Polymer Conjugates Using Polynorbornene Block

Jan 14, 2016 - Alan J. Russell , Stefanie L. Baker , Coray M. Colina , Charles Adrian Figg ... Kevin M. Burridge , Thaiesha A. Wright , Richard C. Pag...
2 downloads 0 Views 3MB Size
Download PDF
Article pubs.acs.org/Biomac

“Graft-to” Protein/Polymer Conjugates Using Polynorbornene Block Copolymers Sergey A. Isarov, Parker W. Lee, and Jonathan K. Pokorski* Department of Macromolecular Science and Engineering, School of Engineering, Case Western Reserve University, Cleveland, Ohio 44106, United States S Supporting Information *

ABSTRACT: A series of water-soluble polynorbornene block copolymers prepared via Ring-Opening Metathesis Polymerization (ROMP) were grafted to proteins to form ROMP-derived bioconjugates. ROMP afforded low-dispersity polymers and allowed for strict control over polymer molecular weight and architecture. The polymers consisted of a large block of PEGylated monoester norbornene and were capped with a short block of norbornene dicarboxylic anhydride. This cap served as a reactive linker that facilitated attachment of the polymer to lysine residues under mildly alkaline conditions. The generality of this approach was shown by synthesizing multivalent polynorbornene-modified viral nanoparticles derived from bacteriophage Qβ, a protein nanoparticle used extensively for nanomedicine. The conjugated nanoparticles showed no cytotoxicity to NIH 3T3 murine fibroblast cells. These findings establish protein bioconjugation with functionalized polynorbornenes as an effective alternative to conventional protein/polymer modification strategies and further expand the toolbox for protein bioconjugates.

■

the conjugated PEG.4,9−13 However, the architecture of PEG polymers also plays an important role. For example, interferon 2α-PEG conjugates have been extensively studied with both branched and linear PEG chains. The first generation of conjugates developed in the early 1990s consisted of linear PEG chains, resulting in increased circulation half-life and improved pharmacokinetics. However, second generation conjugates feature branched PEG chains, further improving pharmacokinetic behavior.11,14,15 There are a number of benefits to branched PEGylation, notably those associated with increased biodistribution and circulation lifetime. A simple change in polymer architecture leads to increased surface masking, and hence decreased antibody recognition over linear PEG of the same molecular weight, as has been seen in many different therapeutic proteins.16−19 A more recent series of studies by Gauthier, Leroux, and co-workers outlined the significance of polymer architecture in modulating the stability and the catalytic activity of enzymes in vivo.20−23 The most compelling results of these studies were that both comb

INTRODUCTION Proteins have recently gained popularity as safe and effective therapeutic agents with a quickly growing variety of clinical applications.1 A number of protein-based therapies are already in use as FDA-approved pharmaceuticals and many others are being investigated in clinical trials.2 Proteins are attractive candidates for therapeutic use since many are endogenous and have known mechanisms of action.3 One major challenge facing the use of such therapies is their short half-life in the circulation where they are susceptible to a variety of degradative processes such as immune-mediated clearance and enzymatic degradation.4 Typically, attaching polymers such as poly(ethylene glycol) (PEG) to proteins (PEGylation) is employed as a means to protect proteins from such degradation.5,6 PEG is a hydrophilic polymer with negligible toxicity that, when attached to proteins, increases circulation half-life by reduction of renal clearance, diminishing immunogenicity, and protection from enzymatic degradation.7 In addition, protein PEGylation has been shown to increase water solubility, reduce aggregation, and improve structural and thermal stability.6−8 PEGylation clearly imparts a number of important benefits to proteins in vivo. It has been extensively reported that one of the primary factors contributing to these benefits is the length of © XXXX American Chemical Society

Received: November 24, 2015 Revised: December 31, 2015

A

DOI: 10.1021/acs.biomac.5b01582 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules polymers and dendrimers outperformed linear PEG in terms of enzymatic activity and decreased immunogenicity. Polymer architecture clearly plays a role in defining the behavior of protein/polymer conjugates. As such, we chose to pursue a unique polymer architecture that may impart new and useful in vivo properties to proteins beyond those made possible by conventional PEG conjugation.24 PEGylation has been the prevailing standard for nearly 40 years,25 however, with the advent of living polymerizations highly functional polymers with low dispersities can be attached to proteins.24 Many of these polymers are derived from controlled radical polymerization (CRP) techniques such as atom transfer radical polymerization (ATRP) and reversible addition−fragmentation chain transfer (RAFT) polymerization.26−28 These synthetic techniques allow for the development of high-performance protein/polymer systems such as “smart” bioconjugates with physical properties that can change in response to a variety of external stimuli such as temperature or pH.29 Among the benefits of synthetic polymer bioconjugation is the ability to accurately control polymer molecular weight and tune chemical functionality. Controlling molecular weight while maintaining low polymer dispersity is of great importance given that the hydrodynamic size of the conjugate can greatly affect its in vivo behavior. For example, the permeability limits of tissues vary greatly and accurate control of particle size can be used to modulate blood residence time and tissue targeting.30 Furthermore, synthetic CRP techniques give access to a multitude of unique chemical functionalities that can be used to control bioconjugation chemistry.26 These techniques allow for the modification of proteinaceous substrates with therapeutically useful polymers bearing a variety of functionalities such as imaging agents or therapeutics.24,31,32 Given the importance of size, architecture, and structural homogeneity to in vivo behavior, protein bioconjugates with ROMP-derived polynorbornenes (PNBs) are of particular interest due to their unique structure and versatility. ROMP polymerizations are known for their speed, fidelity, and ability to yield low-dispersity polymers while maintaining accurate control over polymer molecular weight and architecture. ROMP can be performed with functionalized monomers under both organic and aqueous conditions, and has been used to prepare a wide variety of bioactive materials for diagnostic and therapeutic applications.33−36 In addition, the PNB backbone is unique relative to the acrylate or acrylamide backbones of many CRP-derived polymers, as well as to that of the PEG backbone. A recent study by Wignall and co-workers investigated the conformation of PEGylated monoester polynorbornenes using small angle neutron scattering (SANS) and concluded that such polymers assume rigid cylindrical conformations in aqueous solution.37 Furthermore, Gianneschi et al. modeled peptide norbornene polymers, resulting in globular conformations, where the peptide component interacts with the backbone to rigidify the molecule.38 This is in vast difference to single-site PEGylation, whereby the solved structure shows that the PEG chain has minimal interaction with the protein component.39 As a result, PNB conjugation to protein surfaces may impart different surface masking properties over PEG and CRP-derived polymers. Surprisingly, there are very few studies where protein conjugates are synthesized with PNB-derived polymers. Reports that do exist were limited by multistep syntheses and used for specialty applications, providing impetus to develop a universal graft-to ROMP strategy.40,41

In a previous study, we investigated the use of ROMP to grow oligo(ethylene glycol)-functionalized PNBs directly from the surface of proteins (graft-from ROMP).42 PEGylated PNBs have previously been used as substrates for aqueous ROMP polymerizations and are therefore excellent candidates for polymer bioconjugation chemistry.43,44 These polymers maintain water solubility even at relatively high molecular weights and their associated PEG functionalization can provide a useful comparison to proteins conjugated with free PEG groups. In our previous report regarding graft-from ROMP, control over polymer molecular weight proved difficult. In order to address this issue, we have developed a graft-to strategy as an alternative that would allow for the attachment of low molecular weight and low dispersity polymers. The graft-to strategy is among the simplest and most widely used methods for polymer bioconjugation.26−29 This is, in part, because it allows for the preparation and characterization of polymers prior to bioconjugation. In this report, we describe the graft-to functionalization of lysozyme with PEGylated PNBs. Bioconjugation chemistry proceeds with full conversion of the protein and the resulting protein/polymer conjugate retains native secondary structure. To demonstrate the generality of the process, we have also prepared multivalent bioconjugates using bacteriophage Qβ as a protein scaffold. When expressed in E. coli, the Qβ coat protein self-assembles, forming a ∼30 nm icosahedral capsid or, more commonly, a virus-like particle (VLP).45 Qβ VLPs can be expressed recombinantly and have seen intense development as biological nanomaterials. Polynorbornene conjugation had minimal effect on the structure of the Qβ VLP, while showing no toxicity to murine fibroblasts. The methods reported herein provide a new tool to conjugate proteins with low dispersity polymers of unique backbone architectures.

■

EXPERIMENTAL SECTION

Materials. Methylene chloride (reagent grade), methylene chloride (extra dry), chloroform, and deuterated chloroform were purchased from Acros Organics. Methanol, tetrahydrofuran (dried over molecular sieves), ethyl acetate, pentane, butanol, hexane, Miller LB Broth, Dsucrose, and isopropyl β-D-1-thiogalactopyranoside were purchased from Fisher Scientific. cis-5-Norbornene-exo-2,3-dicarboxylic anhydride, 5-norbornene-exo-2-carboxylic acid, Grubbs’ second generation catalyst, L-cysteine, and 2-mercaptoethanol were purchased from Sigma-Aldrich. Poly(ethylene glycol) monomethyl ether (Mn ≃ 350) and sodium hydride (60% oil dispersion) were purchased from AlfaAesar. 4-Dimethylaminopyridine was purchased from AnaSpec, Inc. N(3-(Dimethylamino)propyl)-N′-ethylcarbodiimide hydrochloride was purchased from Chem-Impex International. Poly(ethylene glycol) (Mn = 8000) was purchased from Amresco. Ellman’s reagent was purchased from Thermo-Scientific. Egg white lysozyme was purchased from bioWORLD. All reagents were used directly, without further purification. Instrumentaton. 1H NMR spectra were obtained using either a 300 MHz Varian Gemini spectrometer or a 600 MHz Varian Inova NMR spectrometer. All NMR spectra were analyzed against residual solvent peaks. Nanostructure-assisted laser desorption-ionization (NALDI) spectra were obtained with a Bruker Autoflex III MALDITOF-TOF mass spectrometer equipped with a 200 Hz Smartbeam II laser system. For all NALDI experiments, neat samples without matrix were analyzed using a Bruker NALDI nanostructured target plate accessory in the range of m/z = 100−1000 Da. Gel permeation chromatography (GPC) was performed on a Shimadzu Prominence GPC instrument equipped with a Shimadzu RID10A differential refractometer detector. Stationary phase was two Phenomenex 10E3A size exclusion columns in sequence maintained at 40 °C. GPC mobile phase was anhydrous THF at a flow rate of 1.0 mL/min. Size exclusion B

DOI: 10.1021/acs.biomac.5b01582 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules chromatography (SEC) was performed using a GE Healthcare AKTAFPLC 900 chromatography system equipped with a Superdex 75 10/ 300 GL size exclusion column for lysozyme samples or a Sephacryl 1000 SF 10/300 size exclusion column for Qβ samples. For all FPLC experiments, the mobile phase was 50 mM phosphate buffer, with 150 mM NaCl (pH 7.4) at a flow rate of 0.4 mL/min. Dynamic light scattering (DLS) experiments were performed on a Wyatt DynaPro NanoStar DLS instrument. Samples were analyzed at 25 °C in plastic disposable cuvettes with a path length of 10 mm. SDS polyacrylamide gel electrophoresis (PAGE) was performed on Novex NuPAGE 4− 12% bis−tris protein gels (1.0 mm × 12 well; 35 min, 200 V, 10X SDS-PAGE running buffer, pH 8.3). Gels were stained with Coomassie SimplyBlue SafeStain (Life Technologies). Transmission electron microscopy (TEM) was performed on an FEI Tecnai F30 microscope. TEM samples were mounted on 400 mesh hexagonal copper grids bearing Formvar support film, stained with 3% uranyl acetate solution, and allowed to dry for 12 h. Microplate measurements were taken with a Biotek Synergy HT microplate reader. Centrifugation was performed with an Eppendorf 5810 R centrifuge. Ultracentrifugation was performed with a Beckman Coulter Optima L100 XP ultracentrifuge. Circular dichroism spectroscopy was performed using an AVIV circular dichroism spectrometer model 215 with a 450-W Suprasil Xenon arc lamp. Exo-poly(ethylene glycol) Monoester Norbornene (1). The monomer was prepared based on a previously reported procedure.42,46 5-Norbornene-exo-2-carboxylic acid (1.38 g, 10 mmol), poly(ethylene glycol) monomethyl ether (3.50 g, 10 mmol), and DMAP (122.17 mg, 1.0 mmol) were dissolved in 25 mL dry CH2Cl2. EDC (1.92 g, 10 mmol) was then added and the mixture was refluxed for 12 h under nitrogen atmosphere. The mixture was then concentrated under reduced pressure to a white sticky oil which was taken up into ethyl acetate (100 mL) and water (50 mL). The organic layer was washed with two aliquots of saturated sodium bicarbonate solution (100 mL) and two aliquots of brine (100 mL). The organic layers were combined, concentrated under reduced pressure, and the clear oily residue was subjected to silica gel chromatography using a gradient of 1−6% MeOH/CH2Cl2 in increments of 100 mL to yield pure product as a clear pale-yellow oil that was soluble in both CH2Cl2 and water (3.4 g, 82%). For all subsequent stoichiometric calculations using this product, an average molecular weight of 571 Da was assumed as given by the highest-intensity peak shown by NALDI-MS. 1H NMR (300.1 MHz, CDCl3, 17 °C, ppm): δ = 6.13 (2H, m), 4.24 (2H, PEG methylene, br t, J = 9.0 Hz), 3.64 (PEG methylene, m), 3.37 (3H PEG methyl, s), 3.04 (1H, br s), 2.91 (1H, br s), 2.26 (1H, m), 1.94 (1H, m), 1.53 (1H, d, J = 6.0 Hz), 1.36 (2H, m). MS (NALDI-TOF) m/z: [M] Calcd for C15H24O5 (n = 3), 284.35; Found, 284.762 (10%); Calcd for C28H24O6 (n = 4), 328.40; Found, 328.848 (37%); Calcd for C19H32O7 (n = 5), 372.46; Found, 372.914 (76%); Calcd for C21H36O8 (n = 6), 416.51; Found, 416.968 (100%); Calcd for C23H40O9 (n = 7), 460.56; Found, 461.023 (95%); Calcd for C25H44O10 (n = 8), 504.62; Found, 505.093 (73%); Calcd for C27H48O11 (n = 9), 548.67; Found, 549.128 (44%); Calcd for C29H52O12 (n = 10), 592.72; Found, 593.175 (20%). Grubbs’ Third Generation Catalyst. Grubbs’ second generation catalyst was converted to the third generation catalyst according to a previously reported procedure.47 To a 20 mL scintillation vial equipped with a magnetic stir bar was added Grubbs’ second generation catalyst, [(H2IMes) (PCy3) (Cl)2RuCHPh] (100 mg, 0.12 mmol). The red powder was dissolved directly in 1.0 mL of pyridine and allowed to stir under a nitrogen atmosphere for 5 min at 25 °C until all red color disappeared to yield a clear, dark green solution. A total of 10 mL of cold pentane was then added, and the solution was allowed to stir for 5 more minutes. The green precipitate that formed was isolated via fine fritted funnel and washed with 20 mL of cold pentane. The green powder was then collected, dried under reduced pressure, and stored as pure Grubbs’ third-generation catalyst of the form [(H2IMes) (py)2(Cl)2RuCHPh], (66 mg, 76%). Poly(norbornene) Block Copolymers (2). A dram vial equipped with a magnetic stir bar was charged with exo-poly(ethylene glycol) monoester norbornene monomer (200 mg, 350 μmol) dissolved in 0.5

mL dry CH2Cl2. The vial was capped with a rubber septum and purged with nitrogen. Grubbs’ catalyst [(5 kDa: 28.33 mg, 38.9 μmol), (10 kDa: 14.16 mg, 19.5 μmol), (15 kDa: 9.8 mg, 13.4 μmol)] dissolved in 0.5 mL of dry CH2Cl2 was then added quickly via syringe. The solution was allowed to stir in the dark for 15 min. exo-Norbornene dicarboxylic anhydride [(5 kDa: 12.78 mg, 77.8 μmol), (10 kDa: 9.58 mg, 58.4 μmol), (15 kDa: 6.6 mg, 40.4 μmol)] dissolved in 0.2 mL of dry CH2Cl2 was then added quickly via syringe and allowed to stir for 15 more minutes. The solution was then quenched with 400 μL of ethyl vinyl ether and allowed to stir for 30 min. The solution was then concentrated and precipitated into cold pentane. The pentane was decanted, and the residue was collected and dried under high vacuum to yield a clear amber oil that was soluble in CH2Cl2 and water (5 kDa, 77%; 10 kDa, 84%; 15 kDa, 83%). Qβ Expression and Purification. Chemically competent BL21(DE3) E. coli cells were transformed with pET28CP (containing the Qβ coat protein sequence) and plated onto lysogeny broth (LB) agar media containing kanamycin (50 μg/mL). The following day, isolated colonies were picked from plates into 100 mL of autoclaved selective LB media and grown to saturation for 12 h at 37 °C. A total of 10 mL of culture was then diluted into 1000 mL of freshly prepared selective LB. Culture growth was monitored by optical density at 600 nm (OD600). When the OD600 of the cultures reached approximately 0.8 (mid log phase), protein expression was induced with the addition of 10 mL of 100 mM IPTG, giving a final IPTG concentration of 1 mM. Shaking was continued at 37 °C for an additional 6 h, at which point cells were collected by centrifugation in an Eppendorf A-4−81 rotor at 4000 rpm (4 °C) for 30 min. The supernatant was decanted, and the cell pellet was frozen at −80 °C for 12 h. Cells were then resuspended in ∼100 mL of 1× Tris Buffered Saline (TBS), pH 7.4. The buffer used for the original resuspension continued to be used for subsequent steps of particle preparation. Samples were chilled on ice and then sonicated with a probe sonicator (10 min total sonication time, 5 s on and 5 s off, 60−70 W power output) in an ice bath to lyse cells. The cell debris was pelleted in an Eppendorf FA-45−6−30 rotor at 10000 rpm for 10 min, and the supernatant was decanted and collected. The Qβ particles were precipitated from the resulting supernatant by the addition of 10% w/v PEG8000 at 4 °C for 12 h on a rotisserie. The precipitated fraction was isolated from the supernatant by centrifugation in an Eppendorf FA-45−6−30 rotor for 10 min (4 °C) at 10000 rpm. The pellet was redissolved in ∼20 mL of TBS and extracted with a 1:1 v/v solution of n-BuOH/CHCl3 to remove excess lipid. The aqueous fraction was collected following centrifugation using a FA-45−6−30 rotor for 10 min, 4 °C at 10000 rpm. Qβ particles were purified on 10−40% sucrose velocity gradients in an SW28 rotor at 28000 rpm for 4.5 h. Approximately 4 mL was pulled from each gradient tube and subsequently pelleted in an ultracentrifuge (50.2Ti rotor, 42K, 3 h). The purified Qβ particles were dissolved in PBS (pH 7.4) and purity was verified via PAGE, FPLC, DLS, and TEM. Preparation of Lysozyme Conjugates (3). Egg white lysozyme (5 kDa, 5.58 mg; 10 kDa, 3.15 mg; 15 kDa, 3.2 mg, 1 equiv), dissolved in 4.5 mL of 10 mM phosphate buffer (pH 7.4), was added to a 20 mL scintillation vial equipped with a magnetic stir bar and stirred at 25 °C for 10 min. Poly(norbornene-PEG)-b-(norbornene anhydride) (5 kDa, 187.3 mg; 10 kDa, 182.0 mg; 15 kDa, 255 mg, 75 equiv) was dissolved in 0.5 mL of DMSO and slowly added to the lysozyme solution via dropwise infusion over 45 min. The pH of the solution was constantly monitored using litmus paper and maintained in the range of pH 8−9 using 100 mM NaOH throughout the entire addition period. After addition was completed, the solution was stirred for an additional 60 min. Lysozyme/PNB conjugates were purified via FPLC by collecting product peak between elution volumes of 6 and 8 mL (Figure S3B−D). Preparation of Qβ Conjugates (4). Qβ (2.16 mg, 1 equiv), dissolved in 4.5 mL of 10 mM phosphate buffer (pH 7.4), was added to a 20 mL scintillation vial equipped with a magnetic stir bar and stirred at 25 °C for 10 min. Poly(norbornene-PEG)-b-(norbornene anhydride) (134.8 mg, 80 equiv/Qβ coat protein) was dissolved in 0.5 mL of DMSO and slowly added to the lysozyme solution via dropwise C

DOI: 10.1021/acs.biomac.5b01582 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules Scheme 1. ROMP Block Copolymerization and Subsequent Grafting to Lysozyme

infusion over 45 min. The pH of the solution was constantly monitored using litmus paper and maintained in the range of pH 7−8 using 10 mM NaOH throughout the entire addition period. After addition was completed, the solution was stirred for an additional 60 min. Qβ/PNB conjugates were purified via spin filtration (100 kDa MWCO, 4k rpm, 6×; Qβ) and characterized via PAGE, FPLC, DLS, and TEM (Figure 3). PAGE Iodine Staining for PEG. Following electrophoresis, the gel was washed according to recommended procedure (Invitrogen SimplyBlue SafeStain microwave protocol), submerged in 100 mL of 5% w/v BaCl2 solution, and allowed to shake on an orbital shaker for 15 min. The gel was then placed in 100 mL of deionized water and allowed to shake for 30 min. Finally, the gel was placed in 100 mL of 50 mM iodine solution and allowed to shake for 15 min to stain. The iodine solution was then decanted and replaced with 100 mL of deionized water to begin destaining. The gel destained quickly, and the deionized water was replaced every 2 min until the desired contrast between analyte bands and gel background was achieved. Ellman’s Protein Quantification Assay. Lysozyme/PNB conjugates were quantified via Ellman’s Assay modified for internal cysteines adapted from a previous protocol.48 An L-cysteine standard curve was prepared from 0.5−0.1 mM in PBS (pH 7.4). Lysozyme standards were also prepared at 0.5 and 0.1 mg/mL as internal standards. A total of 400 μL of sample or standard was placed into a 0.75 mL Eppendorf and denatured at 90 °C for 10 min. Samples were cooled to room temperature and 2 μL of 6 N HCl was added. A 20 μL aliquot of 0.2 mM 2-mercaptoethanol and 100 μL of 2 mM Ellman’s Reagent were then added, and the samples were briefly vortexed. The samples were incubated at room temperature for 30 min and then the absorbance at 412 nm was read in triplicate using a microplate reader. The amount of lysozyme/PNB was quantified via comparison to the cysteine standard curve, and the lysozyme internal standard was used to verify the effectiveness of the assay. Circular Dichroism Spectroscopy. Circular dichroism spectra were collected from PBS buffered samples in the concentration range of 1− 0.2 mg/mL using a 1 mm cuvette. Spectra were collected three times at 25 °C from 180−260 nm with a step of 1 nm, an averaging time of 4 s, and a settling time of 0.333 s. The reported spectra were averaged, baseline corrected, and normalized via the protein molarity of the sample. Cell Viability Assay. NIH 3T3 cells were maintained in advanced DMEM containing 1% penicillin/streptomycin and 10% FBS at 37 °C in a humidified atmosphere with 5% CO2. Cells were plated in 96-well microtiter plates in triplicate (104 cells/well) in 100 μL of complete DMEM. After 24 h, media was replaced with protein solutions (Qβ/ PNB) or Qβ wild-type) prepared in DMEM (100 μL/well) to the indicated concentrations and allowed to incubate for 24 h at 37 °C. Cells were then assayed for viability using the MTT assay. MTT (5 mg/mL, 25 μL/well) was added to each individual well and incubated at 37 °C for approximately 1 h (assay was stopped when significant accumulation of purple formazan crystals were visibly observed in control wells). Media was carefully aspirated and DMSO was added (200 μL/well) to dissolve the purple MTT-formazan crystals. Absorbance of the dissolved formazan was quantified at 570 nm using a UV−vis plate reader and cell viability was determined as a

fraction of absorbance relative to untreated control wells. Data are presented as average values ± standard deviation.

■

RESULTS AND DISCUSSION In this report, we have prepared poly(norbornene-PEG)-bpoly(norbornene anhydride) in three target molecular weights: 5, 10, and 15 kDa. We selected this range of polymer weights as a model that was comparable to conventional protein modifications with PEG for pharmacological and other therapeutic applications.10,49 Additionally, small Mn target increments allowed us to demonstrate the ability to accurately and effectively synthesize and graft a variety of short polymers onto proteins. These polymers are predominately composed of a PEGylated PNB block, and are capped with a short block of ∼2 units of exo-norbornene dicarboxylic anhydride (Scheme 1). The PEGylated block provides a unique backbone architecture for PEG-like polymers, while the cyclic anhydride block is required for reactivity toward primary amines. ROMP proceeded with Grubbs’ third generation catalyst in dry organic conditions, allowing for accurate targeting of polymer molecular weights while maintaining low dispersity. 1H NMR spectra of crude polymer solutions indicated complete monomer conversion by monitoring the disappearance of monomer olefin proton signals in the region of 6.0−6.5 ppm and the appearance of polymer olefin peaks in the region of 5.0−6.0 ppm (Figure 1A). Olefin proton signals from both monomers can be distinguished in this region. 1H NMR of concentrated purified polymer solutions also confirmed that no signals from the ruthenium catalyst at either ∼19 ppm (catalyst alkylidene) or 6.5−8.5 ppm (catalyst fingerprint) were observed, indicating that the majority of residual catalyst, as observable by NMR, was removed after one round of precipitation in pentane (Figure S1). Gel permeation chromatograms (GPC) of polymer samples showed that the synthesized polymers were in good agreement with the targeted molecular weight and maintained dispersities of less than 1.1. A small high-molecular weight shoulder was seen in the GPC tracings of the 10 and 15 kDa polymers. This shoulder is likely attributed to a small amount of incomplete ROMP termination, and showed no bearing on subsequent bioconjugation reactions. 1H NMR end group analysis confirmed that polymers were capped with an average of ∼2 anhydride groups. (Figure 1B, Table 1). The short cap of norbornene anhydride at the end of each polymer served as a reactive linker that facilitated attachment of the polymers to lysine residues on the protein surface. We selected the use of the anhydride functionality over the more commonly used N-hydroxysuccinimidyl ester (NHS) on the basis of the greatly reduced cost of the norbornene anhydride over the norbornene NHS ester, as well as our previous report D

DOI: 10.1021/acs.biomac.5b01582 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

much faster than that of proteinaceous material, it can be used to quickly estimate the presence of polymer impurity. Fast protein liquid chromatography (FPLC) of crude conjugates also indicated the disappearance of wild-type lysozyme (∼14 mL) and the appearance of a new peak centered at 7.0 mL. This peak was well resolved and could be easily distinguished from free PNB, eluting at a slightly higher volume of approximately 10 mL (Figure S3). As a result, FPLC provided an efficient way to purify the conjugates. PAGE gel analysis of purified conjugates, carried out via ImageJ software,55 showed high molecular weight bands focused at 50, 80, and 100 kDa for lysozyme conjugated with 5, 10, and 15 kDa PNB, respectively (Figure 2A). Given the polymer molecular weights determined via GPC, this suggests the attachment of a polymer at all six lysine residues on lysozyme. A distribution of conjugate weights can also be seen, most notably on the 15 kDa conjugate, indicating incomplete conversion of all lysines, as is typical. The PAGE gel was also visualized following staining with barium chloride and iodine.

Figure 1. (A) 1H NMR spectra of PEGylated norbornene monomer (NB-PEG) and PNB block copolymers. (B) GPC chromatograms of PNB block copolymers.

Table 1. GPC Characterization of Block Copolymersa Polymer 5 kDa 10 kDa 10 kDa 15 kDa

(Lys) (Qβ)

[M1]/[M2]/[C]

Mn

Mw

Mw/Mn

9:2:1 18:3:1 18:3:1 26:3:1

6423 11205 10597 15154

7009 11981 11107 15873

1.08 1.07 1.05 1.04

a

M1 = PEGylated norbornene, M2 = norbornene anhydride, C = catalyst. Separate 10 kDa polymers were prepared for lysozyme and Qβ conjugations, as indicated.

showing effective conjugation of the anhydride. The bioconjugation of both lysozyme and Qβ were carried out at pH ranges of 8−9 and 7−8, respectively, in PBS buffer containing 10% v/v DMSO. These pH ranges were selected as the most alkaline extremes of previously reported ranges at which these proteins retain their structure, stability, and activity.50,51 The composition of DMSO was selected based on previously reported protein bioconjugation protocols, as well as our own previous experience performing bioconjugations with these proteins.42,52,53 However, care should be taken in each individual case as DMSO can oxidize proteins and slightly alkaline conditions can lead to deamidation. The graft-to reaction with lysozyme was performed with a 75-fold molar excess of polymer per protein. Since lysozyme possesses six solvent-accessible lysine residues on its surface,54 this corresponds to approximately 12 equiv of polymer per lysine residue. Lysozyme/PNB conjugates were analyzed via polyacrylamide gel electrophoresis (PAGE) to determine approximate purity, conversion efficiency and molecular weight. PAGE analysis of crude conjugates showed complete conversion from wild-type lysozyme, as indicated by the disappearance of the wild-type lysozyme band. The appearance of a high molecular weight band at the top of the gel confirmed the formation of conjugates (Figure S2A). As seen previously, free PNB stains temporarily with Coomassie on PAGE gels, yielding a unique color and pattern (Figure S2B).42 Since this pattern destains

Figure 2. (A) PAGE gel of purified lysozyme/PNB conjugates; L = ladder, lane 1 = wild-type lysozyme, lanes 2−4 represent lysozyme conjugates with 5, 10, and 15 kDa PNB. (B) PAGE gel stained with barium iodide, highlighting PEGylated products. (C) Size exclusion chromatograms of lysozyme/PNB conjugates (colored) and wild-type lysozyme (black). (D) Circular dichroism spectra of lysozyme/PNB conjugates (colored) and wild-type lysozyme (black). E

DOI: 10.1021/acs.biomac.5b01582 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

Figure 3. (A) General structure of Qβ/PNB conjugate. (B) PAGE gel of purified Qβ/PNB; L = ladder, lane 1 = wild-type Qβ coat protein, lane 2 = Qβ coat protein/PNB conjugate. (C) PAGE gel stained with barium iodide; lane 3 = wild-type Qβ coat protein, lane 4 = Qβ coat protein/PNB conjugate. (D) Size exclusion chromatograms of wild-type Qβ (black) and Qβ/PNB conjugate (orange): 280 nm = solid line, 260 nm = dashed line. (E) Dynamic light scattering analysis of wild-type Qβ and Qβ/PNB conjugates. (F) TEM micrograph of wild-type Qβ particles. (G) TEM micrograph of Qβ/PNB conjugates.

Barium iodide forms complexes with PEG chains and allows for selective visualization of PEG content.56 PAGE gels stained with barium iodide confirmed that PEGylated conjugates were present (Figure 2B). FPLC of purified conjugate samples shows a single peak centered at 7 mL, at the void volume of the column (Figure 2C). Purified lysozyme/PNB conjugates all elute at the column void volume; therefore, chromatographic resolution corresponding to differences in conjugate size was not observed. Circular dichroism (CD) spectroscopy was performed on purified conjugates in order to determine whether the secondary structure of lysozyme was retained following conjugation. Protein quantification via UV was unsuccessful since the absorbance spectra of purified conjugates could not be distinguished from that of PNB controls (Figure S4). The Bradford protein assay was also unsuccessful since PNB produced a false positive colorimetric signal. Thus, Ellman’s assay was used to quantify the conjugates via the eight cysteine residues present as internal disulfide bonds and allowed for molar ellipticity determination. Free PNB did not give a false

positive to the Ellman’s reagent and the conjugates were quantified by comparison to an L-cysteine standard. Molar ellipticity profiles of lysozyme/PNB conjugates aligned with equivalent intensity to that of wild-type lysozyme in the range of 200−250 nm, indicating that the protein secondary structure remained intact (Figure 2D). Free PNB gave a small signal between 200−235 nm (Figure S5) that introduced noise into the spectra, likely contributing to any differences between spectra in this range. The polynorbornene modification strategy was next adapted to bacteriophage Qβ to demonstrate the generality of this process toward a multivalent nanoparticle carrier. Qβ has been extensively used as a nanoparticle in a wide variety of therapeutic and diagnostic applications, including as a molecular targeting platform for drug delivery32 and as a scaffold for polymeric MRI contrast agents.57 The Qβ particle consists of 180 coat proteins, each approximately 14.2 kDa in size, with an exterior surface displaying four primary amines per coat protein that could participate in bioconjugation reactions (720 total amines per particle). We selected 10 kDa PEGylated F

DOI: 10.1021/acs.biomac.5b01582 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules PNB for the modification of Qβ, consistent with a variety of existing PEGylated therapeutic agents.49 In order to graft polymers to the surface of Qβ, we employed a slightly modified bioconjugation strategy from that described above to afford Qβ/PNB nanoparticle conjugates (Figure 3A). Qβ has been shown to be most structurally stable in the pH range of 6− 8.50,51 In our experience, alkaline conditions lead to higher degrees of dissociation, likely due to the intricate network of disulfide linkages. As a result, when modifying Qβ care was taken to maintain the pH in the range of 7−8 in order to prevent capsid disassembly. In an attempt to partially compensate for the decrease in pH, a slightly larger excess of PNB was used in the conjugation reaction (20 equiv per amine). Due to the large size of the intact Qβ/PNB conjugates, purification was achieved via centrifugal spin filtration using a high molecular weight spin filter (100 kDa cutoff). Purified conjugates were evaluated via PAGE, revealing that moderate degrees of modification were achieved. A high molecular weight band appeared at ∼24 kDa, corresponding to the attachment of one 10 kDa PNB. Pixel density analysis of PAGE gels revealed an average degree of modification of coat proteins of approximately 50%, where single modified species were predominant. Iodine staining confirmed the presence of PEGylated material in the same location, rather than protein dimer (Figure 3B,C). Unmodified coat protein was still present, likely attributed to the combination of performing the bioconjugation reaction at a slightly lower pH and steric hindrance of the multivalent particle, as can be seen in other Qβ/polymer modifications.57 Concentrated filtrate from multiple rounds of spin filtration was also analyzed via PAGE, and no protein content was observed, suggesting that the Qβ particle did not dissociate throughout the bioconjugation reaction (Figure S6). FPLC of purified Qβ/PNB conjugates showed a single peak centered at ∼20.5 mL with a 260/280 nm ratio of 1.9, indicative of encapsulated RNA. Surprisingly, the nanoparticle conjugate appeared at a higher elution volume than that of the wild-type Qβ (Figure 3D). This behavior is likely attributed to attractive interactions between the size exclusion media and the PNB on the particle surface. Nevertheless, analysis of particle size via dynamic light scattering (DLS) indicated an increase in hydrodynamic diameter from ∼30 to ∼40 nm as a result of conjugation (Figure 3E) and visualization via transmission electron microscopy verified that the particle remained intact following bioconjugation. TEM analysis revealed the appearance of a distinct coating around Qβ/ PNB particles that is not present in the wild-type samples, likely attributable to the polymer. (Figures 3F,G and S7). Finally, cell viability assays were performed to determine cytotoxicity of the polymer conjugates. Cell viability studies were conducted by exposing NIH 3T3 embryonic fibroblasts cells to Qβ/PNB conjugates. This cell line was selected given its broad use as an in vitro standard for mammalian cytotoxicity.29,58 Cells were cultured at a density of 10000 cells per well and exposed to Qβ/PNB conjugates at concentrations of 0.1, 0.5, and 1.0 mg/mL for 24 h. Cell viability was not significantly affected by either Qβ/PNB or wild-type Qβ nanoparticles after 24 h of incubation (Figure 4). These findings demonstrate that the nanoparticle conjugates are noncytotoxic to the NIH 3T3 cell line at concentrations of up to 1 mg/mL.

Figure 4. Cell viability of NIH 3T3 embryonic fibroblast cells after 24 h exposure to Qβ/PNB and Qβ wild-type nanoparticles at varied concentrations, relative to untreated controls.

■

CONCLUSION In this report, we have investigated the graft-to conjugation of PEGylated polynorbornenes (PNBs) derived from ROMP onto proteinaceous substrates. ROMP provided a means to accurately target polymer molecular weight while maintaining low dispersities. Our strategy involved the preparation of watersoluble block copolymers that possessed a large block of oligo(ethylene glycol) monoester norbornene, capped with a short block of amine-reactive norbornene dicarboxylic anhydride. Multiple polymer chains rapidly attached to lysozyme within 45 min to yield lysozyme/PNB conjugates that could be readily purified via size exclusion chromatography and still maintained native secondary structure. Grafting PNBs to the coat protein of bacteriophage Qβ showed moderate polymer attachment with no particle dissociation. Cell viability studies revealed that Qβ/PNB nanoparticle conjugates show no cytotoxicity to NIH 3T3 embryonic fibroblast cells at concentrations of up to 1 mg/mL. Our results have established the use of PNB block copolymers as a viable and effective alternative to conventional protein bioconjugation. PNB bioconjugation allows for a wide variety of chemical modifications and architectures, a unique polymer backbone structure, and the added benefit of a simple yet powerful synthetic technique afforded by ROMP.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.biomac.5b01582. 1 H NMR, UV, and CD spectra PNB block copolymers, FPLC tracings, PAGE gels, and TEM micrographs of protein/polymer conjugates (PDF).

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors acknowledge the Case Western Reserve University Swagelok Center for Surface Analysis of Materials for TEM measurements and the National Science Foundation (CHE 1306447) for supporting this work.

■

REFERENCES

(1) Pavlou, A. K.; Reichert, J. M. Nat. Biotechnol. 2004, 22, 1513.

G

DOI: 10.1021/acs.biomac.5b01582 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules (2) Leader, B.; Baca, Q. J.; Golan, D. E. Nat. Rev. Drug Discovery 2008, 7, 21. (3) Dimitrov, D. S. Methods Mol. Biol. 2012, 899, 1. (4) Pisal, D. S.; Kosloski, M. P.; Balu-Iyer, S. V. J. Pharm. Sci. 2010, 99, 2557. (5) Jevsevar, S.; Kunstelj, M.; Porekar, V. G. Biotechnol. J. 2010, 5, 113. (6) Yang, T. Mod. Chem. Appl. 2013, 1, 1. (7) Milla, P.; Dosio, F.; Cattel, L. Curr. Drug Metab. 2012, 13, 105. (8) Hamed, E.; Xu, T.; Keten, S. Biomacromolecules 2013, 14, 4053. (9) Wieczorek, S.; Schwaar, T.; Senge, M. O.; Börner, H. G. Biomacromolecules 2015, 16, 3308. (10) Bailon, P.; Berthold, W. Pharm. Sci. Technol. Today 1998, 1, 352. (11) Bailon, P.; Palleroni, A.; Schaffer, C. A.; Spence, C. L.; Fung, W.J.; Porter, J. E.; Ehrlich, G. K.; Pan, W.; Xu, Z.-X.; Modi, M. W.; Farid, A.; Berthold, W. Bioconjugate Chem. 2001, 12, 195. (12) DeNardo, S. J.; Yao, Z.; Lam, K. S.; Song, A.; Burke, P. a.; Mirick, G. R.; Lamborn, K. R.; O’Donnell, R. T.; DeNardo, G. L. Clin. Cancer Res. 2003, 9, 3854S. (13) Riebeseel, K.; Biedermann, E.; Lö ser, R.; Breiter, N.; Hanselmann, R.; Mülhaupt, R.; Unger, C.; Kratz, F. Bioconjugate Chem. 2002, 13, 773. (14) Rajender Reddy, K.; Modi, M. W.; Pedder, S. Adv. Drug Delivery Rev. 2002, 54, 571. (15) Harris, J. M.; Martin, N. E.; Modi, M. Clin. Pharmacokinet. 2001, 40, 539. (16) Bruckman, M. A.; Randolph, L. N.; VanMeter, A.; Hern, S.; Shoffstall, A. J.; Taurog, R. E.; Steinmetz, N. F. Virology 2014, 449, 163. (17) Lee, K. L.; Shukla, S.; Wu, M.; Ayat, N. R.; El Sanadi, C. E.; Wen, A. M.; Edelbrock, J. F.; Pokorski, J. K.; Commandeur, U.; Dubyak, G. R.; Steinmetz, N. F. Acta Biomater. 2015, 19, 166. (18) Harris, J. M.; Chess, R. B. Nat. Rev. Drug Discovery 2003, 2, 214. (19) Vugmeyster, Y.; Entrican, C. A.; Joyce, A. P.; LawrenceHenderson, R. F.; Leary, B. A.; Mahoney, C. S.; Patel, H. K.; Raso, S. W.; Olland, S. H.; Hegen, M.; Xu, X. Bioconjugate Chem. 2012, 23, 1452. (20) Fuhrmann, G.; Grotzky, A.; Lukić, R.; Matoori, S.; Luciani, P.; Yu, H.; Zhang, B.; Walde, P.; Schlüter, A. D.; Gauthier, M. A.; Leroux, J.-C. Nat. Chem. 2013, 5, 582. (21) Liu, M.; Johansen, P.; Zabel, F.; Leroux, J.-C.; Gauthier, M. A. Nat. Commun. 2014, 5, 5526. (22) Liu, M.; Tirino, P.; Radivojevic, M.; Phillips, D. J.; Gibson, M. I.; Leroux, J.-C.; Gauthier, M. A. Adv. Funct. Mater. 2013, 23, 2007. (23) Schulz, J. D.; Patt, M.; Basler, S.; Kries, H.; Hilvert, D.; Gauthier, M. A.; Leroux, J. Adv. Mater. 2015, n/a. (24) Pelegri-O’Day, E. M.; Lin, E.-W.; Maynard, H. D. J. Am. Chem. Soc. 2014, 136, 14323. (25) Abuchowski, A.; van Es, T.; Palczuk, N. C.; Davis, F. F. J. Biol. Chem. 1977, 252, 3578. (26) Wallat, J. D.; Rose, K. A.; Pokorski, J. K. Polym. Chem. 2014, 5, 1545. (27) Grover, G. N.; Maynard, H. D. Curr. Opin. Chem. Biol. 2010, 14, 818. (28) Le Droumaguet, B.; Nicolas, J. Polym. Chem. 2010, 1, 563. (29) Broyer, R. M.; Grover, G. N.; Maynard, H. D. Chem. Commun. 2011, 47, 2212. (30) Elsabahy, M.; Wooley, K. L. Chem. Soc. Rev. 2012, 41, 2545. (31) Duncan, R.; Ringsdorf, H.; Satchi-Fainaro, R. Adv. Polym. Sci. 2006, 192, 1. (32) Pokorski, J. K.; Hovlid, M. L.; Finn, M. G. ChemBioChem 2011, 12, 2441. (33) Tomasek, J.; Schatz, J. Green Chem. 2013, 15, 2317. (34) Miki, K.; Oride, K.; Inoue, S.; Kuramochi, Y.; Nayak, R. R.; Matsuoka, H.; Harada, H.; Hiraoka, M.; Ohe, K. Biomaterials 2010, 31, 934. (35) Gueugnon, F.; Denis, I.; Pouliquen, D.; Collette, F.; Delatouche, R.; Héroguez, V.; Grégoire, M.; Bertrand, P.; Blanquart, C. Biomacromolecules 2013, 14, 2396.

(36) Smith, D.; Pentzer, E. B.; Nguyen, S. T. Polym. Rev. 2007, 47, 419. (37) Cheng, G.; Hua, F.; Melnichenko, Y. B.; Hong, K.; Mays, J. W.; Hammouda, B.; Wignall, G. D. Eur. Polym. J. 2008, 44, 2859. (38) Blum, A. P.; Kammeyer, J. K.; Yin, J.; Crystal, D. T.; Rush, A. M.; Gilson, M. K.; Gianneschi, N. C. J. Am. Chem. Soc. 2014, 136, 15422. (39) Cattani, G.; Vogeley, L.; Crowley, P. B. Nat. Chem. 2015, 7, 823. (40) Carrillo, A.; Gujraty, K. V.; Rai, P. R.; Kane, R. S. Nanotechnology 2005, 16, S416. (41) Chen, B.; Metera, K.; Sleiman, H. F. Macromolecules 2005, 38, 1084. (42) Isarov, S. A.; Pokorski, J. K. ACS Macro Lett. 2015, 4, 969. (43) Samanta, D.; Kratz, K.; Zhang, X.; Emrick, T. Macromolecules 2008, 41, 530. (44) Breitenkamp, K.; Emrick, T. J. Polym. Sci., Part A: Polym. Chem. 2005, 43, 5715. (45) Brown, S. D.; Fiedler, J. D.; Finn, M. G. Biochemistry 2009, 48, 11155. (46) Liu, Y.; Piñoń , V.; Weck, M. Polym. Chem. 2011, 2, 1964. (47) Love, J. A.; Morgan, J. P.; Trnka, T. M.; Grubbs, R. H. Angew. Chem., Int. Ed. 2002, 41, 4035. (48) Robyt, J. F.; Ackerman, R. J.; Chittenden, C. G. Arch. Biochem. Biophys. 1971, 147, 262. (49) Alconcel, S. N. S.; Baas, A. S.; Maynard, H. D. Polym. Chem. 2011, 2, 1442. (50) Davies, R. C. C.; Neuberger, a.; Wilson, B. M. M. Biochim. Biophys. Acta - Enzymol. 1969, 178, 294. (51) Feng, Y. Y.; Ong, S. L.; Hu, J. Y.; Tan, X. L.; Ng, W. J. J. Ind. Microbiol. Biotechnol. 2003, 30, 549. (52) Mero, A.; Campisi, M. Polymers (Basel, Switz.) 2014, 6, 346. (53) Jiang, Y.; Liang, M.; Svejkar, D.; Hart-Smith, G.; Lu, H.; Scarano, W.; Stenzel, M. H. Chem. Commun. 2014, 50, 6394. (54) Masuda, T. Chem. Senses 2005, 30, 667. (55) Schneider, C. A.; Rasband, W. S.; Eliceiri, K. W. Nat. Methods 2012, 9, 671. (56) Kurfürst, M. M. Anal. Biochem. 1992, 200, 244. (57) Pokorski, J. K.; Breitenkamp, K.; Liepold, L. O.; Qazi, S.; Finn, M. G. J. Am. Chem. Soc. 2011, 133, 9242. (58) Lu, H.; Wang, D.; Kazane, S.; Javahishvili, T.; Tian, F.; Song, F.; Sellers, A.; Barnett, B.; Schultz, P. G. J. Am. Chem. Soc. 2013, 135, 13885.

H

DOI: 10.1021/acs.biomac.5b01582 Biomacromolecules XXXX, XXX, XXX−XXX