Peptide-polymer vesicles prepared by atom transfer radical polymerization

Peptide–Polymer Vesicles Prepared by Atom Transfer Radical Polymerization ¨ WIK, JAN C. M. van HEST LEE AYRES, P. HANS, J. ADAMS, DENNIS W. P. M. LO Organic Chemistry Department, Institute for Molecules and Materials Institute, Radboud University Nijmegen, Toernooiveld 1, 6525 ED, Nijmegen, The Netherlands

Received 6 July 2005; accepted 25 August 2005 DOI: 10.1002/pola.21107 Published online in Wiley InterScience (www.interscience.wiley.com).

The peptide Ac-Ser-Ala-Gly-Ala-Gly-Glu-Gly-Ala-Gly-Ala-Gly-Ser-Gly-OH was prepared with solid-phase peptide chemistry. Before the removal of the peptide from the solid support, the alcohol side groups of the two serines were functionalized with an a-bromo ester moiety to create a bifunctional initiator. This peptide-based initiator was used in solution for the atom transfer radical polymerization of methyl methacrylate to yield a well-defined ABA triblock copolymer, in which the poly(methyl methacrylate) end blocks had a number-average molecular weight of 1.1 kg/mol (based on 1 H NMR spectroscopy) and a polydispersity of 1.17. The aggregation behavior of this amphiphilic triblock copolymer was then investigated. Upon the suspension of the polymer in a mixture of tetrahydrofuran and water, followed by the removal of tetrahydrofuran, spherical aggregates were formed. By the application of different electron microscopy techniques, it was determined that these aggregates were polymersomes, preC 2005 Wiley Periodicals, Inc. J Polym sumably coexisting with large compound micelles. V ABSTRACT:

Sci Part A: Polym Chem 43: 6355–6366, 2005

Keywords:

atom transfer radical polymerization; block copolymers; peptides; vesicles

INTRODUCTION Block copolymers have been studied extensively in the past decades because of their interesting assembly and phase-separation behavior in both the solid state and solution. Well-defined nanostructured materials have been prepared for a wide variety of applications, from electronics1,2 to drug delivery.3 Recently, there has been a growing interest in a specific class of hybrid block copolymers, in which well-defined peptide sequences are combined with synthetic polymers to create materials with novel secondary architectures.4 There are many examples of the use of peptide–polymer hybrids to create phase-separated Correspondence to: J. C. M. van Hest (E-mail: j.vanhest@ science.ru.nl) Journal of Polymer Science: Part A: Polymer Chemistry, Vol. 43, 6355–6366 (2005) C 2005 Wiley Periodicals, Inc. V

nanostructures on surfaces. Gallot and coworkers5–8 produced one of the first examples by preparing diblock copolymers of polystyrene or polybutadiene with poly(c-benzyl-L-glutamate) and poly(ebenzyloxycarbonyl-L-lysine). These block copolymers formed lamellar structures in solution, and in the dry state, the peptide chains folded and arranged in a hexagonal array on a surface. Recently, the development of controlled N-carboxyanhydride (NCA) polymerization methodologies has enabled researchers to investigate in more detail the effects of polydispersity on phase separation and polymer assembly.9–11 Sogah and coworkers12–15 prepared block copolymers of b-sheet folding peptides, consisting of alanine or alanyl–glycine repeats, and poly(ethylene glycol) (PEG). These structures gave rise to a phase-separated material with a secondary structure similar to that of spider silk, consisting of crystalline b-sheet domains surrounded by an 6355

6356

AYRES ET AL.

amorphous PEG matrix. Smeenk et al.16 recently prepared a similar material, using protein engineering techniques. They produced a b-sheetforming polypeptide, which was subsequently coupled to a synthetic polymer. This triblock copolymer assembled into well-defined fibers on a surface. Amphiphilic peptide–polymer hybrid materials have also been used to form aggregates in solution.17 One interesting series of block copolymers was prepared by Cornelissen and coworkers18–20 by combining polystyrene with polyisocyanides containing peptides in the side chain. These socalled superamphiphiles formed a variety of assemblies in solution, such as helical ribbons and vesicular aggregates, known as polymersomes. Klok et al.21 reported the synthesis of watersoluble, stimuli-responsive vesicles from peptidebased diblock copolymers. They prepared a block copolymer of polybutadiene and poly(glutamic acid) by using an amine-functionalized polybutadiene as a macroinitiator for the polymerization of NCA of glutamic acid. They showed that these block copolymers formed vesicles whose size could be altered by the variation of the pH. In a different approach, the same group22,23 used solid-phase peptide chemistry to produce a coiled coil peptide motif with the sequence G(EAKLAEI)3Y (G ¼ glycine, E ¼ glutamic acid, A ¼ alanine, K ¼ lysine, L ¼ leucine, I ¼ isoleucine, Y ¼ tyrosine), which was subsequently coupled to a PEG chain. They demonstrated that these block copolymers formed discrete supramolecular aggregates of two, three, or four block copolymers instead of the unspecific self-organization normally associated with amphiphilic block copolymers. Recent advances in controlled radical polymerization have given materials scientists an alternative approach with which to synthesize peptide–polymer hybrid materials. The attachment of initiator moieties to the end of a peptide has made it possible to perform controlled radical polymerizations from a peptide while still on the resin. The first examples of this approach were published by Wooley et al.24 for nitroxide-mediated polymerization (NMP) and Mei et al.25 for atom transfer radical polymerization (ATRP). Rettig et al.26 extended this work by showing that it was possible to use a peptide-based initiator in solution. In this case, a well-defined polymer with a polydispersity index of 1.19 was obtained. In a recent article, Wooley et al.27 prepared two initiators based on Tritrpticin, a 13-residue

antimicrobial peptide, one for NMP and one for ATRP. These functionalized peptides were used to initiate the polymerization of t-butyl acrylate, followed by styrene, to produce a triblock copolymer. After cleavage from the resin and hydrolysis of the t-butyl esters with trifluoro acetic acid (TFA), the ABC-type triblock copolymers formed micellar aggregates in solution. Interestingly, the antimicrobial activity of the peptide was enhanced with respect to the free peptide, and the detrimental side effects normally associated with antimicrobial peptides, such as a high hemolytic activity, were reduced. The aim of the work presented in this article was to prepare, in a controlled fashion, a peptidebased amphiphilic triblock copolymer that could form well-defined aggregates in solution. In our approach, ATRP was chosen as the polymerization technique, and polymerization was performed in solution. To obtain a triblock copolymer, a peptide-based bifunctional initiator had to be prepared. The peptide sequence that we have chosen is a hydrophilic sequence also known to form b hairpins.16 This should give rise to an amphiphilic polymer containing a peptide-based hydrophilic middle block flanked by two hydrophobic synthetic blocks. Besides the difference in polarity, the presence of a peptide sequence prone to forming a b-hairpin conformation should further facilitate the controlled self-assembly process. In this article, we describe the successful synthesis of the bifunctional peptide initiator, the controlled polymerization to form triblock copolymers, and the subsequent aggregation of these block copolymers into a mixture of polymersomes and large compound micelles (LCMs).

EXPERIMENTAL General Procedures 1

H and 13C NMR spectra were measured on a Bruker 400-MHz machine with a Varian probe. Infrared (IR) spectra were measured on an ATI Mattson Genesis series Fourier transform infrared (FTIR) spectrometer. Samples were placed on the FTIR crystal as a solid powder and compressed before measurement. Each sample was scanned 32 times. Matrix-assisted laser desorption/ionization timeof-flight (MALDI-TOF) mass spectra were measured on a Bruker Biflex III machine with dihy-

PEPTIDE–POLYMER VESICLES

droxybenzoic acid (DHB) as a matrix. The samples were prepared by the dissolution of 2 mg of an analyte in 1 mL of tetrahydrofuran (THF), after which this solution was mixed in a 1:1 ratio with a solution of 10 mg of DHB in 1 mL of H2O containing 0.1% TFA. This was then placed on a matrix-assisted laser desorption/ionization plate. Gel permeation chromatography (GPC) measurements were performed with a Shimadzu gel permeation chromatograph, with Shimadzu refractive-index and ultraviolet–visible detection, fitted with a Polymer Laboratories PLgel 5-lm mixedD column and a PL 5-lm guard column (molecular weight separation range of 500–500,000) with THF or dimethyl sulfoxide (DMSO) as a mobile phase at 35 or 70 8C, respectively. Polymer Laboratories poly(methyl methacrylate) (PMMA) calibration standards were used. For the transmission electron microscopy (TEM) studies, a JEOL JEM-1010 instrument was used. Samples were prepared by the drying of a drop of the dispersion on a carbon-coated copper grid. When aqueous solutions were used to prepare samples, the excess water was blotted away with filter paper after 2 min. For Pt shadowing, the grids were placed in an Edwards model 306 coater under an angle of 458 with respect to the Pt source. The samples were left to dry for a day in the fume hood before study. Scanning electron microscopy (SEM) was performed on a JEOL JSM T300 operating at 30 kV. The same samples used for TEM were fixed onto a metal stub with carbon glue. This was then placed inside the SEM instrument. Cryo scanning electron microscopy (cryo-SEM) was performed on a JEOL JSM T300 operating at 30 kV. The sample solution was quenched in nitrogen slush. Afterwards, the sample was freeze-fractured with standard procedures and transferred into the cryo-SEM instrument. The sample was sublimed for 5 min before being inserted into the sample chamber.

Reagents CuCl (Aldrich; 97%) was purified by being washed with glacial acetic acid three times and once with diethyl ether.28 p-Alkoxybenzyl alcohol Wang resin (Bachem; 1.14 mmol/g), 9-fluorenylmethoxy carbamate (Fmoc) and 9-fluorenylmethoxy carbamate tritylprotected serine [Fmoc Ser (Trt)-OH; Bachem;
99%], 9-fluorenylmethoxy carbamate glycine

6357

(Fmoc Gly-OH; Bachem;
99%), 9-fluorenylmethoxy carbamate tert-butyl-protected glutamic acid [Fmoc Glu(OtBu)-OH; Bachem;
99%], 9-fluorenylmethoxy carbamate alanine (Fmoc Ala-OH; Bachem;
99%), 2-bromoisobutyric acid (Aldrich; 98%), 2,20 -bipiridyl (Aldrich; 99%), N,N-dicyclohexylcarbodiimide (Fluka; 99%), 4-dimethylaminopyridine (DMAP; Acros; 99%), DMSO-d6 (Aldrich; 99.9%), N,N0 -diisopropylethylamine (DIPEA; Fluka; 99%), 1-hydroxybenzotriazole hydrate (HOBt; Fluka;
98%), N,N-diisopropylcarbodiimide (DIPCDI; Fluka;
98%), TFA (Aldrich; 98%), triisopropyl silane (TIS; Acros; 99%), pentamethyl diethylene triamine (PMDETA; Aldrich; 98%), potassium hydrogen sulfate (KHSO4; Riedel-de Hae¨n; 99%), sodium hydrogen carbonate (NaHCO3; Merck; 99.5%), and anhydrous sodium sulfate (Fluka; 99%) were all used as received. Dichloromethane (DCM) and ethyl acetate (EtOAc) were distilled from calcium hydride, THF was distilled from sodium/benzophenone, and methyl methacrylate (MMA) (Aldrich; 99.0%) was distilled before use. Dimethylformamide (DMF) and isopropyl alcohol were used as received (J.T. Baker). Synthesis of 2-Bromoisobutyric Acid Functionalized Ac-Ser-Gly-Ala-Gly-Ala-Glu-Gly-Ala-Gly-Ala-SerGly-OH (1) 1 was synthesized by standard solid-phase methods with a Wang resin. A suspension of the Wang resin (30 g) in 300 mL of DMF was cooled in an ice bath, after which Fmoc Gly-OH (13.5 g, 45 mmol), 9.20 g (60 mmol) of HOBt, and 4.30 g (34.2 mmol) of DIPCDI were added. This mixture was shaken for 6 h. The functionalized resin was filtered and washed repeatedly with DCM, DMF, and isopropyl alcohol. Unfunctionalized groups on the resin were capped by the addition of 10.2 mL of benzoylchloride and 8.4 mL of pyridine to a suspension of the resin in 300 mL of DCM at 0 8C. This mixture was shaken for 30 min, filtered, and washed repeatedly with DCM, DMF, and isopropyl alcohol. Then, 1.5 g of the Fmoc Gly functionalized Wang resin (loading ¼ 0.65 mmol/g) was swollen and filtered three times in 20 mL of DMF. Next, 20 mL of DMF, containing 20% (v/v) piperidine, was added to remove the Fmoc group. A positive Kaiser test indicated the completeness of this reaction. The next amino acid was coupled by the addition of a mixture of 1.66 g (2.01 mmol) of Fmoc Ser(Trt)-OH, in 30 mL of DMF, to 3.51 mL of a 1 M solution of

6358

AYRES ET AL.

HOBt in DMF and 3.22 mL of a 1 M solution of DIPCDI in DMF. The mixture was shaken for 16 h, after which it was washed with DMF and twice with DCM and isopropyl alcohol. A negative Kaiser test indicated the completeness of the reaction. The Fmoc group was then removed with the aforementioned procedure. This was repeated, with a reaction time of 4 instead of 16 h, with the following 11 amino acids: Fmoc Gly-OH (0.87 g, 2.01 mmol), Fmoc Ala-OH (0.91 g, 2.01 mmol), Fmoc Gly-OH (0.87 g, 2.01 mmol), Fmoc Ala-OH (0.91 g, 2.01 mmol), Fmoc Gly-OH (0.87 g, 2.01 mmol), Fmoc Glu(OtBu)-OH (1.24 g, 2.01 mmol), Fmoc Gly-OH (0.87 g, 2.01 mmol), Fmoc Ala-OH (0.91 g, 2.01 mmol), Fmoc Gly-OH (0.87 g, 2.01 mmol), Fmoc Ala-OH (0.91 g, 2.01 mmol), Fmoc Gly-OH (0.87 g, 2.01 mmol), and, finally, Fmoc Ser(Trt)-OH (1.66 g, 2.01 mmol). While still on the resin, the Fmoc-protecting group on the terminal serine was removed, and the free amine was capped by the addition of 90 mL of DMF containing 1 mL of Ac2O and 2 mL of DIPEA. The trityl-protecting groups were then removed from the serine side chains with 20 mL of a mixture containing 3% TFA and 5% TIS in DCM. This was shaken for 5 min, and then the mixture was removed and replaced with another 20 mL of the same mixture. After 15 min, it was removed, and the resin was rinsed with 20 mL of a solution containing 5% DIPEA in DCM. The free OH groups of the serines were then coupled with 2-bromoisobutyric acid (0.97 g, 3.9 mmol) with DMAP (0.71 g, 3.9 mmol) and DIPCDI (0.60 mL, 3.9 mmol) in 20 mL of DCM. The resin was washed repeatedly with DCM, DMF, and isopropyl alcohol, and this coupling procedure was repeated to obtain 100% conversion. The final bifunctional peptide-based initiator was cleaved from the resin with a 90% TFA/water solution and was precipitated in diethyl ether. Five hundred fifty milligrams of the peptide was obtained. MALDI-TOF before a-bromo ester formation: m/e 954 (M  2H2O þ Hþ), 990 (M þ Hþ), 1012 (M þ Naþ), 1028 (M þ Kþ). MALDI-TOF: m/e 1230 (100%), 1232 (50%), 1233 (48%), 1228 (48%), 1229 (25%), 1234 (19%), (M  Br þ Naþ), 1310 (100.0%), 1312 (65%), 1311 (50%), 1308 (45%), 1309 (25%), 1313 (25%), 1314 (10%), (M þ Naþ). 1 H NMR [SO(CD3)2, d]: 1.2 [CH(CH3), 12H, d], 1.7 (CH2CH2COOH, 2H, m), 1.8 [H3CC(¼ ¼O) NH and (C(CH3)2Br, 15H, m], 2.2 (CH2CH2 COOH, 2H, t), 3.7 (NHCH2C¼ ¼O, 12H, m), 4.1– 4.3 [NHCH(CH3)C¼ ¼O, NHCH(CH2CH2COOH)

¼O, 9H, m], 4.7 C¼ ¼O, and NHCH(CH2OR)C¼ [NHCH(CH2OR)C¼ ¼O, 2H, m], 7.8–8.2 [NHC(R)C ¼ ¼O, 13H, m], 8.4 (AcNHC¼ ¼O, 1H, t). 13C NMR [SO(CD3)2, d]: 17.96, 18.33, 18.57, 22.74, 26.81, 31.34, 40.79, 42.27, 48.13, 48.54, 48.58, 48.84, 52.54, 157.84, 168.03, 168.30, 168.52, 168.60, 169.39, 170.77, 171.49, 172.20, 172.21, 172.29, 172.68. IR: 3277 (NH str.), 2926 (CH str.), 1731 (C¼ ¼O str. a-bromo ester), 1640 (C¼ ¼O str. amide I), 1519 (NH vib. amide II), 1445 cm1 (CH vib.).

ATRP of MMA from Initiator 1 Initiator 1 (30.9 mg, 0.025 mmol) and CuCl (10.1 mg, 0.1 mmol) were placed in a Schlenk vessel. The vessel was then evacuated and purged with Ar three times to remove any air present in the system. MMA (106 lL, 1 mmol) and 11 lL of PMDETA (0.1 mmol) were added to the vessel, along with 1 mL of DMSO-d6. This solution was purged for 10 min with Ar at room temperature, and then the polymerization mixture was heated to 90 8C. The conversion of the polymerization was followed with 1H NMR spectroscopy by the comparison of the integrals of the protons of the methacrylate double bond at d ¼ 6.0 ppm with the peak due to the Ca protons of the alanines and glutamic acids at d ¼ 4.25 ppm. After 1 h 5 min, the polymerization reached 55% conversion and was stopped. The blue polymer solution was precipitated into diethyl ether and stirred for 10 min. The ether layer was then decanted off, and the remaining polymer was left to air-dry. The yield could not be determined because samples were taken throughout the polymerization; however, 55 mg of the polymer was obtained. For GPC analysis, 5 mg of the polymer was placed in a 15:4:1 dioxane/water/NaOH (4 M) mixture and stirred for 1 h. The solvent was then removed, and the resulting crude product was dissolved in 30 mL of DCM. This was then washed three times with 10 mL of demiwater. The DCM layer was then evaporated, and the sample was redissolved in THF. 1 H NMR [SO(CD3)2, d]: 0.6–1.3 [CH2C (CH3)R  and (CH(CH3), m], 1.5–2.0 [CH2  C(CH 3 )R, CH 2 CH 2 COOH, H 3 CC(¼ ¼O)  NH, and C(CH3)2Br, m], 2.2 (CH2CH2COOH, m), 3.0–3.8 [C(O)OCH3 and NHCH2C¼ ¼O, m], 4.1–4.3 [NHCH(CH3)C¼ ¼O, NHCH(CH2CH2CO OH) C¼ ¼O, and NHCH(CH2OR)C¼ ¼O, m], 4.7 [NHCH(CH2OR)C¼ ¼O, m], 7.8–8.2 [NHC(R)

PEPTIDE–POLYMER VESICLES

6359

Scheme 1. Functionalization of Ac-Ser-Gly-Ala-Gly-Ala-Glu-Gly-Ala-Gly-Ala-SerGly-OH on the resin: (a) 20 mL of a mixture of 3% TFA and 5% TIS in DCM; (b) DIPCDI, DIPEA, and 2-bromoisobutyric acid in DCM; and (c) 95% TFA in water.

C¼ ¼O, m]. IR: 3277 (NH str.), 2927 (C H str.), 1726 (C¼ ¼O methacrylate), 1648 (C¼ ¼O str. amide I), 1527 (NH vib. amide II), 1458 cm1 (CH vib.). 1H NMR (CDCl3) for the PMMA block after cleavage (d): 0.6–0.9 [CH2C(CH3)R], 1.2 [CH2C(CH3)R], 3.6 [C(O)OCH3]. GPC: number-average molecular weight (Mn) ¼ 629 g/ mol, polydispersity index ¼ 1.17. 1H NMR: Mn ¼ 1120 g/mol.

RESULTS AND DISCUSSION To synthesize an ABA triblock copolymer, of which the B block is a peptide, it is necessary to prepare a peptide-based bifunctional initiator. The approach that we have chosen is to include serines in the peptide sequence that, when deprotected, contain alcohols in the side chain. These alcohols can then be modified into a-bromo esters

6360

AYRES ET AL.

to give the desired peptide-based ATRP initiator. This peptide can then be used to initiate the polymerization of a monomer such as MMA to give an ABA-type block copolymer with a hydrophilic, peptide-based B block and hydrophobic PMMA A blocks. We have chosen to use ATRP as it is a straightforward polymerization technique for which functional initiators can be easily synthesized. The sequence Ac-Ser-Gly-Ala-Gly-Ala-Glu-GlyAla-Gly-Ala-Ser-Gly-OH was synthesized with conventional solid-phase peptide synthesis. This sequence is based on the crystalline b-sheet region of silk worm silk and represents a full b hairpin.29,30 After the synthesis of the peptide, the serines were deprotected and functionalized with 2-bromoisobutyric acid (Scheme 1). Finally, the peptide was cleaved from the resin with TFA to yield the desired bifunctional initiator. The synthesis of this peptide was not as straightforward as can usually be expected for solid-phase chemistry. During initial experiments, mass spectrometry revealed that some of the couplings did not go to completion, giving rise to missing amino acids. To ensure complete couplings, the reaction times were extended, and this problem was resolved. In addition, the coupling of the a-bromo ester to the free OH only went to 80% completion the first time; therefore, the product was washed and the coupling was repeated until 100% conversion was reached. The final product was characterized fully by NMR, mass spectrometry, and IR, and from 1.5 g of the resin, 550 mg of the peptide was obtained. ATRP of MMA from Bifunctional Initiator 1 The polymerization of MMA from initiator 1 was performed with ATRP. MMA was chosen as it is a commonly used monomer for ATRP and because of its hydrophobic character. The polymerization was carried out in DMSO-d6 as a solvent, which we have shown to be suitable for the preparation of peptide–polymer hybrid materials via ATRP,31–33 with CuCl/PMDETA as a catalyst, because all reagents are quite soluble. After 75 min at 90 8C, the polymerization reached 55% conversion. Although there is a slight jump at the beginning of the polymerization, the semilogarithmic plot of the conversion versus time shows first-order kinetics, indicating that the polymerization was controlled (Fig. 1). The deviation at the beginning of the polymerization is commonly seen with ATRP and can be explained by the combination of a bromide-initiating species with a CuCl-based catalyst. When this type of mixed

Figure 1. Semilogarithmic plot of the conversion, [M0]/[M], versus time for the polymerization of MMA with bifunctional peptide-based initiator 1.

halide system is used, the rate of initiation is faster than the rate of propagation because of the difference in the stability between a CBr bond and a CCl bond; therefore, an increase in the apparent rate of polymerization is observed at the start of the polymerization.34 GPC in DMSO was attempted to analyze the molecular weight of the polymer. However, the polymer was not UV-active above 280 nm (which DMSO below absorbs) and the refractive index of DMSO is very high; this hampered the detection of our polymer. Because this polymer was only soluble in DMSO and mixtures of THF and water, it was not possible to obtain molecular weight data for the hybrid polymer. To overcome this problem, the PMMA blocks were cleaved from the peptide with a mixture of NaOH (4 M), dioxane, and methanol (MeOH; 1:14:5). GPC analysis of the cleaved PMMA block showed an Mn value of 610 g/mol and a polydispersity of 1.17. There is a discrepancy between the Mn value of 1120 g/mol calculated on the basis of the conversion from NMR and the Mn value observed by GPC. This can be attributed to the fact that at low molecular weights, refractive-index detection becomes more molecular-weight-sensitive, making the molecular weight data obtained less accurate. Another explanation for the discrepancy is mass loss caused by the removal of some of the methyl groups from the side chain of MMA, which can occur under these strongly basic conditions, and a difference in the hydrodynamic volume between the PMMA calibrants used and the resulting poly(methacrylic acid). This loss of a fraction of the

PEPTIDE–POLYMER VESICLES

6361

Figure 2. TEM images of PMMA–1–PMMA prepared by the suspension of the triblock copolymer in THF, the addition of water, and the subsequent removal of THF. Both images were made with platinum shadowing.

PMMA methyl esters was confirmed by 1H NMR spectroscopy. However, from the first-order kinetics and the polydispersity of 1.17, we can conclude that the polymerization was reasonably well controlled. Block Copolymer Assembly The assembly behavior of the obtained peptidebased amphiphilic block copolymer was studied with TEM, SEM, and cryo-SEM. Several different

sample preparation methods were used. First, an aqueous solution of 1 mg/mL of the block copolymer was prepared. The hybrid polymer was difficult to solubilize, contrary to the water-soluble peptide sequence, and hence a suspension was obtained. To overcome this insolubility problem, the commonly used method of applying an organic cosolvent to solubilize amphiphilic block copolymer assemblies was used. The polymer was added to two different organic solvents, THF and MeOH. As the block copolymer was also not solu-

6362

AYRES ET AL.

Figure 3. SEM images of PMMA–1–PMMA prepared by the suspension of the triblock copolymer in THF, the addition of water, and the subsequent removal of THF.

PEPTIDE–POLYMER VESICLES

6363

Figure 4. Cryo-SEM image of PMMA–1–PMMA prepared by the suspension of the triblock copolymer in THF, the addition of water, and the subsequent removal of THF.

ble in either of these solvents, suspensions were obtained. However, upon the addition of water to the suspension of the block copolymer in MeOH, the polymer dissolved. Upon the removal of MeOH, it precipitated out again. Upon the addition of water to the polymer suspended in THF and the subsequent removal of THF in vacuo, an opaque solution was obtained, suggesting the formation of colloidal aggregates. TEM measurements were performed of the polymer solutions, as prepared by the three aforementioned methods, by the placement of the solutions onto a carbon grid and the removal of the solvent. Besides the use of unstained samples, platinum shadowing also was applied to better visualize the structures formed. The TEM results are depicted in Figure 2 for the THF/water combination. No defined aggregates could be observed when pure water was used or when the block copolymer was first suspended in MeOH. However, when the samples were first suspended in THF, aggregates were clearly visible (Fig. 2). The TEM images show that the observed aggregates were spherical in nature, with diameters ranging from 100 to 300 nm. It is possible to see through some of the particles in the platinum-shadowed images, and this suggests that a number of the aggregates were hollow. There are also areas around the aggregates that are not shadowed, and this suggests that the

particles collapsed under the vacuum used during TEM. To get a more definite answer about the assembly characteristics, SEM was also used to image the aggregates. As a control, the block copolymer dissolved in water was measured. Again, no welldefined structures were observed. The SEM images of the sample that was prepared with THF as a cosolvent, however, clearly showed the presence of spherical aggregates (Fig. 3). With cryo-SEM, the aggregates were frozen into a water droplet before their introduction into the microscope; this stopped the aggregates from collapsing under the low vacuum necessary for electron microscopy. Also, with this method, the presence of spherical aggregates was clearly shown (Fig. 4), and a more reliable estimate of the particle size could be made, which ranged from 500 nm up to even 10 lm. The formation of these aggregates was caused by phase separation of the hydrophilic peptide block and the hydrophobic PMMA block. This aggregation occurred only when the triblock copolymers were first suspended in THF. It is thought that this was due to the PMMA block of the polymer. In water, these PMMA blocks were not soluble and therefore were too rigid to associate and cause phase separation. When placed in a THF/water mixture, they were more soluble

6364

AYRES ET AL.

and therefore had an improved ability to aggregate, causing phase separation. In an MeOH/ water mixture, it is possible that the solubility and therefore dynamics of the PMMA block were still too low to allow phase separation, and so no well-defined aggregation was observed. The spherical aggregates that were observed could be micelles, vesicles (polymersomes), or LCMs.35–37 In this case, there is strong evidence to suggest that these particles were a mixture of polymersomes and LCMs. The spheres were larger than expected for micelles, ranging from 500 nm to 10 lm according to cryo-SEM, suggesting either LCMs or polymersomes.27,38 With TEM, it was only possible to see through some of the aggregates and not through others, and a few appeared to collapse under the vacuum applied by the electron microscopy measurements; this indicated that some of the particles were hollow. This suggested a mixture of hollow and solid aggregates, and as they were too large for micelles, this indicated that we had a mixture of polymersomes with LCMs. b-Hairpin Characterization of Vesicles To investigate the secondary structure of our peptide-based block copolymer, IR spectroscopy was used. This technique was chosen because it could reveal detailed information about the conformation of the peptide structure under investigation. From the amide I and amide II peaks, it was observed that neither the initiator nor the triblock copolymer assumed a b-hairpin conformation before assembly into vesicles. It was also clear from the IR spectra of the triblock polymer after assembly into vesicles that the peptide remained (partly) in a random coil type of structure (Fig. 5 and Table 1). It was not expected that a b-hairpin conformation would be observed for initiator 1 or for the triblock copolymer, PMMA–1–PMMA, as they were either precipitated from TFA into diethyl ether or dissolved in DMSO and, therefore, not exposed to conditions in which folding could occur. It was surprising, however, that no b-hairpin formation was observed for this peptide sequence when exposed to the more favorable conditions used when the polymer assembled into the observed vesicular structures. It can be supposed that the peptide must have folded in some manner to allow the PMMA tails to aggregate within the vesicle wall. IR clearly shows that during this aggregation process, the peptide did not

Figure 5. Amide I and II regions in IR for (A) 1, (B) PMMA–1–PMMA, and (C) PMMA–1–PMMA in vesicles.

PEPTIDE–POLYMER VESICLES

6365

Table 1. Comparison of Standard Amide I and Amide II Values39 for Antiparallel b Sheets and Amide I and II Values for 1, PMMA–1–PMMA, and Aggregated PMMA–1–PMMA Compound

Amide I (cm1)

Amide II (cm1)

Random coil Antiparallel b sheet 1 PMMA–1–PMMA PMMA–1–PMMA in vesicle

1656 1632–1613 (s)/1685 (w) 1640 1648 1644 (s)

1535 1530 1519 1527 1531

fold into a b hairpin. This could be a result of the steric hindrance introduced by the PMMA chains causing the peptide to twist during the aggregation process, preventing the creation of hydrogen bonds within the loop. It could also be that aggregation occurred too rapidly, trapping the peptide in a random coil structure before it had enough time to form b hairpins.

CONCLUSIONS In this article, a new synthetic method for the preparation of peptide-containing triblock copolymers has been demonstrated. The synthesis of a b-hairpin-containing, bifunctional ATRP initiator has been shown. The inclusion of two serines in the peptide sequence allowed the introduction of an a-bromo ester, in place of the a-bromo amide used by Rettig et al.26 This bifunctional initiator was then used to initiate the ATRP of MMA. Having created a peptide-based amphiphilic polymer, we investigated the structures formed in solution. It was shown that in water no aggregation took place and therefore no structures were discerned; however, upon first suspending the polymer in THF and then adding water, we observed the formation of a mixture of spherical aggregates. These structures were studied with TEM, SEM, and cryo-SEM, which confirmed that we had a mixture of hollow polymersomes and solid LCMs. IR spectroscopy revealed that the peptide assumed a random coil structure within the wall of the vesicle instead of the expected b hairpin. This work highlights the possibilities that exist when peptide-based polymers are used to create nanostructured materials. Altering the peptide that is used, the length of the polymer that is made, or even the monomer that is polymerized could lead to a wide range of fascinating

nanometer-sized structures with functionalities on the surface.

interesting

The authors gratefully acknowledge the Netherlands Technology Foundation for its financial support.

REFERENCES AND NOTES 1. Kim, H. C.; Wilds, J. B.; Kreller, C. R.; Volksen, W.; Brock, P. J.; Lee, V. Y.; Magbitang, T.; Hedrick, J. L.; Hawker, C. J.; Miller, R. D. Adv Mater 2002, 14, 1637–1639. 2. Shin, K.; Leach, K. A.; Goldbach, J. T.; Kim, D. H.; Jho, J. Y.; Tuominen, M.; Hawker, C. J.; Russell, T. P. Nano Lett 2002, 2, 933–936. 3. Kopecek, J. Eur J Pharm Sci 2003, 20, 1–16. 4. Klok, H. A.; Lecommandoux, S. Adv Mater 2001, 13, 1217–1229. 5. Billot, J. P.; Douy, A.; Gallot, B. Makromol Chem Macromol Chem Phys 1976, 177, 1889–1893. 6. Billot, J. P.; Douy, A.; Gallot, B. Makromol Chem Macromol Chem Phys 1977, 178, 1641–1650. 7. Douy, A.; Gallot, B. Polym Eng Sci 1977, 17, 523– 526. 8. Gallot, B. Prog Polym Sci 1996, 21, 1035–1088. 9. Schlaad, H.; Kukula, H.; Smarsly, B.; Antonietti, M.; Pakula, T. Polymer 2002, 43, 5321–5328. 10. Deming, T. J. Adv Drug Delivery Rev 2002, 54, 1145–1155. 11. Aliferis, T.; Iatrou, H.; Hadjichristidis, N. Biomacromolecules 2004, 5, 1653–1656. 12. Rathore, O.; Winningham, M. J.; Sogah, D. Y. J Polym Sci Part A: Polym Chem 2000, 38, 352–366. 13. Rathore, O.; Sogah, D. Y. Macromolecules 2001, 34, 1477–1486. 14. Rathore, O.; Sogah, D. Y. J Am Chem Soc 2001, 123, 5231–5239. 15. Sogah, D. Y.; Claussen, R. C.; Winningham, M. J.; Rathore, O. Abstr Pap Am Chem Soc 1999, 217, 071–BTEC. 16. Smeenk, J. M.; Otten, M. B. J.; Thies, J.; Tirrell, D. A.; Stunnenberg, H. G.; van Hest, J. C. M. Angew Chem Int Ed 2005, 44, 1968–1971. 17. Lowik, D.; van Hest, J. C. M. Chem Soc Rev 2004, 33, 234–245.

6366

AYRES ET AL.

18. Cornelissen, J. J. L. M.; Fischer, M.; Sommerdijk, N. A. J. M.; Nolte, R. J. M. Science 1998, 280, 1427–1430. 19. Vriezema, D. M.; Hoogboom, J.; Velonia, K.; Takazawa, K.; Christianen, P. C. M.; Maan, J. C.; Rowan, A. E.; Nolte, R. J. M. Angew Chem Int Ed 2003, 42, 772–776. 20. Vriezema, D. M.; Kros, A.; de Gelder, R.; Cornelissen, J.; Rowan, A. E.; Nolte, R. J. M. Macromolecules 2004, 37, 4736–4739. 21. Checot, F.; Lecommandoux, S.; Gnanou, Y.; Klok, H. A. Angew Chem Int Ed 2002, 41, 1339–1343. 22. Klok, H. A.; Langenwalter, J. F.; Lecommandoux, S. Macromolecules 2000, 33, 7819–7826. 23. Vandermeulen, G. W. M.; Tziatzios, C.; Klok, H. A. Macromolecules 2003, 36, 4107–4114. 24. Becker, M. L.; Liu, J. Q.; Wooley, K. L. Chem Commun 2003, 180–181. 25. Mei, Y.; Beers, K. L.; Byrd, H. C. M.; Vanderhart, D. L.; Washburn, N. R. J Am Chem Soc 2004, 126, 3472–3476. 26. Rettig, H.; Krause, E.; Borner, H. G. Macromol Rapid Commun 2004, 25, 1251–1256. 27. Becker, M. L.; Liu, J. Q.; Wooley, K. L. Biomacromolecules 2005, 6, 220–228. 28. Keller, R. N.; Wycoff, H. D. Inorg Synth 1946, 2, 1–4.

View publication stats

29. Simmons, A. H.; Michal, C. A.; Jelinski, L. W. Science 1996, 271, 84–87. 30. Krejchi, M. T.; Atkins, E. D.; Waddon, A. J.; Fournier, M. J.; Mason, T. L.; Tirrell, D. A. Science 1994, 265, 1427–1432. 31. Ayres, L.; Adams, P. J. H. M.; Lowik, D. W. P. M.; van Hest, J. C. M. Biomacromolecules 2005, 6, 825–831. 32. Ayres, L.; Vos, M. R. J.; Adams, P. J. H. M.; Shklyarevskiy, I. O.; van Hest, J. C. M. Macromolecules 2003, 36, 5967–5973. 33. Ayres, L.; Koch, K.; Adams, P. J. H. M.; van Hest, J. C. M. Macromolecules 2005, 38, 1699–1704. 34. Matyjaszewski, K.; Shipp, D. A.; Wang, J. L.; Grimaud, T.; Patten, T. E. Macromolecules 1998, 31, 6836–6840. 35. Yu, Y. S.; Zhang, L. F.; Eisenberg, A. Macromolecules 1998, 31, 1144–1154. 36. Zhang, L. F.; Eisenberg, A. Macromol Symp 1997, 113, 221–232. 37. Zhang, L. F.; Eisenberg, A. Science 1995, 268, 1728–1731. 38. Cameron, N. S.; Corbierre, M. K.; Eisenberg, A. Can J Chem 1999, 77, 1311–1326. 39. Krimm, S.; Bandekar, J. Advances in Protein Chemistry; Academic: San Diego, 1986; pp 282–361.