed MTT assays to detect cell viability 48 h after PBS replacement, according to a standard protocol. MR and Fluorescence Imaging. All animal experiments were carried out according to the criteria of Institutional Animal Care and Use Committee (IACUC) at Xiamen University and approved by IACUC. 5 × 106 of MCF-7 cells mixed with matrigel (BD) were subcutaneously injected into female BALB/c nude mice (7−8 week). For MR imaging, different concentrations of MONCs were dispersed in the hot agarose liquid (0.9 w/v %), and then the T1 relaxivity was detected. T1 relaxation rates under 1.5 T field were plotted against the ferric ion contents. The relaxivity of MONCs was finally acquired using a linear fit analysis. BALB/c nude mice bearing MCF-7 tumors were intravenously administrated with 6 mg/kg of MONCs. Subsequently, the treated mice were further monitored with animal MR imaging instrument (9.4 T, Bruker Biospin) at predetermined times by utilizing a T1-weighted sequence, MESE. The parameters of MESE sequence was as followed: FOV (field of view) = 4.0 cm × 4.0 cm, TR (repetition time)/TE (echo time) = 3000/10 ms, Matrix = 256 × 256, Average times = 1, Slice Thickness = 1 mm. To further confirm the in vivo biodistribution of MONCs, we euthanized the treated mice 24 or 48 h postinjection, then took out the major organs as well as tumors and monitored DOX fluorescence with animal optical imaging system (IVIS Lumina II). Combinational Therapy in Vivo. For in vivo chemo- and photodynamic combinational therapy, female nude mice bearing about 90 mm3 of MCF-7 tumors were randomly divided into groups. Mice in each group have similar body weights. Various materials containing equivalent 3 mg/kg dose of DOX were administrated into MCF-7-xenografted mice via tail vein on day 0, 3, 6, 9, and 12 after the first injection. At 24 h following injection, the tumor of treated mice was irradiated with 630 nm laser for 10 min at 200 mW/cm2. The tumor volumes were daily recorded and calculated according to the following formula: Tumor volume = W2 × L/2, in which L and W are respectively the longest and the shortest diameters measured by a caliper. On 20 days following the first treatment, the tumors and major organs of treated mice were collected for hematoxylin and eosin (H&E) staining. Quantification of DOX in Tumor by HPLC Analysis. Briefly, nude mice bearing 700 mm3 of MCF-7 tumors were intravenously received a single dose of MONCs, free DOX·HCl and DOXIL (equivalent 6 mg/kg of DOX) (n = 5/each group). At 24, 48 and 72 h after i.v. injection, the treated mice were put to death. Next, 100 mm3 of central interstitial regions in tumors or whole tumors was dissected and homogenized in phosphate buffered saline together with 0.1 M NaOH and SDS at 4 °C. To extract DOX from tumor tissues, 1 mL of tissue homogenate was mixed with 3 mL of chloroform−methanol (4:1, v/v) solvent for 35 min at 4 °C. The resulting mixture was further centrifuged at 22 000g (4 °C, 30 min) in order to acquire organic phases. The organic phases were collected, then treated with freeze-drying and finally dissolved in 50 to 150 μL of methanol. The prepared samples were further analyzed by HPLC.

Pharmacokinetics and half-life time of MONCs were investigated through detecting the DOX content in blood. Briefly, the blood of mouse was collected from tail vein at different time points after intravenous administration at a DOX dose of 6 mg/kg. Then, the blood was centrifuged at 10 000g for 15 min to acquire plasma, and the plasma was further treated in the same way as tumor tissues for HPLC analysis. Statistical Analysis. Data analysis in the work was carried out using the GraphPad Prism software. P < 0.05 was considered to be statistical significance. The P values were obtained utilizing a twotailed unpaired t test. All data were exhibited via mean ± SD.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.8b09396. TEM, DLS, EDS, XPS, R1 proton relaxivity and circulation half-life of MONC (PDF)



AUTHOR INFORMATION

Corresponding Author

*[email protected] ORCID

Heng Liu: 0000-0002-1238-0683 Gang Liu: 0000-0003-2613-7286 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The work was supported by the National Key Research and Development Program of China (2018YFA0107301 and 2017YFA0205201), the National Natural Science Foundation of China (81422023, 81603015, 81871404, U1705281, and U1505221), the Fundamental Research Funds for the Central Universities (20720160065 and 20720150141), and the Program for New Century Excellent Talents in University, China (NCET-13-0502). All animal experiments were approved by the Animal Management and Ethics Committee of Xiamen University.



REFERENCES

(1) Davis, M. E.; Chen, Z. G.; Shin, D. M. Nanoparticle therapeutics: an emerging treatment modality for cancer. Nat. Rev. Drug Discovery 2008, 7, 771. (2) Das, A.; Wadhwa, S.; Srivastava, A. K. Cross-linked guar gum hydrogel discs for colon-specific delivery of ibuprofen: formulation and in vitro evaluation. Drug Delivery 2006, 13, 139. (3) Li, H. J.; Du, J. Z.; Liu, J.; Du, X. J.; Shen, S.; Zhu, Y. H.; Wang, X.; Ye, X.; Nie, S.; Wang, J. Smart superstructures with ultrahigh pHsensitivity for targeting acidic tumor microenvironment: instantaneous size switching and improved tumor penetration. ACS Nano 2016, 10, 6753. (4) Wang, J.; Tao, W.; Chen, X.; Farokhzad, O. C.; Liu, G. Emerging advances in nanotheranostics with intelligent bioresponsive systems. Theranostics 2017, 7, 3915. (5) Chauhan, V. P.; Jain, R. K. Strategies for advancing cancer nanomedicine. Nat. Mater. 2013, 12, 958. (6) Minchinton, A. I.; Tannock, I. F. Drug penetration in solid tumours. Nat. Rev. Cancer 2006, 6, 583. (7) Chen, H.; Zhang, W.; Zhu, G.; Xie, J.; Chen, X. Rethinking cancer nanotheranostics. Nat. Rev. Mater. 2017, 2, 17024. (8) Stapleton, S.; Jaffray, D.; Milosevic, M. Radiation effects on the tumor microenvironment: Implications for nanomedicine delivery. Adv. Drug Delivery Rev. 2017, 109, 119.

14988

DOI: 10.1021/jacs.8b09396 J. Am. Chem. Soc. 2018, 140, 14980−14989

Article

Journal of the American Chemical Society (9) Wang, J.; Mao, W.; Lock, L. L.; Tang, J.; Sui, M.; Sun, W.; Cui, H.; Xu, D.; Shen, Y. The role of micelle size in tumor accumulation, penetration, and treatment. ACS Nano 2015, 9, 7195. (10) Tang, L.; Yang, X.; Yin, Q.; Cai, K.; Wang, H.; Chaudhury, I.; Yao, C.; Zhou, Q.; Kwon, M.; Hartman, J. A.; Dobrucki, I. T.; Dobrucki, L. W.; Borst, L. B.; Lezmi, S.; Helferich, W. G.; Ferguson, A. L.; Fan, T. M. Investigating the optimal size of anticancer nanomedicine. Proc. Natl. Acad. Sci. U. S. A. 2014, 111, 15344. (11) Waite, C. L.; Roth, C. M. Nanoscale drug delivery systems for enhanced drug penetration into solid tumors: current progress and opportunities. Crit Rev. Biomed Eng. 2012, 40, 21. (12) Jain, R. K.; Stylianopoulos, T. Delivering nanomedicine to solid tumors. Nat. Rev. Clin. Oncol. 2010, 7, 653. (13) Zhao, T.; Wang, P.; Li, Q.; Al-Khalaf, A. A.; Hozzein, W. N.; Zhang, F.; Li, X.; Zhao, D. Near-infrared triggered decomposition of nanocapsules with high tumor accumulation and stimuli responsive fast elimination. Angew. Chem., Int. Ed. 2018, 57, 2611. (14) Chen, J.; Ding, J.; Wang, Y.; Cheng, J.; Ji, S.; Zhuang, X.; Chen, X. Sequentially responsive shell-stacked nanoparticles for deep penetration into solid tumors. Adv. Mater. 2017, 29, 1701170. (15) Han, K.; Zhang, J.; Zhang, W.; Wang, S.; Xu, L.; Zhang, C.; Zhang, X.; Han, H. Tumor-triggered geometrical shape switch of chimeric peptide for enhanced in vivo tumor internalization and photodynamic therapy. ACS Nano 2017, 11, 3178. (16) Li, H. J.; Du, J. Z.; Du, X. J.; Xu, C. F.; Sun, C. Y.; Wang, H. X.; Cao, Z. T.; Yang, X. Z.; Zhu, Y. H.; Nie, S.; Wang, J. Stimuliresponsive clustered nanoparticles for improved tumor penetration and therapeutic efficacy. Proc. Natl. Acad. Sci. U. S. A. 2016, 113, 4164. (17) Lei, Q.; Wang, S. B.; Hu, J. J.; Lin, Y. X.; Zhu, C. H.; Rong, L.; Zhang, X. Z. Stimuli-responsive ″cluster bomb″ for programmed tumor therapy. ACS Nano 2017, 11, 7201. (18) Xu, R.; Zhang, G.; Mai, J.; Deng, X.; Segura-Ibarra, V.; Wu, S.; Shen, J.; Liu, H.; Hu, Z.; Chen, L.; Huang, Y.; Koay, E.; Huang, Y.; Liu, J.; Ensor, J. E.; Blanco, E.; Liu, X.; Ferrari, M.; Shen, H. An injectable nanoparticle generator enhances delivery of cancer therapeutics. Nat. Biotechnol. 2016, 34, 414. (19) Tong, R.; Hemmati, H. D.; Langer, R.; Kohane, D. S. Photoswitchable nanoparticles for triggered tissue penetration and drug delivery. J. Am. Chem. Soc. 2012, 134, 8848. (20) Ma, W.; Cheetham, A. G.; Cui, H. Building nanostructures with drugs. Nano Today 2016, 11, 13. (21) Lock, L. L.; LaComb, M.; Schwarz, K.; Cheetham, A. G.; Lin, Y. A.; Zhang, P.; Cui, H. Self-assembly of natural and synthetic drug amphiphiles into discrete supramolecular nanostructures. Faraday Discuss. 2013, 166, 285. (22) Cheetham, A. G.; Zhang, P.; Lin, Y. A.; Lock, L. L.; Cui, H. Supramolecular nanostructures formed by anticancer drug assembly. J. Am. Chem. Soc. 2013, 135, 2907. (23) MacKay, J. A.; Chen, M.; McDaniel, J. R.; Liu, W.; Simnick, A. J.; Chilkoti, A. Self-assembling chimeric polypeptide-doxorubicin conjugate nanoparticles that abolish tumours after a single injection. Nat. Mater. 2009, 8, 993. (24) McDaniel, J. R.; MacEwan, S. R.; Li, X.; Radford, D. C.; Landon, C. D.; Dewhirst, M.; Chilkoti, A. Rational design of ″heat seeking″ drug loaded polypeptide nanoparticles that thermally target solid tumors. Nano Lett. 2014, 14, 2890. (25) Huang, P.; Wang, D.; Su, Y.; Huang, W.; Zhou, Y.; Cui, D.; Zhu, X.; Yan, D. Combination of small molecule prodrug and nanodrug delivery: amphiphilic drug-drug conjugate for cancer therapy. J. Am. Chem. Soc. 2014, 136, 11748. (26) Ke, Z.; Hou, X.; Jia, X. B. Design and optimization of selfnanoemulsifying drug delivery systems for improved bioavailability of cyclovirobuxine D. Drug Des., Dev. Ther. 2016, 10, 2049. (27) Dai, Y.; Yang, Z.; Cheng, S.; Wang, Z.; Zhang, R.; Zhu, G.; Wang, Z.; Yung, B. C.; Tian, R.; Jacobson, O.; Xu, C.; Ni, Q.; Song, J.; Sun, X.; Niu, G.; Chen, X. Toxic reactive oxygen species enhanced synergistic combination therapy by self-assembled metal-phenolic network nanoparticles. Adv. Mater. 2018, 30, 1704877.

(28) Ke, X.; Ng, V. W.; Ono, R. J.; Chan, J. M.; Krishnamurthy, S.; Wang, Y.; Hedrick, J. L.; Yang, Y. Y. Role of non-covalent and covalent interactions in cargo loading capacity and stability of polymeric micelles. J. Controlled Release 2014, 193, 9. (29) Huang, C.; Chu, C.; Wang, X.; Lin, H.; Wang, J.; Zeng, Y.; Zhu, W.; Wang, Y. J.; Liu, G. Ultra-high loading of sinoporphyrin sodium in ferritin for single-wave motivated photothermal and photodynamic co-therapy. Biomater. Sci. 2017, 5, 1512. (30) Maeda, H. Toward a full understanding of the EPR effect in primary and metastatic tumors as well as issues related to its heterogeneity. Adv. Drug Delivery Rev. 2015, 91, 3. (31) Chu, C.; Lin, H.; Liu, H.; Wang, X.; Wang, J.; Zhang, P.; Gao, H.; Huang, C.; Zeng, Y.; Tan, Y.; Liu, G.; Chen, X. Tumor microenvironment-triggered supramolecular system as an in situ nanotheranostic generator for cancer phototherapy. Adv. Mater. 2017, 29, 1605928. (32) Wang, X.; Hu, J.; Wang, P.; Zhang, S.; Liu, Y.; Xiong, W.; Liu, Q. Analysis of the in vivo and in vitro effects of photodynamic therapy on breast cancer by using a sensitizer, sinoporphyrin sodium. Theranostics 2015, 5, 772. (33) Takahashi, A.; Ozaki, Y.; Kuzuya, A.; Ohya, Y. Impact of coreforming segment structure on drug loading in biodegradable polymeric micelles using PEG-b-poly(lactide-co-depsipeptide) block copolymers. BioMed Res. Int. 2014, 2014, 579212. (34) Zhang, P.; Zhang, L.; Qin, Z.; Hua, S.; Guo, Z.; Chu, C.; Lin, H.; Zhang, Y.; Li, W.; Zhang, X.; Chen, X.; Liu, G. Genetically engineered liposome-like nanovesicles as active targeted transport platform. Adv. Mater. 2018, 30, 1705350. (35) Cheng, T.; Liu, J.; Ren, J.; Huang, F.; Ou, H.; Ding, Y.; Zhang, Y.; Ma, R.; An, Y.; Liu, J.; Shi, L. Green tea catechin-based complex micelles combined with doxorubicin to overcome cardiotoxicity and multidrug resistance. Theranostics 2016, 6, 1277.

14989

DOI: 10.1021/jacs.8b09396 J. Am. Chem. Soc. 2018, 140, 14980−14989