Water-Soluble Reactive Copolymers Based on Cyclic N-Vinylamides

Jul 2, 2015 - Functional and Interactive Polymers, Institute of Technical and Macromolecular Chemistry, RWTH Aachen University, D-52056 Aachen, German...
2 downloads 6 Views 8MB Size
Article pubs.acs.org/Macromolecules

Water-Soluble Reactive Copolymers Based on Cyclic N‑Vinylamides with Succinimide Side Groups for Bioconjugation with Proteins Huan Peng,†,§ Michael Kather,†,§ Kristin Rübsam,§ Felix Jakob,§ Ulrich Schwaneberg,‡,§ and Andrij Pich*,†,§ †

Functional and Interactive Polymers, Institute of Technical and Macromolecular Chemistry, RWTH Aachen University, D-52056 Aachen, Germany ‡ Institute for Biotechnology, RWTH Aachen University, D-52056 Aachen, Germany § DWI Leibniz Institute for Interactive Materials e.V., D-52056 Aachen, Germany S Supporting Information *

ABSTRACT: Reversible addition−fragmentation chain transfer (RAFT) copolymerizations of methacrylic acid Nhydroxysuccinimide ester and cyclic N-vinylamide derivatives (N-vinylpyrrolidone, N-vinylpiperidone, and N-vinylcaprolactam) were successfully performed with methyl 2-(ethoxycarbonothioylthio)propanoate as chain transfer agent (CTA). Effects of different reaction parameters, such as solvent type, temperature, and CTA-to-initiator (C/I) ratio, were studied to optimize the polymerization conditions in order to obtain copolymers with variable chemical composition, controlled molecular weight, and narrow polydispersity index (PDI). The solvent type has a high impact on the polymerization reaction, and a high C/I ratio decreases polydispersity as well as conversion. Increased steric hindrance through an enlarged lactam ring offsets the monomer reactivity. The controlled character of RAFT polymerization was evidenced by the low PDI of the copolymers and a linear relationship between conversion and molecular weight. Biohybrid nanogels were synthesized by direct coupling between reactive copolymers and enhanced green fluorescent protein (EGFP) or cellulase (CelA2_M2) at room temperature in a water-in-oil emulsion. The EGFP-conjugated nanogels were fluorescent, while the CelA2_M2 encapsulated in nanogels retained its catalytic activity, as demonstrated by the hydrolysis of 4-methylumbelliferyl-β-D-cellobioside. polymers with different structures, including block,10,11 star,12,13 brush,14,15 and comb copolymers.16,17 Furthermore, it tolerates a wide range of functional monomers and can be performed in various solvents,18,19 and even in water,20,21 within a wide temperature range. More importantly, RAFT does not require highly rigorous removal of oxygen22,23 and other impurities, which is a great advantage over other polymerization methods. The polymerization of N-hydroxysuccinimide ester is widely studied due to its high reactivity with amine groups in alkaline conditions, which renders it promising for potential biomedical applications. Savariar et al. synthesized copolymers by controlled polymerization of N-isopropylacrylamide with

1. INTRODUCTION In recent years, controlled radical polymerizations, which include atom transfer radical polymerization (ATRP),1,2 nitroxide-mediated polymerization (NMP),3,4 and reversible addition−fragmentation chain transfer (RAFT) polymerization,5,6 have drawn growing attention from industrial and academic areas. The past decades witnessed a tremendous increase in reports on controlled polymerization,7,8 suggesting rapid development and high research interest. ATRP probably is the most thoroughly studied technique, and many publications appeared since its development in 1995.9 However, the usually unresolved challenge of removing heavy metal catalysts employed in ATRP largely restricts its applications, especially in the area of biomedical materials. Recently, the attention of many polymer researchers was shifted to RAFT polymerization, since it enables them to synthesize © XXXX American Chemical Society

Received: May 4, 2015 Revised: June 6, 2015

A

DOI: 10.1021/acs.macromol.5b00947 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

(Acros, 99%), potassium ethyl xanthogenate (Aldrich, >98%), ethylenediamine branched polyethylenimine (PEI, Aldrich, Mn ≈ 600 by GPC), 4-methylumbelliferyl β-D-cellobioside (4-MUC, analytical grade, Aldrich), potassium phosphate buffer (Kpi buffer, pH 7.2, prepared by KH2PO4, K2HPO4, Aldrich), ethanol (VWR, 99.9%), dichloromethane (VWR, 99.9%), sulfate magnesium (VWR, >99%), 1,4-dioxane (1,4-D, Aldrich, anhydrous, 99.8%), N,Ndimethylformamide (DMF, Aldrich, anhydrous, 99.8%), diethyl ether (VWR, 99.9%), chloroform-d1 (CDCl3, Deutero GmbH, 99.8%), and dialysis tubes (MWCO ≈ 12−14 kDa, Carl Roth, Germany; MWCO = 50 and 100 kDa, Spectrum Laboratories, USA) were used as received. N-Vinylcaprolactam (VCL, Aldrich, 98%) and N-vinylpiperidone (VPI, BASF, 98%) were distilled and recrystallized from hexane before use. 2,2′-Azobis(2-methylpropionitrile) (AIBN, Aldrich, 98%) was recrystallized in ethanol before use. EGFP (with 19 solvent-accessible lysine residues on the surface, Mn ≈ 26 kDa) and CelA2_M2 (variant M2 (H288F), with 24 solvent-accessible lysine residues on the surface, Mn ≈ 69 kDa) were produced and purified as reported in previous literature.34,35 Deionized water was used for all experiments. 2.2. Measurements. Nuclear Magnetic Resonance (NMR). 1H NMR and 13C NMR spectra were recorded on a Bruker DPX-400 FT NMR spectrometer at 400 MHz. Deuterated chloroform (CDCl3) was used as solvent. The chemical shifts reported are all relative to the residual solvent peak. Gel Permeation Chromatography (GPC). GPC was performed on a combined GPC system with a high-performance liquid chromatography pump (Bischoff HPLC), a Jasco 2035-plus RI detector, and four MZ-DVB gel columns (30 Å, 100 Å, and 2 × 3000 Å) in series at 30 Å. A solution of DMF containing 1.0 g L−1 LiBr was used as eluent at a flow rate of 1.0 mL min−1. The molecular weights were calculated using a polystyrene (PS) calibration. Fourier Transform Infrared Spectroscopy (FTIR). FTIR spectra (resolution 4 cm−1) were recorded using a Nicolet NEXUS 670 FTIR spectrometer. Samples were prepared on silica plates at room temperature. For each spectrum, more than 200 scans were averaged to enhance the signal-to-noise ratio. Raman Spectroscopy. Raman spectra were recorded on Bruker RFS 100/S spectrometer. The laser used was Nd: YAG at 1064 nm wavelength at a power of 250 mW. On average, 1000 scans were taken at a resolution of 4 cm−1. For sample holding aluminum pans of 2 mm before were used. Software used for data processing was OPIS 4.0. Dynamic Light Scattering (DLS). DLS measurements were performed with a commercial laser light scattering spectrometer (ALV/DLS/SLS-5000) equipped with an ALV-5000/EPP multiple digital time correlator and laser goniometry system ALV/CGS-8F S/N 025 with a helium−neon laser (Uniphase 1145P, output power of 22 mW and wavelength of 632.8 nm) as a light source. Confocal Microscopy. The fluorescence images were recorded on a Leica SP8 confocal microscope (Leica, Germany). Transmission Electron Microscopy (TEM). TEM was measured on Zeiss LibraTM 120 (Carl Zeiss, Oberkochen, Germany). The electron beam accelerating voltage was set at 120 kV. A drop of the sample was trickled on a piece of Formvar carbon-coated copper grid. Before being placed into the TEM specimen holder, the copper grid was air-dried under ambient conditions. Bicinconinic Acid (BCA) Assay for Quantification of Protein in Nanogel. A calibration curve of defined concentrations of a protein solution was generated by standard BCA assay procedure. The biohybrid nanogel solution was measured in a multidetection microplate reader (FLUOstar Omega, BMG Labtech, USA) with absorbance at 562 nm. The absorbance value of the nanogel was fit in the curve to get the protein concentration. Catalytic Activity Assay. The experiments were performed in a fluorescence microtiter plate (MTP) reader (Tecan Infinite M1000 Pro, Tecan Group Ltd., Switzerland). The excitation and emission wavelengths were 330 and 450 nm, respectively, and the gain was 140 manual. Behavior of Copolymers in Aqueous Solutions. Cloud points of aqueous polymer solutions were determined by turbidity measurement in a Cary 100 Bio UV−visible spectrophotometer (Agilent, USA). The

methacrylic acid N-hydroxysuccinimide ester (MNHS) using 2cyano-2-propyl benzodithioate as chain transfer agent (CTA) with a defined molecular weight and low polydispersity.24 Yanjarappa and co-workers presented in their research the RAFT polymerization of N-(2-hydroxypropyl)methacrylamide and N-(methacryloyloxy)succinimide with 2-cyanoispropyl dithiobenzoate as CTA and its application in biofunctionalization.25 Schilli et al. polymerized N-(methacryloyloxy)succinimide by RAFT in dimethylformamide (DMF).26 However, the polydispersity index (PDI) was 1.46 at high CTA concentrations, indicating poor control over the molecular weight at increased CTA concentrations. Polymers based on cyclic N-vinylamide monomers, including N-vinylpyrrolidone, N-vinylpiperidone, and N-vinylcaprolactam, were widely used in various fields, such as controlled drug delivery,27−29 separation processes,30 and tissue engineering.31 Especially poly(N-vinylpiperidone)32 and poly(N-vinylcaprolactam),33 as thermoresponsive polymers or “smart” polymers, have raised much research interest due to their special properties in aqueous solutions. Therefore, the use of cyclic N-vinylamide monomers along with MNHS to design reactive and stimuli-responsive copolymers is a challenging task. Experimental data on controlled copolymerization of cyclic N-vinylamide monomers with MNHS have been scarcely reported in the literature so far. This is mainly due to the very different reactivity ratios of the above-mentioned monomers, leading to huge obstacles in controlling polymerization and polymer architecture. To further explore the family of smart functional polymers, we consider it interesting to synthesize polymers of cyclic Nvinylamide monomers with MNHS by RAFT. Bearing in mind the fast coupling of N-hydroxysuccinimide ester with amine group at physiological conditions and special thermoresponsive properties of N-vinylamide monomers, the copolymers could exhibit unique properties and promise various applications. To avoid disadvantages caused by different reactivity and get a regular copolymer structure, MNHS was supplemented into the reaction solution by a syringe pump at a constant speed. Using methyl 2-(ethoxycarbonothioylthio)propanoate as CTA and AIBN as initiator, the RAFT reactions were carried out in different solvents (anisole, 1,4-dioxane, and DMF) and at different temperatures (60, 70, and 80 °C) to optimize reaction conditions and study polymerization kinetics. Results presented in this paper show that well-defined reactive and stimuliresponsive copolymers could be obtained under a broad range of conditions with low polydispersity. On account of the mild substitution conditions of N-hydroxysuccinimide ester by amine group, biohybrid nanogels based on the direct crosslinking between the reactive copolymer and proteins containing lysine linkages in their structure were synthesized. In these preliminary experiments, two different proteins, enhanced green fluorescent protein (EGFP) and cellulase (CelA2_M2), were used to cross-link water-soluble copolymers in water-in-oil (W/O) emulsion and to produce biohybrid nanogels. The EGFP conjugated nanogel (F-NG) was proved to be fluorescent, as characterized by confocal microscopy, while the CelA2_M2 encapsulated nanogel (E-NG) is a good colloidal catalyst for depolymerization of cellulose.

2. EXPERIMENTAL SECTION 2.1. Materials. N-Vinylpyrrolidone (VP, Aldrich, 99%), methacrylic acid N-hydroxysuccinimide ester (MNHS, Aldrich, 98%), anisole (Aldrich, anhydrous, 99.7%), methyl 2-bromopropionate B

DOI: 10.1021/acs.macromol.5b00947 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules Scheme 1. Synthesis of Chain Transfer Agent Methyl 2-(Ethoxycarbonothioylthio)propanoate

Scheme 2. Copolymerization of N-Vinylamide Monomers and MNHS

were purged with dry nitrogen for 1 h. Before the Schlenk flask was heated to 80 °C, a small amount of the solution was taken out for 1H NMR analysis. Solutions in the vials were inhaled in two 1 mL syringes, which were washed with dry nitrogen before. The AIBN solution was added to the Schlenk flask immediately while the MNHS solution was added continuously with a syringe pump (Harvard Apparatus, Holliston, MA) in 3 h. The reaction proceeded until the monomer conversion did not change. The solution turned pale yellow color during the polymerization. Fast cooling of the reaction mixture using liquid nitrogen stopped the reaction. Approximately 50 μL of solution was dissolved in CDCl3 for 1H NMR analysis to determine the monomer conversion of VP, which should be taken as conversion of the system. The copolymers were precipitated in 400 mL of cold diethyl ether and further washed with diethyl ether for three times. The copolymers were dried at 40 °C under vacuum for 48 h and then characterized by 1H NMR, FTIR, Raman, and GPC. Here the systems using VP, VPI, and VCL are named as PVPS, PVPIS, and PVS system, respectively. 1 H NMR (in CDCl3): (PVPS) δ[ppm] = 4.39−4.48 (−CHCH3,OCH2CH3,end group), 3.73−3.89 (-NCH),3.17−3.23 (-NCH2), 2.81(-COCH2CH2CO−), 2.21−2.35 (-NCOCH2), 1.96−2.04 (NCH2CH2CH2CO), 1.57−1.67 (CCH3), 1.40 (CH2C, CH2CH in backbone, CHCH3, CH2CH3 in end group); (PVPIS) δ[ppm] = 4.58 (-NCH, -OCH2CH3, −CHCH3, end group), 3.65 (−OCH3), 3.05 (-NCH2), 2.83(-COCH2CH2CO−), 2.20−2.35 (-NCOCH2), 1.70−1.88 (NCH2CH2CH2CH2CO, CCH3 in backbone), 1.42 (CH2C, CH2CH in backbone, CHCH3, CH2CH3 in end group); (PVS) δ[ppm] = 4.38 (-NCH, -OCH2CH3, −CHCH3, end group), 3.63 (−OCH3), 3.20 (-NCH2), 2.81(-COCH2CH2CO−), 2.34− 2.49 (-NCOCH2), 1.46−1.76 (NCH2CH2CH2CH2CH2CO, CCH3 in backbone), 1.21−1.25 (CH2C, CH2CH in backbone, CHCH3, CH2CH3 in end group). FTIR (on silica plate): (PVPS) 2948 cm−1 (CH2), 1806, 1776 cm−1 (OCCH2CH2CO), 1737 cm−1(OC-O), 1666 cm−1(NCO), 1492 cm−1(CH3), 730 cm−1[-(CH2CH2CH2)n-]. (PVPIS) 2943, 2868 cm −1 (CH2 ), 1805, 1776 cm −1 (OCCH2 CH2 CO), 1736 cm−1(OC-O), 1612 cm−1(NCO), 1491 cm−1(CH3), 730 cm−1[(CH2CH2CH2)n-]. (PVS) 2925, 2855 cm−1 (CH2), 1806, 1777 cm−1 (OCCH2CH2CO), 1739 cm−1(OC-O), 1632 cm−1(NCO), 1493 cm−1(CH3), 717 cm−1[-(CH2CH2CH2)n-].

temperature-controlled sample holders are connected to a Cary temperature controller allowing six simultaneous measurements. The solutions were scanned with a temperature trend with a fixed wavelength at 400 nm. The solutions of the copolymers were prepared in deionized water and sonicated until all the polymers were dissolved. The heating speed was applied from 20 to 80 °C for PVS and from 30 to 90 °C for PVPIS. The cloud points were determined by the lowest point extracted from the derivative of the transmittance versus temperature curve. Synthesis of the Chain Transfer Agent (CTA). The synthesis of methyl 2-(ethoxycarbonothioylthio)propanoate was performed according to the instructions given by Destarac36 and shown in Scheme 1. A solution of methyl 2-bromopropionate (2 g, 11.7 mmol, 1 equiv) in ethanol (15 mL) was stirred in a double-neck flask immersed in an ice bath. Potassium ethyl xanthogenate (2.16 g, 13.4 mmol, 1.1 equiv) was progressively added with a spatula for 45 min. The solution became yellowish. After addition, the solution was mixed for 3 h at room temperature. The formation of potassium bromide (insoluble in ethanol) was observed. The solution was then filtered with a Buchner filter and concentrated under reduced pressure with a rotary evaporator. Dichloromethane (30 mL) was then added, and the product was extracted in the organic phase after treatment with distilled water (4 × 50 mL). The organic phase was dried over sulfate magnesium overnight in an Erlenmeyer flask. After filtration, dichloromethane was removed in a rotary evaporator. The product was dried under vacuum to yield a yellow bright liquid (2.09 g, yield 91.1%). 1H NMR (in CDCl3): δ[ppm] = 4.56 (-OCH2CH3), 4.32 (CHCH3), 3.69(-OCH3), 1.52 (CHCH3), 1.35 (−OCH2CH3). 13C NMR (in CDCl3): δ[ppm] = 211 (CS), 171 (CO), 70 (-OCH2CH3), 52 (-OCH3), 47 (CHCH3), 17 (CHCH3), 14 (−OCH2CH3). General Polymerization Procedure. All polymerizations were carried out with AIBN as initiator and methyl 2(ethoxycarbonothioylthio)propanoate as CTA under dry nitrogen with standard Schlenk techniques. A typical procedure is described below and shown in Scheme 2. A 25 mL Schlenk flask was charged with 2 g (18 mmol) of VP and 0.042 g (2.1 mmol) of CTA in 3 mL of anisole. Two small vials were charged with 0.067 g (0.37 mmol) of MNHS in 1 mL of anisole and 0.0058 g (0.035 mmol) of AIBN in 0.2 mL of anisole, respectively. The solution in the Schlenk flask was degassed by five freeze−pump−thaw cycles, and solutions in the vials C

DOI: 10.1021/acs.macromol.5b00947 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

Figure 1. Principle of activity quantification system in which 4-MUC is hydrolyzed by the encapsulated cellulase (CelA2_M2) in E-NG nanogels, which results in a continuous fluorescence buildup (4-MU) that is quantified in a fluorescence MTP reader as previously reported (Lehmann et al.35). Raman: (PVPS) 2976, 2926 cm−1(CH2), 1806, 1776 cm−1 (O CCH2CH2CO), 1737 cm−1(OC-O), 1669 cm−1(NCO), 1494 cm−1(CH3), 754 cm−1[-(CH2CH2CH2)n-]. (PVPIS) 2947, 2870 cm−1 (CH2), 1805, 1780 cm−1 (OCCH2CH2CO), 1736 cm−1(OCO), 1627 cm − 1 (NCO), 1491 cm −1 (CH 3 ), 723 cm − 1 [(CH2CH2CH2)n-]. (PVS) 2929, 2856 cm−1 (CH2), 1806, 1776 cm−1 (OCCH2CH2CO), 1737 cm−1(OC-O), 1630 cm−1(NCO), 1446 cm−1(CH3), 702 cm−1[-(CH2CH2CH2)n-]. The theoretical number-average molecular weight based on conversion was calculated with the following equation: M n(theor) =

[Monomer]0 M monomerS + MCTA [CTA]0

stored at 4 °C. The loading amount of CelA2_M2 was 16.17 mg determined by standard BCA assay. The catalytic activity test of the biohybrid nanogel was measured according to a previous report.35 4-MUC was used as a fluorogenic substrate. 1 mL of the prepared nanogel solution was diluted to 20 mL for catalytic assay (CelA2_M2 concentration, 0.0539 mg/mL). Certain amount of the E-NG solution was transferred into a 96-well MTP (flat-bottomed, PS plates, black). Then some Kpi buffer was added to dilute the sample. Afterward 20 μL of 4-MUC solution was added to each well to initiate the enzyme reaction. The reaction mixture was then put in a fluorescence plate reader, incubating at 30 °C, to measure fluorescence every 2 min. The reported values are average values of three measurements with average deviations. The principle of the activity test is that 4-MUC is hydrolyzed to cellobioside and the fluorophore 4-methylumbelliferone (4-MU) catalyzed by E-NG (Figure 1). The water-soluble 4-MU can be detected and recorded by the fluorescence plate reader. To confirm that the catalytic activity is from the encapsulated enzyme, the P-NG and pure Kpi buffer are used as control samples for the fluorometric assay. The samples were measured in the same manner. The amounts of the E-NG and P-NG solutions and Kpi buffer for activity assay are listed in Table1.

(1)

where MCTA is the molecular weight of the chain transfer agent, Mmonomer is the average molecular weight of the monomers for copolymerization, S is conversion, and [Monomer]0 and [CTA]0 are initial concentrations of monomers and chain transfer agent, respectively. Polymerization Kinetics. Samples were taken out periodically, and a small fraction was used for 1H NMR in CDCl3 to determine the conversion, while the major fraction was precipitated in cold diethyl ether. The dry samples were sent for DMF GPC to determine the molecular weight. The conversion of VP (VPI or VCL) was determined by comparing the integration of vinyl proton peaks with that of anisole, which can also be used as internal standard. For polymerizations in DMF and 1,4-D, around 50 μL of anisole was added in the systems as internal standard. Nanogel Synthesis. Nanogels were synthesized by cross-linking between PVPS (4% MNHS, Mn,GPC = 7110 Da) and PEI in W/O emulsion, named as P-NG. The approach is a modified procedure according to previous literature.37 The aqueous phase with 0.1 g of PVPS, 0.008 g of of PEI in 0.5 mL of Kpi buffer (pH 7.2, 0.02 M) and organic phase consisting of 0.14 g of span80 in 6 mL of toluene were sonicated by a Branson Sonifier W450 (Terra Universal Inc., USA) with a 1/4-in. horn at duty cycle of 40% and amplitude of 360 W for 15 min under ice cooling. After sonication, the emulsion was stirred overnight at room temperature. Obtained nanogels were separated by centrifugation at 11 000 rpm for 40 min. The supernatant was decanted, and the precipitate was redispersed in toluene again, which was centrifuged at 11 000 rpm for 40 min to clean the nanogels from span80. This process was repeated three times. Then the nanogels were redispersed in water and further purified by dialysis (MWCO ≈ 12−14 kDa) for 48 h. Nanogel samples were characterized by TEM, DLS, and FTIR spectroscopy. Biohybrid Nanogels Synthesis and Activity Measurements. Two different biohybrid nanogels were synthesized in a similar method as described above. The fluorescence nanogel (named as F-NG) was synthesized using EGFP (5 mg) as cross-link agents. The nanogels were washed with toluene first, then with water by redispersion and centrifugation cycles and afterward by dialysis (MWCO 50 kDa). The enzyme CelA2_M2 (25 mg) acted as cross-link agent, and Kpi buffer (pH 7.2, 0.02 M) was used as aqueous phase to synthesize the enzyme E-NGs. The nanogels were purified in the same manner, and the molecular weight cutoff (MWCO) of the dialysis tube was 100 kDa. The biohybrid nanogels were redispered in 15 mL of Kpi buffer and

Table 1. Amounts of E-NG, P-NG, and Kpi Buffer in Activity Assay well no. E-NG/buffer well no. E-NG/buffer well no. E-NG/buffer well no. P-NG/buffer well no. P-NG/buffer

(V/V) (V/V) (V/V) (V/V) (V/V)

A1 10 μL/0 B1 5 μL/5 μL C1 1 μL/9 μL D1 0/10 μL E1 10 μL/0

A2 10 μL/0 B2 5 μL/5 μL C2 1 μL/9 μL D2 0/10 μL E2 10 μL/0

A3 10 μL/0 B3 5 μL/5 μL C3 1/9 μL D3 0/10 μL E3 10 μL/0

3. RESULTS AND DISCUSSION 3.1. Optimization of Polymerization Conditions. It is widely reported that solvent type and reaction temperature influence polymerization processes and properties of obtained polymers.38 In the present work, we investigated the effect of these reaction parameters on the Mn and PDI of copolymers. Tables 2, 3, and 4 summarize conversions, Mn values, and PDIs of the three synthesized copolymers. The target molecular weight of all the reactions was 10 000 Da, and the ratio of [CTA]/[AIBN] was 6. The molar feed ratio of MNHS was 2%. The polymerizations of VP and MNHS were efficiently carried out in anisole and 1,4-D, leading to high conversions and low PDIs, whereas in DMF, the reaction was disfavored. Anisole is thought to be the most suitable solvent for the polymerization of VP and MNHS. As demonstrated in Table 2, monomer conversion in anisole at 70 °C reaches values up to D

DOI: 10.1021/acs.macromol.5b00947 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

hindrance caused by the 7-membered ring. At the same time, the polydispersity remains rather low, especially in anisole and 1,4-D at 80 °C. Identical to the above systems, polymerization in anisole is more favored than those in 1,4-D and DMF, indicating anisole is again the most suitable solvent for the RAFT polymerization. Temperature was more important than solvent in this system, as demonstrated in Table 4. The

Table 2. RAFT Polymerization Conditions and Properties of VP/MNHS Copolymers Mn (Da) sample

T (°C)

solvent

conversion (%)

theory

GPC

PDI

1 2 3 4 5 6 7 8 9

60 60 60 70 70 70 80 80 80

anisole 1,4-D DMF anisole 1,4-D DMF anisole 1,4-D DMF

83 65 30 84 31 75 68 66 39

8333 6568 3137 8431 3235 7549 6862 6667 4019

7517 6130 2160 7864 2394 6646 6042 5979 3183

1.12 1.33 1.55 1.18 1.33 1.52 1.23 1.23 1.46

Table 4. RAFT Polymerization Conditions and Properties of VCL/MNHS Copolymers Mn (Da)

84%, and the PDIs are quite low (1.12, 1.18, and 1.23 at 60, 70, and 80 °C, respectively). Although high temperature accelerated the polymerization rate, it could also promote unfavorable side reactions and interchain termination reactions, which could be effectively suppressed at a low temperature, resulting in lower PDI and higher conversion at 70 °C. From Table 2, it can be estimated that the strong proton-accepting solvent, DMF, favoring side reactions and leading to rather high PDI and low conversion, is not suitable for the polymerization. Similarly, the polymerization of VPI and MNHS proceeds “better” in anisole and 1,4-D than in DMF. The minor role of temperature in controlling Mn and PDI of copolymers synthesized in anisole or DMF is suggested by little difference in corresponding values, as shown in Table 3. While DMF

Mn (Da) T (°C)

solvent

conversion (%)

theory

GPC

PDI

1 2 3 4 5 6 7 8 9

60 60 60 70 70 70 80 80 80

anisole 1,4-D DMF anisole 1,4-D DMF anisole 1,4-D DMF

62 56 25 60 34 24 59 54 30

6274 5686 2647 6078 3529 2549 5980 5490 3137

6061 5560 2588 5903 3326 2295 5879 5303 2939

1.19 1.25 1.37 1.25 1.27 1.3 1.25 1.33 1.41

T (°C)

solvent

conversion (%)

theory

GPC

PDI

1 2 3 4 5 6 7 8 9

60 60 60 70 70 70 80 80 80

anisole 1,4-D DMF anisole 1,4-D DMF anisole 1,4-D DMF

29 24 17 21 23 23 52 35 33

3039 2549 1862 2255 2451 2451 5294 3627 3431

2621 2583 2043 2148 2420 2365 4587 3635 3475

1.31 1.35 1.34 1.29 1.48 1.37 1.15 1.23 1.64

polymerization results were greatly improved at 80 °C, which could be explained by monomer reactivity being disfavored by the high steric hindrance, which was largely increased at high temperature. As indicated from Table 4, DMF made the polymerization worse due to the high PDI. With 1,4-D as solvent, similar to the polymerization in anisole, the conversion was largely increased, and PDI decreased when temperature was increased to 80 °C. It can be concluded that temperature mainly influences the RAFT polymerization of VCL and MNHS. Similarities and differences among the three systems suggested that solvent is of great importance to the polymerizations. Anisole is the most suitable solvent, while DMF disfavors the reactions. With the enlargement of the lactam ring, monomer reactivity and polymerization controllability are decreased due to the increased steric hindrance. Significance of temperature increases with the enlargement of the lactam ring since high temperature offset the disadvantages caused by steric hindrance to some extent. 3.2. Influence of Chain Transfer Agent-to-Initiator (C/ I) Ratio. It is known that the ratio of C/I is very important for RAFT polymerization in terms of molecular weight and polydispersity control.38 In this work, experiments at different CTA/I ratios at 80 °C in anisole were carried out to investigate this effect. For all the experiments, the targeted Mn is 10 000 Da (calculated by eq 1 when the conversion is 100%), and the molar feed ratio of MNHS is 2%. As demonstrated in Table 5, when the C/I ratio increased from 2 to 10 for PVPS, PVPIS, and PVS systems, the monomer conversions decreased from 71% to 62%, from 59% to 49%, and from 54% to 31%, respectively, while PDI decreased from 1.31 to 1.16, from 1.29 to 1.20, and from 1.30 to 1.12, respectively. For an ideal RAFT polymerization, the number of polymer chain is mainly determined by CTA-derived chains, which are reinitiated by CTA fragments since the number of initiator-derived chains can be ignored compared with the large initial CTA concentration. In the three systems, the molecular weights can be well controlled even if the C/I ratio is only 2, which means many initiator-derived chains existing in the system, indicating good control of the polymerization. As shown in Table 5, conversion

Table 3. RAFT Polymerization Conditions and Properties of VPI/MNHS Copolymers sample

sample

disfavored the polymerization, anisole is finally selected as a suitable solvent. The temperature influence on polymerization in 1,4-D is quite interesting. Polymerization at 60 °C with highest conversion and lowest PDI is thought to run under controlled conditions. Adversely, reaction at 70 °C was disfavored, probably due to side reactions triggered by thermodynamic effects, which caused increased radical reactivity at higher temperature, leading to decreased selectivity during various radical reactions. At higher temperatures, the remarkably increased fragmentation rate constant and gradual decomposition of CTA may disfavor the controlled polymerization. However, further increase of temperature increased conversion but also PDI. It is assumed that monomer reactivity was greatly favored at 80 °C, which contributed to high conversion, whereas unfavorable side reaction caused higher polydispersity. As for the PVS system, the conversions were rather low (all less than 50%), which may be attributed to a high steric E

DOI: 10.1021/acs.macromol.5b00947 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

∼70 °C (for PVPS) or ∼80 °C (for PVS) and a C/I ratio of 6, while the influence of temperature on PVPIS was not significant. 3.3. Kinetics of Polymerization. Polymerization kinetics in anisole at 60, 70, and 80 °C was investigated, keeping C/I = 6, MNHS ratio 2%, and target Mn = 10 000 Da. Timedependent conversions and first-order kinetic plots of the PVPS system at different temperatures are shown in Figure 2 and Figure S2, Supporting Information, respectively. The polymerization conversions are 83%, 84%, and 68% at 60, 70, and 80 °C, respectively. Obviously, polymerization rate increased at higher temperatures, which may be attributed to a favored decomposition of the initiator and fragmentation of intermediate radical, thus accelerating initiation and polymerization rate. The monomer conversions were relatively high (above 60%) at three polymerization temperatures. It is reported that in some RAFT polymerizations the induction period can be observed.38 In the present system, the induction period cannot be clearly observed probably due to the relatively fast polymerization rate at the initial polymerization stage. On the other hand, high temperature also might cause unfavorable side reactions and even CTA decomposition. Nevertheless, a linear increase in experimentally obtained molecular weight from GPC (MGPC) with theoretical molecular weight (Mth) and low PDIs for the three reaction runs are presented in Figure 2a. In addition, the symmetrical unimodal GPC peaks obtained at different reaction times for those reactions (Figure S2) suggest controlled character of the polymerizations. Although the kinetic plots of PVPIS and PVPS system are quite similar, interesting differences were observed. The main difference is that the PVPIS system is not greatly affected by temperature, suggested by little difference in conversions, Mn values, and PDIs between polymerizations at different temperatures (Figure 3 and Figure S3, Supporting Information). We assume that steric hindrance caused by the 6-membered ring reduces the acceleration effect caused by elevated temperatures. The monomers were polymerized in a relatively slow and progressive state. The surprisingly similar conversions, Mn values, and PDIs indicate similar dynamic states in all three polymerizations. The linear increase of MGPC with Mth in Figure 3a and the symmetrical unimodal GPC peaks as demonstrated

Table 5. Properties of Copolymers Synthesized in Anisole at 80 °C with Varied C/I Ratios Mn (Da) sample

C/I (mol/mol)

conversion (%)

theory

GPC

PDI

1 2 3

2 6 10

PVPS 71 68 62

7156 6862 6274

6498 6042 5683

1.31 1.23 1.16

4 5 6

2 6 10

PVPIS 59 59 49

5980 5980 5000

5729 5879 4865

1.29 1.25 1.20

7 8 9

2 6 10

PVS 54 52 31

5490 5294 3235

4853 4587 2484

1.30 1.15 1.12

decreased with the increase of C/I ratio. Especially for PVS system, it decreased more than 20% when C/I ratio is 10 compared with C/I ratio is 2, which means retardation of polymerization rate at high C/I ratio. Another interesting observation is that PDI at C/I = 6 and 10 is apparently lower than that at 2 (for PVPS, PDI = 1.16, 1.23 at C/I = 10 and 6, respectively, whereas PDI = 1.31 at C/I = 2; for PVPIS, PDI = 1.20 and 1.25 at C/I = 10 and 6, respectively, whereas PDI = 1.29 at C/I = 2; for PVS, PDI = 1.12 and 1.15 at C/I = 10 and 6, respectively, whereas PDI = 1.30 at C/I = 2), which may be attributed to the low initiator concentrations that disfavored termination reactions. According to a previous report,38 low PDI values are mainly due to the rapid establishment of the pre-equilibrium and attainment of the equilibrium. The former is efficient reinitiation by R• fragment, while in the latter the population of the macro-CTAS and/or intermediate radicals is much higher than the total number of the propagating chains. Apparently, a high C/I ratio leads to low conversions and low PDI values. Experimental results in Table 5 suggested that the optimal C/I molar ratio is 6. In summary, as optimal polymerization conditions, we selected anisole as solvent with

Figure 2. (a) Kinetic plot of ln([M]0/[M]t) versus time for the RAFT copolymerization of N-vinylpyrrolidone (VP) and methacrylic acid Nhydroxysuccinimide ester (MNHS) at 60 (squares), 70 (circles), and 80 °C (triangles). (b) Plot of experimental molecular weight from GPC (MGPC) and polydispersity index versus theoretical molecular weight (Mth, calculated by eq 1) for RAFT copolymerization of VP and MNHS at 60 (squares), 70 (circles), and 80 °C (triangles). The dashed line is the ideal situation when MGPC = Mth. F

DOI: 10.1021/acs.macromol.5b00947 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

Figure 3. (a) Kinetic plot of ln([M]0/[M]t) versus time for the RAFT polymerization of N-vinylpiperidone (VPI) and methacrylic acid Nhydroxysuccinimide ester (MNHS) at 60 (squares), 70 (circles), and 80 °C (triangles). (b) Plot of experimental molecular weight from GPC (MGPC) and polydispersity index versus theoretical molecular weight (Mth calculated by eq 1) for polymerizations at 60 (squares), 70 (circles), and 80 °C (triangles). The dashed line is the ideal situation where MGPC = Mth.

Figure 4. (a) Kinetic plot of ln([M]0/[M]t) versus time for the RAFT polymerization of N-vinylcaprolactam (VCL) and methacrylic acid Nhydroxysuccinimide ester (MNHS) at 60 (squares), 70 (circles), and 80 °C (triangles). (b) Plot of experimental molecular weight from GPC (M GPC) and polydispersity index versus theoretical molecular weight (Mth calculated by eq 1) for polymerizations at 60 (squares), 70 (circles), and 80 °C(triangles). The dash line is the ideal situation where MGPC = Mth.

Figure S4. The linear increase of MGPC with Mth (Figure 4) plus the unimodal GPC peaks (Figure S4) still indicate rather good control of the polymerizations. As shown in Figure S4, the GPC peaks of PVS are not as smooth as those of PVPS, especially when the conversion is low, which is probably due to side reaction arising from bimolecular termination of the growing polymer chain, indicating the low efficiency of CTA for the PVS system. 3.4. Chemical Structure of Copolymers by NMR and FTIR/Raman Spectroscopy. A series of PVPS copolymers were synthesized at optimal temperature, keeping C/I ratio at 6, and target Mn = 10000 Da with increased MNHS feed ratios as 2%, 4%, 6%, 8%, and 10%, named as PVPS2%, PVPS4%, PVPS6%, PVPS8%, and PVPIS10%, respectively. The same controlled syntheses of PVPIS and PVS copolymers at optimal temperature were also performed, and the copolymers were named in similar way. The polymerization results are summarized in Tables S1−S3, Supporting Information. The monomer conversions decreased with the enlargement of the lactam ring (PVPS, ∼69% to 84%; PVPIS, ∼61% to 67%; PVS,

in Figure S3 suggest that the polymerizations proceeded in a controlled manner. Compared with the two other systems, the conversions of PVS runs are quite low (29%, 21%, and 52% at 60, 70, and 80 °C, respectively). The low conversions may be attributed to low chain transfer efficiency due to the steric hindrance caused by the 7-membered ring. This may disfavor the formation of the macro-CTA while causing an obstacle for the propagation of radicals, resulting in low conversion and large amount of oligomers. However, as shown in Figure 4 and Figure S4, Supporting Information, high temperature can increase the conversion to some extent. At 80 °C, the conversion is more than 50%, which is largely increased compared with those of polymerizations at 60 and 70 °C, which are only 29% and 21%, respectively. Taking account the low PDI at 80 °C (1.15) and the fact that high temperature favors unfavorable side reaction, it can be assumed that high temperature has such a favorable influence on the polymerization that the unfavorable result caused by the side reaction can even be ignored. Conversion increased with increased temperature, as demonstrated in G

DOI: 10.1021/acs.macromol.5b00947 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

Figure 5. (a) FT-IR spectra of PVS copolymers with MNHS contents 2%, 4%, 6%, 8%, and 10%. (b) Peak height ratio of peak at 1737 cm−1 CO from MNHS and 1660 cm−1 CO from VCL vs MNHS:VCL ratio in reaction mixture.

Figure 6. (a) Transmittance curve as a function of temperature of PVS2% (5 mg/mL); the inset image shows reversible turbidity change before and after cloud point. (b−d) Cloud points of PVS (black rectangles) and PVPIS (red circles) as a function of concentrations (b), molecular weights (c), and MNHS molar ratios (d).

∼44% to 57%), which could be explained as the monomer reactivity decreasing with the enlargement of the lactam ring. It

is reported in the literature that the reactivity ratios of VP and methyl methacrylate for bulk polymerization at 50 °C are 0.005 H

DOI: 10.1021/acs.macromol.5b00947 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

physiological temperature at 34−37 °C, although that of PVPS is higher than 100 °C.43−46 Compared with the vast amount of scientific literature on the stimuli-responsive behavior of these two polymers, the properties of PVPIS, although thermosensitive, are seldom reported. The LCST for PVPS has been reported to be between 68 and 87 °C.47 The thermoresponsive behavior of the copolymers in aqueous solution was studied by UV−vis spectroscopy at a fixed wavelength at 400 nm and various temperatures (Figure 6a). The cloud points were determined at the minima from the derivative of the transmittance curve. The results in Figure 6 show that the cloud points shift to lower temperatures when the polymer concentration is increased. The phase separation process can be explained as a coil−globule transition process in which the macromolecule collapses from an expanded state to a collapsed state. When individual collapsed polymer chains are not stabilized in a solvent, they will aggregate, increasing the turbidity. At high polymer concentration, more polymer chain will collapse upon heating, which results in aggregation at lower temperatures, thus causing a shift of the cloud point temperature to lower values. As demonstrated in Figure 6 for synthesized copolymers, the temperature of the transition decreases with the increase of the molecular weight, as expected. As was reported before,48 the radius of gyration for the macromolecules is scaled as threefifths power of its chain length. It shifts from half power at the transition to one-third power in the collapsed state. The polymer chain length is proportional to molecular weight. The collapsed polymer chain with higher molecular weight would have larger globule size, which is easier to aggregate, resulting in lowering of the cloud point temperature. It is also observed that the cloud points decrease with increasing MNHS ratio due to the higher hydrophobicity of MNHS. Larger amounts of MNHS units in copolymer chains will accelerate coil-to-globule transition of polymer chains by enhancing hydrophobic interactions at lower temperatures. Estimated from the structure, the LCST of PVPIS should be between those of its 5-membered and 7-membered ring copolymer analogues. It is reported that the cloud points of PVPS with Mn = 4.5−83 kDa are in the range 68−87 °C at 20 mg/mL, which are much higher than that of poly(Nvinylcarpolactam).47 The cloud point of PVPIS with low molecular weight or low concentration is found to be above 100 °C. So the concentrations of the copolymers for cloud point measurement are higher than those of PVS copolymers. Apparently, the tendency of cloud point change with concentration, Mn, and MNHS ratio is similar to that of PVS copolymer. According to previous reports,47 the cloud point of PVPS with Mn = 10 600 Da at 20 mg/mL is around 81 °C, while block copolymer poly(N-vinylpiperidone)62-block-poly(vinyl acetate)32 with Mn = 11 400 Da possesses a cloud point of around 55 °C at l mg/mL, based on which it could be estimated that the cloud points of PVPIS copolymer should sharply decrease with increase of MNHS content. This is evidenced by the cloud point tendency demonstrated in Figure 6d, in which the cloud point decreased by around 10 °C when MNHS ratio increased by 2%. When MNHS content is higher than 8%, the copolymer is not water-soluble, indicating that its cloud point is lower than room temperature. It is further proved by the cloud points comparison between PVPS (Mn = 6000 Da) and PVPIS2% (Mn = 6061 Da) at 20 mg/mL, which are around 87 and 68 °C, respectively.47 Conversely, the cloud point of PVS is slightly influenced by the MNHS ratio

and 4.7, respectively.39 In the reported work, the input molar ratio of VP was 80%, while its fraction in copolymer was only 3.65%. According to the Mayo−Lewis equation,40,41 when r1 ≫ 1 ≫ r2, a composition drift will happen, in which one monomer is consumed preferentially until this monomer gets depleted, and then the second monomer segments are added. Composition drift can lead to copolymer with significantly different composition compared with the input monomer’s composition. Due to the similar structures of MNHS and methyl methacrylate, it could be estimated that their reactivity ratios are quite close. Thus, it is assumed that the reactivity ratio of MNHS is much higher than those of the cyclic Nvinylamides. Polymerizations with the two monomers simply mixed in anisole, resulting in almost a homopolymer of MNHS, evidenced that composition drift happened. To synthesize random copolymers with controlled MNHS distribution in polymer chains, the high reactivity monomer should be slowly added to the reaction system. With the help of the syringe pump, the MNHS ratio in the copolymer could be well controlled, which is very close to the input ratio as demonstrated in Tables S1−S3. 1H NMR was used to determine the MNHS content in copolymers. Figure S5, Supporting Information, shows the 1H NMR of PVS copolymers with characteristic peaks for VCL (CH2 at 3.2 ppm) and succinimide (CH2 signals at 2.8 ppm). The copolymer composition could be calculated by the equation in Figure S5b; the results are demonstrated in Tables S1−S3. To further characterize the copolymer composition ,we used FTIR and Raman spectroscopy. Figure 5 shows the FTIR spectra of PVS serial copolymers. Peaks at 1660 and 1737 cm−1 could be assigned to CO from VCL and CO from MNHS in the copolymer structure, respectively. In Figure 5, the y-axis is the peak height ratio between two characteristic peaks, while the x-axis is the input ratio of MNHS and VCL. The five points are in a nearly linearly relationship, indicating that the MNHS ratio in the copolymer increases proportionally with the increase of the monomer input ratio, thus evidencing the good control of MNHS composition in the serial copolymers. Similar results could be obtained from the Raman spectra (Figure S6, Supporting Information). Since MNHS is quite hydrophobic, the copolymer would become water insoluble when MNHS content increases to certain amount. The PVPS copolymers are still soluble in water even when MNHS content increases to 10.15%, while PVPIS and PVS copolymers are not water-soluble when the MNHS content increases to 7.98% and 6.27%, respectively. It is easy to understand considering that the monomer hydrophobicity increases with the enlargement of the lactam ring. As demonstrated in Tables S1−S3, PDI increases with the increase of MNHS in copolymers. One possible explanation for this phenomenon is the high reactivity ratio of MNHS compared to other monomers. On the other hand, it is known that xanthate CTAs possess low transfer constant but are more effective with less activated monomer.42 Therefore, in our system, methyl 2-(ethoxycarbonothioylthio)propanoate as CTA is more effective for N-vinylamides than MNHS, which also contributed to the increased PDI when MNHS concentration increased. 3.5. Behavior of Copolymers in Aqueous Solutions. Thermoresponsive polymers can change their conformation and properties in solution with temperature variation. PVS has a lower critical solution temperature (LCST) close to I

DOI: 10.1021/acs.macromol.5b00947 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

Figure 7. (a) TEM images (scale bar in the inset image is 200 nm) and (b) DLS size distribution histogram of P-NG nanogel synthesized by crosslinking of PVPS and PEI.

Figure 8. (a) FTIR spectra of nanogel, PEI, and PVPS (a1 and a2 are magnified regions for selected spectra areas, the arrow in a1 points to the new formed peak of −NH −, the arrow in a2 points to the disappeared peak from N-succinimide group). (b) Scheme of reaction between PVPS copolymer and primary amine groups of PEI.

compared with that of PVPIS copolymer. This difference is hard to explain and may be attributed to the symmetry and steric hindrance difference between the 6-membered and 7membered rings. The influence of concentration and molecular weight on cloud point is similar to that of the PVS copolymer. As the copolymer PVPS is more hydrophilic compared to PVPIS and PVS, simply increasing temperature even until 100 °C is not sufficient to make the copolymer precipitate from water. It is reported that poly(vinylpyrrolidone) demonstrated clouding behavior in salt solution,49,50 we believe that the

copolymer PVPS should have similar properties due to the similar structure. 3.6. Application of Reactive Copolymers in Synthesis of Biohybrid Nanogels. A considerable amount of literature has been published on postmodification of polymers containing the N-succinimide group.51−54 Due to the fast reaction of Nsuccinimide group with amine group at mild conditions, different biomacromolecules like proteins, peptides, or carbohydrates rich in primary amine groups could be conjugated with the reactive copolymers. In the present study we used PEI as a model multifunctional cross-linking agent to J

DOI: 10.1021/acs.macromol.5b00947 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

Figure 9. Confocal microscopy images: (a) bright field, (b) fluorescent field, and (c) overlay of (a) and (b) (scale bar 2 μm). (d) DLS size distribution histogram of EGFP-containing nanogel (F-NG) (scale bar 200 nm).

Figure 10. (a) TEM images of E-NG nanogel cross-linked by CelA2_M2 (scale bar of the inset image is 200 nm). (b) Particle size distribution histogram by DLS.

buffer in water droplets was adjusted to 7.2, and a high excess amount of PEI was used to favor the synthesis of nanogel by cross-linking. As demonstrated in TEM images in Figure 7, the obtained particles were with nearly homogeneous size and rough surface, which seemed to be covered with a thin soft layer. We assume that this surface layer could originate from the conjugated PEI and the hydrolysis copolymers. The FTIR spectra of the nanogel, PEI, and PVPS copolymer clearly show the disappearance of the N-succinimide group (Figure 8a2) and the appearance of the new peak at 1560−1570 cm−1 after crosslink reaction (Figure 8a1), which should belong to the −NH−

explore the cross-link conditions with synthesized reactive copolymers to perform nanogel synthesis in W/O emulsion. In following experiments we synthesized biohybrid nanogels by conjugation of proteins with reactive copolymers in a similar system. These preliminary experiments were done to demonstrate the applicability of synthesized copolymers as reactive building blocks for design of complex polymer colloidal structures. Since the N-succinimide ester can be hydrolyzed in slightly alkaline conditions, the cross-link reaction will compete with hydrolysis, which leads to very soft nanogel. The pH of the Kpi K

DOI: 10.1021/acs.macromol.5b00947 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

Figure 11. (a) Variation of of 4-MU fluorescence over 4-MUC conversion time with varied amounts of E-NG, P-NG, and varied buffer ratios. (b) Specific activity of E-NG and original CelA2_M2.35 One unit was defined as the amount of CelA2_M2 that catalyzes the conversion of 1 μmol of 4MUC per minute.

of the amide. The sharp double peaks at 3000 cm−1 suggest unreacted amine groups in the particles, which should originate from the conjugated PEI, indicating possibility of further modification of the particles. Figure 7b demonstrates the particle size distribution by DLS: the particle size is around 127 nm in radius, which is larger than that measured by TEM (around 200 nm in diameter) due to particle collapse in the dry state. The PDI is around 0.17 from DLS, indicating a nearly homogeneous particle size distribution. After successful optimization of the cross-linking conditions in W/O emulsion and successful synthesis of nanogels by conjugation of reactive copolymers and PEI, we used fluorescent protein EGFP instead of the synthetic aminefunctionalized polymer (PEI) for nanogel synthesis. The confocal microscopy images of nanogels in Figure 9 prove the successful incorporation of the fluorescent EGFP protein. The particles show a broader size distribution compared with P-NG sample; no aggregations could be observed in the obtained images. The average particle size is 343.3 nm in radius with a PDI of 0.21 by DLS (Figure 9d). Nanogels (E-MG) with encapsulated CelA2_M2 were synthesized in a similar manner to the F-NG sample. TEM images demonstrate that the nanogels exhibit a broad size distribution, which is good agreement with DLS results (Figure 10). The PDI is 0.45, and the radius is 265.4 nm (determined by DLS), which is larger than particle sizes determined by TEM. The difference can be attributed to the fact that nanogels are measured in swollen and in dry states by DLS and TEM, respectively. The IR spectroscopy (Figure S10, Supporting Information) proved successful substitution of the succinimide group by vanishing of the peak at 1780 cm−1 and appearance of new peak at 1560 cm−1, assigning to the newly formed amide bond. However, the amide group peak in IR is relatively weak due to the low cross-link density, which corresponds well to the formation of soft particles (see TEM images.) Results in Figure 11a reveal that the fluorescence signal in ENG solutions of different concentrations linearly increased with time, proving that E-NG is catalytically active. In P-NG solution and pure Kpi buffer, no changes were detected. This indicates that 4-MUC cannot be hydrolyzed by P-NG or pure Kpi buffer. Therefore, P-NG and pure Kpi buffer were used as control samples for the 4-MUC assay system to confirm that the catalytic activity of the E-NG nanogels originates from the

conjugated CelA2_M2. The specific activity of the CelA2_M2 in the nanogels calculated according to previous report based on the BCA assay and catalytic activity assay results is 3.64 U/ mg.35 Compared with the reported specific activity of CelA2_M2 as 15.1, the catalytic activity of the biohybrid nanogel was weaker. It is assumed that the cross-linked protein in a nanogels was in an arrested or half-folding form, which leads to the reduction of the enzymatic activity. Furthermore, from the slopes of fitting curves in Figure 11a, it can be concluded that the overall catalytic activity increases with the increase of E-NG concentration in an approximately linear manner, indicating that the CelA2_M2 is evenly conjugated within nanogels.

4. CONCLUSION In this work, we systematically studied the RAFT copolymerization of cyclic N-vinylamides (N-vinylcaprolactam, Nvinylpiperidine, and N-vinylpyrrolidone) and methacrylic acid N-hydroxysuccinimide ester at varied temperatures and in three solvents. The experiments have explored the optimal polymerization conditions, which are found to be anisole as solvent, ∼70 °C (for PVPS) or ∼80 °C (for PVS) as reaction temperature, and chain transfer agent/initiator (C/I, molar/ molar) ratio of 6. The temperature influence on PVPIS synthesis is not significant. The narrow polydispersity of copolymers, linear relationship between conversion and molecular weight, and symmetrical unimodal GPC peaks prove the controlled character of the polymerization process. Our experimental data indicate that solvent greatly influences the polymerization process, while increased C/I ratio can decrease PDI as well as conversion. Increasing the reaction temperature caused acceleration of the polymerization rate while favoring side reactions. The experimental results suggest that steric hindrance originating from the monomer structure of cyclic N-vinylamides is the main factor influencing copolymerization with MNHS. With the enlargement of the lactam ring, the monomer reactivity and controllability of the polymerization decreased, and increased temperature turned out to be favorable for polymerization. Synthesized water-soluble reactive copolymers were used as reactive building blocks for conjugation with amine-functionalized polymers or proteins in W/O emulsions to synthesize nanogels with variable chemical structures. Cross-linking L

DOI: 10.1021/acs.macromol.5b00947 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

(19) Zehm, D.; Ratcliffe, L. P. D.; Armes, S. P. Macromolecules 2013, 46, 128−139. (20) McCormick, C. L.; Lowe, A. B. Acc. Chem. Res. 2004, 37, 312− 325. (21) Chaduc, I.; Lansalot, M.; DAgosto, F.; Charleux, B. Macromolecules 2012, 45, 1241−1247. (22) Albertin, L.; Cameron, N. R. Macromolecules 2007, 40, 6082− 6093. (23) Zhang, Z. B.; Zhu, J.; Cheng, Z. P.; Zhu, X. L. Polymer 2007, 48, 4393−4400. (24) Savariar, E. N.; Thayumanavan, S. J. Polym. Sci., Part A: Polym. Chem. 2004, 42, 6340−6345. (25) Yanjarappa, M. J.; Gujraty, K. V.; Joshi, A.; Saraph, A.; Kane, R. Biomacromolecules 2006, 7, 1665−1670. (26) Muller, A. H. E.; Schilli, C.; Rizzardo, E.; Thang, S. H.; Chong, B. Y. K. Polym. Prepr. 2002, 43, 687. (27) Schmaljohann, D. Adv. Drug Delivery Rev. 2006, 58, 1655−1670. (28) Bajpai, A. K.; Shukla, S. K.; Bhanu, S.; Kankane, S. Prog. Polym. Sci. 2008, 33, 1088−1118. (29) Onaca, O.; Enea, R.; Hughes, D. W.; Meier, W. Macromol. Biosci. 2009, 9, 129−139. (30) Kikuchi, A.; Okano, T. Prog. Polym. Sci. 2002, 27, 1165−1193. (31) Chan, G.; Mooney, D. Trends Biotechnol. 2008, 26, 382−392. (32) Ieong, N. S.; Redhead, M.; Bosquillon, C.; Alexander, C.; Kelland, M.; O'Reilly, R. K. Macromolecules 2011, 44, 886−893. (33) Liang, X.; Kozlovskaya, V.; Chen, Y.; Zavgorodnya, O.; Kharlampieva, E. Chem. Mater. 2012, 24, 3707−3719. (34) Noor, M.; Dworeck, T.; Schenk, A.; Shinde, P.; Fioroni, M.; Schwaneberg, U. J. Biotechnol. 2012, 157, 31−37. (35) Lehmann, C.; Sibilla, F.; Maugeri, Z.; Streit, W. R.; Dominguez de Maria, P.; Martinez, R.; Schwaneberg, U. Green Chem. 2012, 14, 2719−2726. (36) Beija, M.; Marty, J. D.; Destarac, M. Chem. Commun. 2011, 47, 2826−2828. (37) Singh, S.; Bloehbaum, J.; Moeller, M.; Pich, A. J. Polym. Sci., Part A: Polym. Chem. 2012, 50, 4288−4299. (38) Mori, H.; Matsuyama, M.; Sutoh, K.; Endo, T. Macromolecules 2006, 39, 4351−4360. (39) Greenley, R. Z. J. Macromol. Sci., Chem. 1980, 14, 445−515. (40) Tidwell, P. W.; Mortimer, G. A. J. Polym. Sci., Part A: Gen. Pap. 1965, 3, 369−387. (41) Skeist, I. J. Am. Chem. Soc. 1946, 68, 1781. (42) Harrisson, S.; Liu, X.; Ollagnier, J. N.; Coutelier, O.; Marty, J. D.; Destarac, M. Polymers 2014, 6, 1437−1488. (43) Maeda, Y.; Nakamura, T.; Ikeda, I. Macromolecules 2002, 35, 217−222. (44) Laukkanen, A.; Valtola, L.; Winnik, F. M.; Tenhu, H. Macromolecules 2004, 37, 2268−2274. (45) Lau, A. C. W.; Wu, C. Macromolecules 1999, 32, 581−584. (46) Meeussen, F.; Nies, E.; Berghmans, H.; Verbrugghe, S.; Goethals, E.; Prez, F. Polymer 2000, 41, 8597−8602. (47) Ieong, N. S.; Redhead, M.; Bosquillon, C.; Alexander, C.; Kelland, M.; O'Reilly, R. K. Macromolecules 2011, 44, 886−893. (48) Grosberg, A. Y.; Kuznetsov, D. V. Macromolecules 1992, 25, 1970−1979. (49) Maeda, Y.; Nakamura, T.; Ikeda, I. Macromolecules 2002, 35, 217−222. (50) Dan, A.; Ghosh, S.; Moulik, S. P. J. Phys. Chem. B 2008, 112, 3617−3624. (51) Pollak, A.; Blumenfeld, H.; Wax, M.; Baughn, R. L.; Whitesides, G. M. J. Am. Chem. Soc. 1980, 102, 6324−6336. (52) Li, H. M.; Bapat, A. P.; Li, M.; Sumerlin, B. S. Polym. Chem. 2011, 2, 323−327. (53) Relogio, P.; Bathfield, M.; Haftek-Terreau, Z.; Beija, M.; Favier, A.; Giraud-Panis, M. J.; D'Agosto, F.; Mandrand, B.; Farinha, J. P.; Charreyre, M. T.; Martinho, J. M. G. Polym. Chem. 2013, 4, 2968− 2981. (54) Steinhauer, W.; Keul, H.; Moeller, M. Polym. Chem. 2011, 2, 1803−1814.

reactions between proteins and the reactive copolymers in aqueous droplets successfully led to formation of biohybrid nanogels. The nanogels based on EGFP exhibited fluorescent properties, and cellulase-loaded nanogels demonstrated relatively high catalytic activity, indicating that reactive copolymers are suitable building blocks for conjugation with proteins and design of interesting biohybrid structures.



ASSOCIATED CONTENT

* Supporting Information S

1

H and 13C NMR spectra, kinetics and GPC traces, results of RAFT polymerizations, Raman and FTIR data, transmittance vs temperature curves, and standard calibration curve for BCA assay. The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.macromol.5b00947.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS H.P. thanks China Scholarship Council for financial support. A.P. thanks Volkswagen Foundation and Deutsche Forschungsgemeinschaft (DFG) with Collaborative Research Center SFB 985 “Functional Microgels and Microgel Systems” for financial support.



REFERENCES

(1) Matyjaszewski, K.; Xia, J. H. Chem. Rev. 2001, 101, 2921−2990. (2) Matyjaszewski, K.; Tsarevsky, N. V. Nat. Chem. 2009, 1, 276− 288. (3) Nicolas, J.; Guillaneuf, Y.; Lefay, C.; Bertin, D.; Gigmes, D.; Charleux, B. Prog. Polym. Sci. 2013, 38, 63−235. (4) Sciannamea, V.; Jérôme, R.; Detrembleur, C. Chem. Rev. 2008, 108, 1104−1126. (5) Boyer, C.; Bulmus, V.; Davis, T. P.; Ladmiral, V.; Liu, J. Q.; Perrier, S. Chem. Rev. 2009, 109, 5402−5436. (6) Shipp, D. A. J. Am. Chem. Soc. 2008, 130, 10448−10448. (7) Bruno, A. Macromolecules 2010, 43, 10163−10184. (8) Barbey, R.; Lavanant, L.; Paripovic, D.; Schüwer, N.; Sugnaux, C.; Tugulu, S.; Klok, H. A. Chem. Rev. 2009, 109, 5437−5527. (9) Wang, J.; Matyjaszewski, K. J. Am. Chem. Soc. 1995, 117, 5614− 5615. (10) Jennings, J.; Beija, M.; Kennon, J. T.; Willcock, H.; O’Reilly, R. K.; Rimmer, S.; Howdle, S. T. Macromolecules 2013, 46, 6843−6851. (11) York, A. M.; Zhang, T. L.; Holley, A. C.; Guo, Y. L.; Huang, F. Q.; McCormick, C. L. Biomacromolecules 2009, 10, 936−943. (12) Zhang, C. L.; Miao, M.; Cao, X. T.; An, Z. S. Polym. Chem. 2012, 3, 2656−2664. (13) Rosselgong, J.; Williams, E. G. L.; Le, T. P.; Grusche, F.; Hinton, T. M.; Tizard, M.; Gunatillake, P.; Thang, S. H. Macromolecules 2013, 46, 9181−9188. (14) Li, S. P.; Gao, C. Polym. Chem. 2013, 4, 4450−4460. (15) Li, C. Z.; Benicewicz, B. C. Macromolecules 2005, 38, 5929− 5936. (16) Zhang, X. W.; Lian, X. M.; Liu, L.; Zhang, J.; Zhao, H. Y. Macromolecules 2008, 41, 7863−7869. (17) Houillot, L.; Bui, C.; Farcet, C.; Moire, C.; Raust, J. A.; Pasch, H.; Save, M.; Charleux, B. ACS Appl. Mater. Interfaces 2010, 2, 434− 442. (18) Benaglia, M.; Rizzardo, E.; Alberti, A.; Guerra, M. Macromolecules 2005, 38, 3129−3140. M

DOI: 10.1021/acs.macromol.5b00947 Macromolecules XXXX, XXX, XXX−XXX