Article Cite This: Biomacromolecules 2019, 20, 2809−2820
pubs.acs.org/Biomac
Redox-Responsive Fluorescent Polycarbonates Based on Selenide for Chemotherapy of Triple-Negative Breast Cancer Li Yu,†,§ He-Liang Ke,‡,§ Fu-Sheng Du,*,† and Zi-Chen Li*,† †
Downloaded via BUFFALO STATE on July 23, 2019 at 07:57:58 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
Beijing National Laboratory for Molecular Sciences (BNLMS), Key Laboratory of Polymer Chemistry and Physics of Ministry of Education, Department of Polymer Science and Engineering, College of Chemistry and Molecular Engineering, Peking University, Beijing 100871, China ‡ Emergency Center, First Affiliated Hospital of Wenzhou Medical University, Wenzhou 325000, China S Supporting Information *
ABSTRACT: Transient increase of reactive oxygen species (ROS) is vital for some physiological processes, whereas the chronic and sustained high ROS level is usually implicated in the inflammatory diseases and cancers. Herein, we report the innovative redoxresponsive theranostic micellar nanoparticles that are able to load anticancer drugs through coordination and hydrophobic interaction and to fluorescently monitor the intracellular redox status. The nanoparticles were formed by the amphiphilic block copolymers composed of a PEG segment and a selenide-containing hydrophobic polycarbonate block with a small fraction of coumarin-based chromophore. Under the alternative redox stimulation that might be encountered in the physiological process of some healthy cells, these nanoparticles underwent the reversible changes in size, morphology, and fluorescence intensity. With the assistance of small model compounds, we clarified the chemistry behind these changes, that is, the redox triggered reversible transformation between selenide and selenoxide. Upon the monotonic oxidation similar to the sustained high ROS level of cancer cells, the nanoparticles could be disrupted completely, which was accompanied by the drastic decrease in fluorescence. Cisplatin and paclitaxel were simultaneously coloaded in the nanoparticles with a moderate efficacy, and the coordination between selenide and platinum improved the stability of the drug-loaded nanoparticles against dilution. The naked nanoparticles are cytocompatible, whereas the dual drug-loaded nanoparticles exhibited a concentration dependent and synergistic cytotoxicity to triple-negative breast cancer (TNBC) cells. Of importance, the drug-loaded nanoparticles are much more toxic to TNBC cells than to normal cells due in part to ROS overproduction in the former cell lines.
■
INTRODUCTION Some physiological activities of healthy cells cause transient increase in reactive oxygen species (ROS), which would quickly return to normal levels under the regulation of the antioxidant system in order to maintain the intracellular redox homeostasis. Such redox reactions are ubiquitous in living organisms, playing an important role in cell signaling, gene transcription, immune response, and metabolism.1 However, overproduction of ROS induced endogenously or by exogenous stimuli may damage biomolecules and cause pathological disorders. Growing evidence indicates that chronic oxidative stress is usually implicated in various diseases such as cancers, inflammation, atherosclerosis, and the agingassociated diseases.2 In the past two decades, varied ROSresponsive fluorescent probes, prodrugs, polymers, or supramolecular assemblies have been developed, showing great potential for biomedical or pharmaceutical applications, in particular for the diagnosis or therapy of the oxidative stressrelated diseases.3 Among the ROS-responsive polymers, the majority of them consist of the sensitive motifs that, upon © 2019 American Chemical Society
physiologically available oxidative stimulation, undergo irreversible transformation of the chemical structure, which results monotonically in the change of the polymer properties.4 These monotonically responsive polymers could be the suitable vehicles of delivery systems targeting the diseases with persistent oxidative stress.5 However, when they are applied for the delivering of toxic chemotherapeutic agents, the transient increase of ROS level in normal cells may also irreversibly disrupt the polymeric carriers and trigger the undesired drug release, causing harmful side effect on the healthy cells or tissues. Therefore, it is essential to construct reversibly redox-responsive polymeric carriers that could sense the physiological redox state and change the chemical structure reversibly. Triple-negative breast cancer (TNBC) is a highly metastatic and invasive subtype of breast cancers (BCs) with the features Received: April 28, 2019 Revised: June 9, 2019 Published: June 11, 2019 2809
DOI: 10.1021/acs.biomac.9b00583 Biomacromolecules 2019, 20, 2809−2820
Article
Biomacromolecules Scheme 1. Synthesis of the Fluorescent Copolymer and Its Reversible Redox-Responsive Transformation
water and fluorescence change of this copolymer under physiological redox conditions, (2) to demonstrate the ability of the copolymer micelles to load cisplatin and paclitaxel simultaneously, and (3) to reveal the different cytotoxicity of the dual drug-loaded micelles to TNBC cells and normal cells.
of molecular heterogeneity and relatively poor prognosis, accounting for about 15% of BC cases. TNBC is usually defined as the BC that lacks estrogen receptor, progesterone receptor, and human epidermal growth receptor 2 (HER2) and thus cannot be successfully treated with endocrine therapy or HER2-trageting therapy.6 Currently, systemic chemotherapy is the mainstay for the treatment of TNBCs, in particular the recurrent or metastatic one. Various small molecule therapeutic agents including paclitaxel and platinum agents have been used alone or combined with each other to treat TNBC.7 Most of TNBC cells are deficient in breast cancer type-1 susceptibility gene (BRCA1) and usually sensitive to DNA-cross-linking agents such as cisplatin and mitomycin C.8 Despite the promise of using dose dense polychemotherapy, the progress in TNBC treatment is limited in part due to the inherently harmful effects of current systemic chemotherapy, such as liver injury, kidney failure, and marrow suppression. It would be attractive to develop new chemotherapeutic nanomedicines that target TNBC cells but show diminished adverse effect on normal cells or healthy tissues. Selenium is an essential element in human body, and a growing interest has been paid to the selenium-containing organic compounds that show promise as anticancer agents to selectively kill cancer cells.9 In the recent decade, varied selenium-based ROS-responsive polymers have been developed and investigated to explore their potential application in chemotherapy, photodynamic therapy, radiotherapy, or their combination.10 Both thioether and selenide can undergo a twoelectron oxidation to afford their oxides. However, the sulfoxide and selenoxide show significant divergence in their reactivity toward chemical reduction due most probably to very weak π-bonding in selenoxide. A rapid and reversible redox transformation is able to occur between selenide and its selenoxide derivative under physiological redox conditions but not for the thioether/sulfoxide pair.9a Therefore, various selenide-based fluorescent probes were developed to monitor intracellular or in vivo redox status.11 Although few papers reported the redox-controlled reversible assembly and disassembly of the selenide-containing amphiphilic copolymers or surfactants, their pharmaceutical applications as smart drug carriers were not explored.12 Recently, we have synthesized an amphiphilic copolymer composed of a PEG block and a polycarbonate segment with pendent ethylseleno group and demonstrated its potential as an ROS-responsive vehicle for the combination of chemotherapy and photodynamic therapy.13 To further explore its pharmaceutical usage, herein we have incorporated a selenide-containing fluorescent chromophore group into the above selenide-based polycarbonate amphiphilic block copolymer (Scheme 1). Our purposes are (1) to study the reversible assembly profiles in
■
EXPERIMENTAL SECTION
Materials. 3-Methyl-3-oxetanemethanol, diethyl diselenide, 7diethylaminocoumarin, ascorbic acid (VC), vitamin E (VE), paclitaxel (PTX), and ruthenium red (RR) were purchased from J&K Chemical Ltd. and used as received. Poly(ethylene glycol) monomethyl ether with molecular weight of 5 kDa (mPEG113, Fluka), pyrene (Acros), hydrogen peroxide (H2O2, 30 wt %, Beijing Chemical Works), regenerated cellulose dialysis membrane (MWCO: 50 kDa, Beijing Huamei Scientific Co.), cisplatin (TagerMol), glutathione (GSH, TagerMol), 2′,7′-dichlorodihydrofluorescein diacetate (DCFH-DA, AAT Bioquest), Fura-2/AM (Dojindo), metallothionein (TCI), glutathione reductase (GR) and its coenzyme NADPH (Sigma), αlipoic acid, uric acid, and cysteine (Aladdin) were used as received. Triethylamine (TEA) and deuterated chloroform (CDCl3) were dried over potassium hydroxide and potassium carbonate, respectively, prior to use. 1,8-Diazabicyclo[5.4.0]undec-7-ene (DBU, J&K Chemical Ltd., 98%) was distilled over CaH2 under reduced pressure. Dichloromethane (DCM), toluene, and tetrahydrofuran (THF) were distilled over CaH2 after refluxing for 10 h. Deuterated phosphate buffer (PB) with pH = 7.4 was prepared from NaOD (40 wt % in D2O, Alfa) and D3PO4 (85 wt % in D2O, Alfa). Hydrogen bromide (HBr, 40 wt %), ethyl bromide, ethyl chloroformate, benzoic acid, sodium borohydride (NaBH4), sodium hypochlorite (NaOCl), tert-butyl hydroperoxide (t-BuOOH), bromine, other reagents, and solvents were purchased from Beijing Chemical Reagent Co. and used as received. The block copolymer of mPEG113 and poly(trimethylene carbonate), that is PEG-b-PTMC (Mn GPC = 11 700, degree of polymerization of PTMC block = 35), was prepared in our lab. The intermediate compound 1 (2-bromomethyl-2-methylpropane-1,3diol) for monomer synthesis, compound 7 (2-((ethylselanyl)methyl)-2-methylpropane-1,3-diol) for coordination study, and the cyclic carbonate monomer M1 were synthesized following the published procedure.13 Measurements. The detailed procedures and equipment for the measurements of NMR, Fourier transform infrared (FT-IR), electrospray ionization mass spectroscopy (ESI-MS), gel permeation chromatography (GPC), laser light scattering (LLS), transmission electron microscopy (TEM), high-performance liquid chromatography (HPLC), and inductively coupled plasma-atomic emission spectrometry (ICP-AES) are given in the Supporting Information. Syntheses. The detailed synthetic protocols of fluorescent monomer F1, model compounds, and block copolymers (PF series) are presented in the Supporting Information. Preparation of Block Copolymer Nanoparticles. The solvent evaporation method was applied for the preparation of PF nanoparticles. Briefly, the copolymer (1.0 mg) was dissolved in THF (0.5 mL). To this solution 10 mL of PB (pH 7.4, 10 mM) was added dropwise for 60 s under sonication. This solution was stirred 2810
DOI: 10.1021/acs.biomac.9b00583 Biomacromolecules 2019, 20, 2809−2820
Article
Biomacromolecules Table 1. Characterization of PF Block Copolymersa polymer
[M]0:[I]0b
M1/F1
DPc
Mnc
Mnd
Đd
CACe (mg/L)
Rhf (nm)
Rg/Rhf (nm)
PF1 PF2 PF3
80:1 40:1 20:1
98.6/1.4 98.6/1.4 98.7/1.3
72 36 19
22200 14100 9500
27800 21400 16100
1.07 1.06 1.06
0.42 0.79 0.94
36.7 24.2 28.7
0.77 0.72 0.76
ROP conditions: [M]0 was ∼2.0 M, DBU was 5.0 mol % of monomer, 25 °C. bInitial molar ratio of monomer to macroinitiator. cCalculated by H NMR. dMeasured by GPC using polystyrene standards in THF. eCritical aggregation concentration measured by fluorescence method using pyrene as a probe, 37 °C. fMeasured by LLS in 10 mM PB (pH 7.4) with a polymer concentration of 0.1 mg/mL, 37 °C.
a
1
Table 2. Characterization of the Drug-Loaded Nanoparticles PTX nanoparticlea
PTX/Pb (wt %)
Pt/Pc (wt %)
PF3 PF3@PTX PF3@Pt PF3@PTX/Pt50 PF3@PTX/Pt25 PF3@PTX/Pt10 PTMC@PTX/Pt25
0 25 0 25 12 5 12
0 0 25 50 25 10 25
Pt
LCd (wt %)
LEe (wt %)
7.8 ± 0.4
31 ± 1.6
1.6 2.7 2.1 3.0
± ± ± ±
0.1 0.2 0.2 0.2
6.4 22 42 25
± ± ± ±
0.4 1.6 4.0 1.6
LCd (wt %)
8.6 21 6.8 3.3 5.8
± ± ± ± ±
0.3 1.3 0.6 0.3 0.3
LEe (wt %)
Rhf (nm)
Rg/Rhf
± ± ± ± ±
55.0 61.5 50.8 51.4 55.2 58.9 60.4
0.75 0.80 0.72 0.72 0.74 0.76 0.75
34 42 27 33 23
1.2 2.6 2.4 3.2 1.2
a
PTMC denotes PEG-b-PTMC; Pt denotes cisplatin. bWeight ratio of PTX/polymer in feed. cWeight ratio of cisplatin/polymer in feed. dLoading content defined as drug in nanoparticle/polymer in nanoparticle (×100). eLoading efficiency defined as drug in nanoparticle/drug in feed (×100). f Measured by LLS in 10 mM PB (pH 7.4) at 37 °C with a polymer concentration of 1.0 mg/mL. overnight at 37 °C to evaporate THF completely and adjusted to a final polymer concentration of 0.1 mg/mL. This stock solution was stored at 4 °C and used for the measurements of critical aggregation concentration (CAC), LLS, and TEM. The characterization results of the copolymers and their nanoparticles are shown in Table 1. Redox Reaction of M1, Compound 5, and PF2 in Homogeneous Solutions. The redox reaction of M1 and PF2 were conducted with the similar procedures. Take M1 as an example. One milliliter of M1 solution (5 mg/mL, ∼20 mM of selenide) in CD3CN was charged into an NMR tube. After incubation at 25 °C for 10 min, 1H NMR spectrum was recorded as that of 0 min time point. After 30 min of the addition of H2O2 (19 mM in the reaction solution), 1H NMR spectrum was measured. Then, 7 mg of VC (40 mM in the reaction solution) was added to the above solution, which was mixed rapidly and 1H NMR spectrum was recorded after an additional 30 min. For MS characterization, 50 μL of the sample was taken out from the NMR tube after 30 min of oxidation and measured directly. For FT-IR, the reaction solution taken out was dried under reduced pressure and submitted for preparation of the IR specimen. For the 1H NMR measurement of compound 5, DMSO-d6 was used as the solvent with an initial concentration of ∼3 mM. H2O2 and VC, used for the redox reactions, were 1 equiv and 3 equiv of 5, respectively. The oxidation and reduction were carried out at 25 °C for 30 and 60 min, respectively. To monitor the redox reaction products of compound 5 by FT-IR, the reaction was performed in methanol instead of DMSO. UV−vis Absorption and Fluorescence Spectra of 5 and 6. UV−vis absorption spectrum of compound 5 or 6 (10 μM) in 10% (v/v) DMF aqueous solution was recorded at 25 °C on a PE Lambda 35 UV−vis spectrometer using a 1 cm quartz cell. Ten percent (v/v) DMF aqueous solution was applied as a reference. The fluorescence spectra were measured on a Hitachi F-7000 spectrometer equipped with a temperature controller. The parameters used were 1200 nm/ min of scanning rate, 900 V of excitation voltage, and 440 nm of excitation wavelength. The fluorescence spectra were recorded from 450 to 660 nm. Both excitation and emission slit widths were 2.5 nm. For the redox reaction of 5 and 6 monitored by fluorescence, H2O2 (30 μM) was added to the solution of 5 (25 μM) and mixed quickly, and the emission spectra were recorded at different times. After 30 min, 60 μM VC was added to the above solution and mixed quickly. The spectra were monitored continually until the emission intensity did not change.
Effects of Oxidants, Reducing Agents, and Metal Ions on the Fluorescence of PF2 Nanoparticles. PF2 nanoparticle solution with a concentration of 1.0 mg/mL (36 μM of chromophore, ∼2.8 mM of selenide, in 10 mM PB, pH = 7.4) was equilibrated at 37 °C. The fluorescence of the solution was first measured and normalized as 100%. After the addition of oxidant (3.0 mM), the solution was incubated for 60 min and the fluorescence at 510 nm was measured again. In the case of H2O2 as the oxidant, after 60 min of oxidation a different reducing agent (6.0 mM) was added, the solution was incubated for an additional 90 min, and the fluorescence was recorded. To study the effect of metal ions, the redox reactions were conducted in the presence of 10 mM metal ion, using H2O2 as the oxidant and VC as the reductant. The oxidants used include H2O2, NaOCl, t-BuOOH, hydroxyl radical (•OH), and singlet oxygen (1O2). Hydroxyl radical was quantitatively produced through the Fenton reaction. 1O2 was generated in situ by the reaction of NaOCl with H2O2. VC, GR/NADPH, VE, GSH, metallothionein, α-lipoic acid, uric acid, and cysteine were applied as the reducing agents. Metal ions include Fe3+, Al3+, Mg2+, Na+, Zn2+, Cu2+, K+, and Ca2+. Redox-Responsive Change in Fluorescence of Different PF Nanoparticles. The solution of PF nanoparticles with a concentration of 0.1 mg/mL (0.2−0.32 mM of selenide) was equilibrated for 10 min at 37 °C, and the fluorescence spectrum was recorded for 0 min time point. Upon the addition of H2O2 (100 μM in the reaction solution), fluorescence spectra were taken at the desired time points. After 120 min, VC (100 μM) was added to the mixture, and the fluorescence was monitored for additional 120 min. For the drugloaded PF3 nanoparticles, a polymer concentration of 1.0 mg/mL (2.0−3.2 mM of selenide) was used. Loading and Release of Cisplatin and PTX. The drug-loaded PF3 nanoparticles were prepared following the aforementioned solvent evaporation method. The feed ratios of PTX/cisplatin to PF3 were 5.0/10.0 wt %, 12.5/25.0 wt %, and 25.0/50.0 wt %.. Cisplatin (Pt) dissolved in PB (pH 7.4, 10 mM) with a concentration of 2.0 mg/mL was used as a stock solution. Take the preparation of PF3@PTX/Pt50 as an example. Ten milligrams of PF3 and 2.5 mg of PTX were dissolved in 1.0 mL of THF. To this solution, 2.5 mL of Pt stock solution and 7.5 mL of PB solution (pH 7.4, 10 mM) were added, and the mixture was stirred overnight at 37 °C to remove THF. After centrifugation, the supernatant was dialyzed against the same PB solution for 72 h at 37 °C and changing the dialysis medium 2811
DOI: 10.1021/acs.biomac.9b00583 Biomacromolecules 2019, 20, 2809−2820
Article
Biomacromolecules thrice. The final concentration of copolymer nanoparticles was tuned to be 1.0 mg/mL. The loading content and efficiency of cisplatin were determined by ICP-AES. Briefly, 5 mL of aqua regia (1 + 1) was added to 0.5 mL of the drug-loaded PF3 nanoparticle solution, which was heated to dissolve Pt sufficiently and dried up. To the cooled residue, an additional 5 mL of aqua regia was added, and the mixture was heated to boiling and cooled down again. The final Pt-containing sample was tuned to 50 mL for ICP-AES analysis. The content of cisplatin was determined based on a calibration curve. The loading content and efficiency of PTX were determined by HPLC. Ten mM H2O2 (1.0 mL) was added into 1.0 mL of the drugloaded nanoparticle solution, which was stirred for 24 h to disrupt the nanoparticle. Three milliliters of acetonitrile was added to this solution and mixed thoroughly. The final sample was tuned to 5 mL and filtered through a Millipore 0.22 μm PVDF membrane prior to measurement. The peak at 9.6 min was used to determine the concentration of paclitaxel in the sample based on a calibration curve that was obtained from a series of solutions containing different concentrations of PTX in CH3CN/H2O (60/40, v/v). All the measurements were conducted in triplicate. Rh and Rg of the drugleaded or empty nanoparticles were characterized by LLS. The characterization results of these drug-loaded nanoparticles are shown in Table 2. The drug-loaded PF3@PTX/Pt25 nanoparticles with 6.8 wt % of cisplatin and 2.7 wt % of PTX was used to study the in vitro release profiles under two stimulation modes. The experiments were performed by a dialysis method. In the case of monotonic oxidation mode, 1.0 mL of PF3@PTX/Pt25 nanoparticle solution was added into a dialysis tubing (MWCO: 50 kDa) that was immersed in 5.0 mL of PB solution (pH 7.4, 10 mM). Then, H2O2 was added to reach an initial concentration of 1.0 mM. The system was stirred gently for 12 h at 37 °C. During this process, 0.5 mL of the dialysis medium was taken out for HPLC analysis and replenished with the same volume of fresh medium at the desired time points. The released PTX was determined by HPLC, and cisplatin by ICP-AES after treating with aqua regia through the aforementioned procedure. For the alternative oxidation/reduction triggering mode, H2O2 (1.0 mM) was first added. After 60 min of oxidation, VC with a final concentration of 1.0 mM was added for additional 120 min of reduction. The redox process was repeated four times. All the measurements were conducted in triplicate. In Vitro Cytotoxicity Assay. All the cell lines were kindly obtained from the Institute of Basic Medical Sciences, Chinese Academy of Medical Sciences. The pristine PF3 nanoparticle solution was prepared following the aforementioned solvent evaporation method with a relatively high concentration (5.0 mg/mL). PEG-bPTMC nanoparticles were prepared by the same procedure and used as a nonresponsive control. mPEG113 and branched PEI (Aldrich, 25 kDa) were used as the negative and positive controls, respectively. CCK-8 assay was used to assess cytotoxicity of the samples to the normal human breast cell (HBL-100) and human TNBC cell (MDAMB-231). The cells were seeded in 96-well plates (3 × 104 cells per well) and cultured in DMEM (Dulbecco’s modified Eagle’s medium) supplemented with 10% fetal bovine serum for 24 h. The conditions of humidified atmosphere, 5% CO2, and 37 °C were used for cell culture and attachment. Then, 10 μL of the sample solution with a specific concentration was added to each well, and the cells were cultured for another 24 h prior to CCK-8 assay. The absorbance of the solution in each well was detected on a PerkinElmer EnSpire multimode microplate reader at 450 nm. Cell viability (%) was defined as (Asample/Acontrol) × 100. The data were obtained in triplicate. To access the cytotoxicity of the drug-loaded polymer nanoparticles, three human TNBC cell lines (MDA-MB-231, MDA-MB468, MDA-MB-453) and three normal cell lines (HBL-100, L-02, 293T) were used. The drug-loaded nanoparticles were prepared following the aforementioned procedures but with a relatively high polymer concentration (5.0 mg/mL). For the normal cells with Ca2+ stimulation, the concentration of cisplatin in the well was 15 μg/mL.
The cells were incubated with the samples for 24 h prior to CCK-8 assay. During this period, Ca2+ stimulation was performed from 0 to 4 times, to mimic the transient up-and-down regulation of the intracellular ROS level. For Ca2+ stimulation, 5.0 mM of CaCl2 was added to the culture medium. After 30 min of incubation, the original culture medium was carefully removed, and the wells were replenished with fresh DMEM. For the TNBC cells, the procedure of Ca2+ stimulation was not applied because their inherently high intracellular ROS level.
■
RESULTS AND DISCUSSION Molecular Design and Selection of Cells. Dialkyl selenide and its selenoxide derivative could undergo the reversible transformation under mild redox conditions.12 This feature was applied to design reversible redox-responsive fluorescent probes that were capable of sensing intracellular or in vivo redox status.11 We have reported the amphiphilic PEG/ polycarbonate block copolymer with pendent ethylseleno groups and found that this novel selenide-containing copolymer and its selenoxide derivative were cytocompatible to human breast epithelial HBL-100 cells. The copolymer micellar nanoparticles were sensitive to the physiologically available concentration of H2O2 or cell-compatible red light irradiation when loaded with chlorin e6, showing potential as a smart vehicle for the combined chemotherapy and photodynamic therapy.13 On these bases, we select this polycarbonate copolymer as the reversible redox-responsive framework (Scheme 1). Moreover, a selenide-based chromophore is incorporated onto the polycarbonate as a pendent group. The redox triggered reversible transformation between the selenide and selenoxide would induce a fluorescence up and down due to the reversed alteration of the intramolecular change transfer state.14 This kind of redox-responsive fluorescent amphiphilic copolymer could be a promising theranostic nanoplatform for delivering active agents to the ROS-overproduced cells and monitoring the intracellular redox status simultaneously. Three breast cancer cell lines, MDA-MB-231, MDA-MB453, and MDA-MB-468, were used as the model TNBC cells.7a,15 Three normal cell lines, HBL-100, hepatocyte cell (L02), and embryonic kidney cell (293T), were applied as the controls. Calcium ion (Ca2+) stimulation of the three normal cell lines was used to mimic the transient change (up and down) of intracellular ROS level during the normal physiological processes.16 The intracellular ROS level was evaluated by using DCFH-DA as a ROS-activatable fluorescent probe.17 We optimized the experimental conditions and confirmed that Ca2+ stimulation of the three normal cell lines could transiently increase the intracellular ROS to a maximum level in 30 min, which was followed by a significant ROS downregulation in 90 min. On the other hand, all three TNBC cell lines showed the persistent higher ROS levels, which is consistent with the published results (Figures S2− S5).18 The stimulation under the optimized concentration of Ca2+ did not cause significant apoptosis, and the Ca2+-treated cells remained healthy after a redox cycle (Figures S7). Synthesis of Block Copolymers and Preparation of Their Nanoparticles. Monomer M1 and F1 are both cyclic carbonates containing selenide. M1 was synthesized following the reported procedure.13 The fluorescent monomer F1 was synthesized by the cyclization of 1,3-diol using ethyl chloroformate as a ring-closing reagent (Scheme S1). The structure of F1 was confirmed by 1H/13C NMR (Figure S11). As a redox-responsive motif with a large proportion in the copolymers, M1 is responsible for the regulation of particle 2812
DOI: 10.1021/acs.biomac.9b00583 Biomacromolecules 2019, 20, 2809−2820
Article
Biomacromolecules
Figure 1. 1H NMR spectra of M1 (5 mg/mL, 20 mM) in CD3CN with the sequential addition of 19 mM H2O2 and 40 mM VC at 25 °C. The peaks marked with yellow triangles are assigned to VC.
Figure 2. Fluorescence change of compound 5 (25 μM) in 10% (v/v) DMF aqueous solution at 25 °C in the cycle of (A) oxidation with H2O2 (30 μM) and (B) reduction with VC (60 μM). λex = 440 nm.
morphology, while F1 can be a fluorescent motif to reflect the intracellular redox state, as well as the position and morphology of nanoparticle. Three amphiphilic block copolymers (PF1−PF3) were synthesized by the ring-opening polymerization (ROP), which was catalyzed by a strong basic amine (DBU) in DCM at 25 °C, using mPEG113 as a macroinitiator. They are different in hydrophobic block length but similar in the contents of the fluorescent unit (1.3−1.4% in molar percentage). Since the initiation activity of mPEG113 is relatively low, the oligomers initiated by trace water could be seen in the GPC curve of the unpurified PF1 (Figure S15A). These oligomers could be removed by precipitation from acetonitrile. Each of the purified copolymers shows a unimodal distribution. Their structures and molecular weights were clarified by 1H NMR and GPC (Figures S14 and S15B). By the solvent evaporation approach, the block copolymers could form micellar nanoparticles in aqueous PB solution (pH 7.4, 10 mM) with critical aggregation concentrations (CACs)
ranging from 0.42 to 0.94 mg/L (Figure S16). According to the results of LLS, Rh values of the nanoparticles are in the range of 24−37 nm with a narrow distribution (Figure S17). All the data are summarized in Table 1. Reversible Redox Reaction of Selenide in Organic Solutions. To better understand the redox reaction of selenide used in this work, we characterized the reaction product of M1 in CD3CN after the sequential oxidation with H2O2 and reduction with vitamin C (VC). As shown in Figure 1, the selenide was quickly oxidized to its selenoxide derivative by stoichiometric H2O2 under mild condition, and the selenoxide could be rapidly reduced back to M1 by VC. The products in the redox process were confirmed by MS and FTIR spectra (Figures S18 and S19A). The similar redox profiles of PF2 block copolymer in CD3CN were observed, showing the reversible transformation between selenide and selenoxide (Figures S19B and S20). 2813
DOI: 10.1021/acs.biomac.9b00583 Biomacromolecules 2019, 20, 2809−2820
Article
Biomacromolecules
Figure 3. Effects of oxidants, reductants, or metal ions on the fluorescence of PF2 nanoparticles (1.0 mg/mL) in PBS (pH 7.4, 10 mM) at 37 °C. (A) The fluorescence percentage after 1 h with various oxidants (3.0 mM). (B) The fluorescence percentage after 1 h of oxidation with H2O2 and additional 1.5 h of reduction with different reductants (6.0 mM). For GR/NADPH, 100 units/ml was used. (C) Reversible redox triggered fluorescence change in the presence of 10 mM metal ions. λex = 440 nm, λem = 510 nm. All data was normalized with respect to the initial intensity and presented as the mean ± standard deviation in triplicate.
Figure 4. (A,B) Plots of fluorescence percentage of PF1−PF3 nanoparticles versus time under the sequential addition of 100 μM H2O2 (A) and 100 μM VC (B). All data were normalized with respect to the initial intensity. (C) The fluorescence intensity of PF3 nanoparticles during four redox cycles. The results are presented as the mean ± standard deviation in triplicate. Polymer concentration, 0.1 mg/mL in PB solution (pH 7.4, 10 mM); λex/λem, 440/510 nm.
components on the fluorescence of PF2 nanoparticles in PB solution. As shown in Figure 3A, all of the five oxidants could significantly reduce the fluorescence intensity. Among them, H2O2, NaOCl, and 1O2 showed the stronger oxidation efficacy, making the fluorescence decreased to less than 10% of the initial value in 1 h. This result indicates that PF2 is capable of detecting a wide range of intramolecular ROS. Using the H2O2-oxidized samples, we further investigated the fluorescence recovery of PF2 nanoparticles in the presence of different reducing agents (Figure 3B). Of the tested reductants, VC and glutathione reductase (GR)/NADPH greatly recovered the fluorescence with a recovering degree of ∼100% by VC, which is attributed to the reductive transformation from selenoxide to selenide in the chromophore. The other reductants did not trigger the recovering in fluorescence, demonstrating that these reductants are unable to reduce the selenoxide of the oxidized chromophore. When H2O2 and VC were applied as the oxidant and the reductant, respectively, the metal ions usually found in cells did not interfere the fluorescence change during the redox cycle (Figure 3C). Compared to those Se-containing small molecule probes reported recently, which response specifically to one or two of the intracellular ROS such as HOCl, peroxynitrite, or H2O2, PF2 is responsive to a broad spectrum of intracellular ROS.11 Currently, although we do not know what causes the difference in specificity of the Se-based fluorescent probes to different ROS, the broad response of PF2 nanoparticles is
To clarify the redox-responsive fluorescence change of the chromophore, a Se-containing small molecule model compound (5) was synthesized (Figure S12). Under the aforementioned redox conditions, 5 and its selenoxide derivative 6 underwent the same reversible transformation, which was proven by 1H NMR, FT-IR, and MS spectra (Figures S21−S23). Compound 5 showed green fluorescence with a maximum emission at 510 nm. The concentration related fluorescence self-quenching was observed over 25 μM (Figure S25). Upon oxidation with H2O2 (1.2 equiv), the fluorescence intensity decreased to ∼90% of the initial value in 30 min. The addition of VC (2.4 equiv) recovered the fluorescence completely within 70 min (Figure 2). This redox regulated fluorescence change is attributed to the reversible transformation between selenide 5 and the selenoxide 6. Because the ethylseleno group is directly linked to the coumarin ring, the conversion from selenide to the more electron-withdrawing selenoxide strengthens the degree of intramolecular charge transfer, resulting in the drastic decrease of fluorescence intensity.14 These results reveal that this chromophore is sensitive enough to the physiologically available H2O2 and vitamin C and could be used for imaging the intracellular redox status. Reversible Redox-Responsive Change in Fluorescence of PF Nanoparticles. It is well-known that there are varied oxidative species, reducing agents, and metal ions in cells.1,19 We next studied the effect of these intracellular 2814
DOI: 10.1021/acs.biomac.9b00583 Biomacromolecules 2019, 20, 2809−2820
Article
Biomacromolecules
Figure 5. Time-dependent changes of scattered light intensity, Rh, Rg, and Rg/Rh of PF nanoparticles (0.1 mg/mL) in PB solution (pH 7.4, 10 mM) upon the sequential addition of 100 μM H2O2 and 100 μM VC at 37 °C. Oxidation time, 60 min; reduction time, 120 min; detection angle, 90°.
Figure 6. TEM photographs of PF3 nanoparticle (0.1 mg/mL) during the redox process.
attractive for drug delivery to the ROS-overproduced cells. Moreover, VC and GR/NADPH pair are widely present in various cells, and PF2 could be applied as a broad spectrum probe to monitor the intracellular redox status. We further studied the redox-responsive fluorescence changes of the PF nanoparticles with different hydrophobic block lengths under triggering of H2O2 and VC with the intracellularly available concentration. For all the PF1−PF3 nanoparticles, the fluorescence intensity decreased gradually upon the addition of 100 μM H2O2. The changing rate in fluorescence followed the order of PF3 > PF2 > PF1, indicating a faster oxidation for the copolymer with a shorter hydrophobic block length (Figure 4A). Under the used conditions, the molar ratio of H2O2 to the selenide group was in the range of 0.3 for PF1 to 0.5 for PF3, which is consistent approximately with the declining degree in fluorescence. This result reveals that the oxidation of the selenide groups in the copolymer nanoparticles could be well reflected by the change of the fluorescence intensity. After the addition of 100 μM VC, the fluorescence recovered completely in ∼80 min for all three copolymers. Again, the reduction rate decreased with increasing the hydrophobic block length (Figure 4B). In addition, the redox reaction could be carried out at least four cycles without an obvious loss in fluorescence intensity (Figure 4C). In the absence of oxidant, the fluorescence of the nanoparticle solutions did not change for at least 2 h at 37 °C (Figure S26A). We observed the same phenomenon for the partly oxidized nanoparticles (Figure S26B). These results imply that the PF nanoparticles and their selenoxide derivatives could undergo the reversible transformation under the triggering of suitable oxidant and reductant but remain stable without redox stimulation. Morphology Change of PF Nanoparticles under Oxidation or Redox Stimulation. Stimuli-responsive deformation or dissociation of a polymeric nanoparticles is a key step for the triggered drug release. Herein, we established two stimulation modes, the monotonic oxidation and the alternative oxidation/reduction, to study the responsive
deformation of PF nanoparticles. As shown in Figure S3A, the intracellular ROS level in MDA-MB-231 cells was approximately half of that in HBL-100 cell treated with H2O2 of 200 μM. Therefore, 100 μM of H2O2 was applied for the monotonic oxidation in the following in vitro experiments, to mimic the sustained high ROS level in TNBC cells. On the other hand, 100 μM of H2O2 and 100 μM of VC was used alternately to simulate the transient up-and-down status of ROS in normal cells. The responsive changes in morphology of the PF nanoparticles were monitored by LLS and TEM. For the monotonic oxidation mode, upon stimulation with H2O2 for 12 h, all three nanoparticles underwent a rapid swelling and dissociation as proven by the decrease in scattered light intensity and the increase of Rh, Rg, and Rg/Rh, which was caused by the transformation from hydrophobic selenide to hydrophilic selenoxide (Figure S27). Consistent with the aforementioned fluorescence results, the oxidative deformation rate followed the order of PF3 > PF2 > PF1. TEM images also clearly showed the swelling and dissociation process of PF3 nanoparticles (Figure S28). After 60 min of oxidation, not only did the particle size increase but also the initial micellar nanoparticles transformed to a vesicular morphology. Further oxidation of the selenide groups resulted in a complete disruption of the PF3 nanoparticles. For the alternative oxidation/reduction mode, the initial oxidation by H2O2 (100 μM) resulted in the similar LLS results as discussed above until 60 min when VC (100 μM) was added. After 70 min, a partial recovery of the scattered light intensity was observed, which was accompanied by the slight decrease in both Rh and Rg, particularly for PF2 and PF3 nanoparticles (Figure 5). These results demonstrate the oxidation induced swelling or deformation at the initial stage (
