Controlled Molecular Organization of Surface Macromolecular

Dec 3, 2008 - E-mail: [email protected]., †. Genomics ... Shape-Persistent, Thermoresponsive Polypeptide Brushes Prepared by Vapor Deposit...
0 downloads 0 Views 3MB Size
58

Biomacromolecules 2009, 10, 58–65

Controlled Molecular Organization of Surface Macromolecular Assemblies Based on Stimuli-Responsive Polypeptide Brushes Chih-Tsung Yang,† Yuli Wang,‡ Susan Yu,† and Ying-Chih Ingrid Chang*,† Genomics Research Center, Academia Sinica, 128 Sec 2 Academic Road, Taipei, Taiwan, and Department of Chemical Engineering and Materials Science, University of California, Irvine, California 94027-2575 Received July 16, 2008; Revised Manuscript Received November 13, 2008

End-tethered cationic polypeptide brushes of poly(L-lysine) (t-PLL) were combined with three anionic polymers, poly(acrylic acid) (PAA), poly(L-glutamic acid) (PLGA), and poly(L-aspartic acid) (PLAA), to form reversible polyelectrolyte complex films at surfaces at neutral pH. The polyelectrolyte complex formation was confirmed by an in situ zeta-potential study and by positive fluorescent images after adding prelabeled anionic polymers. The secondary conformations of the t-PLL complex films depend upon the specific polyelectrolyte with which t-PLL was coupled as studied by circular dichroism and FTIR. Specifically, the random coil chain configuration of the t-PLL film was converted to an R-helical, β-sheet, or random coil structure after forming complexes with PAA, PLGA, or PLAA, respectively. Each of these complexes could be returned to the original random coil t-PLL structure by a dilute acid rinse. Additional thickness and morphological studies from ellipsometry and atomic force microscopy have further shown that the corresponding film thicknesses of the individual solvated films were affected more by the secondary structures in films than by the adsorbed mass or surface net charges. The solvated thickness was reduced significantly after the random coil t-PLL film was coupled with polyanions in forming compact regulated structures in films. This biomimetic approach provides a new opportunity for controlling the molecular organization in surface macromolecular assemblies and may provide a model for structural study of protein complexes on a chip.

Introduction When polymeric macromolecules are densely attached terminally to a surface, the polymer chains are forced to assume an elongated structure, resulting in a macromolecular assembly that has been dubbed a “polymer brush.” Polymer brushes have drawn wide scale scientific interest,1-4 in part because of the array of potential applications of polymer brushes in areas such as colloidal stabilization, adhesion, lubrication, microelectronics, and nanobiotechnology.5-9 Polymer brushes have been fabricated by attaching polymeric chain ends to a solid surface through so-called “grafting to” or “grafting from” techniques.10-15 The resulting polymeric chain configuration and orientation are the results of competition between segment-segment, and segment-solution interactions. Although one end of the polymer chain is fixed to the substrate, the molecular chain retains sufficient flexibility to undergo structural transitions or segmental aggregation without film dissolution. As a result, it is possible to reversibly switch the molecular conformations or segmental aggregated states of polymer brushes by external stimuli.16 Stimuli such as pH,17-20 solvents,21-23 temperature,24 salts,25 electrical field,26 and light27,28 were applied to control the molecular organization of polymer brushes. The formation of polymer-polymer complexes provides another approach to adjust the molecular conformation of polymers, as seen in nature. For example, DNA molecules form double-stranded DNA complexes via intermolecular interactions of complementary strands, resulting in a conformational transition from the original random coils of individual, single-stranded * To whom correspondence should be addressed. E-mail: yingchih@ gate.sinica.edu.tw. † Genomics Research Center. ‡ University of California.

DNA chains to double helices.29,30 Similarly, with 20 different naturally occurring and many more synthetic amino acids as the molecular building blocks, one can practically design and synthesize all kinds of protein macromolecules, which are able to utilize a series of hierarchical forces, such as electrostatic, polar, and hydrophobic forces, to form macromolecular assemblies that have tertiary and quaternary structures that are responsive to environmental stimulants by forming organized structures with well-defined, long-range spatial distribution. Previously, we and others have confirmed the stability and responsiveness of brushes made with polypeptide biomacromolecules.17,31-34 We showed that the secondary structures of polypeptide brushes are interconvertible among R-helix, β-sheet, and random coil configurations when stimulated by pH, ions, surfactants, solvent, and so on.10,11,35-37 In this work, we demonstrate that polymer-polymer complexes can also adjust the conformation of biomacromolecules in a brush configuration. We selected to study polypeptide brushes, the synthetic version of naturally occurring proteins, because of their relative simplicity in molecular compositions and synthesis. In contrast to previous studies on polypeptides, which focused primarily on how their conformation can be affected by small molecules or ions, we characterized the stimuli-responsiveness of polypeptide brushes in response to the addition of macromolecular chains. Similar to research on layer-by-layer deposition (LbL),20,38-41 where the surface layers are formed based on the counterionic polyelectrolyte adsorption, we utilized surface initiated vapor deposition polymerization to fabricate cationic poly(L-lysine) (PLL) with one end of the PLL attached to a solid substrate, forming a surface-tethered PLL brush (t-PLL). We selected three weakly charged anionic polymers, poly(acrylic acid) (PAA), poly(L-glutamic acid)

10.1021/bm8007956 CCC: $40.75  2009 American Chemical Society Published on Web 12/03/2008

Stimuli-Responsive Polypeptide Brushes Scheme 1. Molecular Formula of (a) t-PLL on Amine-Modified Silicon Oxide Surface, (b) PAA, (c) PLGA, and (d) PLAA

(PLGA), and poly(L-aspartic acid) (PLAA; Scheme 1), as the encountering macromolecular species to guarantee the formation of t-PLL/anionic polymer complexes via electrostatic attraction. These three anionic polymers were selected for their identical carboxylic acid side chains but different molecular structures (Scheme 1) to provide further insight into the ability of the encountering macromolecule to affect the polypeptide brush conformation via a polymer-polymer complex.

Materials and Methods Materials. N-carbobenzoxyl-L-lysine, triphosgene, hydrobromic acid solution (33% in glacial acetic acid), benzene, anhydrous tetrahydrofuran, dichloroacetic acid, chloroform, PAA (DP ∼ 160), PLGA (DP ∼ 110-365), PLAA (DP ∼ 100-330), and PLL (DP ∼ 144-335) were purchased from Sigma-Aldrich (St. Louis, MO) and used directly. Deionized water (DI water, pH ) 6.86) was purified by a Milli-Q purification system. For the fluorescence measurements, PLGA and PLAA tagged with Alexa Fluor 488 were synthesized in the following manner: Alexa Fluor 488 carboxylic acid succinimidyl ester (50 µL; concn ) 1 mg/mL) was added to a mixture of polymer solution (5 mg in 0.5 mL DI water) and phosphate buffered saline solution (1 mL; pH ) 7). The reaction was stirred for 1 h at room temperature in darkness. The resulting mixture was centrifuged with a Microcon centrifugal filter device (nominal molecular weight limit in Daltons ) 30000). Then the sample reservoir was held upside down in a vial, followed by transferring the concentrate to the vial by centrifuging three times for 3 min at 1300 rpm. The solution was diluted to 1 mg/mL and ready to use. Preparation of t-PLL Brushes. The procedure for forming t-PLL brushes followed the previously published protocol35 with a slight modification. N-Carboxyl anhydride (NCA) of N-carbobenzoxyl-Llysine was synthesized by phosgenating N-carbobenzoxyl-L-lysine with triphosgene in a moisture free environment. The NCA was then purified by rephosgenation according to the procedure published by Dorman et al.42 Silicon wafers, glass slides, and quartz substrates (approximately 1 × 3 cm in dimension) were cleaned with a freshly prepared piranha solution (3:1 concentrated H2SO4/30% H2O2 by volume) at 120 °C for 30 min. (Caution: piranha solution is highly corrosiVe. Extreme care should be taken when handling it.) They were then washed with a large amount of DI water and subsequently rinsed with acetone. The substrates were blown dry under a stream of nitrogen and immediately used for the silanization reaction, in which 150 µL of 3-aminopropyltriethoxysilane (APS) was placed together with the cleaned substrates on a Petri dish inside a vacuum glass desiccator (Wheaton dry-seal desiccator, 100 mm) and was evacuated down to about 0.3 Torr with a vacuum pump. The desiccator was then disconnected from the vacuum pump, kept tightly sealed, and left for about 16 h. After completion of silanization, the substrates were removed from the desiccator. They were ultrasonicated in fresh acetone for 5 min, repeated five times, and finally dried under a stream of nitrogen. Surface-initiated polymerization of poly(N-carbobenzoxyl-L-lysine) (PCBL) was performed by using a vapor deposition polymerization (VDP) chamber.35 The monomer (typically, 8 mg of NCA) was first dissolved in 0.3 mL of

Biomacromolecules, Vol. 10, No. 1, 2009

59

anhydrous tetrahydrofuran and then loaded in an aluminum holder. The solution was quickly dried in vacuo so that a thin and uniform solid layer of NCA was spread on the bottom of the holder. The APS modified substrates and NCA containers were placed in the VDP chamber. The pressure and temperature of the VDP chamber were controlled at about 1.0 × 10-4 Torr and 102 °C, respectively, for 20 to 30 min. After the polymerization, the PCBL sample was cleaned with a mixture of dichloroacetic acid and chloroform (20/80 (v/v)) in an ultrasonic bath for 5 min before being rinsed with fresh chloroform and dried under a stream of nitrogen. To convert surface-tethered PCBL to PLL, N-carbobenzyloxyl protection groups were removed by immersing the PCBL sample in a hydrobromic acid/benzene mixture (1:1 by volume), and then the resulting PLL-grafted sample was ultrasonicated for 40 min. The final PLL-grafted sample was washed with toluene, acetone, and DI water and then dried under a stream of nitrogen. To prepare the polyelectrolyte patterns for AFM and profilometry measurements, the silicon wafer was first treated with APS and then photolithographically patterned using S1813 photoresist, according to the procedure reported previously.43 The lithographic pattern consisted of a series of parallel, 10 µm wide stripes, combined with an additional 100 µm feature for ellipsometric measurements. Polyelectrolyte Complex Formation and Deposition. The 1 mg/ mL sodium salt forms of the anionic polyelectrolytes (PAA, PLGA, and PLAA) solutions were prepared. The pH was adjusted to ∼7 by adding small amount of 1 M NaOH (aq) or1 M HCl (aq). The t-PLL film was protonated by soaking in a neutral tris(hydroxymethyl)aminomethane (Tris) buffer solution (0.1 M, pH ) 7) for 1 min and then rinsed with running DI water. To form polyelectrolyte complex film at surface, the t-PLL film was dipped in the polyelectrolyte solution for 10 min, subsequently rinsed with running DI water for 2 min, and then dried with a stream of nitrogen. To regenerate the t-PLL film for new experiments, the t-PLL/polyelectrolyte sample was immersed in 0.1 M HCl(aq) for 5 min to desorb anionic polyelectrolyte from the surface, and then rinsed with running DI water for another 2 min, followed by drying with nitrogen. Characterization. Tapping mode AFM (TM-AFM) images were recorded by an Asylum Research AFM (MFP-3D) with an Olympus biolever (resonant frequency ) 37 kHz; spring constant ) 30 pN/nm). All the experiments were observed on the patterned silicon substrate (10 µm in width) in both air and DI water at a scale of 70 × 70 µm2. The cross-sectional height profiles were obtained by drawing a line crossing the surface patterns (Figure 5 and Supporting Information). The thickness value is based on the average cross-sectional height of patterned polymers calculated by Metamorph program on each AFM image. The zeta-potential data were measured by an Anton Paar electrokinetic analyzer. Zeta-potential data were recorded after each polyelectrolyte adsorbed on or desorbed from the t-PLL. All zeta-potential measurements were conducted in the following manner: The streaming potential of the t-PLL on silicon substrate filled with 10-3 M KCl was measured. Then, the t-PLL was coupled with PAA (1 mg/mL, pH ) 7) to form a t-PLL/PAA complex, and the surface potential was measured. The zeta-potential of the t-PLL surface was then determined by desorption of PAA in a 0.1 M HCl(aq) solution for 5 min followed by abundant DI water wash. The process of the adsorption and desorption of polyelectrolyte counterions was also observed via fluorescence microscopy on a glass substrate (2.5 × 1 × 0.1 cm). The t-PLL-modified glass slide was partially covered with an adhesive tape (0.2 cm in width) and then immersed in the Alexa Fluor-488-labeled polyanions (PAA, PLGA, and PLAA) for adsorption, followed by a DI water wash and a nitrogen purge. The covered tape was subsequently removed before the fluorescent imaging. Fourier transform infrared spectroscopy (FTIR) measurements of the treated substrates primarily followed the previously published protocol.35 The spectra of dry t-PLL and its complexes on wedge silicon

60

Biomacromolecules, Vol. 10, No. 1, 2009

wafers were recorded using a Nicolet Magna-IR 860 spectrometer in transmission mode with a clean silicon wafer as the reference. Before and during the measurements, the sample chamber was purged with dry air (Whatman FT-IR purge gas generator) to offset IR absorption from ambient moisture. Spectra were recorded at 4 cm-1 resolution, and 32 scans were collected per spectrum. Circular dichroism (CD) was used to characterize the conformations of t-PLL and its complexes on quartz substrates when the films were soaked in DI water. The spectra were recorded with a JASCO J-60 spectropolarimeter at a bandwidth of 1 nm and a scanning rate of 20 nm/min. A rectangular demountable liquid cell (Starna Cells Inc.) with a 1 mm path length was used. A cell holder was used to fix the quartz plate with the cell. The setup of the liquid cell was the same as described previously.35 The polymer films were coated on one side of the interior of the quartz cubicle on the beam path, such that we could record the conformational changes that occurred when the cell was filled with solution. Ellipsometric measurements of the t-PLL and its complexes on silicon substrates, for both dry and solvated films, were performed with a Gaertner LSE stokes ellipsometer with a He-Ne laser (λ ) 632.8 nm) and a fixed incident angle of 70°. The film thickness and refractive index were calculated using the Gaertner Ellipsometer Measurement Program. To measure the thickness and refractive index of the solvated films at the solid-water interface, a liquid cell (Gaertner Inc.) was used for the ellipsometric measurements. The setup of the liquid cell was the same as described previously.35 A total of 60 mL of DI water was added into the liquid cell, and the ellipsometric angles Ψ and ∆ were recorded as a function of time until equilibrium was reached. The measured Ψ and ∆ were analyzed by using an isotropic three-phase model (substrate/film/water) to calculate the film thickness and refractive index. The refractive indices of the aqueous phase (including all buffers) were simply taken as 1.333. The thickness of the polypeptide film was obtained by subtracting the thickness of SiO2 (11.4 Å) and APS (17.6 Å) from the calculated thickness. At least five different spots were measured for the average. The modeled structures of the PLL/PAA, PLL/PLGA, and PLL/ PLAA complexes were simulated by performing an optimized geometry calculation in Mechanics using Augmented MM2 parameters implemented in CAChe (Fujitz, Japan). The optimization of the complex structure was performed by 1000 iterations of steepest descent to where the conjugate gradient converged at 0.1 kcal/mol. The initial model of each molecule represented by a seven-repeated unit was created by CAChe. In the step to obtain the complex structure, the initial carboxylic groups of anionic polymers (both starting structures and bond angles are in their extended states) were brought to contact (within ∼5 Å) with the modeled structure for the amine side chains of PLL, which was based on the conformation suggested by CD analysis. Subsequently, each complex model was subjected to a molecular dynamics trajectory run at 1000 K to sample a time-ordered sequence of random conformations. Several low energy conformations were selected, their structures were further optimized, and then their potential energies were compared to determine the final, simulated structure.

Yang et al.

Figure 1. Transmission FTIR spectra of (a) t-PLL on silicon wafers and (b-d) its complexes with PAA, PLGA, and PLAA, measured in dry films.

Figure 2. CD spectra of (a) t-PLL on quartz and its complexes with (b) PAA, (c) PLGA, and (d) PLAA, measured in solvated films.

Results Complex Formation and Desorption. The FTIR spectrum of dry t-PLL film (Figure 1a) showed the amide I and amide II characteristic peaks at 1652 cm-1 and 1539 cm-1; the CD spectrum of t-PLL on quartz immersed in water (Figure 2a) showed the maxima at ∼220 nm and minima at ∼197 nm. Both CD and FTIR identified the t-PLL films (in air and water) were protonated and random coiled.37 Continuing the analysis of the t-PLL brush beyond what was done in previous work, the zetapotential measurement of the t-PLL brushes in this work showed that the t-PLL brush had a net positive charge of approximately 9 mV, as shown in Figure 3. The t-PLL zeta-potential value confirmed the net positive charge of the t-PLL brush, which was due primarily to its protonated -NH3+ side chain groups.

Figure 3. Evolution of zeta-potential (mV) during the alternate adsorption/desorption of polyelectrolyte counterions. PAA (DP ∼ 160), PLGA (DP ∼ 110-365), and PLAA (DP ∼ 100-330) were prepared at 1 mg/mL in a 0.1 M Tris-HCl buffer (pH ) 7).

In addition to analyzing the t-PLL brush, zeta-potential measurements were used to monitor the charge accumulation and alteration at surfaces and to confirm the successful adsorption of the anionic polymer species by the cationic t-PLL brushes

Stimuli-Responsive Polypeptide Brushes

Figure 4. Fluorescent images of a t-PLL surface (×60) after (a) partially adsorbing Alexa Fluor-488 labeled PLAA complex (green), followed by (b) cleaning with HCl(aq) and DI water, the surface was further shown to be capable of (c) partially adsorbing Alexa Fluor488-labeled PLAA (green).

at neutral pH, as shown in Figure 3. After coupling with PAA and thoroughly rinsing with DI water, the zeta-potential changed to a negative value, -14 mV. This indicated that the counterions provided by the PAA overcompensated the positive charge of the t-PLL brush. The zeta-potential measurements were also useful for monitoring the desorption process of the complexed polyanion. The t-PLL/PAA complex was rinsed with 0.1 M HCl(aq) and DI water, resulting in a complete charge reversal in the zetapotential measurement, as shown in Figure 3. The charge reversal indicated that the acid/water treatment effectively removed the anionic polymer from the t-PLL surface. As a result, the regenerated t-PLL surface can further adsorb/desorb other anionic polymers following the same protocols. Figure 3 shows the zeta-potential changes through the sequential adsorption and desorption of PLGA (-17 mV) and PLAA (-15 mV). The effectiveness of the polyanion adsorption and desorption were further confirmed by fluorescence experiments. In this case, a partially covered t-PLL bound glass substrate was used to adsorb an Alexa Fluor-488-labeled PLAA, as shown in Figure 4a. The boundary line of covered/exposed t-PLL areas can be clearly visualized by the black/green colors, indicating the adsorption of PLAA only onto the exposed t-PLL area. The boundary line vanished after a 0.1 M HCl(aq) and water rinse (Figure 4b), indicating that the fluorescent dye labeled PLAA was removed from the t-PLL brush. The same results were found when the adsorption/desorption cycle was repeated with Alexa Fluor-488-labeled PLGA (Figure 4c). The fluorescent experiments suggested the effectiveness of using acid/water wash to regenerate a fresh t-PLL surface, which is capable of repeatedly adsorbing polyanionic species. Complexation-Induced Conformational Transition. The conformations of t-PLL and its surface complexes were examined by using FTIR and CD measurements in a complementary way.37 In neutral aqueous solution at room temperature, t-PLL primarily adopts a random coil conformation, as the backbone hydrogen bonds are disrupted by the electrostatic repulsion of the side chain amines (pKa ) 10.5). The random coil configuration was confirmed by the characteristic broad amide I peak at 1652 cm-1 in FTIR (Figure 1a), a positive maximum at 220 nm, and a minimum at 197 nm in the CD spectrum (Figure 2a). In a similar way, the combination of FTIR and CD measurements enabled the determination of whether the chain conformations of the t-PLL surface complexes were random coils, R-helical, or β-sheet conformations. First, when t-PLL was coupled with PAA, the amide I peak at 1652 cm-1 could be assigned to either a random coil or an R-helical conformation of t-PLL (PAA has no absorption between 1610 and 1700 cm-1, Figure 1b). However, the CD spectrum of the t-PLL/PAA complex (Figure 2b) clearly showed the characteristic peaks of R-helices (double minima at 208 and 222 nm and a maximum

Biomacromolecules, Vol. 10, No. 1, 2009

61

at 196 nm), ruling out the random coil conformation. Second, for the t-PLL/PLGA complex, in the FTIR, the amide I peak was located at 1623 cm-1 (strong) and 1695 cm-1 (weak; Figure 1c). In the CD, the CD spectrum shows the minimum at 216 nm and the maximum at 196 nm (Figure 2c). Both indicated that the t-PLL/PLGA complex was predominantly in the antiparallel β-sheet conformation. The increase in amide adsorption peak at 3300 cm-1 in the FTIR spectra was a further proof of the PLGA addition. Third, for the t-PLL/PLAA complex, the amide I peak located at 1652 cm-1 in the FTIR (possibly to be either R-helix or random coil, Figure 1d) and the minimum at 198 nm in CD (random coil, Figure 2d) concluded that the surface t-PLL/PLAA complex primarily adopted a random coil conformation. Complexation-Induced Thickness Changes. As shown in Table 1a, dry film ellipsometric measurements of the thickness of the t-PLL brush, as well as of the complexes it formed with each of the anionic polyelectrolytes, revealed that the films were thicker in complex forms as a result of added mass of PAA, PLAA, or PLGA. While all samples greatly expanded from the dry state to the wet state, the degree of expansion, as suggested by the thickness of the films, showed no clear dependence with surface mass and surface net charge (as measured by zetapotential, Figure 3); the t-PLL (∼230 nm) alone swells more, thus being the thickest, among all four samples. For the three complex films built upon the same t-PLL film, while t-PLL/ PLGA complex (∼210 nm) had a thickness that approached that of t-PLL in water, the t-PLL/PAA (∼116 nm) and t-PLL/ PLAA (∼146 nm) complexes had thickness reduced almost by 30∼50%. As all these complexes were built upon the same t-PLL sample, the molecular grafting density and the initial process conditions are identical. The zeta-potential measurements also suggested similar net charges (∼-15 mV) in all three complexes. Therefore, the thickness reduction in water is better explained by the molecular conformations of the resulting films, where the R-helix of t-PLL/PAA adopted the most compact molecular structures, and antiparallel β-sheet of t-PLL/PLGA is in more extended form. A similar finding was observed by AFM analysis. In addition, AFM further provided visualization of surface morphology. Figure 5 shows one example of AFM image on t-PLL 10 µm patterns in water (Figure 5a) and the subsequent measurement after coupling with PAA and immersing in water (Figure 5b). The average thickness of the patterned t-PLL brush and the t-PLL/PAA complex were estimated by the average of crosssectional heights. The thickness is tabulated in Table 1 (all AFM images and cross sectional profiles were in Supporting Information). During the experiment, it was found that after coupling with the PAA, the measurement noise became much more significant. Nevertheless, we found all films swell dramatically in water (e.g., t-PLL swells from original 48 nm in air to 207 nm in water), with the t-PLL alone being the thickest. In addition, for the three complexes, the thickness was increased in the order of t-PLL/PAA, t-PLL/PLAA, and t-PLL/PLGA, which is consistent with its internal molecular structure from the most compact R-helix, “regulated” random coil, to antiparallel β-sheet structure. More data interpretation will be offered in the following Discussion.

Discussion Complex Formation and Conformation. Both zeta-potential and fluorescence measurements have confirmed the feasibility of our protocols in reversibly adsorbing polyelectrolytes with ionic t-PLL brushes. The CD and FTIR spectra added to these

62

Biomacromolecules, Vol. 10, No. 1, 2009

Yang et al.

Figure 5. Tapping mode AFM images and height profiles for (a) t-PLL patterns and (b) t-PLL/PAA patterns on an APS-modified silicon substrate in water. Image size: 70 × 70 µm2. The cross-sectional height profiles corresponding to the red lines in the AFM images are plotted below the AFM images. Table 1. Film Thickness of t-PLL and its Complexes with PAA, PLGA, and PLAAa ellipsometryb dry films

t-PLL t-PLL/PAA t-PLL/PLGA t-PLL/PLAA

AFMc

films in water

dry films

d (nm)

Nf

d (nm)

Nf

68.5 ( 1.2 72.6 ( 0.3 85.2 ( 0.6 75.1 ( 0.1

1.545 1.535 1.523 1.540

229.4 ( 2.3 116.5 ( 1.9 208.4 ( 1.2 146.3 ( 1.4

1.365 1.446 1.361 1.434

d (nm) 48 ( 3 54 ( 1 60 ( 1 55 ( 1

CD and FTIRd

films in water

dominating secondary conformation

d (nm) 207 ( 5 81 ( 3 126 ( 4 117 ( 5

random coil R-helix β-sheet random coil

a Measured in both dried and wetted films by ellipsometry (d: film thickness; Nf: refractive index of the film) and AFM. b For comparison purpose, all of three complexes were fabricated with the same t-PLL sample. To dissociate the complexes, the complexes between t-PLL and anionic polyelectrolyte was soaked in 0.1 M HCl(aq) for 2 min and rinsed with DI. Anionic polyelectrolyte was released from t-PLL film, as confirmed by FTIR. The t-PLL brush was then used to form new complex with a different anionic polyelectrolyte. The average and error were estimated based on at least five different spots of each sample. c All complexes were fabricated with the sample patterned t-PLL sample, followed by the regeneration method and experimental sequence described in ref 1. d The films may contain multiple conformations. The secondary conformation of each film was assigned by CD and FTIR spectra as the most prominent one.

results by showing the conformational conversions of the complexes as a result of the complex formation. Compared to previous studies37 in which the conversions of conformations were accomplished primarily by coupling small molecules and ions, here we have shown that t-PLL films can undergo conformational transition by coupling macromolecular species, primarily through the NH3+s-OOC interaction. Interestingly, these complex films exhibit dramatically different molecular organizations due to relatively minor differences in the monomer structure of both t-PLL and the three anionic polyelectrolytes. First, it is anticipated that the cationic t-PLL chains were in a fully extended configuration at neutral pH to minimize the internal charge density. In the case of the t-PLL/PAA complex, the anionic PAA (Scheme 1b), which contains a vinyl backbone, is considered a flexible polymer. The PAA can interact with and interpenetrate the t-PLL film, thereby minimizing the charge density of the t-PLL film and converting the majority of t-PLL secondary conformation from its original, fully extended state to a compressed yet stable R-helical conformation. Comparatively, the PLGA and PLAA, both consisting of more rigid amide backbones than that of PAA, are polar and capable of forming both inter- and intramolecular hydrogen bonds through either their backbones or carboxylic acidic side chains (Schemes 1c and 1d). In the case of t-PLL/PLGA, CD and FTIR analysis identified intermolecular antiparallel β-sheet structures as the most predominant secondary structure of the complex. This might be explained by the identical amide backbones of

both t-PLL and PLGA, which allowed for pairwise intermolecular hydrogen bonding. Interestingly, in the case of t-PLL/ PLAA, the t-PLL/PLAA complex could not form a regulated β-sheet structure as the counterpart of t-PLL/PLGA. Instead, an overall random coil conformation was observed. This could be explained by the reduced side chain length of the PLAA, which contains one fewer methylene group than that of the PLGA. In order to form both ammonium-carboxylic acid salt bridges between the side chains and pairwise hydrogen bondings between the backbones of t-PLL and PLAA, larger potential energy and bond strains are required. In this case, the longrange molecular order could not be maintained, as evidenced by both FTIR and CD experiments. Both the ellipsometry and AFM thickness measurements of the complexes in water reflect this interpretation of the conformational conversion of the different complexes. When measured in water, the thicknesses were not increased directly with the addition of polyelectrolyte mass, but instead were affected consistently by the overall structures of the macromolecular complexes. For example, we have observed degrees of thickness shrinkage from t-PLL to t-PLL/PAA of 50% by ellipsometry and 60% by AFM (Table 1). Although the quantitative discrepancy might be due to instrumental differences and sample variations (such as initial process conditions, PLL molecular weight and molecular weight distribution, and molecular grafting density), qualitatively, the results point to the same process. As stated above, the t-PLL chains probably were fully stretched in water to minimize the internal charge density. As a result of

Stimuli-Responsive Polypeptide Brushes Scheme 2. Schematic Drawings of (a) Extended, Cationic t-PLL Brushes Encountering Anionic Polyelectrolytes To Form t-PLL Complexesa

a

The internal structures of t-PLL complexes are based on CD and FTIR data and are simplified by computer generated pairwise cartoons: (b) helices of PLL/PAA, (c) antiparallel β-sheets of PLL/PLGA, (d) kinked coils of PLL/PLAA (symbols in b-d: rigid silver rod, PLL backbone; yellow line, anionic polymer backbone; blue bar, side chain N; red bar: O; green dot line: salt bridge).

the extended chain configuration, which is preferentially perpendicular to the surfaces, the film thickness was approximately proportional to the molecular chain length. The solvated thicknesses of the t-PLL/PAA (Table 1a) were about 55∼65% of the corresponding t-PLL/PLGA thicknesses. This change in thickness roughly corresponds to the ratio of the average molecular length per amino acid in R-helix (0.15 nm) and β-sheet (0.34 nm) conformations. Such expansion/collapse characteristics resulting from the conformational changes were similar to our previous findings, where the solvated t-PLL brush thickness decreased by 30 to 50% when its conformation underwent a transition from fully extended random coils to compact structures of β-sheets or R-helices, induced by pH, surfactants, or ions.37 Independent computer simulations of the molecular structures of freely solvated PLL complexes concur with the CD and FTIR experimental findings, and provide a helpful visualization of the complex conformation. The molecular complexes generated by the computer simulation are shown in Scheme 2. In the PLL/ PAA complex, a structure consisting of PAA wrapping around R-helical PLL was generated. This structure might be explained by the flexibility and hydrophobicity of the backbone of PAA, which would favor wrapping around the hydrophobic groove formed by the PLL side chains. Together with the acidic side chains of PAA, the wrapping enables intramolecular hydrogen bond formation of the PLL main chain. In the case of the PLL/ PLGA and PLL/PLAA complexes, however, the backbone amides of PLGA and PLAA can hydrogen bond with the backbone amides of PLL. The intermolecular antiparallel β-sheet structure optimized the hydrogen bonding interaction of PLGA and PLL, which have identical backbones and comparable side chain length. PLAA, with one fewer methylene in its side chain, seemed to be more prone to participate in interactions with the -NH group of the main chain of PLL, thus interfering with the extended backbone conformation of PLL, yielding an unclassified secondary structure. Aside from the computer simulation, we have mixed equal parts of freely solvated PLL and counterionic polyelectrolyte

Biomacromolecules, Vol. 10, No. 1, 2009

63

(PAA, PLGA, or PLAA) in solution for comparison. The CD and FTIR results indicated the conformational transition from the random coils of PLL to be predominantly R-helix, β-sheet, and random coil structures in PLL/PAA, PLL/PLGA, and PLL/ PLAA, respectively.44 These experimental results are qualitatively consistent with the findings from t-PLL- formed complexes at surfaces. However, we may suggest that the solution phase study has encountered several shortcomings when compared with the surface based assays, including the difficulty to recover the products due to the changes of complex solubility (e.g., β-sheet complex tends to precipitate), as well as the tedious purification from uncoupled polyions.45-48 On the other hand, the configuration of ionic polypeptide brushes tethered to the surface at a single point, such as t-PLL in the present study, clearly provides a sufficient degree of freedom for undergoing conformational transition, while simultaneously maintaining chemical stability for the assemblies remaining on the interfaces, such that the surface carries the entire scope of assemblies for study. Comparison of Film Stability and Conformation with Films Made by LbL Deposition. The preparation of the three polyelectrolyte complexes in this work, t-PLL/PAA, t-PLL/ PLGA, and t-PLL/PLAA, is similar to the nonstoichiometric LbL coupling process,38-41,49 except that the first layer, t-PLL, is chemically tethered and in a brush-like configuration in this case. The zeta-potential and fluorescence results paint a clear picture of the reversible coupling of the counterionic polyelectrolyte (PAA, PLAA, or PLGA) with the t-PLL brush due to its high chemical stability. Nevertheless, identical to the LbL method, the charge overcompensation of each adsorption layer, as seen in the zeta-potential measurements, allowed further addition of counterionic polyelectrolytes for the buildup of additional layers. In our case, up to 12 alternating layers were compiled, as determined by ellipsometry (Supporting Information). Recently, LbL deposition of polyelectrolytes for the fabrication of thin films has received great attention, in part due to its ease in coupling functional molecules, nanoparticles, etc., in films through nonstoichiometric electrostatic interactions of oppositely charged polyions.50,51 Macroscopic layer ordering in LbL films has been extensively discussed,52-54 but the transition of higher ordered inter- or intramolecular structures induced by the ionic coupling process were relatively less investigated. Biologically relevant materials such as the weak polyelectrolytes PLL and PLGA or PLAA have been employed as LbL building blocks, and the effects of pH, ionic strength, and molecular weight to the resulting film stability and conformation have been examined.44,55,56 Similar to our findings, β-sheet and random coils were found to be the dominant structures in PLGA/PLL and PLAA/PLL LbL films, respectively, at neutral pH.57-59 The thickness measurements permit for a further comparison of the film stability of t-PLL-based polyelectrolyte complexes with the PLL/polyelectrolyte LbL films. The PLL/polyelectrolyte LbL films which were prepared without adding NaCl in the buffer solution could be easily removed by rinsing with DI water. In contrast, the t-PLL/polyelectrolyte complexes could tolerate much more stringent conditions, such as 5 min of ultrasonication in a water bath, without any film depletion. This result has further confirmed the stability of t-PLL based complexes as a result of the covalent bonding to the underlying solid substrates.

64

Biomacromolecules, Vol. 10, No. 1, 2009

Conclusion In this work, our primary focus was to capture the conformational transition events on a t-PLL surface caused by the interaction of t-PLL with three chemically similar but structurally different anionic polyelectrolytes: PAA, PLGA, and PLAA. The use of the t-PLL brush was advantageous in that it provided a fully reversible system for the adsorption of the anionic polyelectrolytes and the resulting conformational changes. Our findings have shown that supramolecular assemblies were formed directly from the t-PLL surfaces, resulting in predominant secondary conformations ranging from R-helices to antiparallel β-sheets to random coils. The complexes were primarily arranged by electrostatic interactions, while weaker forces (H-bonds, dispersion, etc.) and the chain hindrance further regulated the higher-ordered molecular organization. Interestingly, the t-PLL brush layer was sensitive to minor changes of the monomer structures of coupling polymers, resulting in distinctive conformational and thickness/refractive index changes; it could potentially be used as a stimuli-responsive “smart layer” with interesting electrical and optical applications.60,61 Additionally, the tethered polyelectrolyte approach might serve as a new bioassay platform for studying topics related to conformational transitions, or as a new category of biochip platforms in studying protein conformational transition in vitro. Besides t-PLL, surface-tethered PLGA (t-PLGA), which complexes with cationic polyions to form different conformations, such as helical structures in the t-PLGA/poly(allylamine) and the t-PLGA/chitosan, β-sheet structures in the t-PLGA/PLL, and the t-PLGA/poly(L-ornithine), was confirmed in our laboratory and a manuscript including this data is under preparation. This universal approach might provide a fresh view to the protein conformational transition that is associated with the diseases, such as β-amyloids transforming from solvated helices to β-sheet pleated sheets in Alzheimer’s disease. Acknowledgment. This work was supported by University of California, Bio-Star and National Science Council, Taiwan (No. 96-2113-M-001-002). Dr. Ying-Ta Wu at Genomics Research Center, Academia Sinica, is acknowledged for computational modeling and interpretation. Supporting Information Available. Surface pattern measured by KLA-Tencor Surface-Profiler, ellipsometric measurements of LbL deposition of 12 layers on t-PLL brushes, and AFM images of patterned t-PLL and t-PLL/polyelectrolyte complexes in both air and water. This material is available free of charge via the Internet at http://pubs.acs.org.

References and Notes (1) Milner, S. T. Science 1991, 251 (4996), 905–914. (2) Halperin, A.; Tirrell, M.; Lodge, T. P. AdV. Polym. Sci. 1992, 100, 31–71. (3) Biesheuvel, P. M. J. Colloid Interface Sci. 2004, 275 (1), 97–106. (4) Piech, M.; Bell, N. S. Macromolecules 2006, 39 (3), 915–922. (5) Alarcon, C. D. H.; Farhan, T.; Osborne, V. L.; Huck, W. T. S.; Alexander, C. J. Mater. Chem. 2005, 15 (21), 2089–2094. (6) Kim, J.; Liu, Y.; Ahn, S. J.; Zauscher, S.; Karty, J. M.; Yamanaka, Y.; Craig, S. L. AdV. Mater. 2005, 17 (14), 1749–1753. (7) Minko, S. Polym. ReV. 2006, 46 (4), 397–420. (8) Ohno, K.; Morinaga, T.; Takeno, S.; Tsujii, Y.; Fukuda, T. Macromolecules 2006, 39 (3), 1245–1249. (9) Senaratne, W.; Andruzzi, L.; Ober, C. K. Biomacromolecules 2005, 6 (5), 2427–2448. (10) Chang, Y. C.; Frank, C. W. Langmuir 1996, 12 (24), 5824–5829. (11) Chang, Y. C.; Frank, C. W. Langmuir 1998, 14 (2), 326–334. (12) Zhao, B.; Brittain, W. J. Prog. Polym. Sci. 2000, 25 (5), 677–710.

Yang et al. (13) Ruhe, J.; Knoll, N. J. Macromol. Sci., Polym. ReV. 2002, C42 (1), 91–138. (14) Advincula, R. C.; Brittain, W. J.; Caster, K. C.; Ruehe, J. Polymer Brushes; Wiley-VCH: Weinheim, Germany, 2004. (15) Edmondson, S.; Osborne, V. L.; Huck, W. T. S. Chem. Soc. ReV. 2004, 33 (1), 14–22. (16) Russell, T. P. Science 2002, 297 (5583), 964–967. (17) Ito, Y.; Ochiai, Y.; Park, Y. S.; Imanishi, Y. J. Am. Chem. Soc. 1997, 119 (7), 1619–1623. (18) Park, Y. S.; Ito, Y.; Imanishi, Y. Chem. Mater. 1997, 9 (12), 2755– 2758. (19) Ionov, L.; Houbenov, N.; Sidorenko, A.; Stamm, M.; Luzinov, I.; Minko, S. Langmuir 2004, 20 (23), 9916–9919. (20) Tokareva, I.; Minko, S.; Fendler, J. H.; Hutter, E. J. Am. Chem. Soc. 2004, 126 (49), 15950–15951. (21) Zhao, B.; Brittain, W. J.; Zhou, W. S.; Cheng, S. Z. D. J. Am. Chem. Soc. 2000, 122 (10), 2407–2408. (22) Minko, S.; Muller, M.; Usov, D.; Scholl, A.; Froeck, C.; Stamm, M. Phys. ReV. Lett. 2002, 88 (3), (23) Lupitskyy, R.; Roiter, Y.; Tsitsilianis, C.; Minko, S. Langmuir 2005, 21 (19), 8591–8593. (24) Balamurugan, S.; Mendez, S.; Balamurugan, S. S.; O’Brien, M. J.; Lopez, G. P. Langmuir 2003, 19 (7), 2545–2549. (25) Zhang, H. N.; Ruhe, J. Macromolecules 2005, 38 (11), 4855–4860. (26) Lokuge, I. S.; Bohn, P. W. Langmuir 2005, 21 (5), 1979–1985. (27) Park, Y. S.; Ito, Y.; Imanishi, Y. Macromolecules 1998, 31 (8), 2606– 2610. (28) Samanta, S.; Locklin, J. Langmuir 2008, 24 (17), 9558–9565. (29) Reishus, D.; Shaw, B.; Brun, Y.; Chelyapov, N.; Adleman, L. J. Am. Chem. Soc. 2005, 127 (50), 17590–17591. (30) Strand, S. P.; Danielsen, S.; Christensen, B. E.; Varum, K. M. Biomacromolecules 2005, 6 (6), 3357–3366. (31) Whitesell, J. K.; Chang, H. K. Science 1993, 261 (5117), 73–76. (32) Oosterling, M.; Willems, E.; Schouten, A. J. Polymer 1995, 36 (23), 4485–4490. (33) Wieringa, R. H.; Schouten, A. J. Macromolecules 1996, 29 (8), 3032– 3034. (34) Wieringa, R. H.; Siesling, E. A.; Geurts, P. F. M.; Werkman, P. J.; Vorenkamp, E. J.; Erb, V.; Stamm, M.; Schouten, A. J. Langmuir 2001, 17 (21), 6477–6484. (35) Wang, Y. L.; Chang, Y. C. Langmuir 2002, 18 (25), 9859–9866. (36) Wang, Y. L.; Chang, Y. C. Macromolecules 2003, 36 (17), 6503– 6510. (37) Wang, Y. L.; Chang, Y. C. Macromolecules 2003, 36 (17), 6511– 6518. (38) Hoogeveen, N. G.; Stuart, M. A. C.; Fleer, G. J.; Bohmer, M. R. Langmuir 1996, 12 (15), 3675–3681. (39) Caruso, F.; Mohwald, H. J. Am. Chem. Soc. 1999, 121 (25), 6039– 6046. (40) Ladam, G.; Schaad, P.; Voegel, J. C.; Schaaf, P.; Decher, G.; Cuisinier, F. Langmuir 2000, 16 (3), 1249–1255. (41) Quinn, A.; Such, G. K.; Quinn, J. F.; Caruso, F. AdV. Funct. Mater. 2008, 18, 17–26. (42) Dorman, L. C.; Shiang, W. R.; Meyers, P. A. Synth. Commun. 1992, 22 (22), 3257–3262. (43) Wang, Y. L.; Chang, Y. C. AdV. Mater. 2003, 15 (4), 290–293. (44) Zhao, W. H.; Zheng, B.; Haynie, D. T. Langmuir 2006, 22 (15), 6668– 6675. (45) Davidson, B.; Fasman, G. D. Biochemistry 1969, 8 (10), 4116–4126. (46) Fasman, G. D.; Hoving, H.; Timashef, S. Biochemistry 1970, 9 (17), 3316–3324. (47) Adler, A. J.; Fasman, G. D. J. Phys. Chem. 1971, 75 (10), 1516– 1526. (48) Carroll, D. Biochemistry 1972, 11 (3), 426–433. (49) Benkirane-Jessel, N.; Lavalle, P.; Hubsch, E.; Holl, V.; Senger, B.; Haikel, Y.; Voegel, J. C.; Ogier, J.; Schaaf, P. AdV. Funct. Mater. 2005, 15 (4), 648–654. (50) Decher, G. Science 1997, 277 (5330), 1232–1237. (51) Hammond, P. T. AdV. Mater. 2004, 16 (15), 1271–1293. (52) Arys, X.; Laschewsky, A.; Jonas, A. M. Macromolecules 2001, 34 (10), 3318–3330. (53) Boulmedais, F.; Schwinte, P.; Gergely, C.; Voegel, J. C.; Schaaf, P. Langmuir 2002, 18 (11), 4523–4525. (54) Zhang, H. N.; Ruhe, J. Macromolecules 2005, 38 (26), 10743–10749. (55) Lvov, Y. Protein Architecture: Interfacial Molecular Assembly and Immobilization Biotechnology; Deckker: New York, 2000; pp 125167. (56) Haynie, D. T.; Zhang, L.; Rudra, J. S.; Zhao, W. H.; Zhong, Y.; Palath, N. Biomacromolecules 2005, 6 (6), 2895–2913.

Stimuli-Responsive Polypeptide Brushes (57) Cooper, T. M.; Campbell, A. L.; Crane, R. L. Langmuir 1995, 11 (7), 2713–2718. (58) Boulmedais, F.; Bozonnet, M.; Schwinte, P.; Voegel, J. C.; Schaaf, P. Langmuir 2003, 19 (23), 9873–9882. (59) Haynie, D. T.; Balkundi, S.; Palath, N.; Chakravarthula, K.; Dave, K. Langmuir 2004, 20 (11), 4540–4547.

Biomacromolecules, Vol. 10, No. 1, 2009

65

(60) Schrof, W.; Rozouvan, S.; Van Keuren, E.; Horn, D.; Schmitt, J.; Decher, G. AdV. Mater. 1998, 10 (4), 338–341. (61) Bellomo, E. G.; Davidson, P.; Imperor-Clerc, M.; Deming, T. J. J. Am. Chem. Soc. 2004, 126 (29), 9101–9105.

BM8007956