Polyoxazoline: Chemistry, Properties, and Applications in Drug

Apr 1, 2011 - The conjugation of biopharmaceuticals to PEG by a process called “PEGylation” has led to the clinical and marketing success of a num...
1 downloads 11 Views 2MB Size
ARTICLE pubs.acs.org/bc

Polyoxazoline: Chemistry, Properties, and Applications in Drug Delivery Tacey X. Viegas,*,† Michael D. Bentley,† J. Milton Harris,† Zhihao Fang,† Kunsang Yoon,† Bekir Dizman,† Rebecca Weimer,† Anna Mero,§ Gianfranco Pasut,§ and Francesco M. Veronese§ †

Serina Therapeutics, Inc., 601 Genome Way, Huntsville, Alabama 35806, United States Department of Pharmaceutical Sciences, University of Padua, Via F. Marzolo 5, 35131 Padua, Italy

§

ABSTRACT: Polyoxazoline polymers with methyl (PMOZ), ethyl (PEOZ), and propyl (PPOZ) side chains were prepared by the living cationic polymerization method and purified by ion-exchange chromatography. The following properties of polyoxazoline (POZ) were measured: apparent hydrodynamic radius by aqueous size-exclusion chromatography, relative lipophilicity by reverse-phase chromatography, and viscosity by coneplate viscometry. The PEOZ polymers of different molecular weights were first functionalized and then conjugated to model biomolecules such as bovine serum albumin, catalase, ribonuclease, uricase, and insulin. The conjugates of catalase, uricase, and ribonuclease were tested for in vitro activity using substrate-specific reaction methods. The conjugates of insulin were tested for glucose lowering activity by injection to naïve SpragueDawley rats. The conjugates of BSA were injected into New Zealand white rabbits and serum samples were collected periodically and tested for antibodies to BSA. The safety of POZ was also determined by acute and chronic dosing to rats. The results showed that linear polymers of POZ with molecular weights of 1 to 40 kDa can easily be made with polydispersity values below 1.10. Chromatography results showed that PMOZ and PEOZ have a hydrodynamic volume slightly lower than PEG; PEOZ is more lipophilic than PMOZ and PEG; and PEOZ is significantly less viscous than PEG especially at the higher molecular weights. When PEOZ was attached to the enzymes catalase, ribonuclease, and uricase, the in vitro activity of the resultant bioconjugates depended on the extent of protein modification. POZ conjugates of insulin lowered blood glucose levels for a period of 8 h when compared to 2 h for insulin alone. PEOZ, like PEG, was also able to successfully attenuate the immunogenic properties of BSA. The POZ polymers (10 and 20 kDa) are safe when administered intravenously to rats, and the maximum tolerated dose (MTD) was greater than 2 g/kg. Blood counts, serum chemistry, organ weights, and the histopathology of key organs were normal. These results conclude that POZ has the desired drug delivery properties for a new biopolymer.

’ INTRODUCTION Biocompatible polymers are used in drug delivery in a number of complex formats. They can be attached directly to pharmaceutical actives where they are part of the drug substance, conjugated to lipids where they become part of the liposomal delivery system, and linked with cationic molecules where they become part of a nucleic acid complexing polymer. Some examples of biostable and biodegradable polymers that have been tested in medicines are poly(ethylene glycol) (PEG), hyaluronic acid (HA), polysialic acid (PSA), hydroxyethyl starch (HES), polyglutamic acid (PGA), polylactic and polyglutaric acid (PLGA), N-(2-hydroxypropyl) methacrylamide (HPMA), polyvinylpyrrolidone (PVP), and polydextrans. From this list, PEG is the one polymer that has had wide success in delivering proteins, aptamers, and small molecule drugs. The conjugation of biopharmaceuticals to PEG by a process called “PEGylation” has led to the clinical and marketing success of a number of blockbuster macromolecule drugs for the treatment of hepatitis C, neutropenia, and anemia.1 The polymer has the ability to negate several limitations of therapeutic proteins, such as rapid clearance, proteolysis, and immunogenicity. It provides a camouflaging shield around the protein and produces an apparent hydrodynamic size that reduces proteolysis and renal glomerular r 2011 American Chemical Society

filtration. This shield also protects the drug from possible protein antigenantibody-like reactions. Although PEG has been clinically proven and more then ten conjugates are already on the market, this polymer is not completely devoid of technical and commercial challenges. Recent observations of specific antibodies against PEG were detected in the serum of patients treated with PEG-asparaginase2 and PEG-uricase3 that have resulted in a neutralizing effect with loss of therapeutic efficacy. These reports also suggest that preexisting anti-PEG antibodies, of the IgM and IgG types, were identified in over 25% of patients, during screening, who never received prior treatment with PEG drugs.4,5 This observation can be correlated to the large use of PEG in food, beauty, and body care products. Another controversy relates to possible kidney cell vacuolization observed in animals following repeated administration of PEG drug conjugates of hemoglobin,6 TNF-R bp,7 and leptin.8 While marked vacuolation has been reported with the conjugates itself, these observations were not noted when PEG was administered alone. There are numerous patents on the Received: January 24, 2011 Revised: March 7, 2011 Published: April 01, 2011 976

dx.doi.org/10.1021/bc200049d | Bioconjugate Chem. 2011, 22, 976–986

Bioconjugate Chemistry

ARTICLE

that are yellow in color and with polydispersity indices of >1.3. In addition, the termination with the weak nucleophile water does not always give the desired product of 5-position attack (the “thermodynamic” product), but rather gives an attack in the 2-position (the “kinetic” product). This kinetic product is not stable and can rearrange to give an ester product or undergo reversal to the cation.31 These polymers have a high MW shoulder of approximately 510% and significant low MW tailing, when tested and detected by gel permeation chromatography.32 It is generally stated in the literature that this broadening of the MW distribution is due to the chain transfer process through an eliminationdimerization mechanism, although structural details and experimental support for this process are scarce.33 To the extent that chain transfer reactions do occur, such reactions then cannot be considered to be truly living polymerizations. In this paper, the chemistry and properties of POZ are described and compared against PEG. In addition, the chemical and biological properties of POZ conjugates of some model proteins and peptides were determined and compared against similar PEG conjugates. Conjugates of ribonuclease (RNase), uricase, and catalase were prepared and their activity measured by in vitro assays. The activity of POZ insulin conjugates was determined in vivo by measuring the glucose lowering effect in normal rats. The ability of POZ to camouflage the immunogenicity of bovine serum albumin (BSA) was evaluated in a 72 day rabbit study. The acute toxicity of POZ was determined following single and repeated intravenous administration to rats.

Figure 1. Chemical structures of three types of polyalkyloxazoline (POZ).

compositions and applications of PEG in drug delivery that hamper the ability to introduce new PEGylated therapies into the clinic. This hindrance in “freedom to operate” has forced drug companies to seek and test alternative polymers such as HPMA,9 PSA,10 and PVP.11 Another class of polymers called polyoxazolines (POZ) is described in this paper. These polymers can be made with high quality, of different architectures, and with different functional groups. They not only possess the key beneficial properties of PEG, but have characteristics that are novel and unique for different drug delivery applications. They are nonionic, stable, and highly soluble in water and organic solvent. The generic chemical structures of polymethyloxazoline (PMOZ), polyethyloxazoline (PEOZ), and polypropyloxazoline (PPOZ) are shown in Figure 1. A review of the literature shows that the interest in POZ has existed for the past thirty years. It was first developed as a food additive in the early 1980s for which elaborate animal safety studies were completed and documented.12 Studies reported by other scientists showed that POZ can be used in multiple pharmaceutical and medical applications.13,14 It can be conjugated to protein and small molecule drugs,15 grafted onto liposomal bilayers,16,17 formulated into micelles,18 and applied onto surfaces.19 In all these applications, POZ has the same “stealth” properties as PEG without any adverse effects in animal models. Recent studies completed with radiolabeled POZ suggest that the polymer is rapidly excreted by the kidney with no significant accumulation in tissues.20,21 Polyoxazoline polymers are prepared by the living cationic polymerization method. A number of processes have been reported in the literature. One method discusses the stoichiometric addition of an electrophile initiator such as an alkyl tosylate or alky triflate to the oxazoline monomer that is dissolved in a dry organic solvent and in an inert atmosphere.2224 The propagation phase is conducted at 80 °C for approximately 1 to 3 days. In the second method, the polymerization is carried out with microwave energy to reduce the propagation time from days to hours.25,26 In both cases, the living cation is terminated by the introduction of a nucleophile such as OH, NH, COO, or S. Termination is conducted with aqueous sodium carbonate to give a hydroxyl terminal group or by reacting with a secondary amine such as morpholine or piperidine to give a terminal tertiary amine. POZ can be monofunctionalized at one end of the chain during the initiation step27 or at the termination step with the use of carboxylate ions.28 POZ has also been synthesized with pendant functional aldehyde and amine groups and terminated with inert groups.29,30 These chemistries allow for higher drug loading which is not possible with linear polymers like PEG. The synthetic methods described in the above-mentioned references appear to be straightforward, but in reality, they do have a number of process difficulties in relation to scaling up and reproducibility. The various side reactions that occur have led to impurities that are difficult to remove, thereby creating products

’ MATERIALS AND METHODS Materials. Bovine albumin, uricase, catalase, ribonuclease, cytidine 20 ,30 -phosphate, uric acid, and hydrogen peroxide, N-hydroxysuccinimide (NHS), dicyclohexylcarbodiimide (DCC), 2,4,6-trinitrobenzene sulfonic acid (TNBS), trifluoroacetic acid (TFA), methyl trifluoromethylsulfonate (MeOTf), disuccinimidyl carbonate (DSC), p-nitrophenyl chlroformate (PNPC), potassium t-butoxide, and sodium hydride were purchased from Sigma-Aldrich, St. Louis, MO. The organic solvents acetonitrile, dichloromethane, dimethylacetamide, dichloromethane, dimethylformamide, chlorobenzene, chloroform, methanol, and diethyl ether were obtained from EMD Chemicals. The monomers of 2-methyl-2-oxazoline, 2-ethyl-2-oxazoline and 2-propyl-2-oxazoline were obtained from Sigma Chemicals and Acros Chemicals. Methyl 3-mercaptopropionate was obtained from Sigma Chemicals and Chevron-Philips, Belgium. Amberlite desalting resins IRA67 and IR120H were purchased from Dow Chemicals, Midland, MI, and ion-exchange media such as DEAE sepharose FF resin and SP Sepharose were purchased from GE Healthcare (Amersham Biosciences). Insulin and insulin glargine were obtained from Biocon Ltd., Bangalore, India. The bicinchoninic acid (BCA) assay kit is from Pierce. Poly(ethylene glycol) (PEGs) reagents of 5 and 10 kDa were purchased from LaysanBio, Huntsville, AL. POZ Synthesis and Purification. The monomers and organic solvent were first dried by refluxing over calcium hydride and then distilled and collected in a dry flask containing molecular sieves. In a clean and dry flask and under an argon atmosphere, the organic solvent was first siphoned in. Different organic solvents were tested during the polymerization process. They were acetonitrile, chloroform, chlorobenzene, and dimethylacetamide. A measured amount of methyl or ethyl or propyl oxazoline M0 was 977

dx.doi.org/10.1021/bc200049d |Bioconjugate Chem. 2011, 22, 976–986

Bioconjugate Chemistry

ARTICLE

Figure 2. Synthetic schemes used to polymerize and functionalize POZ-OH and POZ-COOH.

80 °C for 2830 h for the 5 kDa polymer and 4042 h for the 10 kDa polymer. The termination of the polymerization was carried out as follows: (a) When a hydroxy group terminus was required, 40 mL of a 5% sodium carbonate solution was added to the polymer solution (5 kDa or 10 kDa) and the mixture was stirred for 30 min. The aqueous layer was separated and the organic layer was extracted once more with an additional 40 mL of a 5% sodium carbonate solution. The aqueous layer was separated, combined and stirred overnight at room temperature. The cloudy mixture formed (∼80 mL) was acidified with 40 mL of 0.5 M HCl acid in order to give a clear solution of pH < 6. The polymer was then extracted into methylene chloride (3  200 mL), and the combined organic layers were dried with anhydrous magnesium sulfate for one hour, concentrated to 25 mL, and precipitated by dropwise addition to 250 mL of diethyl ether. The resulting white solid was filtered and dried overnight in a vacuum oven. (b) When a carboxylic acid group terminus was required, the polymer solution (5 kDa or 10 kDa) was transferred into a 500 mL round-bottom flask containing a mixture of chlorobenzene (100 mL), methyl 3-mercaptopropionate (1.3 g, 8 mmol, 2 equiv), and potassium t-butoxide (0.67 g, 6 mmol, 1.5 equiv), and stirred overnight at room temperature. The chlorobenzene in the mixture was evaporated out, and the residue of the thioester was hydrolyzed by dissolving in 100 mL of 0.1 M sodium hydroxide solution and stirring at room temperature for 4 h. The crude acid (POZ-COOH) was desalted in an Amberlite ion-exchange resin IRA67/IR120H bed and then purified by ion-exchange chromatography (IEC)

next added, followed by an accurate measurement and transfer of the MeOTf initiator I0. The molecular weight of the polymer being synthesized depends on the molar ratio of M0/I0. The contents in the flask were then heated at temperatures between 60 and 90 °C for extended periods of time, depending on the desired molecular weight of the polymer prepared and the boiling point of the solvent used. The flask was then cooled and the solution containing the living polymer was slowly transferred to a mixture of the terminating agent in dry organic solvent and mixed overnight with constant stirring. The terminating agent used depended on the end terminal group desired. When a hydroxy group was required, the terminating mixture was either a methanolic potassium hydroxide solution or a 5% sodium carbonate solution in water. When a carboxylic acid group was desired, the terminating mixture was methyl 3-mercaptopropionate and potassium t-butoxide. The solvent was then stripped from the polymer by rotary evaporation. The thiopropionate-terminated residue was dissolved in an aqueous solution containing 0.1 N sodium hydroxide and mixed for a couple of hours to hydrolyze the thiopropionate ester. Figure 2 outlines a general polymerization scheme. The following example outlines the specific procedures used for the synthesis of a 5 kDa PEOZ polymer. Chlorobenzene (100 mL) and 2-ethyl-2-oxazoline (39.7 g, 400 mmol, 50 equiv) were mixed at room temperature under argon and in a 250 mL round-bottom flask. The initiator, methyl triflate (MeOTf, 880 μL, 8 mmol, 1 equiv), was carefully added into the flask with constant stirring. The mixture was then heated at 80 °C for 1.75 h and then rapidly cooled to room temperature by immersing the flask in an ice bath. When a 10 kDa PEOZ polymer was made, the amount of MeOTf used was less (440 μL, 4 mmol, 0.5 equiv) and the polymerization time was 3.5 h. When acetonitrile was used as the solvent, the polymerization was carried out at 978

dx.doi.org/10.1021/bc200049d |Bioconjugate Chem. 2011, 22, 976–986

Bioconjugate Chemistry

ARTICLE

(ε = 1.7  104 M1 cm1). The degree of activation, evaluated on the basis of p-nitrophenol release, was 87%. (b) Activation of POZ-OH with disuccinimidylcarbonate: PEOZ-OH (5.3 kDa, 0.05 mmol, 1 equiv) was dissolved in acetonitrile and dehydrated by azeotropic distillation to give a small amount of concentrated solution. After cooling to room temperature, disuccinimidyl carbonate (DSC, 0.2 mmol, 4 equiv) and pyridine were added and the mixture was stirred overnight. The solvent was removed using a rotary evaporator, and anhydrous dichloromethane was added to dissolve the residue. Any insoluble material was removed by filtration, and the filtrate was washed with a pH 4.5 sodium chloride saturated acetate buffer. The dichloromethane phase was separated, dried over anhydrous sodium sulfate, and then concentrated to a small volume before it was precipitated by slow addition to diethyl ether. The PEOZ-SC precipitate was collected and dried under vacuum to give a yield of 0.40 g or 80%. To determine the degree of SC substitution, an accurate amount of reagent was reacted with glycineglycine (Gly-Gly) to evaluate the degree of amino group modification.34 Briefly, a Gly-Gly solution of 0.285 mg/mL was prepared in borate buffer 0.2 M, pH 8. An equimolar amount of PEOZ-SC (10.8 mg) was added and the reaction mixture was stirred for 30 min. The assay was conducted by preparing two solutions: solution A: 950 μL of borate buffer (0.1 M, pH 9.3), 25 μL of the reaction mixture and 25 μL of 2,4,6-trinitrobenzene sulfonic acid solution (TNBS, 1% w/v in DMF); and solution B: 950 μL of borate buffer (0.1 M, pH 9.3), 25 μL of Gly-Gly, and 25 μL of TNBS solution. After 30 min, the absorbance of both solutions was measured by UV spectrophotometry at 420 nm, i.e., the amount of trinitrophenyl glycineglycine complex formed from the coupling of unreacted Gly-glyamino group with TNBS was measured. The polymer activation was calculated by applying the following formula:

using a column packed with DEAE sepharose FF resin. The aqueous layer was extracted, precipitated, and dried as described above. POZ Characterization. The purified POZ-OH and POZCOOH of different molecular weights were analyzed by nuclear magnetic resonance (NMR), matrix-assisted laser desorption/ ionization time-of-flight (MALDI-TOF), gel filtration chromatography (GFC), gel permeation chromatography (GPC), reverse-phase chromatography (RP), and viscometry. Proton NMR experiments were performed on a 500 MHz Varian NMR spectrometer. Samples (1020 mg) were first dissolved in 0.75 mL of CDCl3 or D2O and then analyzed. Mass spectrometry measurements were performed on a Bruker Microflex MALDI time-of-flight instrument (Bruker-Daltonics, Billerica, MA, USA). A stock solution was prepared mixing equal volumes of 0.1% TFA in water and 0.1% TFA in acetonitrile. Separate solutions of the polymer, the matrix (R-cyano-4-hydroxycinnamic acid) and the salt (sodium iodide) solution were prepared in the stock solution. Equal and micromolar amounts of each solution were spotted on the target for mass and polydispersity measurements. High-performance liquid chromatography (HPLC) was used for the both size exclusion methods. The GFC method was used to determine the relative hydrodynamic radius of each polymer type and size. A Bio-Sep-SEC-S-400 column (Phenomenex) was used and the mobile phase was 1 mM HEPES (pH 7.0) buffer. A GPC method was used to determine the size and polydispersity of each polymer. Two Phenogel columns (5 μ, 500A°, 300  7.8 mm, Phenomenex) were used in series and maintained at 60 °C, and the mobile phase was N,N-dimethylformamide (DMF). In the case of the lipophilicity method, a C-8 column was used and an isocratic mobile phase was used that comprised of a mixture of 70:30 methanol/water. In all these methods, an ultraviolet and refractive index detector were used in series. The viscosity of PEOZ-OH of molecular weight 1040 kDa was measured on a Brookfield LVDV-II cone and plate viscometer fitted with a temperature-controlled jacketed plate. The polymer sample (0.5 mL of a 10%, 20%, 30%, and 40% w/w solution in water) was placed on the center of the plate, which was attached to the main drive of the instrument. The cone (CPE-40) was rotated at different rates (rpm) and the viscosity (mPas) was recorded each time at 25 °C. The viscosities of similar concentrations of PEG-OH and branched PEOZ2-OH were also measured for comparison. Functionalizing POZ. POZ-OH and POZ-COOH were functionalized as follows. Figure 2 illustrates the different synthetic schemes for these reagents. (a) Activation of POZ-OH with p-nitrophenylchloroformate: PEOZ-OH (5.3 kDa, 0.05 mmol, 1 equiv) was dissolved in toluene and dehydrated by azeotropic distillation to give a small amount of concentrated solution. It was cooled to room temperature before dichloromethane, p-nitrophenylchloroformate (pNPC, 0.95 mmol, 6 equiv), and pyridine were added. The mixture was stirred overnight at room temperature and then precipitated by dropwise addition to prechilled diethyl ether. The PEOZ-pNPC precipitate was collected and dried under vacuum to give a yield of 0.45 g or 95%. To determine the degree of pNPC substitution, an accurate amount of the reagent was dissolved in a known volume of 0.2 M NaOH solution. After 2 h at room temperature, the amount of p-nitrophenol released was measured by spectrophotometry at 410 nm

polymer activation ¼ ðabsorbancesolution A =absorbancesolution B Þ100 The percentage of activation of PEOZ-SC was 85%. (c) Activation of POZ-COOH with H-hydroxysuccinimide: The purified PEOZ-COOH (5.3 kDa, 600 mg, 0.11 mmol) was dissolved in anhydrous dichloromethane, and N-hydroxysuccinimide (NHS, 104 mg, 0.9 mmol) and dicyclohexylcarbodiimide (DCC, 185 mg, 0.9 mmol) were added. The reaction mixture was allowed to react for 24 h, and the product (PEOZ-NHS) was precipitated in diethyl ether, collected, and dried. The degree of NHS activation was evaluated by the method described above and the percentage of activation was 90%. (d) Hydrolysis and reactivity of PEOZ-NHS, PEOZ-p-NPC, and PEOZ-SC: In the hydrolysis experiments, the activated polymer was accurately weighed and dissolved in dioxane. 50 μL of this stock solution was withdrawn and added to 950 μL of borate buffer 0.1 M, pH 8.0, and at room temperature (5% of dioxane in the mixture). The rates of hydrolysis of the POZ reagents were evaluated by the release of N-hydroxysuccinimide and its absorbance measurement at 260 nm using a UV spectrophotometer 979

dx.doi.org/10.1021/bc200049d |Bioconjugate Chem. 2011, 22, 976–986

Bioconjugate Chemistry

ARTICLE

Catalase Assay. The catalase activity was determined according to the method of Beers and Sizer.37 Briefly, 100 μL aliquots of free and conjugated enzyme (0.051 mg/mL) were added to 0.05% (v/v) of hydrogen peroxide in 0.01 M phosphate buffer pH 7.0. The solution was immediately transferred into a 1 cm cuvette and the hydrolysis of the substrate was measured by UV for 2 min at 240 nm. The residual activity of the conjugates was calculated as reported above. Conjugation of POZ-NHS to Insulin. A suspension of insulin (250 mg, MW 5807 Da) in 5.6 mL of boric acid (50 mM) was slowly acidified by the addition of 1 N HCl until a clear solution was obtained. The solution pH was then adjusted to 10.5 by the addition of 1 N NaOH solution. To this solution was slowly added 6.9 mL of a solution of PEOZ-NHS 10K in HCl solution (pH 2.5, 70 mg/mL, 4.1  105 mol). The solution pH was adjusted to 10.5 by adding 0.5 N NaOH solution. The clear solution was allowed to mix gently at room temperature for 57 min, before it was acidified to 2.4 by the addition of 1 N HCl acid. This reaction was monitored by reverse phase HPLC using a C-18 column. The chromatogram showed that there was 48.5% monoconjugated PEOZ-Insulin and 10.6% diconjugated PEOZInsulin and the rest native insulin. The monoconjugated PEOZInsulin was separated primarily as the lysine (B29) conjugated species using ion-exchange chromatography (IEC, Sepharose SP column) and the eluent was desalted by buffer exchange into deionized water.38 The product was analyzed for purity by SDS-PAGE and reverse-phase chromatography and for mass by MALDI-TOF. Conjugation of POZ-NHS to BSA. BSA (90 mg, 1.344 μmol, with 60 available amino groups) was dissolved in 0.1 M sodium borate buffer pH 8.5, and a 5 molar excess of activated polymer, PEOZ-NHS 5 kDa, PEG-NHS 5 kDa, PEOZ-NHS 10 kDa, or PEG-NHS 10 kDa, was added. Each was mixed for several seconds until a homogeneous solution was obtained, then allowed to sit at room temperature for 5 h. The reaction was terminated by the addition of a 5 molar excess of glycine. An ultrafiltration cell with a 30 kDa molecular weight cutoff membrane (Amicon XM-30) was used to remove unreacted polymer. Cationic exchange chromatography was next used to separate the conjugate from the protein and polymer. A SP-Sephadex C-25 column (3  18 cm) was equilibrated with 20 mM phosphate buffer pH 3.9 and eluted with a 0 to 0.5 M sodium chloride gradient at a flow rate of 1 mL/min. The fractions containing the conjugated protein were collected and concentrated and desalted by ultrafiltration. The samples were then lyophilized to give a dry powder. The protein content in each sample was determined by the BCA assay, and the degree of protein modification was evaluated by the trinitrobenzensulfonic acid reaction method. Erythrocyte Compatibility Study. One milliliter of freshly collected heparinized mouse blood was centrifuged at 2000 rpm for 10 min, and the precipitated erythrocytes were washed three times with PBS buffer. The red blood cells (RBCs) were resuspended in the buffer and diluted to have an optical density between 0.6 and 0.7 at 650 nm. This stock solution dispersion was freshly prepared and used within 48 h after preparation. Erythrocyte suspension (300 μL) was added to 300 μL of phosphated buffered saline solutions (PBS buffer) containing either PEOZ or PEG (0.210 mg/mL) at different molecular weights (5, 10, and 20 kDa), and the samples were incubated at 37 °C for 60 min under constant shaking. After centrifugation at 2000 rpm for 10 min, the released hemoglobin was measured by optical density (OD) at 414 nm. Complete hemolysis, as a

(Perkin-Elmer instruments, Northwolk, CT, USA). In the reactivity experiment, the stock solution was added to glycyl-glycine dissolved in borate buffer 0.1 M, pH 8.0, and at room temperature (ratio of polymer reagent/ glycyl-glycine was 1:1). The rate at which the primary amine of glycyl-glycine reacted with the POZ reagents to release of N-hydroxysuccinimide, was measured at 260 nm. The spectrophotometer was programmed to take a reading every 5 s, in order to plot the rate of the reaction. From these plots, the half-lives of hydrolysis and reactivity were calculated and reported in minutes. POZ and PEG Conjugation to Enzymes. Three different enzymes, RNase, uricase, and catalase, were reacted with PEOZ-NHS (5.3 kDa) and PEG-NHS (5.5 kDa) reagents. The RNase, uricase, and catalase concentrations in the test solutions were measured by UV spectrophotometry using molar extinction coefficients of 9.45  103, 13  103, and 1.67  105 M1 cm1 at 280 nm, respectively. RNase has a molecular weight of 13.7 kDa and 11 amino groups available for conjugation; uricase has a molecular weight of 130 kDa and 100 amino groups for conjugation; and catalase has a molecular weight of 240 kDa and 112 available amino groups for conjugation. The ratios of polymer to the number of amino groups on each enzyme were 2:1 and 1:1. The conjugation reactions were carried out for 30 min in a 0.2 M borate buffer (pH = 8.5). To remove the unreacted polymer, the free N-hydroxysuccinimide (NHS), and any remaining native protein, the crude polymerenzyme mixtures were purified using a Sephadex G-75 column, eluting with phosphate buffer 0.1 M, pH 7.4, at the flow rate of 1.5 mL/min. The effluent from the column was monitored at 226 nm. The first eluting fractions containing only the conjugated proteins were collected, concentrated, and desalted by dialysis. The samples were then lyophilized to give a dry powder. The purity of the collected fractions was verified by the gel filtration column. The protein content in each sample was determined by bicinchoninic acid reaction (BCA assay) using a Pierce protein assay kit and the degree of protein modification was evaluated spectrophotometrically by the trinitrobenzensulfonic acid reaction method described by Habeeb.34 The results are expressed as a percentage of modified amino groups relative to the total amount of amine groups present on the enzyme. Enzymatic Assays. RNase Assay. The enzymatic activity of the native and conjugated RNase was evaluated as described by Crook et al.35 Briefly, 100 μL aliquots of free and conjugated enzyme (0.20.4 mg/mL) were added to 1 mL of 1.5 mg/mL of cytidine 20 ,30 -cyclic monophosphate dissolved in 0.1 M Tris/ acetate buffer pH 7.0. The solution was immediately transferred in a 1 cm cuvette and the hydrolysis of the substrate was measured by UV for 20 min at 287 nm. The data were plotted with a first-order rate equation, and the rate of the reaction was calculated. This kinetic rate is directly proportional to the activity of the enzyme. The activity of the POZylated or PEGylated enzyme is reported as a percentage of the native enzyme. Uricase Assay. The method described by Mahler36 was followed for uricase. Briefly, 100 μL of free and conjugated enzyme (0.51 mg/mL) was added to 1 mL of 0.002 mg/mL of uric acid dissolved in 0.05 M borate buffer pH 9.0. The solution was immediately transferred into a 1 cm cuvette, and the hydrolysis of the substrate was measured by UV for 3 min at 292 nm. The residual activity of the conjugates was calculated as reported above. 980

dx.doi.org/10.1021/bc200049d |Bioconjugate Chem. 2011, 22, 976–986

Bioconjugate Chemistry

ARTICLE

positive control, was achieved by using a 1% v/v solution of Triton X-100 in water, a surfactant known to lyse RBCs. The RBC-PBS solution was used as the negative control. The percentage RBC lysis was calculated according the following formula: %lysis ¼ ½ðApolymers  Ablank Þ=ðA100%lysis  Ablank Þ100 where Apolymer is the absorbance value of the hemoglobin released from the RBCs treated with the polymer solution; Ablank is the absorbance value of the hemoglobin released from the RBCs treated with PBS buffer; and A100%lysis is the absorbance value of the hemoglobin released from RBCs treated with 1% Triton X-100. Efficacy of POZ-Insulin in Rat Model. The effect of POZ conjugated insulin on blood glucose lowering in naïve rats was the efficacy model used to study the activity of the conjugate. An in-house institutional animal care and use committee (IACUC) reviewed and approved the study protocol. Four groups of female SpragueDawley rats (89 weeks, 200 g) with four animals per group were subcutaneously dosed with normal saline (control), insulin, insulin glargine, or PEOZ 10K insulin. The insulin and insulin glargine samples were dissolved in acidified normal saline. The dose of each insulin formulation was 10 IU/kg. At time intervals of 0, 0.5, 1, 2, 4, 6, 8, and 24 h, blood drops of 510 μL volume were removed from the tail vein through a needle prick. Each drop was analyzed with the aid of a blood glucose meter (OneTouch) and measurements were recorded in mg/dL of glucose. Immunogenicity of POZ-BSA in Rabbit Model. The 70 day standard rabbit protocol was used in this study. An IACUC at the contract research laboratory reviewed and approved the study protocol. New Zealand white rabbits were divided into 5 groups of 2 animals per group. The test formulation for each group contained BSA, PEOZ 5 kDa BSA, PEG 5 kDa BSA, PEOZ 10 kDa BSA, and PEG 10 kDa BSA dissolved in normal saline. On day 0, 10 mL of blood was bled from each animal and spun to collect a baseline serum sample. The animals then received their primary immunization dose of 500 μg of antigenic protein injected subcutaneously. At the same time, Freund’s adjuvant was also injected subcutaneously through a different site on the skin. Booster doses of 250 μg antigen and Freund’s adjuvant were administered on days 14, 28, and 42. At time points of 35, 58, and 72 days, a 5 mL sample of blood was collected from each animal for serum sample analysis. Each serum sample was assayed by ELISA for antibodies to BSA. 96 well flat-bottom high binding polystyrene plates (Immulon 2HP, ThermoFisher) were coated with BSA dissolved in borate buffer and allowed to incubate overnight in a refrigerator. The next day, the excess buffer was discarded and each well was washed. A measured amount of rabbit serum was added to the first well, mixed, and then appropriately diluted to the adjacent well by a factor of 2 each time. The plate was then incubated overnight in a refrigerator. The secondary antibody, HRP-SA 2° (goat antirabbit) dissolved in blocking buffer, was added to each well and the plate allowed to stand at room temperature for 3 h. The developer solution consisted of 2,20 -azino-bis(3-ethylbenzthiazoline-6-sulfonic acid) in citrate buffer with 30% hydrogen peroxide. A measured volume of the developer solution was added to each well, and after 20 min, the color produced was measured at 405 nm in a plate reader.

Figure 3. 1H NMR spectrum of PEOZ (CDCl3 as solvent).

Toxicity Studies. Two acute dose studies were performed in female SpragueDawley rats (89 weeks, ∼200 g) to determine the maximum tolerated dose (MTD) of a PEOZOH 10 kDa polymer. The animals (3 per dose group) were intravenously administered the test article dissolved in 0.9% sodium chloride injection. A control group was administered 0.9% sodium chloride injection. In the single acute dose tolerance study, the doses were 0.5, 1, and 2 g/kg. Blood was collected 1 and 7 days postdose for serum and complete blood count (CBC) analysis. Animals sacrificed on day 7 for gross examination of organs. In the repeat dose study, the animals received 4 weekly doses of 0.5, 1, and 2 g/kg. Blood was collected 1 and 7 days after the last day of dosing for serum and CBC analysis. Animals were sacrificed for gross examination of organs. One multiple dose study was performed in female Sprague Dawley rats (89 weeks, ∼200 g) to determine the maximum tolerated dose (MTD) of a PEOZ-COOH 20 kDa polymer. Eight animals were dosed intravenously (tail vein) at a dose of 50 mg/kg on days 0, 2, 4, 7, 9, 11, and 14. The dose volume was between 0.3 and 0.5 mL. Blood was collected on days 15 and 21 for serum chemistry analysis and comprehensive CBC. Half the number of animals was sacrificed on day 15 and the rest on day 21. The kidneys (left and right), liver, and spleen were harvested from each animal and stored in 10% neutral buffered formalin. Each organ was imbedded in wax, sectioned, fixed on slides, and examined under a light microscope.

’ RESULTS AND DISCUSSION Polymer Properties. Different solvents were tested as suitable polymerization media in the preparation of PEOZ, PMOZ, and PPOZ. Acetonitrile and chloroform were the desired solvents for PMOZ, PEOZ, and PPOZ, while chlorobenzene and dimethylacetamide worked well with PEOZ and PPOZ only. The polymerization times ranged from 1 to 40 h and this was dependent on the molecular weight being synthesized, the solvent selection, and the temperature of the reaction. Each batch of POZ-OH and POZ-COOH was characterized by 1H NMR spectroscopy, MALDI-TOF mass spectrometry, and size exclusion chromatography (SEC). The 1H NMR data of a typical PEOZ sample showed the pendent peaks at 1.12 ppm CH3CH2CO); 2.41 ppm (CH3CH2CO), and the backbone peak of 3.47 ppm (-NCH2CH2N-) as shown in Figure 3. When methyltriflate was used as the initiator, the initiating methyl peak appeared as two singlets at 2.9 ppm (small) and 3.05 ppm (large) (CH3-NCH2CH2). When the polymer was terminated with a 981

dx.doi.org/10.1021/bc200049d |Bioconjugate Chem. 2011, 22, 976–986

Bioconjugate Chemistry

ARTICLE

Figure 4. Example of a MALDI-TOF spectrum of PEOZ detected in reflection mode with a mass cutoff at 1000 m/z. Meox = 99.1 g/mol.

Figure 5. Correlation of hydrodynamic volume and molecular weight of PEG, PEOZ, and PMOZ determined by aqueous size-exclusion chromatography. The hydrodynamic volume is directly related to the distribution coefficient (Kav).

Figure 6. Correlation of the hydrophilicitylipophilicty balance of PEG, PMOZ, PEOZ, and PPOZ as determined by reverse phase chromatography.

hydroxyl group, the terminal methylene group appeared at 3.8 ppm (-CH2-OH). MALDI-TOF mass spectrometry was also performed to evaluate the polymer distribution. The polymers showed a monomodal polymer distribution with main mass signals spaced by Δm/z = 99 in agreement with the monomer unit mass of ethyloxazoline = 99.13 g/mol. Figure 4 is an example of a 5 kDa PEOZ-OH polymer. GPC chromatographic analyses were used to confirm that purified POZ reagents had a defined molecular weight (MW) with polydispersity indices 1.021.10. Their retention times were comparable to PEG reagents of similar molecular weights. However, GFC analyses showed that PEG has a hydrodynamic volume slightly larger than that of PMOZ and PEOZ. Figure 5 compares the molecular weight (log scale) against the distribution coefficient (Kav) for each polymer type and size. The value of Kav is calculated from the solute elution volume, the void volume, and the volume of the SEC media in the columns used.39 The data also suggest that POZ is less hydrated than PEG, a property that can be verified by microcalorimetry. This observation suggests that POZ is less of a desiccant than PEG.

RP-HPLC chromatography of PMOZ, PEOZ and PPOZ showed differences in retention times relative to polymer type and size. Figure 6 shows that, as the molecular weights of PMOZ and PEG increase, their retention times decrease thereby suggesting increased hydrophilic behavior. On the other hand, the retention time of PEOZ and PPOZ increase with increasing molecular weights. This suggests that PEOZ is relatively more lipophilic than PMOZ and PEG. It is important to note that PMOZ and PEOZ are highly water-soluble, while PPOZ is sparingly water-soluble. The viscosity of aqueous PEG and PEOZ solutions is illustrated in Figure 7. The plot compares the viscosity of a 40% w/v solution of PEG and PEOZ at molecular weights of 10, 20, 30, and 40 kDa. The results show that linear and branched PEOZ have a significantly lower viscosity than PEG especially at the higher molecular weights of 20, 30, and 40 kDa. This property is important, as it allows for ease of formulation, filterability, and syringeability of POZ conjugates at higher concentrations. 982

dx.doi.org/10.1021/bc200049d |Bioconjugate Chem. 2011, 22, 976–986

Bioconjugate Chemistry

ARTICLE

Figure 7. Viscosity of PEG and PEOZ of different molecular weights at 40% w/v solution and at 25 °C.

Table 1. Hydrolysis and Reactivity Rates of PEOZ and PEG Reagents half- life, t1/2, at pH 8.0 (min) polymer

hydrolysis

reactivity with glycyl-glycine

PEOZ-SC (5 kDa)

5.40

3.28

PEG-SC (5 kDa)

5.90

3.02

PEOZ-p-NPC (5 kDa)

31.02

19.10

PEG-p-NPC (5 kDa)

40.03

25.14

PEOZ-NHS (10 kDa)

15.47

7.54

PEOZ-NHS (5 kDa)

14.08

5.75

PEG-NHS (5 kDa)

20.08

6.02

Figure 8. Panel A: HPLC-GFC profiles of uricase, PEOZ 5 kDa uricase, and PEG 5 kDa uricase in reaction mixtures after 30 min of mixing. Panel B: HPLC-GFC profiles of native uricase, PEOZ 5 kDa uricase, and PEG 5 kDa uricase after purification with Sephadex G-75.

Functional POZ. In order to investigate the different strategies

employed in POZ conjugation to proteins or peptides, several functional POZ reagents had to be made and characterized. The procedures previously described for other polymer activation were found to be suitable for the synthesis of electrophilic reagents such as PEOZ-NHS, PEOZ-p-NPC, and PEOZ-SC. Kinetic investigation of these POZ reagents showed that they had comparable reactivity towards amino groups on proteins and peptides, but slightly weaker stability in alkaline water, when compared to similar PEG reagents. The half-lives (t1/2) for the hydrolysis of PEOZ-NHS, PEOZ-p-NPC, and PEOZ-SC in pH 8 borate buffer are 14.08, 31.02, and 5.4 min as shown in Table 1. The t1/2 hydrolysis values were about 75% the value of their PEG counterparts. The t1/2 reactivity values of PEOZ and PEG SC and NHS reagents were similar, but the PEOZ-p-NPC reagent was about 75% that of the corresponding PEG reagent. A comparison of t1/2 values of each reagent showed that the reactivity rates with glycyl-glycine were twice as fast as their corresponding hydrolysis values. Even though one cannot rule out the effect of alkaline pH conditions on the premature release of N-hydroxysuccinimide, one must note that the fast aminedirected reactivity rate would be an important property when describing the functionality of these POZ reagents. Enzyme Conjugation. To study the conjugation of PEG and POZ to enzymes, we selected three different candidates, namely, RNase, catalase, and uricase. They were chosen because their natural molecular conformation, i.e., a monomeric form for RNase, the dimeric form for catalase, and the tetrameric form for uricase. The molar ratios of polymer to available amino groups on each enzyme were 1:1 and 2:1. Each conjugation mixture was analyzed by gel filtration HPLC and then purified to remove free polymer or proteins. As an example, Figure 8 shows

the chromatographic elution of uricase conjugates before (panel A) and after purification (panel B). Uricase almost completely reacts with the functional polymers, and after purification, the excess of PEOZ and free N-hydroxysuccinimide (NHS) was removed. The degree of conjugation was determined by the Habeeb method,34 and each enzyme had its own biological activity tested as mentioned in the Methods and Materials section. The results showed that the degree of modification and the in vitro bioactivity for each enzyme conjugate were similar for PEOZ and PEG of 5 kDa and 10 kDa molecular weights (Table 2). As expected, when the degree of enzyme modification was lower (as in the case of the 1:1 conjugates), the enzyme activity was higher. Insulin Conjugation. The pure lyophilized PEOZ 10 kDa insulin product was analyzed by reverse phase chromatography and shown to have a purity of >97%. SDS-PAGE was used to verify a single monoconjugated species, and the mass was measured by MALDI-TOF as 16 496 Da. The IEC method was able to identify the desired B29 conjugated isomer from the A1 and B1 conjugated isomers. BSA Conjugation. BSA was chosen as the model antigenic protein because of its known immunogenicity in different animal models. The method of conjugation employed uses a large excess of PEOZ-NHS or PEG-NHS at molecular weights of 5 and 10 kDa.40 The titration of the free amino groups in the conjugated samples showed that both POZ and PEG provided similar degrees of modification. They were about 60% for the 5 kDa polymers and about 74% for the 10 kDa polymers 983

dx.doi.org/10.1021/bc200049d |Bioconjugate Chem. 2011, 22, 976–986

Bioconjugate Chemistry

ARTICLE

Table 2. Characteristics of PolymerEnzyme Conjugates polymer:NH2 molar % of modified amino compounds

%

ratio

groups

PEOZ 5 kDa RNase

1:1

50

30

PEOZ 5 kDa RNase

2:1

70

25

PEG 5 kDa RNase

1:1

48

35

PEG 5 kDa RNase

2:1

72

RNase

activity 100

uricase

27 100

PEOZ 5 kDa uricase PEOZ 5 kDa uricase

1:1 2:1

38 50

70 45

PEG 5 kDa uricase

1:1

35

72

PEG 5 kDa uricase

2:1

53

catalase

48 100

PEOZ 5 kDa catalase

1:1

35

90

PEOZ 5 kDa catalase

2:1

37

88

PEG 5 kDa catalase

1:1

38

85

PEG 5 kDa catalase

2:1

40

79

Figure 9. Effect of subcutaneous injection of insulin, insulin glargine, and PEOZ 10 kDa insulin on the blood glucose levels in male SpragueDawley rats (dose = 10 U/kg; n = 4; mean ( SEM).

Table 3. Characteristics of Polymer Conjugates for the Immunogenic Studies protein contenta

% of modified amino

Mwb

compounds

(wt)%

groups

(kDa)

BSA PEOZ 5 kDa BSA

0.18

60

66 151.8

PEG 5 kDa BSA

0.11

66

145.2

PEOZ 10 kDa BSA

0.034

74

158.4

PEG 10 kDa BSA

0.016

72

168.3

a

Protein content evaluated in 1 mg of conjugates. b Estimated by SEC using protein standards.

Figure 10. Relative immunogenicity of BSA, PEG BSA, and PEOZ BSA in rabbits treated as measured by anti-BSA antibody levels. Animals with treated with immune booster on days 1, 14, 28, and 42.

(Table 3). Each PEOZ conjugate was analyzed by GPC for apparent molecular weights and for traces of free polymer or BSA. After purification, the amounts of free BSA and unreacted PEOZ were below quantifiable limits. Erythrocyte Compatibility Study. The effect of POZ on the integrity of the erythrocyte membrane was investigated by the hemolysis study. The percent of cell lysis, as observed by absorbance measurements at 414 nm, showed that the Apolymers value for POZ was negligible even at the polymer concentration of 10 mg/mL. The Ablank value, which corresponded to the solution of RBCs incubated with PBS, was slightly higher than that of the A polymers solution. This observation is consistent with reports on other polymers like PEG, where it is suggested that nonionic polymers might have a protective effect on cell membranes. It is important to note that the plasma was washed out in this study. This would make the RBCs more sensitive to cell-to-cell contact, external stress, and lytic agents. Therefore, POZ acts like a protective agent in the same way that plasma proteins behave in providing hemocompatibility against external stress and foreign bodies.41 Efficacy of POZ-Insulin in Rat Model. The blood glucose lowering in naïve rats is the quickest and most efficient way to determine the activity of polymer conjugated insulin. It provides two parameters, the initial onset of action in lowering of blood glucose levels (an indirect tmax parameter) and the sustainability of blood glucose lowering (a steady-state pharmacodynamic

parameter). Figure 9 illustrates the glucose lowering effect of insulin, insulin-glargine, and PEOZ 10 kDa insulin after subcutaneous administration in naïve female SpragueDawley rats. The results showed that the test compounds were active and were able to lower the blood glucose levels from 120 to about 40 mg/dL. This effect was achieved within 1 h, 2 h, and 4 h for insulin, insulin-glargine, and PEOZ 10 kDa insulin, respectively. The effect was reversed after 2 and 4 h for insulin and insulin-glargine, respectively. However, the blood glucose lowering effect was sustained for about 8 h in the case of PEOZ 10 kDa insulin. This suggests that PEOZ was able to protect insulin from rapid clearance and proteolysis. Immunogenicity Studies of BSA Conjugates. The ability of POZ to camouflage the immunogenicity of BSA was determined in a 72 day rabbit protocol study. The animals were dosed on days 1, 14, 28, and 42 and the serum samples collected on days 0, 35, 58, and 72. Figure 10 shows that antibody titer values for BSA were extremely high and the maximum limit of the ELISA assay. PEOZ BSA and PEG BSA dosed animals had significantly lower titer values and the 10 kDa conjugate was better than the 5 kDa conjugates because of the higher degree of amine modification and the larger hydrodynamic volume of the conjugate (Table 3). POZ had a slightly higher attenuation of BSA immunogenicity than PEG at each molecular weight tested. 984

dx.doi.org/10.1021/bc200049d |Bioconjugate Chem. 2011, 22, 976–986

Bioconjugate Chemistry

ARTICLE

Table 4. Differences in Properties between PEG and POZ PEG

POZ

Difficult polymerization process and there are limited

POZ easily synthesized with standard glassware

suppliers of quality raw PEG

and with nonexplosive materials

Forms peroxides (need to add antioxidant)

Does not form peroxides

Stable only at temperatures < 20 °C

Stable at room temperature and in water

Diol content of 26%

No Diol

Highly viscous when in aqueous solutions

Low viscosity

Low drug loading

High drug loading

Difficult to actively target Can accumulate in some organs and form vacuoles because

Active targeting possible with pendent polymers Readily cleared from body and is not hygroscopic

of desiccant nature of PEG

Toxicity Studies. In the acute dose studies, PEOZ 10 kDa was administered intravenously to SpragueDawley rats, in single and multiple injections at doses of 500, 1000, and 2000 mg/kg. No adverse events were observed. The body weights and food intake levels of each animal were normal. Blood counts and serum chemistry were normal when compared to the control group that received 0.9% (w/v) sodium chloride. The key organ (liver, kidney, lung, spleen, heart, and gastrointestinal tract) sizes, weights, morphology, and texture were also normal. In the multiple dose study, rats received seven intravenous doses of 50 mg/kg of PEOZ 20 kDa, over a period of 14 days. The blood counts and serum chemistry at days 15 and 21 were normal when compared to the control group. The key organ sizes, weights, morphology, and texture were also normal, and the histopathology of the kidney and spleen showed no microscopic differences when compared against the organs from the control animals. PEOZ was well-tolerated at the molecular weights of 10 and 20 kDa.

POZ is a suitable alternative to PEG. Table 4 list the areas where POZ has many advantages over PEG. Another advantage not discussed in this paper is the ability to link several small drug molecules to a single polymer backbone, sometimes called “pendant drug attachment”. The terminal end of the polymer remains free and can be functionalized for the attachment of targeting moieties. Hence, POZ not only has the ability to increase drug loading, but also the ability to carry this payload to specific molecular receptors.

’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected].

’ REFERENCES (1) Veronese, F. M., Mero, A., and Pasut, G. (2009) Protein PEGylation, basic science and biological applications. PEGylated protein drugs: basic science and clinical applications, Milestones in drug therapy series (Veronese, F. M., Ed.) pp 1131, Birkhauser Verlag, Berlin. (2) Armstrong, J. K., Hempel, G., Koling, S., Chan, L. S., Fisher, T., Meiselman, H. J., and Garratty, G. (2007) Antibody against poly(ethylene glycol) adversely affects PEG-asparaginase therapy in acute lymphoblastic leukemia patients. Cancer 110, 103–11. (3) Sherman, M. R., Saifer, M. G., and Perez-Ruiz, F. (2008) PEGuricase in the management of treatment-resistant gout and hyperuricemia. Adv. Drug Delivery Rev. 6, 59–68. (4) Leger, R. M., Arndt, P., Garratty, G., Armstrong, J. K., Meiselman, H. J., and Fisher, T. C. (2001) Normal donor sera can contain antibodies to polyethylene glycol (PEG). Transfusion 41, 29S. (5) Armstrong, J. K., Leger, R., Wenby, R. B., Meiselman, H. J, Garratty, G., and Fisher, T. C. (2003) Occurrence of an antibody to poly(ethylene glycol) in normal donors. Blood 102, 556A. (6) Bendele, A., Seely, J., Richey, C., Sennello, G., and Shopp, G. (1998) Short communication: renal tubular vacuolation in animals treated with polyethylene-glycol-conjugated proteins. Toxicol. Sci. 42, 152–157. (7) Conover, C., Lejeune, L., Linberg, R., Shum, K., and Shorr, R. G. L. (1996) Transitional vacuole formation following a bolus infusion of PEG-hemoglobin in the rat. Artific. Cells, Blood Subst., Immob. Biotechnol. 24, 599–611. (8) Kinstler, O., and Gegg, C. (2002) Site directed dual pegylation of proteins for improved bioavailability and biocompatibility. U.S. Patent No. 6420339B1. (9) Tao, L., Liu, J., and Davis, T. P. (2009) Branched polymerprotein conjugates made from mid-chain-functional P(HPMA). Biomacromolecules 12, 2847–2851. (10) Jain, S., Hreczuk-Hirst, D. H., McCormack, B., Mital, M., Epenetos, A., Laing, P., and Gregoriadis, G. (2003) Polysialylated insulin:

’ CONCLUSIONS Three types of POZ polymers were prepared and characterized. Linear POZ with molecular weights of 140 kDa were synthesized, with high batch to batch reproducibility, low polydispersity values, and yields of ∼6080%. The polymers prepared had no chain transfer biproducts and no high molecular weight species. Characterization studies show that both PMOZ and PEOZ are highly water-soluble; they have hydrodynamic volumes that are comparable and slightly lower than PEG; PEOZ is slightly more lipophilic than PMOZ and PEG, and is also significantly less viscous than PEG. The latter observation allows for the delivery of high dose proteins and peptides where viscosity is a hindrance in product formulation and administration. POZ can be easily functionalized from the terminal hydroxyl or carboxyl group to allow for conjugation to amine, thiol, aldehyde, ketone, and acetylene groups. POZylation of enzymes, proteins, and peptides produced conjugates with retained bioactivity that were comparable to similar conjugates modified with PEG. The attachment of POZ 10 kDa to lysine B29 of insulin produced a monoconjugate product that was able to lower blood glucose levels and also sustain its activity for a period of 8 h. In future papers, we will discuss other types of POZ attachments using click chemistry and enzyme catalyzed techniques. Toxicity studies in rats showed that PEOZ is safe when administered intravenously and the maximum tolerated dose (MTD) was greater than 2 g/kg. In addition, PEOZ can shield the immunogenic sites on an antigenic protein such as BSA. 985

dx.doi.org/10.1021/bc200049d |Bioconjugate Chem. 2011, 22, 976–986

Bioconjugate Chemistry

ARTICLE

synthesis, characterization and biological activity in vivo. Biochim. Biophys. Acta 1622, 42–49. (11) Veronese, F. M., Sartore, L., Caliceti, P., Schiavon, O., Ranucci, E., and Ferruti, P. (1990) Low molecular weight end-functionalized poly(N-vinylpyrrolidinone) for the modification of polypeptide amino groups. J. Bioact. Compat. Polym. 5, 167–178. (12) Food additive petition for Polyethyloxazoline, The Dow Chemical Company, May 8, 1985. (13) Adams, N., and Schubert, U. S. (2007) Poly(2-oxazolines) in biological and biomedical application contexts. Adv. Drug Delivery Rev. 59, 1504–1520. (14) Hoogenboom, R. (2009) Poly(2-oxazoline)s: A polymer class with numerous potential applications. Angew. Chem., Int. Ed. 48, 7978–7994. (15) Mero, A., Pasut, G., Dalla, Via L., Fijten, M. W., Schubert, U. S., Hoogenboom, R., and Veronese, F. M. (2008) Synthesis and characterization of poly(2-ethyl 2-oxazoline)-conjugates with proteins and drugs: suitable alternatives to PEG-conjugates? J. Controlled Release 2, 87–95. (16) Zalipsky, S., Hansen, C. B., Oaks, J. M., and Allen, T. M. (1996) Evaluation of blood clearance rates and biodistribution of poly (2-oxazoline)-grafted liposomes. J. Pharm. Sci. 85, 133–137. (17) Woodle, M. C., Engbers, C. M., and Zalipsky, S. (1994) New amphipatic polymer-lipid conjugates forming long-circulating reticuloendothelial system-evading liposomes. Bioconjugate Chem. 5, 493–496. (18) Wang, C. H., and Hsiue, G. H. (2005) Polymeric micelles with a pH-responsive structure as intracellular drug carriers. J. Controlled Release 108, 140–149. (19) Konradi, R., Pidhatika, B., Hlebach, A. M., and Textor, M. (2008) Poly-2-methyl-2-oxazoline: a peptide-like polymer for protein-repellent surfaces. Langmuir. 24, 613–616. (20) Goddard, P., Hutchinson, L. E., Brown, J., and Brookman, L. J. (1989) Soluble polymeric carriers for drug delivery. Part 2. Preparation and in vivo behaviour of N-acylethylenimine copolymers. J. Controlled Release 10, 5–16. (21) Gaertner, F. C., Luxenhofer, R., Blechert, B., Jordan, R., and Essler, M. (2007) Synthesis, biodistribution and excretion of radiolabeled poly(2-oxazoline)s. J. Controlled Release 119, 291–300. (22) Kobayashi, S., Mureo, K., Sawada, S., and Saegusa, T. (1985) Synthesis of poly(2-methyl-2-oxazoline) macromers. Polym. Bull. 13, 447–451. (23) Kobayashi, S., Masuda, E., Shoda, S. I., and Shimano, Y. (1989) Synthesis of acryl- and methacryl-type macromonomers and telechelics by utilizing living polymerization of 2-oxazolines. Macromolecules 22, 2878–2884. (24) Liu, Q., Konas, M., and Riffle, J. S. (1993) Investigations of 2-ethyl-2-oxazoline polymerizations in chlorobenzene. Macromolecules 26, 5572–5576. (25) Hoogenboom, R., Leenen, M. A. M., Wiesbrock, F., and Schubert, U. S. (2005) Microwave accelerated polymerization of 2-phenyl-2-oxazoline: Microwave or temperature effects? Macromol. Rapid Commun. 26, 1773–1778. (26) Wiesbrock, F., Hoogenboom, R., Leenen, M. A. M., Meier, M. A. R., and Schubert, U. S. (2005) Investigation of the living cationic ring-opening polymerization of 2-methyl, 2-ethyl, 2-nonyl, and 2-phenyl2-oxazoline in a single-mode microwave reactor. Macromolecules 38, 5025–5034. (27) Gross, A., Maier, G., and Nuyken, O. (1996) Synthesis and copolymerization of macromonomers based on 2-nonyl- and 2-phenyl2-oxazoline. Macromol. Chem. Phys. 197, 2811–2826. (28) Miyamoto, M., Naka, K., Tokumizu, M., and Saegusa, T. (1989) End capping of growing species of poly(2-oxazoline) with carboxylic acid: a novel and convenient route to prepare vinyl- and carboxyterminated macromonomers. Macromolecules 22, 1604–1607. (29) Taubmann, C., Luxenhofer, R., Cesana, S., and Jordan, R. (2005) First aldehyde-functionalized poly(2-oxazoline)s for chemoselective ligation. Macromol. Biosci. 5, 603–612. (30) Cesana, S., Auernheimer, J., Jordan, R., Kessler, H., and Nyuken, O. (2006) First poly(2-oxazoline)s with pendant amino groups. Macromol. Chem. Phys. 207, 183–192.

(31) Nuyken, O., Maier, G., and Gross, A. (1996) Systematic investigations on the reactivity of oxazolinium salts. Macromol. Chem. Phys. 197, 83–85. (32) Park, J. S., and Kataoka, K. (2006) Precise control of lower critical solution temperature of thermosensitive poly(2-isopropyl-2oxazoline) via gradient copolymerization with 2-ethyl-2-oxazoline as a hydrophilic comonomer. Macromolecules 39, 6622–6630. (33) Litt, M., Levy, A., and Herz, J. (1975) Polymerisation of cyclic imino ethers. X. Kinetics, chain transfer, and repolymerisation. J. Macromol. Sci.-Chem. A9, 703–727. (34) Habeeb, A. F. (1966) Determination of free amino groups in protein by trinitrobenzensulphonic acid. Anal. Biochem. 14, 328–336. (35) Crook, E., Mathias, A., and Rabin, B. (1960) Spectrophotometric assay of bovine pancreatic ribonuclease by the use of cytidine 20 ,30 -phosphate. Biochem. J. 74, 234. (36) Mahler, J. L. (1970) A new bacterial uricase for uric acid determination. Anal. Biochem. 38, 65–84. (37) Beers, R. F., and Sizer, I. W. (1952) A spectrophotometric method for measuring the breakdown of hydrogen peroxide by catalase. J. Biol. Chem. 195, 133–140. (38) Hinds, K. D., and Kim, S. W. (2002) Effects of PEG conjugation on insulin properties. Adv. Drug Delivery Rev. 54, 505–530. (39) Fee, C. J., and Van Alstine, J. M. (2004) Prediction of viscosity radius and the size exclusion chromatography behavior of PEGylated proteins. Bioconjugate Chem. 15, 1304–1313. (40) Abuchowski, A., Van Es, T., Palczuk, N. C., and Davis, F. (1977) Alteration of immunological properties of bovine serum albumin by covalent attachment of polyethylene glycol. J. Biol. Chem. 252, 3576– 3581. (41) Allison, B. C., Applegate, B. M., and Youngblood, J. P. (2007) Hemocompatibility of hydrophilic antimicrobial copolymers of alkylated 4-vinylpyridine. Biomacromolecules 8, 2995–2999.

986

dx.doi.org/10.1021/bc200049d |Bioconjugate Chem. 2011, 22, 976–986