Cooperative Orthogonal Macromolecular ... - ACS Publications

Mar 17, 2016 - PE stains DENV NS1 glycoprotein that is expressed in the DENV-infected cells. ..... with Giemsa stain, modified solution (Sigma, St. Lo...
3 downloads 0 Views 4MB Size
Download PDF
Article pubs.acs.org/Macromolecules

Cooperative Orthogonal Macromolecular Assemblies with Broad Spectrum Antiviral Activity, High Selectivity, and Resistance Mitigation Koji Ichiyama,†,¶ Chuan Yang,‡,¶ Lakshmi Chandrasekaran,† Shaoqiong Liu,‡ Lijun Rong,§ Yue Zhao,§ Shujun Gao,‡ Ashlynn Lee,‡ Kenji Ohba,† Youichi Suzuki,† Yoshiyuki Yoshinaka,∥ Kunitada Shimotohno,⊥ Kei Miyakawa,# Akihide Ryo,# James Hedrick,*,○ Naoki Yamamoto,*,† and Yi Yan Yang*,‡ †

Translational ID Lab, Department of Microbiology, Yong Loo Lin School of Medicine, National University of Singapore, 14 Medical Drive, #15-02 Centre for Translational Medicine (MD6), Singapore 117599, Singapore ‡ Institute of Bioengineering and Nanotechnology, 31 Biopolis Way, Singapore 138669, Singapore § Department of Microbiology and Immunology (M/C 790), University of Illinois at Chicago, 835 S. Wolcott, Chicago, Illinois 60612, United States ∥ Department of Molecular Virology, Tokyo Medical and Dental University, Tokyo 113-8510, Japan ⊥ The Research Center for Hepatitis and Immunology, National Center for Global Health and Medicine, 1-7-1, Kohnodai, Ichikawa, Chiba 272-8516, Japan # Department of Microbiology, Yokohama City University School of Medicine, Kanagawa 236-0004, Japan ○ IBM Almaden Research Center, 650 Harry Road, San Jose, California 95120, United States S Supporting Information *

ABSTRACT: Treatment of viral infections continues to be elusive owing to the variance in virus structure (RNA, DNA, and enveloped and nonenveloped viruses) together with their ability to rapidly mutate and garner resistance. Here we report a general strategy to prevent viral infection using multifunctional macromolecules that were designed to have mannose moieties that compete with viruses for immune cells, and basic amine groups that block viral entry through electrostatic interactions and prevent viral replication by neutralizing the endosomal pH. We showed that cells treated with the antiviral polymers inhibited TIM receptors from trafficking virus, likely from electrostatic and hydrogenbonding interactions, with EC50 values ranging from 2.6 to 6.8 mg/L, depending on the type of TIM receptors. Molecular docking computations revealed an unexpected, and general, specific hydrogen-bonding interactions with viral surface proteins, and virus and cell binding assay demonstrated a significant reduction in infection after incubating virus or cells with the antiviral polymers. Moreover, the mannose-functionalized macromolecules effectively prevented the virus from infecting the immune cells. Representative viruses from each category including dengue, influenza, Chikungunya, Enterovirus 71, Ebola, Marburg, and herpes simplex were surveyed, and viral infection was effectively prevented at polymer concentrations as low as 0.2 mg/L with very high selectivity (>5000) over mammalian cells. The generality of these cooperative orthogonal interactions (electrostatic and hydrogen-bonding) provides broad-spectrum antiviral activity. As the antiviral mechanism is based on nonspecific supramolecular interactions between the amino acid residues and mannose/cationic moieties of the macromolecule, the ability to form the virus−polymer and polymer−cell assemblies can occur regardless of viral mutation, preventing drug resistance development.

■

INTRODUCTION

decades have imposed an enormous economic burden. In addition, several new viral pathogens like Nipah, Chikungunya (CHIKV), and mutated avian influenza A(H7N9) virus have been found in the human population. More recently, the Ebola virus (EBOV) has become epidemic in West Africa, resulting in the loss of more than 11 thousand lives.1 Consequently,

Viral diseases continue to be one of the leading causes of morbidity and mortality since ancient times. In recent years, viral infections have emerged as an eminent global public health problem mainly because of a rapid increase in human population, aging, global warming, and medical treatments that suppress the immune system, including irradiation therapy, anticancer chemotherapy and organ transplantation. For example, the SARS outbreak in 2003 worldwide, dengue fever and bird flu (e.g., H1N1) outbreaks in Asia over the last two © 2016 American Chemical Society

Received: January 13, 2016 Revised: February 20, 2016 Published: March 17, 2016 2618

DOI: 10.1021/acs.macromol.6b00091 Macromolecules 2016, 49, 2618−2629

Article

Macromolecules

charges under the physiological conditions (pH 7.4) as there are about 20% of the amine groups protonated.26 The unprotonated amine groups can neutralize the endosomal pH. In addition, the large number of primary and secondary amine groups provides exquisite synthetic handles for subsequent transformations and installation of a wide variety of functional groups. In this way, polymer−virus and polymer− cell interactions can be tuned, as well as toxicity associated with the primary amines can be mitigated. Described herein is a unique approach that combines the use of an electrostatic targeting strategy toward both virus and cells with the neutralization of endosomal pH using PEI functionalized with mannose groups with the objective of exploiting competitive immune cell binding with virus. During the course of the study, we discovered through computational molecular docking studies and virus binding assays unexpected specific interactions between virus surface proteins and the mannosefunctionalized macromolecules that appear to be broadly applicable to a range of viruses. These interactions appear to be through multivalent hydrogen-bonding, which together with cooperative electrostatic interactions between the cationic polymer and the anionic region of virus surface (e.g., anionic phosphatidylserine, PS, which binds TIM-1/TIM-3 receptors on cell surface for entry2−4) prevented virus from infecting cells. In addition, the mannose-functionalized macromolecule was shown to bind TIM-1/TIM-3 receptors, thus inhibiting virus infection. Furthermore, the cationic residuals on the macromolecule can interact with anionic virus binding receptors such as heparan sulfate proteoglycans (for DENV,6 HSV7 and EV71,8) and sialic acid (for influenza virus9,10 and EV71,11). The two supramolecular interactions (hydrogenbonding and electrostatic interactions) are orthogonal (noninteracting and independent of one another), but cooperative in their ability to collectively bind the virus/cell. These supramolecular polymer-virus and polymer-cell interactions provide broad spectrum antiviral activity with high selectivity (i.e., negligible toxicity toward mammalian cells). As the antiviral mechanism is based on nonspecific supramolecular interactions between the amino acid residues on virus and cell surface and mannose/cationic moieties of the macromolecule, the ability to form the polymer−virus and polymer−cell assemblies can occur regardless of viral mutationsidestepping mutations that could lead to the onset of resistance. Exploiting cooperative hydrogen-bonding and electrostatic interactions of polymer− virus and polymer−cell as well as neutralization of the endosomal pH is believed to be a new concept that exploits supramolecular chemistry to prevent viral infection.

significant effort has been directed to develop vaccines and antiviral drugs to control and eradicate viral infections. However, the rapid mutation of viruses, due to inherent genomic instability, makes vaccinations inefficient. Moreover, for many viral infections (e.g., dengue virus (DENV), CHIKV, EBOV, and Marburg virus (MARV)), there are no drugs available in clinic. Since there are so many types and subtypes of pathogenic viruses, in particular RNA viruses that easily mutate to form drug-resistant strains, it is difficult to deal with them individually. There is a pressing need to develop general and safe strategies to prevent viral infections and mitigate the consequences of genetic mutations and the onset of resistance. Viruses can be classified into RNA (e.g., DENV, influenza, CHIKV, Enterovirus 71- EV71, EBOV, and MARV) and DNA (e.g., herpes simplex virus-HSV) viruses based on their genomes, or enveloped (e.g., DENV, influenza, CHIKV, EBOV, MARV, and HSV) and nonenveloped (e.g., EV71) viruses based on whether the viral particles are wrapped in a host-derived membrane or not, showing the complexity in attempting to design a general solution for prevention and treatment of viral infections. Most emerging and re-emerging viruses are RNA viruses. Viral particles infect cells by specific interactions between virus surface proteins and cell surface receptors including T-cell immunoglobulin and mucin domain (TIM-1/TIM-3) proteins for EBOV, MARV2,3 and DENV,4 mannose receptors (immune cells),5 heparan sulfate proteoglycans for DENV,6 HSV7 and EV71,8 sialic acid for influenza virus9,10 and EV71,11 followed by endocytosis internalization.12 Viruses spread viral genome and infection from cell to cell. Conceptually, the presence of electrostatically charged (both cationic and anionic) regions on the viral surface and negative charges on the cell surface receptors (e.g., heparan sulfate proteoglycans and sialic acid) makes the utilization of electrostatic interactions an attractive strategy to exploit in the targeting of both viral particles and cells. There are several reports that employed this strategy using anionic polymers such as sulfated polysaccharides, dextran and heparin,13,14 xylofuranan, ribofuranan, and Curdlan.15,16 Viral infection prevention was believed to result from nonspecific electrostatic interactions with the cationic charges on the viral surface. Cationic polymers including acrylate polymers, 17 linear polyethylenimines (PEIs)18 and poly(phenylene ethynylene)19 are also believed to prevent viral infections by electrostatic interaction. The cationic polymers such as linear PEI exhibit high nonspecific cytotoxicity toward mammalian cells and induce hemolysis. In addition, low pH in the endosome is required for the replication of a number of virus infections.12 Recently, niclosamide, an FDA approved antihelminthic compound, was reported to prevent infections of pH-dependent viruses by neutralizing the endosomal pH (5.0−6.5).20 However, its highest selectivity was only ∼24 against influenza virus (PR8) and human rhinovirus (HRV14). Ammonium chloride21,22 and chloroquine23 having pH neutralization ability were also reported to prevent viral infections, but they are highly toxic, limiting clinical applications. As an alternative to low molecular weight therapeutic agents, our attention was drawn to macromolecules where the pH neutralization capacity can be amplified via the molecular weight and architecture. Branched polyethylenimine is a hyperbranced polymer that is synthesized by the ring-opening polymerization of aziridine leading to a high number of primary, secondary and tertiary amine groups with a theoretical molar ratio of 1:2:1,24 whereas commercially available PEIs have a ratio closer to 1:1:1.25 PEIs carry cationic

■

EXPERIMENTAL SECTION

Materials. Branched PEI (Mn 10 kDa), oxalyl chloride, 2,3;5,6-diO-isopropylidene-D-mannofuranose, dry solvents, and other chemicals were obtained from Sigma-Aldrich and used without any further purification unless otherwise noted. Other solvents of analytical grade were purchased from Fisher Scientific and used as received. Synthesis of Protected Mannose-Functionalized Cyclic Carbonate (MTC-ipman). 5-Methyl-5-carboxyl-1,3-dioxan-2-one (MTC−OH) and the monomer MTC-ipman were synthesized as reported previously.27,28 Briefly, 50 mL of MTC−OH/dry tetrahydrofuran (THF) solution (2.75 g, 17.2 mmol) was added to 50 mL of oxalyl chloride/dry THF solution (2.48 mL, 19.0 mmol) drop by drop. To this mixture, were three drops of anhydrous dimethylformamide (DMF) added over 30 min under N2 atm, which was then stirred for 1 h while volatiles were removed by using N2 bubbles. The intermediate product MTC-Cl was harvested as solid after the solvent was removed 2619

DOI: 10.1021/acs.macromol.6b00091 Macromolecules 2016, 49, 2618−2629

Article

Macromolecules

Figure 1. Synthesis and characterization of antiviral polymers. (a) Synthetic scheme and chemical structure of mannose-functionalized carbonate modified PEI polymers. (b) Antiviral activity (EC50, effective concentration at which the polymer protects 50% LLC-MK2 cells from DENV-2 infection; determined by MTT assay), cytotoxicity (CC50, cytotoxic polymer concentration, at which 50% cells are killed; determined by MTT assay), selectivity index (SI, CC50/EC50), and pH neutralization capacity of unmodified and mannose-functionalized PEI polymers (i.e., percentage of amines required to neutralize the endosomal pH from 5.0 to 7.4). Prevention of DENV-2 infection in human primary peripheral blood mononuclear cells (PBMCs) (c) and macrophages (d) by PEI-man65. Both PBMCs and macrophages express mannose receptor, which mediates DENV infection. PE stains DENV NS1 glycoprotein that is expressed in the DENV-infected cells. Therefore, the fluorescence intensity of the infected cells is higher than that of the control cells without virus infection. After the treatments with PEI-man65 at 2 or 10 mg/L, the fluorescence intensity of the infected cells decreases to the level of the control cells, demonstrating that PEI-man65 is able to prevent the cells from DENV-2 infection. The mannose molecules in the polymer possibly compete with DENV for binding the mannose receptor, hence inhibiting infection. under vacuum, and dissolved in 50 mL of dry chloroform. Dry chloroform (50 mL) containing 2,3;5,6-di-O-isopropylidene-D-mannofuranose (ipman, 4.13 g, 15.8 mmol) and triethylamine (2.8 mL, 20.6 mmol) was dropped into the solution over 30 min at room temperature. The mixture was then heated to 40 °C and reacted for 48 h. The reaction solution was cooled down to room temperature and then concentrated before triethylamine salt was precipitated using 100 mL of THF. The filtrate was evaporated to yield the crude product, which was purified using a silica gel column (gradient elution of ethyl acetate and hexane from 20/80 to 50/50). The product as a sticky colorless oil slowly solidified to a white solid (5.85 g, 85%). 1H NMR (400 MHz, CDCl3, 22 °C): δ 6.22 (s, 1H, H-a), 4.89 (dd, 1H, H-b), 4.72 (d, 2H, H-c), 4.66 (m, 2H, H-c), 4.41 (m, 1H, H-d), 4.22 (m, 2H, H-e), 4.11 (dd,, 2H, H-e), 4.03 (m, 2H, Hf + H-g), 1.50−1.33 (5 s, 15H, −CH3 of ipman and H-i). Synthesis of MTC-Mannose Modified PEI (Figure 1, Figure S1). The synthesis of PEI-man65 (molar ratio of PEI to mannose: 1:65) was given as an example. MTC-ipman (0.302 g, 0.75 mmol) was added to 2 mL of PEI/dry dichloromethane solution (Mn = 10 kDa, 0.1 g, 0.01 mmol) ) in a glovebox, which was stirred for 1 h. Methanol and 1 M HCl(aq) (10 mL each) were added. The mixture was heated to reflux for 2 h, cooled down to room temperature. The crude product was purified via ultrafiltration using a Vivaspin 20 concentrator (MWCO = 5 k, Sartorius AG, Goettingen, Germany), followed by 3 washes with deionized (DI) water, and freeze-dried (0.19 g, 48%). 1H NMR (400 MHz, D2O, 22 °C) (Figure S1a): δ: 4.11 (s, 130H, H-d and H-e), 2.56−3.70 (br, m, 1190H, H-e, H-f, H-g, and H of bPEI), 1.10 (s, 195H, H-i). 13C NMR (100 MHz, D2O, 22 °C) (Figure S1b): δ 179.13 (Cj of MTC-man), 157.55 (Ck of carbamate

linkage), 62.36−71.96 (Ca to Cf of MTC-man), 35.39−52.66 (−CH2CH2− of PEI), 16.81 (−Ci of MTC-man). Using a similar protocol, PEI-man28, PEI-man51, PEI-man86, PEIman104, and PEI-man143 were synthesized at the PEI to MTC-mannose carbamate molar ratios of 1:28, 1:51, 1:86, 1:104, and 1:143 respectively. The successful synthesis of the polymers was demonstrated by 1H NMR and 13C NMR analyses. PEI-man28. Yield: 55%. 1H NMR (400 MHz, D2O, 22 °C): δ 4.08 (s, 56H, H-d and H-e), 2.55−3.65 (br, m, 1042H, H-e, H-f, H-g and H of bPEI), 1.06 (s, 84H, H-i). 13C NMR (100 MHz, D2O, 22 °C): δ 178.93 (C-j of MTC-man), 158.50 (C-k of carbamate linkage), 93.93 (C-a of MTC-man), 62.36−72.28 (C-b to C-f of MTC-man), 35.32− 52.89 (−CH2CH2− of PEI), 16.76 (−CH3 of MTC-man). PEI-man51. Yield: 63%. 1H NMR (400 MHz, D2O, 22 °C): δ 4.08 (s, 102H, H-d and H-e), 2.50−3.70 (br, m, 1134H, H-e, H-f, H-g and H of bPEI), 1.06 (s, 153H, H-i). 13C NMR (100 MHz, D2O, 22 °C): δ 180.31 (C-j of MTC-man), 158.57 (C-k of carbamate linkage), 62.50− 71.60 (C-b to C-f of MTC-man), 35.39−52.69 (−CH2CH2− of PEI), 16.92 (−CH3 of MTC-man). PEI-man86. Yield: 52%. 1H NMR (400 MHz, D2O, 22 °C): δ 4.07 (s, 172H, H-d and H-e), 2.54−3.78 (br, m, 1274H, H-e, H-f, H-g and H of bPEI), 1.04 (s, 258H, H-i). 13C NMR (100 MHz, D2O, 22 °C): δ 179.51 (C-j of MTC-man), 158.56 (C-k of carbamate linkage), 93.61 (C-a of MTC-man), 62.36−70.11 (C-b to C-f of MTC-man), 36.82− 52.67 (−CH2CH2− of PEI), 16.88 (−CH3 of MTC-man). PEI-man104. Yield: 47%. 1H NMR (400 MHz, D2O, 22 °C): δ 4.08 (s, 208H, H-d and H-e), 2.57−3.71 (br, m, 1346H, H-e, H-f, H-g and H of bPEI), 1.07 (s, 312H, H-i). 13C NMR (100 MHz, D2O, 22 °C): δ 178.82 (C-j of MTC-man), 158.54 (C-k of carbamate linkage), 93.93 2620

DOI: 10.1021/acs.macromol.6b00091 Macromolecules 2016, 49, 2618−2629

Article

Macromolecules Table 1. Broad-Spectrum Antiviral Activity of Polymers RNA virus

virus family

cell

polymer

DENV-1 DENV-2 DENV-4 CHIKV EV71 influenza virus (A/H3N2) Ebola Ebola Marburg DNA virus HSV-1 HSV-2

flaviviridae flaviviridae flaviviridae alpha virus enteroviridae orthomyxoviridae filoviridae filoviridae filoviridae

LLC-MK2 LLC-MK2 LLC-MK2 vero RD MDCK A549 A549 A549

PEI-man65 PEI-man65 PEI-man65 PEI-man65 PEI-man65 PEI-man51 PEI-man28 PEI-man143 PEI-man143

herpesviridae herpesviridae

vero vero

PEI-man65 PEI-man65

EC50a (mg/L)

CC50b (mg/L)

selectivity index (CC50c/EC50d)

± ± ± ± ± ± ± ± ±

>1000 >1000 >1000 >1000 >1000 >1000 42 >1000 >1000

>5000 >3225 >3125 >143 >909 >909 3.5 >213 >909

>1000 >1000

>625 >196

0.20 0.31 0.32 7.0 1.1 1.1 12.1 4.7 1.1

0.17 0.06 0.02 0.5 0.1 0.3 6.0 0.8 0.2

1.6 ± 0.2 5.1 ± 0.2

a

EC50: effective concentration, at which the polymer protects 50% cells from viral infection. bCC50: cytotoxic polymer concentration, at which 50% cells are killed. cCC50, determined by MTT assay. dEC50, determined by plaque forming assay. previously,29 using unmodified PEI (Mn 10 kDa) as a control. Briefly, the polymer was dissolved in NaCl solution (150 mM), and pH was brought down to 2 using 0.01 N HCl. The polymer solution was then titrated to pH 10 using 0.01 N NaOH solution in an auto titrator (Spectralab Instruments). The pH neutralization capacity is defined as the percentage of amines required to neutralize the endosomal pH from 5.0 to 7.4 (extracellular pH), and is calculated by the formula: percentage of amines (%) = 100 × (ΔVNaOH × 0.01)/N. Here ΔVNaOH is the volume of NaOH solution (0.01N), which is required to increase the pH from 5.0 to 7.4, and N is the moles of the total amine groups in each polymer molecule. Cells. LLC-MK2 cells were cultured in Eagle’s minimum essential medium (EMEM) (Sigma, St. Louis, MO) supplemented with 10% fetal bovine serum (FBS) (Invitrogen, Carlsbad, CA), 100 units/mL penicillin and 100 μg/mL streptomycin at 37 °C in 5% CO2. Aedes albopictus C6/36 cells were maintained in RPMI-1640 medium (Invitrogen, Carlsbad, CA) with 25 mM HEPES supplemented with 10% FBS, 100 U/mL penicillin and 100 μg/mL streptomycin, and incubated at 28 °C. A549 replicon cells (containing DENV-2 NS genes), Vero, MDCK and RD cell lines were grown in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% FBS, 100 U/mL penicillin and 100 μg/mL streptomycin at 37 °C in 5% CO2. Human PBMCs were isolated from K2EDTA anticoagulated blood, which was obtained from healthy donors, by density gradient centrifugation using Ficoll-Paque (GE-Healthcare, Piscataway, NJ) according to the manufacturer’s instructions. PBMCs were stimulated by PHA-P (Sigma, St. Louis, MO) and maintained in RPMI-1640 supplemented with Interleukin 2 (Sigma, St. Louis, MO), 20% FBS, 100 U/mL penicillin and 100 μg/mL streptomycin at 37 °C in 5% CO2. CD14+ monocytes were isolated from PBMCs by negative selection using the EasySep human monocyte enrichment kit (STEMCELL Technologies, Vancouver, Canada) according to the manufacturer’s instructions. Monocyte purity was >90% as determined by flow cytometry using anti-human CD14 Clone M5E2 (STEMCELL Technologies, Vancouver, British Columbia). Macrophages (MDM) were derived from the monocytes by culturing them in RPMI-1640 medium containing 50 ng/mL M-CSF (BioLegend, San Diego, CA), 20% FBS, 100 U/mL penicillin, and 100 μg/mL streptomycin at 37 °C in 5% CO2 for 7 days. Viruses. DENV-1 and DENV-4 (isolated from clinical samples) as well as DENV-2 (laboratory-adapted New Guinea C (NGC) strain) were kindly given by Dr. Justin JH Chu (National University of Singapore). HSV-1, HSV-2, and influenza A virus (H3N2) were obtained from Lars Heinig (National University of Singapore). The viruses were propagated in various cell lines as described in Table 1. The supernatant from infected cells was centrifuged to remove cell debris, then aliquoted and stored at −80 °C. Antiviral Activity Based on Cytotoxicity Assay. Antidengue virus activity and cytotoxicity of polymers were investigated by

(C-a of MTC-man), 60.85−72.29 (C-b to C-f of MTC-man), 35.38− 52.67 (−CH2CH2− of PEI), 16.77 (−CH3 of MTC-man). PEI-man143. Yield: 59%. 1H NMR (400 MHz, D2O, 22 °C): δ 4.08 (s, 286H, H-d and H-e), 2.53−3.72 (br, m, 1502H, H-e, H-f, H-g and H of bPEI), 1.08 (s, 429H, H-i). 13C NMR (100 MHz, D2O, 22 °C): δ 179.13 (C-j of MTC-man), 158.54 (C-k of carbamate linkage), 64.23− 70.11 (C-b to C-f of MTC-man), 36.83−52.67 (−CH2CH2− of PEI), 16.85 (−CH3 of MTC-man). 1 H NMR Spectroscopy. 1H NMR spectra of all polymers were recorded on a Bruker Advance 400 NMR spectrometer at 400 MHz at room temperature. The 1H NMR measurements were carried out with an acquisition time of 3.2 s, a pulse repetition time of 2.0 s, a 30° pulse width, 5208-Hz spectral width, and 32 K data points. Chemical shift was referred to the solvent peak (δ = 4.70 for D2O). 13 C NMR Spectroscopy. 13C NMR spectra of all polymers were recorded on a Bruker AV-400 NMR spectrometer at 100 MHz at room temperature. The 13C NMR analysis was performed by composite pulse decoupling with an acquisition time of 0.82 s, a pulse repetition time of 5.0 s, a 30° pulse width, 20 080-Hz spectral width, and 32 K data points. FTIR. FTIR analysis was conducted on a PerkinElmer FTIR 2000 spectrometer in the region of 4000−600 cm−1 at a resolution of 2 cm−1 at room temperature for all the polymers to further confirm the chemical transformation, and a total of 64 interferograms were signalaveraged. Samples were prepared by depositing the samples on the surface of a KBr plate. Figure S2 shows a typical FTIR spectrum of PEI-man65. The characteristic peak of PEI at 1536 cm−1 (−N−H stretching) and that of mannose at 1044 cm−1 (−C−O stretching) are clearly seen. The peak at 1710 cm−1 is attributed to − CO stretching of carbamate linkages formed between PEI and MTCmannose, and the ester groups of MTC-mannose. A similar spectrum was also obtained for the rest polymers. This analysis further confirms the successful synthesis of MTC-mannose modified PEI polymers. Fluorescence Spectroscopy Analysis. To evaluate if the polymers aggregate in aqueous solution, the critical micelle concentration (CMC) of the polymers in DI water was analyzed by using a Fluoromax-4 spectrofluorometer at room temperature. Polymer concentration ranged from 0.01 mg/L to 2000.0 mg/L with pyrene concentration fixed at 6.16 × 10−7 M. The polymer/ pyrene solution was equilibrated for 24 h at room temperature (~22 °C). The excitation spectrum of the solution was obtained from 300 to 360 nm at an emission wavelength of 395 nm, and both excitation and emission bandwidths were 2.5 nm. As shown in Figure S3, there was no shift in the peak at 337 nm in the tested concentration range, indicating that there is no change in vibration structure of pyrene emission, and thus PEI-man143 does not aggregate or form micelles up to 2000 mg/L, well above the effective concentrations of the polymer. A similar phenomenon was also observed for the rest polymers. Titration Experiments. pH neutralization capacity of mannosefunctionalized PEIs was examined according to a protocol reported 2621

DOI: 10.1021/acs.macromol.6b00091 Macromolecules 2016, 49, 2618−2629

Article

Macromolecules

which 50% inhibition is achieved. The viability of noninfected cells after 48 h of incubation with polymers at the corresponding concentrations was analyzed by MTT assay using EnVision 2103 Multiplate Reader (PerkinElmer Inc., U.S.A.) to obtain CC50. Replicon Luciferase Assay. A549 cells having a luciferasereporting replicon of DENV-2 were seeded in 6-well plates at a cell density of 25 × 104 cells/well. The cells were treated with 50 mg/L of PEI-man65, 20 μM and 50 μM of NITD008 (as a control compound), or a medium containing DMSO, and incubated for 48 h at 37 °C in 5% CO2. NITD008 was not water-soluble, and a small amount of DMSO was used to promote its dissolution in the cell culture medium. Luciferase activity was analyzed with the Renilla luciferase assay system (Promega, Madison, WI), and results were normalized by the amount of protein. Time of Polymer Addition. LLC-MK2 cells were seeded in 6-well plates at 3 × 105 cells/well. The cells were incubated with DENV-2 (100 pfu per well) at 4 °C for 90 min under shaking. Subsequently, the cells were washed three times with cold PBS, and cultured in a 37 °C incubator for 5 days. Culture medium containing 50 mg/L of PEIman65 or 100 mg/L of heparin (as a control compound) was added to the cells at predetermined time points (−1.5, 0, 1, 2, 3, 4, and 5 h). Viral inhibition activity was studied by the plaque assay. Experiments were performed in triplicates. Mean percentage inhibition was determined relative to the control, which was performed under the same conditions but without polymers. Polymer−Virus Binding Assay. DENV-2 was incubated with PBS (pH 7.4) or PBS containing PEI-man65 (50 mg/L, 1 × 106 pfu) at 4 °C for 1 h to allow the polymer to bind onto the virus. Unbound polymer molecules were then removed by filtration through a Vivaspin 500, 100 kDa molecular weight cut off (GE Healthcare, Buckinghamshire, U.K.) at 6000g for 15 min. The virus was subjected to the plaque forming assay as described above to determine titers. qPCR for Viral Entry Assay. PEI-man65 (0.4 and 100 μg/mL) and growth media (as a control) were incubated with DENV-2 and LLCMK2 cells for 1 h at 4 or 37 °C. The cells were washed thrice with cold PBS, followed by RNA extraction using RNAeasy kit (QIAGEN, Germantown, MD), cDNA synthesis (Bio-Rad, Hercules, CA) and real-time PCR (CFX96 Real-Time PCR detection system, Bio-Rad, Hercules, CA). Primers used for qPCR were as follows:31 DENV-F (5′-TCAATATGCTAAAACGCGCGAGAAACCG-3′) DENV2-R (5′-CGCCACAAGGGCCATGAACAG-3′) Polymer−Cell Binding Assay. LLC-MK2 cells in 6-well plates (seeding density: 3 × 105 cells/well) were treated with PEI-man65 (50 mg/L) at 37 °C for 2 h, 1 h, 30 min, and 15 min and washed with PBS gently. The cells were then infected with DENV-2 (100 pfu per well) for 90 min and inoculated at 37 °C for 5 days prior to plaque forming assay. Control infection without the polymer was set to 100%. The data were expressed as the mean of three individual experiments ± SD. Pretreatment of Cells with Polymer. The respective cells were seeded into a 6-well plate. After 24 h when 80−90% confluence was reached, the cells were treated with PEI-man65 at various concentrations for 1 h, and washed with PBS. The cells were then infected with virus suspension (100 pfu/well). Plaque reduction assay was performed as described above to calculate the EC50 of PEI-man65 under the pretreatment conditions. Cell Fusion Inhibition Assay. This assay was performed to study if polymers can inhibit cell fusion at low pH. C6/36 cells were seeded in 6-well plates one day before the assay. DENV-2 was added to the seeded C6/36 cells at MOI of 0.03 together with either 50 mg/L of PEI-man65 or medium, and incubated for 90 min at 4 °C. The plates were then washed 3 times with PBS. Fresh medium containing the polymer was added to the plates, and the cells were incubated at 28 °C for 2 days. Then, 50 μL of 0.5 M 2-(N-morpholin) ethanesulfonic acid (MES) (pH 5.0) (Sigma, St. Louis, MO) was added to acidify the medium, followed by 2-day incubation at 28 °C. Cells were stained with Giemsa stain, modified solution (Sigma, St. Louis, MO) according to manufacturer’s protocol. The stained cells were analyzed under a light microscope (CKX 31 microscope, Olympus, Tokyo, Japan). In another set of experiments, virus and cells were incubated for 90 min at 4 °C before the polymer was added.

infecting LLC-MK2 cells with DENV. Infected (MOI: 0.5) and noninfected cells were exposed to polymers, heparin (sodium salt from porcine intestinal mucosa, Sigma-Aldrich, Product No. H3393, Batch No. SLBD0213 V, as a control compound) or ribavirin (SIGMA, St. Louis, MO, as a control compound) at different concentrations (cell seeding density in 96-well plate: 2000 per well), and allowed to proliferate for 5 days. Cell viability was analyzed by the 3-[4,5dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) (SIGMA, St. Louis, MO) assay to obtain EC50 (from infected cells) and CC50 (from noninfected cells). Prevention of PBMCs/Macrophages from DENV-2 Infection Investigated by FACS Analysis. PBMCs/macrophages, DENV-2 infected PBMCs/macrophages and DENV-2 infected PBMCs/macrophages that were treated with PEI-man65 at 2 or 10 mg/L were fixed with 4% paraformaldehyde in PBS (pH 7.4, SIGMA, St. Louis, MO) for 5 min and treated with 0.5% saponin and 2% Fc-receptor blocking solution (Human Trustain FcX, BioLegend, San Diego, CA) in PBS at 4 °C for 30 min. Immunostaining of the cells was performed with 1/50 dilution of mouse monoclonal antibody to Dengue virus NS1 glycoprotein (AB41623, Abcam, Cambridge, U.K.) at 4 °C for 30 min, followed by staining with 1/100 dilution of goat antimouse IgG/ PE conjugate (SC-3738, Santa Cruz Biotechnology, Santa Cruz, CA) at 4 °C for 30 min. The cells were analyzed with a flow cytometer (Becton-Dickinson, Franklin Lakes, NJ). Plaque Forming Assay. Each cell monolayer with 80−90% confluence was infected in a 6-well plate with 10-fold serial dilutions of virus for 90 min at 37 °C in 5% CO2 in the respective serum-free medium with gentle shaking every 15 min. The medium was replaced with 0.8% methylcellulose (CALBIOCHEM, San Diego, CA) in the respective maintenance medium supplemented with 10% FBS, 100 U/ mL penicillin and 100 μg/mL streptomycin at 37 °C in 5% CO2. At 3 (MDCK and RD cells) or 5 (LLC-MK2, Vero and A549 cells) days post-inoculation, cells were fixed using 4% paraformaldehyde in PBS (pH 7.4) at room temperature for 20 min. After that, the cells were washed with water before being stained using 1 mL of 1% crystal violet (Sigma, St. Louis, MO) at room temperature for 20 min. The plates were washed and dried for counting the plaque forming units per millilitre (pfu/mL). Plaque Reduction Assay. For EC50 determination of polymers, heparin (as a control compound) or oseltamivir phosphate (Tamiflu, Hoffmann-La Roche Inc., as a control compound), virus was added to respective cells in 6-well plates (80−90% confluence) with polymers, heparin or oseltamivir phosphate at various concentrations (100 pfu per well), and incubated at 37 °C for 90 min. The medium was then removed, and the cells were incubated with medium containing polymers, heparin or oseltamivir phosphate at the corresponding concentrations at 37 °C. Following the plaque forming assay as described above, the plaque forming units per milliliter (pfu/mL) were counted as a function of polymer, heparin or oseltamivir phosphate concentration and calculated to obtain EC50. Inhibition of EBOV and MARV Infections. Pseudotyped HIV-1 with envelope substituted by Ebola or Marburg virus glycoprotein was used to evaluate the antiviral activity of polymers against EBOV and MARV entry following a previously reported protocol.30 Briefly, human A549 cells were seeded at 9 × 104 cells/well in a 24-well plate and cultured in complete DMEM without phenol red 1 day prior to the infection experiment. Polymer was dissolved in PBS (pH 7.4) at a stock concentration of 10000 mg/L, and then diluted to various concentrations between 0.6 and 10000 mg/L. The growth medium was replaced with 495 μL of virus in fresh medium and 5 μL of polymer solution. The final polymer concentrations range from 0.006 to 100 mg/L. PBS (5 μL) without polymer was used as a control. The cells were incubated for 48 h at 37 °C. The cell culture medium was removed, and rinsed with PBS before 100 μL of lysis buffer was added to each well. The mixture was placed on a shaker for 30 min. The supernatant (5 μL) was mixed with 25 μL of luciferase substrate by briefly votexing for the measurement of relative light units using FB 12 Tube Luminometer (Titertek-Berthold, Germany). Inhibition of viral infection was obtained using the formula: [(RLUcontrol − RLUpolymer)/ RLUcontrol] × 100%. EC50 is defined as the polymer concentration, at 2622

DOI: 10.1021/acs.macromol.6b00091 Macromolecules 2016, 49, 2618−2629

Article

Macromolecules

in the 1H NMR spectrum (Figure S1a) were compared to determine the number of MTC-mannose groups. When excess MTC-ipman relative to the PEI primary amines was used, the secondary amines might also react with the cyclic carbonate, as in the case of functionalization with 86, 104, and 143 mannosecarbamate. Molecular weights obtained from 1H NMR analysis are listed in Table S1. The chemical transformation of the mannose-functionalized PEI polymers was further corroborated by 13C NMR (Figure S1b) and Fourier transform infrared spectroscopy (FTIR, Figure S2). The mannose-functionalized PEI polymers are stable when stored at room temperature (∼22 °C). From the fluorescence spectroscopy analysis (Figure S3), there was no aggregation of polymers in aqueous solution. To evaluate pH neutralization capacity of PEI polymers modified with a varying degree of mannose functional groups, acid−base titration experiments were performed. The percentage of amines required to neutralize the endosomal pH from 5.0 to 7.4 increased with increasing mannose substitution degree due to a reduced number of amine groups, implying a lower pH neutralization capacity (Figure 1b). Antiviral activity of PEI and mannose-functionalized PEI polymers was investigated by MTT assay using LLC-MK2 cells infected with dengue virus-2 (i.e., DENV-2). The polymers inhibited DENV-2 replication in a dose-dependent manner with tremendously low EC50 (i.e., effective concentration, at which the polymer protects 50% cells from viral infection) ranging from 0.12 to 50.5 mg/L (Figure 1b). The control compounds heparin16 and ribavirin,34,35 which were reported to have activity against DENV-2, had EC50 values of 6.6 ± 1.6 and 6.5 ± 1.2 mg/L respectively, indicating that PEI-man28, PEI-man51, PEI-man65, PEI-man86, and PEI-man104 were more effective. Polymers with higher mannose substitution degrees had less cationic charges and lower capacity to neutralize the endosomal pH (Figure 1b), which reduced antiviral activity. Although PEI had lower EC50 as compared to mannosefunctionalized PEI, it was the most cytotoxic with CC50 (i.e., cytotoxic polymer concentration, at which 50% cells are killed) of 5.44 mg/L. Mannose substitution effectively mitigated PEI cytotoxicity. For example, CC50 increased to higher than 1000 mg/L when PEI was functionalized with 65 mannose groups (PEI-man65). Among all polymers, PEI-man65 had the highest selectivity index (SI, i.e. CC50/EC50) of 3333 for DENV-2. The antiviral activity of PEI-man65 against DENV-2 was further evaluated in clinically relevant human primary peripheral blood mononuclear cells (PBMCs) and macrophages. The polymer effectively prevented the cells from DENV-2 infection (Figure 1, parts c and d). Both macrophages and PBMCs express mannose receptor, which mediates DENV infection.5,36,37 The mannose molecules in the polymer possibly compete with DENV for binding the mannose receptor, hence inhibiting infection. Broad spectrum antiviral activity of the mannose-functionalized PEI polymers was assessed by testing them in other serotypes of dengue virus (DENV-1 and 4), HSV-1, HSV-2, CHIKV, influenza virus (A/H3N2), EV 71, and pseudotyped HIV-1 with envelope substituted by EBOV or MARV glycoproteins. PEI-man65 was very active against DENV-1, DENV-2, and DENV-4 with EC50 values of 0.20, 0.32, and 0.31 mg/L, and selectivity indices of >5000, >3225, and >3125, respectively (Table 1). In addition, PEI-man65 also effectively inhibited infections of respective target cells with HSV-1, HSV2, CHIKV, and EV 71. Although PEI-man65 was ineffective against influenza infection, PEI-man51 was more potent and

Inhibition of Virus Infection in TIM-1/TIM-3 Expressing 293T cells. For establishment of the 293T cells stably expressing the TIM-1 or TIM-3 protein, lentiviral vectors were generated as previously described.32 293T cells were transfected with pseudoviruses carrying the desired ORF. Cells with the expression of TIM-1 or TIM-3 were selected by blasticidin (InvivoGen, Toulouse, France). 293T cells stably express TIM-1 or TIM-3 due to integration of TIM-1 or TIM-3 expression genes. Parental (with empty vector), TIM-1 and TIM-3 expressing 293T cells were challenged with CHIKV at multiplicity of infection (MOI) 0.1 along with PEI-man65 (50 mg/L). Supernatants were collected 48 h post infection. Virus titers were determined by plaque assay as described above and expressed as average plaque forming unit per ml (pfu/mL) of three replicates ± SD. Evaluation of Drug Resistance. Drug resistance was studied by passaging DENV-2 on LLC-MK2 cells or EV 71 on RD cells in the presence of PEI-man65 at EC50. Cells (seeding density: 3 × 105 per well for LLC-MK2; 1.2 × 106 per well for RD) in 6-well plates were infected by DENV-2 or EV 71 (obtained from the previous passaging) at MOI of 0.1 and MOI of 0.01, respectively, in the presence of PEIman65 at EC50 (0.31 mg/L for DENV-2; 1.1 mg/L for EV 71). After 3 (RD cells) or 5 (LLC-MK2 cells) days postinfection, the cells and virus were collected and spun down using a high-speed temperatureregulated microcentrifuge (MX-305, TOMY, Tokyo, Japan) at 4 °C, 46000 g for 30 min. The virus supernatant was then collected and kept at −80 °C. After plaque forming assay was done to determine the viral titers in the supernatant, the virus supernatant collected was employed to infect the respective cells in the presence of PEI-man65 at 0.31 mg/L for DENV-2 and 1.1 mg/L for EV 71. Plaque reduction assay was performed as described above to determine EC50 of PEI-man65 against the virus of first passage. The virus of first passage was used to infect the cells in the presence of PEI-man65 at 0.31 mg/L for DENV-2 and 1.1 mg/L for EV 71 to give second passage of virus for the determination of EC50. The virus was subjected to PEI-man65 for 5 times, EC50 was determined for each passaging to evaluate if there is any change in EC50 value after repeated treatment with PEI-man65.

■

RESULTS AND DISCUSSION

A selected number of the primary and/or secondary amine groups on branched PEI (Mn 10 kDa) were functionalized with a protected mannose-substituted cyclic carbonate (denoted as MTC-ipman) (Figure 1a). The ring-opening reaction generates a carbamate linkage by the nucleophilic addition of the amine on PEI, a free hydroxyl group and an appended protected mannose functionality. PEI polymers were decorated with the carbohydrates after deprotection of the mannose groups in a mixture of methanol and 1 M HCl(aq) (PEI-mannose carbamate). An additional benefit of this functionalization is a significant reduction in toxicity.27,33 The number of amine groups substituted with mannose carbamate was varied from 28, 51, 65, 86, 104 to 143 (PEI-man28, PEI-man51, PEI-man65, PEI-man86, PEI-man104, and PEI-man143) to systematically study its effect on antiviral activity. It is expected that this substitution is increasingly affected by steric hindrance with an increased substitution degree, requiring optimized conditions to obtain the desirable substitution degree. For synthesis of PEIman28, PEI-man51 and PEI-man65, the reaction of the nucleophilic PEI with the cyclic carbonate was conducted for 1 h under the ambient conditions in an atom-economy transformation without using a catalyst, whereas longer reaction time and higher temperature (at 40 °C and 3 h) were required for the synthesis of PEI-man86, PEI-man104, and PEI-man143. The substitution was monitored and quantified using 1Hnuclear magnetic resonance (1H NMR). For example, relative integral intensities of a broad peak (2.56−3.70 ppm), which are attributed to the protons of PEI methylene and H-e, H-f, and H-g of mannose moieties, and MTC methyl signals (1.10 ppm) 2623

DOI: 10.1021/acs.macromol.6b00091 Macromolecules 2016, 49, 2618−2629

Article

Macromolecules

Figure 2. Polymer-viral protein binding interactions predicted by the blind docking computation study. Proposed binding model between one representative unit of polymer and virus proteins (left column) and predicted hydrogen binding sites of one representative unit of polymer with amino acids in the virus proteins (right column), obtained by MVD. Binding between PEI and DENV-2 envelope protein is nonspecific as the ligand (one representative unit of PEI) binds to the protein at various sites in the same range of energy, while binding between PEI-man65/ PEI-man28 and the virus envelope proteins is specific as the ligand (one representative unit of PEI-man65/ PEI-man28) binds to the proteins at a specific binding site with the least energy through cooperative multivalent hydrogen bonding, which facilitates polymer−virus assembly and prevents viral entry.

were confirmed by detecting the cavities using the cavity detection algorithm in Molegro Virtual Docker (MVD, 2008.2.4.0 Molegro ApS Aarhus, Denmark) (Figure S5). The protein−polymer interaction was analyzed through the flexible receptor docking study. Interaction between a ligand (one representative unit of polymer, Figure 2) and a protein is defined as specific when the ligand binds to the protein at a specific binding site with the least energy, while interaction is defined as nonspecific when the ligand binds to the protein at various sites in the same range of energy. Five binding poses in each grid were obtained from the flexible docking, and by ranking Moldock and rerank scores of the poses, the one with the least score was considered to be the best docking pose. The best pose is predicted to be present in one of the five cavities detected by the MVD software, where the hydrogen bonds were estimated. For example, the amino acid residues involved in forming hydrogen bond between PEI-man65 and DENV-2 E protein are Thr 70, Thr 115, Asp 154, Gln 248, Gly 266, Ala 267, Thr 268, and Leu 277 (Figure 2, right panel). The specific binding site was predicted to be at the interface of the domain II and domain III of the E protein dimer and near the fusion loop of the E protein monomer (Figure 2, left panel). The docking results showed generality of these polymer-virus interactions, where PEI-man65 has specific hydrogen-binding

prevented influenza infection with a low EC50 concentration of 1.1 mg/L (Selectivity index: > 909). In contrast, heparin was ineffective against HSV-1, HSV-2, CHIKV, influenza virus (A/ H3N2), or EV 71 even up to 10 000 mg/L. Oseltamivir phosphate (Tamiflu) had lower activity against influenza virus A/H3N2 as compared to PEI-man51, and its EC50 was measured to be 7.0 ± 1.5 mg/L. Treatment with either PEIman28 or PEI-man143 successfully inhibited EBOV glycoproteinmediated infection with EC50 values of 12.1 and 4.7 mg/L, and selectivity indices of 3.5 and >213, respectively (Table 1, Figure S4). PEI-man143 with a higher number of mannose groups had stronger activity, indicating that except for cationic charge and pH neutralization, mannose also plays a role in preventing EBOV infection. MARV infection was effectively impeded by PEI-man143 with EC50 value of 1.1 mg/L and selectivity index of >909 (Table 1, Figure S4). These data demonstrate the broad-spectrum antiviral activity of the polymers and that potency can be tuned with mannose functionalization. Importantly, there was no significant cytotoxicity at the effective polymer concentrations. The antiviral mechanism of the polymers was explored. Polymer−viral protein or polymer−cell surface protein binding interactions were investigated by molecular docking computations. The binding sites predicted by the blind docking study 2624

DOI: 10.1021/acs.macromol.6b00091 Macromolecules 2016, 49, 2618−2629

Article

Macromolecules

Figure 3. Inhibition of viral entry into LLC-MK2 cells by PEI-man65. (a) Study on inhibition of DENV-2 subgenomic replicon cells. PEI-man65 or the control compound heparin did not inhibit replication of the DENV-2 subgenomic replicon, indicating that the antiviral polymer did not function at a late stage of DENV-2 life cycle. (b) Inhibition of DENV-2 infection during the viral attachment step and early postattachment step. No inhibition was observed for both polymer and heparin after the cells were infected with the virus for 2 h. (c) Polymer binding onto DENV-2 inhibits virus infection. (d) Pretreatment of the cells with the polymer at various time points inhibited DENV-2 infection. Dose-dependent effect of the pretreatment of LLC-MK2 and RD cells with the polymer on prevention of DENV-2 (e) and EV 71 infection (f), respectively. (g) Prevention of low pH-induced fusion of virus-infected C6/36 Aedes albopictus cells. Cells were incubated with DENV-2 for 90 min at 4 °C with or without PEI-man65 (50 mg/L), followed by incubation at 28 °C for 2 days to allow infection and at pH 5.0 for another 2 days. Unlike the control group without any treatment, no fused cells were found in the presence of the polymer (top images). The polymer effectively inhibited syncytia formation even when it was added at the point of viral infection (i.e., incubation at 28 °C) (bottom images), suggesting the ability of the polymer to prevent low pH-induced virus−cell membrane fusion and viral infection.

interactions with various viral proteins including E of DENV-2, E1 of CHIKV, VP1 of EV 71 and GD of HSV-1/HSV-2, PEI-

man51 with HA of influenza virus, and PEI-man28 with GP of EBOV (Figure 2, Figure S6, Table S2). In addition, unmodified 2625

DOI: 10.1021/acs.macromol.6b00091 Macromolecules 2016, 49, 2618−2629

Article

Macromolecules

Figure 4. Inhibition of viral infection in cells expressed with TIM-1 and TIM-3 receptors by PEI-man65. Prevention of CHIKV infection in 293T cells without TIM1/TIM3 receptor expression (a), 293T cells with TIM1 (b) and TIM3 (c) receptor expression, and DENV-2 infection in A549 cells that naturally express TIM1 receptor by PEI-man65 (d). Inserts in parts b and c show specific hydrogen-bonding interactions between PEI-man65 and TIM1/TIM3 proteins, respectively, as estimated by the computational molecular docking study. The polymer effectively prevents CHIKV infection in TIM-1 and TIM-3 expressing 293T cells, and inhibits DENV-2 infection in A549 cells that naturally express TIM-1 receptor, suggesting that the polymer is capable of inhibiting PS/TIM receptor binding through specific multivalent hydrogen-bonding interactions between the polymer and TIM-1/TIM-3 proteins, and electrostatic interaction between the cationic charges on the polymer and the negative charges on PS.

PEI-viral protein interaction was nonspecific (Figure 2). These computational results suggest that the mannose-functionalized polymer is capable of interacting with viral surface proteins through multivalent hydrogen-bonding interactions that facilitate polymer−virus assembly, collectively preventing viral entry. To study if the polymer exerts its antiviral activity at the early or late stage of virus life cycle, the effect of the macromolecule on the replication of DENV-2 subgenomic replicon, encoding only nonstructural viral proteins, was studied in luciferasereplicon transfected A549 cells. PEI-man65 and heparin did not inhibit replication of the DENV-2 subgenomic replicon, whereas the control replication inhibitor NITD00838 showed significant inhibition of replication (Figure 3a), suggesting that the antiviral polymer did not function at a late stage of DENV-2 life cycle. To examine if the polymer blocked the initial attachment step in cells or a downstream event in the viral entry process, the polymer or heparin was added together with DENV-2 to LLC-MK2 cells at 50 and 100 mg/L respectively for 90 min at 4 °C. Unbound polymer/heparin molecules were then removed by phosphate-buffered saline (PBS, pH 7.4) wash (three times) before the cells were incubated for 5 days at 37 °C. The polymer or heparin inhibited DENV infection completely (Figure 3b). The viral infection was also effectively prevented when the polymer or heparin was added to the cells just after the cells and virus were incubated at 4 °C for 90 min, and shifted to a 37 °C incubator for infection. When the polymer or heparin was added to the cells at 1 h postinfection, inhibition was significantly reduced with heparin being more effective. No

considerable inhibition was observed when the polymer or heparin was introduced at 2−5 h postinfection. These results demonstrated that the polymer was effective in preventing viral infection when added during the viral attachment step or early postattachment step, confirming the point of action at DENV entry (i.e., DENV−cell membrane fusion). To further understand antiviral mechanism, the effect of polymer-virus binding on the prevention of viral infection was investigated. DENV-2 was incubated with 50 mg/L of PEIman65 for 1 h at 4 °C for binding, and viral particles were then separated to measure virus titers by plaque assay. Virus binding with the polymer lowered virus titers by more than 40% (Figure 3c). In another study, the binding of polymer-treated DENV-2 to the cells was monitored by qPCR, showing that viral binding was significantly reduced by the polymer treatment (Figure S7). These data demonstrated that the treatment of virus with the polymer inhibited virus binding onto cells, preventing infection. In the cell culture medium (pH 7.4), the polymer carries positive charges as the majority of primary and secondary amine groups are protonable. In addition to the specific viral protein−polymer multivalent hydrogen-bonding interactions (Figure 2), the cationic charges in the polymer interact with the anionic charges on the envelope of virus through electrostatic interaction. These cooperative noncovalent interactions mask the virus, preventing the cells from viral infection. Conversely, concurrent incubation of the polymer, virus and cells at 50 mg/L, the polymer yielded complete inhibition of viral infection. The incomplete inhibition of virus titers after the pretreatment of the virus with the polymer at the same concentration (Figure 3c) implied 2626

DOI: 10.1021/acs.macromol.6b00091 Macromolecules 2016, 49, 2618−2629

Article

Macromolecules

TIM-3 expressing cells was high (>345 and >370 respectively). Moreover, PEI-man65 also effectively inhibited DENV-2 infection in A549 cells that naturally express TIM-1 receptor with an EC50 value of 6.8 mg/L and SI of >147 (Figure 4d). These findings suggest that the polymer is capable of inhibiting PS/TIM receptor binding. A computational molecular docking study on PEI-man65-TIM-1 and TIM-3 protein binding predicted specific hydrogen-bonding interactions (Figure 4, Table S3). In addition, it is also anticipated that there is electrostatic interaction between the cationic charges on the polymer and the negative charges on PS. Both electrostatic and hydrogen-bonding interactions might play roles in polymervirus and polymer-cell supramolecular assembly, which effectively prevented viral infection. Taken together, polymer−virus binding, polymer−cell membrane interaction and neutralization of the endosomal pH all play a role in the prevention of viral infection. Since the polymer has multiple modes of antiviral mechanisms, it is anticipated that resistance development can be mitigated. In order to investigate if repeated use of polymer develops drug resistance, DENV-2 or EV 71 was exposed to a sublethal dose of PEI-man65 (DENV-2:0.31 mg/L; EV 71:1.1 mg/L) 5 times. After each treatment, the viruses were passaged and used to determine EC50 value. EC50 values remained similar (DENV-2:0.31−0.35 mg/L; EV 71:0.79−1.1 mg/L). In addition, after 5 times of polymer treatment, the viruses were treated with the polymer at 100 mg/L, which completely inhibited DENV-2 and EV 71 infections. These results indicate that repeated treatment with PEI-man65 did not mediate resistance in DENV-2 or EV 71, showing the advantage of therapeutic activity through multiple antiviral mechanisms that employ multivalent hydrogen-bonding and electrostatic interactions. The in vivo toxicity of the antiviral polymers was investigated in mice. LD50 of polymer PEI-man51 and PEI-man65 (lethal dose at which half the mice are killed) was determined to be 313 and 463 mg/kg respectively via i.v. injection, indicating that the polymers have low toxicity. In addition, from the analysis of the levels of alanine transaminase (ALT), aspartate transaminase (AST), total bilirubin (TBIL), creatinine, urea nitrogen, and sodium and potassium ions in blood samples taken from PEI-man65-treated mice at 48 h and 14 days after i.v. injection, there was no significant difference in any of the parameters between the control group and the group receiving iv injection of PEI-man65 (Table S4), indicating that the polymer treatment did not lead to any acute liver or kidney damage, nor did it interfere with the electrolyte balance in the blood. It was also observed that the polymer treatment did not induce any abnormal color change in the serum or urine samples or cause lethality in mice. These results demonstrate that the polymer treatment was not toxic in mice even after i.v. injection, it is safe to handle for topical/external use, and has potential for in vivo application.

that there were other factors like binding between the polymer and anionic components on the cell membrane (e.g., anionic heparan sulfate proteoglycans,6 sialic acid11) and neutralization of the endosomal pH, which might contribute to the inhibition of viral infection. Although 65 primary/secondary amine groups were substituted by mannose group, PEI-man65 may still carry cationic charges at pH 7.4 as there are still a number of amine groups left, which are protonable as the microscopic pKa values of primary, secondary and tertiary amine groups are 9.64, 8.6, and 7.5 respectively.39 The pretreatment of LLC-MK2 cells using 50 mg/L PEI-man65 for 2 h or even 15 min, followed by removal of unbound or loosely bound polymer molecules, effectively prevented DENV-2 infection (Figure 3d). The EC50 value of the polymer against DENV-2 infection was 6.0 mg/L when the cells were pretreated with the polymer for 1 h before the virus was added (Figure 3e), which was higher than that when the polymer, cell and virus were incubated at the same time (EC50:0.30 mg/L, Figure 1b). The pretreatment of Vero and RD cells was also highly effective against CHIKV (Vero), HSV-1 (Vero), and EV 71 (RD) infections at EC50 values of 23.5, 1.8, and 0.80 mg/L, respectively (Figure 3f, Figure S8). This is possibly because the polymer interacts with virus binding receptors on the cell surface, which inhibits viral entry. For example, the polymer may compete with HSV-1 and EV71 for binding the anionic cell surface receptors such as heparan sulfate proteoglycans for HSV-17,40 and EV71,8 and/or sialic acid for EV7111 through electrostatic interaction, hence preventing virus infection. These collective findings indicate that both polymer−virus and polymer−cell supramolecular assemblies driven by the electrostatic and multivalent hydrogen-bonding interactions play an important role in blocking viral entry. In addition, the PEI polymers showed pH neutralization capacity (Figure 1b). They can adsorb protons in the endosomal environment after entering the cells together with virus, neutralizing the endosomal pH. This process is known to inhibit pH-dependent viral infections like DENV,41 CHIKV,42 influenza virus,43 HSV,44 EV71,45 and EBOV.46 To demonstrate that the polymer is able to inhibit low pH-induced virus infection, prevention of low pH-induced virus-infected cell membrane fusion was investigated. Virus-infected cell fusion is due to fusogenic viral proteins that are expressed on the cell surface after infection. C6/36 Aedes albopictus cells were incubated with DENV-2 for 90 min at 4 °C with or without PEI-man65 (50 mg/L), followed by incubation at 28 °C for 2 days to allow infection and at pH 5.0 for another 2 days. Unlike the control group without any treatment, no fused cells were found in the presence of the polymer (Figure 3g, top). The polymer inhibited syncytia formation even when it was added at the point of viral infection (i.e., incubation at 28 °C) (Figure 3g, bottom), suggesting that the polymer was indeed capable of preventing low pH-induced virus-cell membrane fusion and viral infection. TIM receptors on cells facilitate virus entry by directly interacting with phosphatidylserine (PS) on the viral envelope.2−4,47,48 Although lentiviral vector-mediated expression of TIM-1 and TIM-3 receptors in 293T cells significantly enhanced CHIKV infection (Figure S9), PEI-man65 prevented the infection with a comparable EC50 (2.9 and 2.7 mg/L, respectively) to the control without the expression of the receptors (EC50: 2.6 mg/L) (Figure 4a−c). In addition, the selectivity toward preventing CHIKV infection in TIM-1 and

■

CONCLUSION Functionalization of branched PEI with mannose−cyclic carbonate at optimal molar ratios through the rapid and facile nucleophilic addition chemistry produces polymers with strong and broad-spectrum antiviral activity, negligible toxicity and high selectivity toward virus particles over mammalian cells. The incorporation of mannose residues on the polymer was designed to reduce PEI toxicity and target the mannose receptor on the immune cells. In addition, the computational 2627

DOI: 10.1021/acs.macromol.6b00091 Macromolecules 2016, 49, 2618−2629

Article

Macromolecules

(5) Miller, J. L.; de Wet, B. J.; Martinez-Pomares, L.; Radcliffe, C. M.; Dwek, R. A.; Rudd, P. M.; Gordon, S. The mannose receptor mediates dengue virus infection of macrophages. PLoS Pathog. 2008, 4, e17. (6) Thaisomboonsuk, B. K.; Clayson, E. T.; Pantuwatana, S.; Vaughn, D. W.; Endy, T. P. Characterization of dengue-2 virus binding to surfaces of mammalian and insect cells. Am. J. Trop. Med. Hyg. 2005, 72, 375−383. (7) Navaratnarajah, C. K.; Miest, T. S.; Carfi, A.; Cattaneo, R. Targeted entry of enveloped viruses: measles and herpes simplex virus I. Curr. Opin. Virol. 2012, 2, 43−49. (8) Tan, C. W.; Poh, C. L.; Sam, I. C.; Chan, Y. F. Enterovirus 71 uses cell surface heparan sulfate glycosaminoglycan as an attachment receptor. J. Virol. 2013, 87, 611−620. (9) Couceiro, J. N.; Paulson, J. C.; Baum, L. G. Influenza virus strains selectively recognize sialyloligosaccharides on human respiratory epithelium; the role of the host cell in selection of hemagglutinin receptor specificity. Virus Res. 1993, 29, 155−165. (10) Pearce, M. B.; Jayaraman, A.; Pappas, C.; Belser, J. A.; Zeng, H.; Gustin, K. M.; Maines, T. R.; Sun, X.; Raman, R.; Cox, N. J.; Sasisekharan, R.; Katz, J. M.; Tumpey, T. M. Pathogenesis and transmission of swine origin A(H3N2)v influenza viruses in ferrets. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 3944−3949. (11) Su, P. Y.; Liu, Y. T.; Chang, H. Y.; Huang, S. W.; Wang, Y. F.; Yu, C. K.; Wang, J. R.; Chang, C. F. Cell surface sialylation affects binding of enterovirus 71 to rhabdomyosarcoma and neuroblastoma cells. BMC Microbiol. 2012, 12, 162−173. (12) Mercer, J.; Schelhaas, M.; Helenius, A. Virus entry by endocytosis. Annu. Rev. Biochem. 2010, 79, 803−833. (13) Mitsuya, H.; Looney, D. J.; Kuno, S.; Ueno, R.; Wong-Staal, F.; Broder, S. Dextran sulfate suppression of viruses in the HIV family: inhibition of virion binding to CD4+ cells. Science 1988, 240, 646− 649. (14) Baba, M.; Pauwels, R.; Balzarini, J.; Arnout, J.; Desmyter, J.; De Clercq, E. Mechanism of inhibitory effect of dextran sulfate and heparin on replication of human immunodeficiency virus in vitro. Proc. Natl. Acad. Sci. U. S. A. 1988, 85, 6132−6136. (15) Nakashima, H.; Yoshida, O.; Tochikura, T. S.; Yoshida, T.; Mimura, T.; Kido, Y.; Motoki, Y.; Kaneko, Y.; Uryu, T.; Yamamoto, N. Sulfation of polysaccharides generates potent and selective inhibitors of human immunodeficiency virus infection and replication in vitro. Jpn. J. Cancer Res. 1987, 78, 1164−1168. (16) Ichiyama, K.; Gopala Reddy, S. B.; Zhang, L. F.; Chin, W. X.; Muschin, T.; Heinig, L.; Suzuki, Y.; Nanjundappa, H.; Yoshinaka, Y.; Ryo, A.; Nomura, N.; Ooi, E. E.; Vasudevan, S. G.; Yoshida, T.; Yamamoto, N. Sulfated polysaccharide, Curdlan sulphate, efficiently prevents entry/fusion and restricts antibody-dependent enhancement of dengue virus infection in vitro: a possible candidate for clinical application. PLoS Neglected Trop. Dis. 2013, 7, e2188. (17) Alasino, R. V.; Bianco, I. D.; Vitali, M. S.; Zarzur, J. A.; Beltramo, D. M. Characterization of the inhibition of eveloped virus infectivity by the cationic acrylate polymer Eudragit E100. Macromol. Biosci. 2007, 7, 1132−1138. (18) Spoden, G. A.; Besold, K.; Krauter, S.; Plachter, B.; Hanik, N.; Kilbinger, A. F.; Lambert, C.; Florin, L. Polyethylenimine is a strong inhibitor of human papillomavirus and cytomegalovirus infection. Antimicrob. Agents Chemother. 2012, 56, 75−82. (19) Wang, Y.; Canady, T. D.; Zhou, Z.; Tang, Y.; Price, D. N.; Bear, D. G.; Chi, E. Y.; Schanze, K. S.; Whitten, D. G. Cationic phenylene ethynylene polymers and oligomers exhibit antiviral activity. ACS Appl. Mater. Interfaces 2011, 3, 2209−2214. (20) Jurgeit, A.; McDowell, R.; Moese, S.; Meldrum, E.; Schwendener, R.; Greber, U. F. Niclosamide is a proton carrier and targets acidic endosomes with broad antiviral effects. PLoS Pathog. 2012, 8, e1002976. (21) Ohkuma, S.; Poole, B. Fluorescence probe measurement of the intralysosomal pH in living cells and the perturbation of pH by various agents. Proc. Natl. Acad. Sci. U. S. A. 1978, 75, 3327−3331.

molecular docking study suggests that the mannose-functionalized polymers specifically bind to virus and cell surface proteins via multivalent hydrogen-bonding interactions. These macromolecules are effective against RNA, DNA, and enveloped or nonenveloped viral infections, i.e. DENV, HSV1, HSV-2, CHIKV, EV71, influenza virus, EBOV, and MARV at low concentrations, and prevent viral infection through multiple mechanisms including binding to viral and cell surfaces through cooperative noncovalent (hydrogen-bonding and electrostatic) interactions (inhibiting virus entry) and neutralizing the endosomal pH (mitigating low pH-induced virus-cell membrane fusion). More importantly, by targeting both viral proteins and host−virus interactions, the antiviral polymers mitigate drug resistance. Mannose-functionalized PEI polymers have no in vivo toxicity even via i.v. injection at high doses. This polymer−virus and polymer−cell assembly is believed to be a new concept that exploits supramolecular chemistry to prevent viral infection and mitigate resistance development.

■

ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.macromol.6b00091. Blind docking study of polymers with viral proteins and cell surface receptors, and in vivo toxicity studies (PDF)

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected] (J.L.H.). *E-mail: [email protected] (N.Y.). *E-mail: [email protected] (Y.Y.Y.). Author Contributions ¶

These authors contributed to the work equally.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This research was financially supported by the Institute of Bioengineering and Nanotechnology (Biomedical Research Council and JCO, Agency for Science, Technology and Research, Singapore), National University of Singapore, and IBM Almaden Research Center (U.S.A.).

■

REFERENCES

(1) Ebola situation report. World Health Organization. 28 October 2015. (2) Kondratowicz, A. S.; Lennemann, N. J.; Sinn, P. L.; Davey, R. A.; Hunt, C. L.; Moller-Tank, S.; Meyerholz, D. K.; Rennert, P.; Mullins, R. F.; Brindley, M.; Sandersfeld, L. M.; Quinn, K.; Weller, M.; McCray, P. B., Jr; Chiorini, J.; Maury, W. T-cell immunoglobulin and mucin domain 1 (TIM-1) is a receptor for Zaire Ebolavirus and Lake Victoria Marburgvirus. Proc. Natl. Acad. Sci. U. S. A. 2011, 108, 8426−8431. (3) Moller-Tank, S.; Kondratowicz, A. S.; Davey, R. A.; Rennert, P. D.; Maury, W. Role of the phosphatidylserine receptor TIM-1 in enveloped-virus entry. J. Virol. 2013, 87, 8327−8341. (4) Meertens, L.; Carnec, X.; Lecoin, M. P.; Ramdasi, R.; GuivelBenhassine, F.; Lew, F.; Lemke, G.; Schwartz, O.; Amara, A. The TIM and TAM families of phosphatidylserine receptors mediate dengue virus entry. Cell Host Microbe 2012, 12, 544−557. 2628

DOI: 10.1021/acs.macromol.6b00091 Macromolecules 2016, 49, 2618−2629

Article

Macromolecules (22) Greber, U. F.; Willetts, M.; Webster, P.; Helenius, A. Stepwise dismantling of adenovirus 2 during entry into cells. Cell 1993, 75, 477−486. (23) Blanchard, E.; Belouzard, S.; Goueslain, L.; Wakita, T.; Dubuisson, J.; Wychowski, C.; Rouillé, Y. Hepatitis C virus entry depends on clathrin-mediated endocytosis. J. Virol. 2006, 80, 6964− 6972. (24) Jones, G. D.; Langsjoen, A.; Neumann, S. M. M. C.; Zomlefer, J. The polymerization of ethylenimine. J. Org. Chem. 1944, 9, 125−147. (25) von Harpe, A.; Petersen, H.; Li, Y.; Kissel, T. Characterization of commercially available and synthesized polyethylenimines for gene delivery. J. Controlled Release 2000, 69, 309−322. (26) Suh, J.; Paik, H. J.; Hwang, B. K. Ionization of poly(ethylenimine) and poly(allylamine) at various pH′s. Bioorg. Chem. 1994, 22, 318−327. (27) Cheng, W.; Yang, C.; Hedrick, J. L.; Williams, D. F.; Yang, Y. Y.; Ashton-Rickardt, P. G. Delivery of a granzyme B inhibitor gene using carbamate-mannose modified PEI protects against cytotoxic lymphocyte killing. Biomaterials 2013, 34, 3697−3705. (28) Pratt, R. C.; Nederberg, F.; Waymouth, R. M.; Hedrick, J. L. Tagging alcohols with cyclic carbonate: a versatile equivalent of (meth)acrylate for ring-opening polymerization. Chem. Commun. 2008, 114−116. (29) Teo, P. Y.; Yang, C.; Hedrick, J. L.; Engler, A. C.; Coady, D. J.; Ghaem-Maghami, S.; George, A. J.; Yang, Y. Y. Hydrophobic modification of low molecular weight polyethylenimine for improved gene transfection. Biomaterials 2013, 34, 7971−7979. (30) Wang, J.; Cheng, H.; Ratia, K.; Varhegyi, E.; Hendrickson, W. G.; Li, J.; Rong, L. A comparative high throughput screening protocol to identify entry inhibitors of enveloped viruses. J. Biomol. Screening 2014, 19, 100−107. (31) Tripathi, N. K.; Shrivastava, A.; Dash, P. K.; Jana, A. M. in Diagnostic Virology Protocols; Methods in Molecular Biology 665; Stephenson, J. R., Warnes, A., Eds.; Springer: Berlin, Germany, 2011, Ch. 4. (32) Tahara-Hanaoka, S.; Sudo, K.; Ema, H.; Miyoshi, H.; Nakauchi, H. Lentiviral vector-mediated transduction of murine CD34hematopoietic stem cells. Exp. Hematol. 2002, 30, 11−17. (33) Yang, C.; Cheng, W.; Teo, P. Y.; Engler, A. C.; Coady, D. J.; Hedrick, J. L.; Yang, Y. Y. Mitigated cytotoxicity and tremendously enhanced gene transfection efficiency of PEI through facile one-step carbamate modification. Adv. Healthcare Mater. 2013, 2, 1304−1308. (34) Medical Management of Biological Casualties Handbook; United States Government Printing Office: 2011. p 115. (35) Takhampunya, R.; Ubol, S.; Houng, H. S.; Cameron, C. E.; Padmanabhan, R. Inhibition of dengue virus replication by mycophenolic acid and ribavirin. J. Gen. Virol. 2006, 87, 1947−1952. (36) Schaeffer, E.; Flacher, V.; Papageorgiou, V.; Decossas, M.; Fauny, J. D.; Krämer, M.; Mueller, C. G. Dermal CD14(+) Dendritic Cell and Macrophage Infection by Dengue Virus Is Stimulated by Interleukin-4. J. Invest. Dermatol. 2015, 135, 1743−1751. (37) Ampel, N. M.; Nelson, D. K.; Li, L.; Dionne, S. O.; Lake, D. F.; Simmons, K. A.; Pappagianis, D. The mannose receptor mediates the cellular immune response in human coccidioidomycosis. Infect. Immun. 2005, 73, 2554−2555. (38) Yin, Z.; Chen, Y. L.; Schul, W.; Wang, Q. Y.; Gu, F.; Duraiswamy, J.; Kondreddi, R. R.; Niyomrattanakit, P.; Lakshminarayana, S. B.; Goh, A.; Xu, H. Y.; Liu, W.; Liu, B.; Lim, J. Y.; Ng, C. Y.; Qing, M.; Lim, C. C.; Yip, A.; Wang, G.; Chan, W. L.; Tan, H. P.; Lin, K.; Zhang, B.; Zou, G.; Bernard, K. A.; Garrett, C.; Beltz, K.; Dong, M.; Weaver, M.; He, H.; Pichota, A.; Dartois, V.; Keller, T. H.; Shi, P. Y. An adenosine nucleoside inhibitor of dengue virus. Proc. Natl. Acad. Sci. U. S. A. 2009, 106, 20435−20439. (39) Borkovec, M.; Koper, G. J. M. Proton binding characteristics of branched polyelectrolytes. Macromolecules 1997, 30, 2151−2158. (40) Spear, P. Herpes simplex virus: receptors and ligands for cell entry. Cell. Microbiol. 2004, 6, 401−410. (41) van der Schaar, H. M.; Rust, M. J.; Chen, C.; van der EndeMetselaar, H.; Wilschut, J.; Zhuang, X.; Smit, J. M. Dissecting the Cell

Entry Pathway of Dengue Virus by Single-Particle Tracking in Living Cells. PLoS Pathog. 2008, 4, e1000244. (42) Bernard, E.; Solignat, M.; Gay, B.; Chazal, N.; Higgs, S.; Devaux, C.; Briant, L. Endocytosis of chikungunya virus into mammalian cells: role of clathrin and early endosomal compartments. PLoS One 2010, 5, e11479. (43) Wiley, D. C.; Skehel, J. J. The structure and function of the hemagglutinin membrane glycoprotein of influenza virus. Annu. Rev. Biochem. 1987, 56, 365−394. (44) Nicola, A. V.; Hou, J.; Major, E. O.; Straus, S. E. Herpes simplex virus type 1 enters human epidermal keratinocytes, but not neurons, via a pH-dependent endocytic pathway. J. Virol. 2005, 79, 7609−7616. (45) Hussain, K. M.; Leong, K. L.; Ng, M. M.; Chu, J. J. The essential role of clathrin-mediated endocytosis in the infectious entry of human enterovirus 71. J. Biol. Chem. 2011, 286, 309−321. (46) Nanbo, A.; Imai, M.; Watanabe, S.; Noda, T.; Takahashi, K.; Neumann, G.; Halfmann, P.; Kawaoka, Y. Ebolavirus is internalized into host cells via macropinocytosis in a viral glycoprotein-dependent manner. PLoS Pathog. 2010, 6, e1001121. (47) Mercer, J.; Helenius, A. Vaccinia virus uses macropinocytosis and apoptotic mimicry to enter host cells. Science 2008, 320, 531−535. (48) Morizono, K.; Chen, I. S. Role of phosphatidylserine receptors in enveloped virus infection. J. Virol. 2014, 88, 4275−4290.

2629

DOI: 10.1021/acs.macromol.6b00091 Macromolecules 2016, 49, 2618−2629