Interactions between Homopolypeptides and Lightly Cross-Linked

Dec 5, 2008 - ReceiVed September 12, 2008. ReVised Manuscript ReceiVed October 21, 2008. The relative importance of electrostatic and nonelectrostatic...
0 downloads 0 Views 404KB Size
522

Langmuir 2009, 25, 522-528

Interactions between Homopolypeptides and Lightly Cross-Linked Microgels Helena Bysell* and Martin Malmsten Department of Pharmacy, Uppsala UniVersity, P.O. Box 580, SE-751 23 Uppsala, Sweden ReceiVed September 12, 2008. ReVised Manuscript ReceiVed October 21, 2008 The relative importance of electrostatic and nonelectrostatic interactions in peptide-microgel systems was evaluated by micromanipulator-assisted light microscopy, confocal microscopy, and circular dichroism. For this purpose, the interaction of various homopolypeptides with lightly cross-linked polyelectrolyte gel particles (∼70 µm in diameter) was studied with focus on peptide-induced microgel deswelling and its relation to peptide distribution within the microgel particles. Negatively charged poly-L-glutamic acid (pGlu) and poly-L-aspartic acid (pAsp), as well as uncharged poly-L-proline (pPro) and poly-L-threonine (pThr), were found to not bind to negatively charged poly(acrylic acid) microgels under the conditions investigated, but were instead depleted from the microgel particles. Positively charged poly-L-arginine (pArg), poly-L-histidine (pHis), and poly-L-lysine (pLys), on the other hand, interacted strongly with the oppositely charged microgel particles and caused significant deswelling of these. In parallel, cationic acrylamidopropyltriethylammoniumchloride (APTAC) microgels bound negatively charged polypeptides to a much higher extent than positively charged and uncharged ones. These findings suggest that electrostatic interactions dominate peptide binding and resulting microgel deswelling in these systems. Nevertheless, although the amount of cationic peptide bound to the anionic microgel particles was similar for cationic pLys, pArg, and pHis, peptide-induced gel deswelling differed significantly, as did the change in peptide conformation after microgel binding and the peptide distribution within the microgels. These effects, as well as pH dependent binding and release of titrable pHis, are discussed in terms of the effects of the charge density of, and structural differences between, the cationic homopolypeptides on the interaction with the oppositely charged microgel particles.

Introduction Stimuli-responsive microgels are receiving increasing attention due to their ability to bind and store substances as well as to release them in response to changes in the external environment.1-3 Such systems have potential, for example, as protective and functional carriers in protein and peptide drug delivery.4-6 Although the osmotic deswelling of charged microgels has been studied for surfactants,7-12 drugs,13-17 and proteins/peptides,18-21 * Corresponding author. (1) Pelton, R. AdV. Colloid Interface Sci. 2000, 85(1), 1–33. (2) Tan, B. H.; Tam, K. C. AdV. Colloid Interface Sci. 2008, 136(1-2), 25–44. (3) Vinogradov, S. V. Curr. Pharm. Des. 2006, 12(36), 4703–4712. (4) Morishita, M.; Goto, T.; Nakamura, K.; Lowman, A. M.; Takayama, K.; Peppas, N. A. J. Controlled Release 2006, 110(3), 587–594. (5) Peppas, N. A. Int. J. Pharm. 2004, 277(1-2), 11–17. (6) Besheer, A.; Wood, K. M.; Peppas, N. A.; Mader, K. J. Controlled Release 2006, 111(1-2), 73–80. (7) Andersson, M.; Råsmark, P.-J.; Elvingson, C.; Hansson, P. Langmuir 2005, 21, 3773–3781. (8) Go¨ransson, A.; Hansson, P. J. Phys. Chem. B 2003, 107, 9203–9213. (9) Hansson, P.; Schneider, S.; Lindman, B. J. Phys. Chem. B 2002, 106, 9777–9793. (10) Khokhlov, A. R.; Kramarenko, E. Y.; Makhaeva, E. E.; Starodubtzev, S. G. Macromolecules 1992, 25, 4779–4783. (11) Nilsson, P.; Hansson, P. J. Phys. Chem. B 2005, 109, 23843–23856. (12) Bradley, M.; Vincent, B.; Burnett, G. Langmuir 2007, 23(18), 9237– 9241. (13) Costa, D.; Miguel, M. G.; Lindman, B. J. Phys. Chem. B 2007, 111(29), 8444–8452. (14) Eichenbaum, G. M.; Kiser, P. F.; Shah, D.; Simon, S. A.; Needham, D. Macromolecules 1999, 32, 8996–9006. (15) Eichenbaum, G. M.; Kiser, P. F.; Dobrynin, A. V.; Simon, S. A.; Needham, D. Macromolecules 1999, 32, 4867–4878. (16) Hoare, T.; Pelton, R. Langmuir 2008, 24(3), 1005–1012. (17) Tan, J. P. K.; Zeng, A. Q. F.; Chang, C. C.; Tam, K. C. Int. J. Pharm. 2008, 357(1-2), 305–313. (18) Karabanova, V. B.; Rogacheva, V. B.; Zezin, A. B.; Kabanov, V. A. Polym. Sci. 1995, 37(11), 1138–1143. (19) Kabanov, V. A.; Skobeleva, V. B.; Rogacheva, V. B.; Zezin, A. B. J. Phys. Chem. B 2004, 108, 1485–1490. (20) Zezin, A.; Rogacheva, V.; Skobeleva, V.; Kabanov, V. Polym. AdV. Technol. 2002, 13, 919–925.

much is still relatively poorly understood on the interaction of proteins and polypeptides with such microgels, and on how factors such as peptide/protein binding to, and distribution within, microgel particles depend on factors such as interaction strength and network properties. In this respect, weakly cross-linked microgels are significantly less investigated than more highly cross-linked systems, of interest, for example, in chromatography, where numerous investigations have been reported concerning investigations on adsorption kinetics and distribution of proteins within polymer gel particles.22-27 In a couple of previous studies,28,29 we therefore investigated the effect of salt concentration, pH, and peptide size on the interaction of positively charged poly-L-lysine and oppositely charged poly(acrylic acid) microgels, with focus on peptide transport, binding, and distribution within the microgels, as well as on microgel deswelling. The aim of this present study is to extend these previous studies through investigations on the relative importance of electrostatic and nonelectrostatic interactions occurring in microgel-peptide systems. For this purpose, we here report results for on a range of homopolypeptides, that is, uncharged polyL-threonine and poly-L-proline, negatively charged poly-L(21) Johansson, C.; Hansson, P.; Malmsten, M. J. Colloid Interface Sci. 2007, 316, 350–359. (22) Dziennik, S. R.; Belcher, E. B.; Barker, G. A.; Lenhoff, A. M. Biotechnol. Bioeng. 2005, 91(2), 139–153. (23) Hubbuch, J.; Linden, T.; Kneips, E.; Ljunglo¨f, A.; Tho¨mmes, J.; Kula, M.-R. J. Chromatogr., A 2003, 1021, 93–104. (24) Hubbuch, J.; Linden, T.; Kneips, E.; Tho¨mmes, J.; Kula, M.-R. Biotechnol. Bioeng. 2002, 80(4), 360–368. (25) Linden, T.; Ljunglo¨f, A.; Hagel, L.; Kula, M.-R.; Tho¨mmes, J. Sep. Sci. Technol. 2002, 37, 1–32. (26) Ljunglo¨f, A.; Larsson, M.; Knuuttila, K.-G.; Lindgren, J. J. Chromatogr., A 2000, 893, 235–244. (27) Malmsten, M.; Xing, K.; Ljunglof, A. J. Colloid Interface Sci. 1999, 220(2), 436–442. (28) Bysell, H.; Hansson, P.; Malmsten, M. J. Colloid Interface Sci. 2008, 323, 9–60. (29) Bysell, H.; Malmsten, M. Langmuir 2006, 22(12), 5476–5484.

10.1021/la8029984 CCC: $40.75  2009 American Chemical Society Published on Web 12/05/2008

Interactions in Peptide-Microgel Systems

glutamic acid and poly-L-aspartic acid, and positively charged poly-L-arginine, poly-L-lysine, and poly-L-histidine, together with negatively charged poly(acrylic acid) (AA) and positively charged acrylamidopropyltriethylammoniumchloride (APTAC) microgels. In addition, we evaluated possible differences between cationic homopolypeptides on the interaction with AA microgels, as such peptides are known to interact differently with oppositely charged polyions.30-32 By focusing on microgel particles with diameter ∼70 µm, gel deswelling response and polypeptide distribution within microgel particles could by straightforwardly monitored by micromanipulator-assisted light microscopy and confocal laser scanning microscopy, respectively.

Experimental Section Materials. Poly-L-arginine (pArg, Mw ) 13 kDa), poly-L-histidine (pHis, Mw )6.3 kDa), poly-L-lysine (pLys, Mw ) 9.2 kDa and 15 kDa), and poly-L-proline (pPro, Mw ) 5.8 kDa) were from SigmaAldrich (St. Louis, MO), while poly-L-threonine (pThr, Mw )7.6 kDa), poly-L-glutamic acid (pGlu, Mw )13 kDa), and poly-L-aspartic acid (pAsp, Mw )11 kDa) were from Sigma-Aldrich (Steinheim, Germany). Additionally, unlabeled and Alexa-488 labeled monodisperse (>95% purity) polypeptides consisting of 24 repeated units of lysine (Lys 24, Mw ) 3.1 kDa), arginine (Arg 24, Mw ) 3.8 kDa), and histidine (His 24, Mw ) 3.3 kDa) were obtained from Biopeptide Co. (San Diego, CA). For microgel synthesis, N,N′-methylenebisacrylamide, N,N,N′,N′-tetramethyl-ethylenediamine (TEMED), and acrylamidopropyltrimethylammoniumchloride (APTAC) were obtained from Sigma-Aldrich (Steinheim, Germany). In addition, ammonium persulfate, acrylic acid, and poly(acrylic acid) were from Aldrich (Steinheim, Germany), while sorbitan monostearate (Span 60) was from Carl ROTH (Karlsruhe, Germany). Fluorescent tags Alexa Fluor 488 carboxylic acid succinimidyl ester and BODIPY 493/503 (4,4-difluoro-1,3,5,7-tetramethyl-4-bora-3a,4a-diaza-s-indacene-8-propionic acid, succinimidyl ester) were both from Invitrogen (Eugene, OR), and the bisinchoninic acid (BCA) assay kit was from Pierce (Rockford, IL). Poly(styrene sulfonate) (pSS, Mw )194 200) was from Polysciences (Warrington, PA). All other chemicals were of analytical grade. For pH control, 5 mM buffer solutions of sodium acetate/acetic acid, sodium phosphate monobasic/ sodium phosphate dibasic, and sodium carbonate/sodium bicarbonate were used for pH 5.5, 7.0, and 9.5, respectively. Sodium chloride was added to obtain the appropriate ionic strength. Purified Milli-Q water was used throughout. Preparation of Microgels. Poly(acrylic acid) microgel particles were synthesized as described previously.29 In brief, 0.05 g of Span 60 was dissolved in 20 mL cyclohexane, preheated to 45 °C and stirred at 1000 rpm under nitrogen atmosphere. A solution of 2.6 g of acrylic acid, 0.1 g of N,N′-methylenebisacrylamide, 20 g of NaOH (2 M), 4 g of NaCl, and 60 µL of TEMED was prepared and 10 mL of this reaction mixture mixed with 0.5 mL of 0.18 M ammonium persulfate solution and added to the preheated cyclohexane solution. The polymerization was performed at 65 °C under nitrogen atmosphere. The reaction was stopped after 30 min by addition of 40 mL of methanol. The gel particles were repeatedly washed with methanol and water and then sieved using a Retsch 5657 test sieve (Haan, Germany). Fractions below the mesh size of 300 µm were collected and stored in water. The dry mass of these microgels was determined from freeze-drying experiments to be 1.7 mg gel/g gel solution using a Flexidry µP freeze-dryer (Kinetics Thermal Systems, Stone Ridge, NY). The swelling/deswelling of the microgels was totally reversible when changing pH and/or ionic strength and reversibility persisted beyond one swelling/deswelling cycle. The fully swollen gel network mesh size was estimated to be ∼10 nm, based on theoretical modeling of the swelling/deswelling behavior of gels when varying pH and salt concentration. (30) Wagner, K. G.; Arfmann, H.-A. Eur. J. Biochem. 1974, 46, 27–34. (31) Morpurgo, M.; Radu, A.; Bayer, E.; Wilchek, M. J. Mol. Recognit. 2004, 17(6), 558–566. (32) Mita, K.; Ichimura, S.; Zama, M. Biopolymers 1978, 17, 2783–2798.

Langmuir, Vol. 25, No. 1, 2009 523 Quaternary ammonium salt microgels from cationic APTAC were also synthesized with the emulsion polymerization method described above. A solution of 0.001 g of N,N′-methylenebisacrylamide, 2.5 g of purified water, 5 g of APTAC (75 w/w % in water), and 375 µL of TEMED was prepared. One milliliter of this solution was mixed with 90 µL of ammonium persulfate (0.18 M) and added to preheated (70 °C) and strongly stirred (1200 rpm) cyclohexane solution (0.09 g of Span 60 in 40 mL of cyclohexane). Gelation occurred almost immediately, and the reaction was stopped after 30 min. Gel particles were repeatedly washed with acetone/water on a P4 vacuum filter. Deswelling/Swelling Kinetics. Microgel deswelling/swelling kinetics was monitored by micromanipulator-assisted light microscopy,28 using an Olympus Bx-51 light microscope (Olympus, Tokyo, Japan) equipped with an ONM-1 manipulator (Narishige, Tokyo, Japan) and a DP 50 digital camera (Olympus, Tokyo, Japan). Micropipets were prepared with a PC-10 puller and a MF-9 forger (both Narishige, Tokyo, Japan). An IM-5A injector (Narishige, Tokyo, Japan) was used to capture gel particles. By micromanipulation, these were placed inside a flow pipet and flushed with peptide solution using a Peristaltic pump P-1 (Pharmacia, Uppsala, Sweden) at a flow rate of 1.8 mL/min. The gel particles were photographed every 10-30 s, depending on deswelling rate, using Viewfinder, Studio 3.0.1 software (Pixera, San Jose, CA). The diameter of the gel particles were measured using Olympus DP-soft (Olympus, Tokyo, Japan), and the deswelling ratios were expressed as V/V0, where V is the volume of a gel particle after exposure to peptide for a certain time and V0 is the volume of the gel particle before addition of peptide in pure buffer at the experimental condition investigated. The volume responses of single gel particles on exposure to negatively charged (pGlu and pAsp) and neutral (pThr and pPro) homopolypeptides were studied at a peptide concentration of 100 mg/L at pH 7.0 and salt concentration of 20 mM. The interaction of AA microgels with positively charged homopolypeptides (pArg, pLys, and pHis) was studied in more detail, and for pArg and pLys the deswelling kinetics was investigated with respect to concentration (10, 30, 60, 100 mg/L) at pH 7.0 and salt concentration 20 mM. In order to study the possible release of pArg and pLys, the gel particles were flushed with high salt concentration (220 mM) for extended times. pHis binding and resulting microgel deswelling was studied at lower pH (pH 5.5) due to the limited solubility of this polypeptide at higher pH g pKa (∼6.0 for the isolated imidazole group).33 For comparison, also pArg and pLys were studied at pH 5.5 at a peptide concentration of 100 mg/L. pH-induced release was studied through raising the pH of the flow solution to 9.5 following peptide binding to microgels at this lower pH. To exclude possible effects from the molecular weight polydispersity of the positively charged peptides investigated on the deswelling behavior of AA gels, complementary studies was performed with the monodisperse 24-mer peptides Lys 24, His 24, and Arg 24. A minimum of three gel particles of diameter ∼70 µm were studied for each homopolypeptide at each condition. For pArg and pLys, experimental deswelling curves were fitted to

V ) (1 - k(t - t0))6 V0

(1)

where V/V0, k, and t and t0 represent the deswelling ratio, the apparent deswelling rate constant, and time, respectively.8 Equation 1 is an approximate expression for the deswelling kinetics of spherical microgel particles consisting of a swollen core surrounded by a thin surface phase at conditions where the kinetics is controlled by stagnant layer diffusion. Since this only might be true for some of the systems investigated, the obtained k values were not interpreted quantitatively, and instead the approach was used merely to reduce the data to a more convenient effective rate constant, kapp.28 Turbidometric Titration. Peptide solutions of pLys, pHis, and pArg at pH 5.5 and 7.0 were titrated with linear poly(acrylic acid) and poly(styrenesulfonate), and complex formation followed tur(33) Patchornik, A.; Berger, A.; Katchalski, E. J. Am. Chem. Soc. 1957, 79(19), 5227–5230.

524 Langmuir, Vol. 25, No. 1, 2009 bidometrically at 420 nm using a Spectronic Genesys5 spectrophotometer (Milton Roy, Rochester, NY). Uptake. The uptake of pArg, pLys, and pHis into microgel particles was studied by equilibrating gel solutions with peptide solutions for ∼1 week. Gel particles were then separated from the peptide solution by centrifugation at 5000 rpm for 15 min. The peptide concentration in the supernatant was determined by complexation with bisinchoninic acid,34 and absorbance determined by using a Saphire2 plate reader (Tecan, Ma¨nnedorf, Switzerland). The peptide concentrations used (500 and 1000 mg/L) were chosen based on previous experiments29 to correspond to maximum peptide uptake in the gels. Duplicate experiments at each concentration were performed at pH 5.5 and pH 7.0. Confocal Laser Scanning Microscopy. Peptide Labeling. Homopolypeptides were labeled with Alexa Fluor 488 dye according to a standard protocol recommended by the supplier. In brief, ∼5-10 µg of dye per milligram of polypeptide was reacted for 1 h at room temperature in basic conditions (150 mM carbonate buffer at pH 8.5). For pHis, the labeling reaction was performed at more acidic conditions (10 mM phosphate buffer at pH 6) due to the limited solubility of pHis at pH > pKa. Unreacted dye was removed by repeated size-exclusion chromatography using PD-10 columns (GE Health Care, Uppsala, Sweden), and the concentration of homopolypeptide measured spectrophotometrically in duplicate experiments after complexation with bisinchoninic acid.34 Absorbance measurements were performed on a Saphire2 plate reader (Tecan, Ma¨nnedorf, Switzerland) at 562 nm. As the Alexa Fluor 488 dye, both in itself and when tagged to cationic homopolypeptides, showed affinity for APTAC gels; the electronically neutral and smaller fluorescent dye, BODIPY 493/503, was used for this microgel, with the homopolypeptides being labeled with the same procedure as described above. Although the BODIPY-labeled pLys was depleted from the APTAC gels, free BODIPY 493/503 dye displayed some affinity for the cationic gels, precluding quantitative analysis of the results obtained from peptides pThr, pArg, pHis, and pPro, which all showed a pronounced affinity for these gels. Whether the latter is an effect of nonelectrostatic interactions between the gel network and the homopolypeptides or between the gel network and the fluorescent tag therefore remains unclear. As shown in Figure S1b in the Supporting Information, neither Alexa Fluor 488 nor BODIPY 493/503 displayed any affinity for negatively charged AA gels. For cationic homopolypeptides, the fluorescent label probably also interacts with the polypeptide side chains (as well as the peptide end group), which could influence the peptide-gel interactions. However, as the Alexa 488-labeled monodisperse peptides Lys 24, Arg 24, and His 24 were all synthesized by solid phase synthesis, giving a labeling degree of exactly one label/peptide molecule located on the end group with the distribution pattern of these correlating well with the polydisperse peptides, this effect was concluded to be of minor importance for this investigation. Peptide Distribution. Ten microliters of microgel solution was equilibrated for at least 48 h with 200 µL of fluorescently labeled homopolypeptide solution to a final concentration of 100 mg/L. The distribution of the labeled homopolypeptides within the microgel particles was monitored with a Confocal Leica DM IRE2 laser scanning microscope (CLSM; Leica Microsystems, Wetzlar, Germany) equipped with an Ar laser, using a 63 × 1.2 water objective and Leica TCS SL software (Leica Microsystems, Wetzlar, Germany). To ensure that peptide distribution within microgel particles was not influenced by a possible molecular weight polydispersity of the peptides, complementary studies were performed with the monodisperse peptides Lys 24, His 24, and Arg 24. Region of interest (ROI) analysis was performed on confocal images to evaluate the extent of peptide incorporation in microgels compared to the background (I/I0). Peptide Conformation. The R-helix content of positively charged homopolypeptides (pArg, pLys, pHis) in buffer solution, as well as after interaction with microgels, was evaluated by (34) Smith, P. K.; Krohn, R. I.; Hermanson, G. T.; Mallia, A. K.; Gartner, F. H.; Provenzano, M. D.; Fujimoto, E. K.; Goeke, N. M.; Olson, B. J.; Klenk, D. C. Anal. Biochem. 1985, 150, 76–85.

Bysell and Malmsten circular dichroism using a Jasco J-810 spectropolarimeter (Jasco, Tokyo, Japan). Ten scans in the range 210-250 nm were collected for each sample, using a quartz cuvette of 1 cm path length. Reference spectra for 100% R-helix and 100% random coil were obtained using pLys 58000 in 0.1 M HCl and 0.1 M NaOH, respectively.35 The R-helix content was calculated using the CDsignal recorded at 225 nm as reported earlier.36 The R-helix content of pArg, pLys, and pHis was analyzed at pH 5.5 and 7.0 using an initial peptide concentration of 100 mg/L and a 1:1 gel/peptide solution to ensure that no free peptides were left in solution. Duplicate measurements were performed at 20 °C.

Results and Discussion Binding. Figure 1a,b shows that the negatively charged (pGlu and pAsp) or uncharged (pPro and pThr) homopolypeptides do not cause any deswelling of negatively charged AA microgels. Instead, CLSM results indicate that these peptides are depleted from the microgels (I/I0 < 1) (Figure 1d). Positively charged pArg, pHis, and pLys, on the other hand, all caused significant deswelling of the oppositely charged microgel particles (Figure 1c). In analogy, neither positively charged (pLys, pArg, and pHis) nor uncharged (pPro and pThr) homopolypeptides caused any volume reduction of positively charged APTAC microgels, whereas negatively charged pAsp and pGlu did (Figure 2). Together, Figures 1 and 2 suggest that an attractive electrostatic driving force is a prerequisite for peptide-induced deswelling in the systems investigated. The deswelling response of AA microgels caused by oppositely charged homopolypeptides can be summarized to be in the order pLys > pArg > pHis, where pLys is able to deswell microgel particles faster and to a larger extent (Figure 1c). pHis has a lower degree of charge (∼0.7) at pH 5.5 than pLys and pArg (both have degree of charge ) 1), which contributes to a reduced microgel deswelling induced by the former peptide. However, even when taking into account the differences in degree of charge of these polypeptides, and comparing peptide-induced deswelling at an equal amount of charges (Figure S2a, Supporting Information), pHis still has the weakest impact on microgel deswelling. Furthermore, although pArg and pLys are both fully charged at the experimental conditions investigated (pH 5.5 and pH 7) (pKa for Lys and Arg being 10.8 and 12.5, respectively),37 the final deswelling ratios of the AA microgels are ∼4 times smaller for pLys compared to pArg (Figure 1). Figure 3 further shows that the deswelling rate increases with the bulk peptide concentration and that the difference in deswelling rates obtained between pArg and pLys is more significant at high peptide concentrations. Monodisperse peptides, Arg 24, Lys 24, and His 24, showed the same deswelling pattern (Figure S2, Supporting Information), and artifacts due to differences in the homopolypeptide molecular weight can therefore be excluded as causing the effects observed. Instead, some further light to the observed differences in deswelling response of AA microgels caused by various cationic peptides can be shed by looking at the peptide distribution in the microgel particles. When swollen in solution, the mesh size of the microgel network is in the order of ∼10 nm, while the radius of gyration of the polypeptides is estimated to be ∼1 nm. Peptide exclusion of the gel network meshes should therefore not be an issue initially. However, after binding of oppositely charged peptides, the gel network mesh size decreases, in some cases becoming sufficiently dense to prevent the peptide from reaching the microgel core. Figure 4 shows that both pHis and pArg display (35) Greenfield, N.; Fasman, G. D. Biochemistry 1969, 8, 4108–4116. (36) Sjo¨gren, H.; Ulvenlund, S. Biophys. Chem. 2005, 116, 11–21. (37) McKee, T.; McKee, J. Biochemistry - an introduction, 2nd ed.; WCB/ McGraw-Hill: New York, 1999; p 86.

Interactions in Peptide-Microgel Systems

Langmuir, Vol. 25, No. 1, 2009 525

Figure 1. Volume response of AA microgels on interaction with (a) negatively charged pGlu and pAsp, (b) uncharged pPro and pThr, at pH 7 and salt concentration 20 mM, and (c) positively charged pLys, pArg, and pHis homopolypeptides at pH 5.5 and salt concentration 20 mM. (d) Fluorescence intensity ratio I/I0 obtained from ROI analysis of CLSM images, where I is the average intensity in the entire gel particle and I0 is the background intensity. No detectable AA microgel deswelling was observed for pAsp and pPro also at pH 5.5 (results not shown).

Figure 2. Volume response of APTAC microgels on interaction with (a) negatively charged pGlu and pAsp, (b) uncharged pPro and pThr, and (c) positively charged pLys, pArg, and pHis homopolypeptides at pH 7.0 (5.5 for pHis) and ionic strength 20 mM.

a shell-like distribution, while pLys distributes more homogenously in the microgel particles. Even when decreasing the electrostatic interaction strength by increasing the salt concentration, the pArg distribution was still limited to the outermost surface layer of microgels (Figure S3, Supporting Information). The surface layer formation of pArg can also be visualized from light microscopy images (Figure S6, Supporting Information).

Again, the monodisperse peptides Arg 24, Lys 24, and His 24 display essentially the same distribution pattern (Figure S4, Supporting Information), demonstrating that differences in peptide distribution are not due to differences in peptide molecular weight. One exception from this is the more pronounced surface localization found for Lys 24 compared to pLys at pH 5.5, reflecting a higher tendency for the monodisperse peptide to

526 Langmuir, Vol. 25, No. 1, 2009

Figure 3. Concentration dependence of the deswelling rate constant, kapp, of pArg and pLys interacting with AA microgels at pH 7.0 and salt concentration 20 mM.

Figure 4. Representative CLSM images and corresponding intensity profiles through the middle section of gel particles showing the distribution of BODIPY-labeled pArg, pLys, and pHis at 100 mg/L for (a) pH 5.5 and (b) pH 7.0 at salt concentration 20 mM.

create a faster collapse of the gel-peptide diffusion front, thereby reducing the size limitation for peptides entering into the gel core at this condition, in analogy to the previously found heterogeneous distribution of pLys (10 kDa) at pH 4.5.29 The differences observed between pArg and pLys are not likely to be due to any difference between the two peptides in terms of linear charge density. Instead, it is likely that other structural differences between these cationic peptides are important for the interactions with oppositely charged AA gels, suggesting also nonelectrostatic interactions to be important for the finer details of interaction between different cationic peptides with the AA microgels. In analogy, nonelectrostatic properties of cationic drugs were previously shown to influence the interaction with anionic microgels, and a decreased uptake was observed for more hydrophobic substances.16 This illustrates the importance of nonelectrostatic interactions also in oppositely charged systems. In parallel, the difference in interaction strength of pHis compared

Bysell and Malmsten

to other cationic polypeptides may be related to the ability of this peptide to interact also hydrophobically through unprotonated side chains of the peptide at the experimental conditions investigated. This is supported by this peptide displaying significantly higher binding affinity to hydrophobic pSS compared to pLys and pArg (Figure S5a, Supporting Information). Furthermore, pArg and pLys are known to interact differently with oppositely charged ions due to differences in their structure. Thus, pArg contains a guanidinium moiety, known to form characteristic pairs of organized and strong hydrogen bonds with carboxylates.38 This guanidinium group is believed to be responsible for the increased binding strengths observed between different nucleotides and pArg,30,38 for the higher binding affinity for anionic surfactants of pArg compared to pLys,39 and for the increased ability of pArg to enter cells40 compared to other basic homopolypeptides. Although, in this present study, turbidometric titration could not detect any difference in binding affinity of pLys and pArg to linear pAA (Figure S5, Supporting Information), the differences observed between these peptides upon interaction with lightly cross-linked AA microgels could still possibly be reflected in the structural differences. At first glance, one might expect this stronger binding of pArg to lead to stronger deswelling response of the AA microgel particles than for pLys. However, the situation is more complex than this. Thus, when the interaction of microgel and oppositely charged peptide is sufficiently strong, gel-peptide complex initiated at the gel surface, “the skin layer”, is so condensed that it prevents further diffusion of peptides into the gel core, also slowing down the rate of microgel deswelling (Figure 1c). This is in analogy with CLSM results found in this study, displaying the heterogeneous distribution of pArg in AA microgels (Figure 4). Corresponding results were previously found for higher molecular weight pLys and/or pLys at a lower degree of microgel swelling at the point of peptide addition.29 It was also noted for cationic drugs binding to microgels with surface-localized functional groups.16 Despite the differences observed between cationic homopolypeptides in peptide-induced AA microgel deswelling, as well as in the resulting peptide distribution within the microgel particles, there is no significant difference between pArg, pLys, or pHis regarding the amount of peptide bound to the gel particles (Figure 5). Thus, the differences observed between the peptides in terms of distribution and peptide-induced microgel deswelling does not translate to differences in the extent of peptide binding. The origin of this effect may be due to the formation of the dense surface skin in the case of pArg preventing peptide binding to the entire microgel particles, thus balancing the higher affinity with a lower binding volume. In analogy to previous observations for pLys,29 peptide binding is lower (Figure 5a) at pH 5.5 than at pH 7.0, as a consequence of the lower charge density of the AA microgels at the lower pH. Quantitatively, the degree of charge of the gel network at pH 5.5 is ∼0.5,29 which corresponds well to the uptake being about half as high at this pH compared to pH 7.0. These findings therefore emphasize that peptide uptake to microgel particles in this system is mainly controlled by the charge of the gel network, and less so by the peptide charge at conditions of fully charged peptides (pH , pKa). The secondary structure in solution of the positively charged peptides investigated was predominantly of random coil con(38) Schug, K. A.; Lindner, W. Chem. ReV. 2005, 105, 67–113. (39) Esson, J. M.; Ramamurthy, N.; Meyerhoff, M. E. Anal. Chim. Acta 2000, 404(1), 83–94. (40) Mitchell, D. J.; Kim, D. T.; Steinman, L.; Fathman, C. G.; Rothbard, J. B. J. Pept. Res. 2000, 56, 318–325.

Interactions in Peptide-Microgel Systems

Figure 5. (a) Saturation uptake of peptides at pH 5.5 and pH 7.0. (/) indicates that no data could be obtained due to the limited solubility of pHis at this pH. (b) R-Helix content before and after binding to AA microgel particles at pH 5.5 and salt concentration 20 mM. (//) No detectable helix content.

formation (Figure 5b), in agreement with previous findings.41-44 On interaction with the oppositely charged microgels, helix induction occurs for both pArg and pLys. Quantitatively, however, the change in R-helix content upon gel binding was significantly higher for pArg than for pLys, and due to the higher binding strength of pArg effectively reducing the intramolecular electrostatic penalty to overcome in the helix formation process. These results correspond well to previous findings on the conformation of pArg in polyanion complexes,32,42 where it has been inferred that the carboxylate anion forms a doubly hydrogenbounded ring structure with the guanidinium group of pArg, and that this ring structure is more stable than the one occurring between the ammonium group of pLys and polyions such as poly(acrylic acid). Release. Results displayed in Figure 6 show that pHis molecules bound to AA microgels can be released through a change in pH of the external solution. With increasing pH, the charge degree of the gel network increases, thereby inducing microgel swelling. Simultaneously, the charge degree of pHis decreases drastically, causing peptide molecules to detach from the microgel network, in turn causing further gel reswelling (Figure 6a,b). However, when lowering the pH to 5.5 again, the microgel volume is still smaller than that before peptide incorporation (V/V0 < 1) (Figure 6c), clearly indicating that not all pHis was released from the microgels. This was also confirmed by CLSM experiments (Figure 6d). In contrast, pH-induced microgel swelling on raising pH after pLys and pArg binding at low pH was completely reversible, leading to the conclusion that only ionization of the gel network caused gel swelling on increasing pH after the initial peptide binding, and not the actual release of peptides (Figure S6, Supporting Information). Another approach that could, in principle, be used to release electrostatically bound proteins or peptides from microgels is to increase the salt concentration in the external solution. This was previously adopted for proteins such as cytochrome C bound to (41) Holzwarth, G.; Doty, P. J. Am. Chem. Soc. 1965, 87(2), 218–228. (42) Ichimura, S.; Mita, K.; Zama, M. Biopolymers 1978, 17, 2769–2782. (43) Robert, W.; McCord, E. W. B. J. W. L.M. Biopolymers 1977, 16(6), 1319–1329. (44) Peggion, E.; Cosani, A.; Terbojevich, M.; Scoffone, E. Macromolecules 1971, 4(6), 725–731.

Langmuir, Vol. 25, No. 1, 2009 527

Figure 6. (a) pHis-induced deswelling of AA microgels at pH 5.5 and salt concentration 20 mM and (b) gel reswelling at pH 9.5 induced by pH-induced release of preadsorbed pHis (salt concentration 20 mM). (c) Exemplifying light microscopy images showing gel particles before and after interaction with pHis, and after pH-induced pHis release. (d) CLSM images and corresponding intensity profiles showing the distribution of pHis before and after pH-induced release from a gel particle.

macroscopic poly(acrylic acid) gels.20 Although being of the same size range as these proteins, however, neither pLys nor pArg could be released from the presently investigated AA microgels by raising the salt concentration from 20 to 220 mM. Most likely, this is an effect of a high fraction of polypeptide segments being in close contact with the opposite microgel charges, resulting in a low probability of simultaneous detachment of all polypeptide segments, thus resulting in slow release kinetics. This is in analogy, for example, with exceedingly slow desorption of polypeptides from oppositely charged surfaces on increasing salt concentration for systems containing such polyelectrolytes.45 For globular and rigid proteins such as cytochrome C, on the other hand, such large fractions of close contacts are precluded due to structural constraints given by the protein structure, and hence, the overall number of contacts is lower, and electrolyteinduced desorption faster. In analogy, results from our group show that a smaller pLys (4.5 kDa) is partially released from the AA microgels by increasing the salt concentration (Supporting Information Figure S7), a consequence of a lower number of peptide-microgel contacts having to be detached simultaneously for peptide desorption to occur.

Conclusions Attractive electrostatic interaction is a prerequisite for peptide incorporation and gel deswelling to occur in the presently investigated microgel systems. The interaction strengths in these systems, influencing both peptide distribution and microgel deswelling, are however also affected by nonelectrostatic contributions, as demonstrated by the different abilities of different monodisperse cationic homopolypeptides to interact with the oppositely charged microgel. Results obtained in this investigation (45) Fleer, G. J.; Cohen Stuart, M. A.; Scheutjens, J. M. H. M.; Cosgrove, T.; Vincent, B. Polymers at Interfaces; Chapman & Hall: London, 1993.

528 Langmuir, Vol. 25, No. 1, 2009

further demonstrate the complexity of interactions occurring in such oppositely charged systems and that additional studies are needed to fully outline the details of the mechanisms involved. Furthermore, promising results were obtained demonstrating the possibility of pH-triggered release of histidine-rich protein and peptide drugs from anionic microgels. Acknowledgment. Dr. Per Hansson is acknowledged for fruitful discussions, as are Martin Andersson and Christian Johansson. This work was financed by the Swedish Foundation for Strategic Research.

Bysell and Malmsten

Supporting Information Available: Light microscopy images and results from CLSM showing the affinity of different fluorescent tags to microgels, deswelling results, and CLSM-images of monodisperse peptides Arg 24, Lys 24, and His 24, as well as pArg distribution within AA microgels at high and low salt concentration, turbidometric titration results, and light microscopy images showing the salt and pH cycling for attempted release of pLys and pArg from microgels and the gel volume ratios before and after salt-attempted pLys release. This material is available free of charge via the Internet at http://pubs.acs.org. LA8029984