当前位置:首页 >> 生物学 >>

Calcium Binding Peptide Motif from CalmodulinConferDivalent Selectivity to Elastin-Like Polypeptides


Article pubs.acs.org/Biomac

Calcium Binding Peptide Motifs from Calmodulin Confer Divalent Ion Selectivity to Elastin-Like Polypeptides
Wafa Hassouneh, Michelle L. Nunalee, M. Coleman Shelton, and Ashutosh Chilkoti*
Department of Biomedical Engineering, Campus Box 90281, Duke University, Durham, North Carolina 27708, United States
S Supporting Information *

ABSTRACT: Calcium-sensitive elastin-like polypeptides (CELPs) were synthesized by periodically interspersing a calcium-binding peptide sequence from calmodulin within an elastin-like polypeptide (ELP) with the goal of creating thermal and calcium responsive peptide polymers. The CELPs exhibit high sensitivity to calcium compared to monovalent cations but do not exhibit the exquisite selectivity for calcium over other divalent cations, such as magnesium, that is displayed by calmodulin. The CELPs were further used as a building block for the synthesis of calcium-sensitive nanoparticles by fusing a hydrophilic, noncalcium-sensitive ELP block with a CELP block that becomes more hydrophobic upon calcium binding. We show that addition of calcium at concentrations between 50 and 500 mM imparts su?cient amphiphilicity to the diblock polypeptide between 33 and 46 °C to trigger its self-assembly into monodisperse spherical micelles with a hydrodynamic radius of ?50 nm.

INTRODUCTION Elastin-like polypeptides (ELPs) are thermally responsive biopolymers consisting of the repeating amino acid sequence Val-Pro-Gly-X-Gly, where X, the guest residue, is any amino acid except proline.1?3 ELPs are soluble in water below a lower critical solution temperature (LCST) but become insoluble and aggregate above their LCST, referred to herein as their transition temperature (Tt). Given the ability to precisely engineer ELPs at the genetic level, additional functionality, beyond thermal responsiveness, can be directly embedded within the sequence through the choice of the guest residues,4,5 by fusing a peptide or protein at one of the termini of an ELP,6,7 or by periodically interspersing functional motifs between ELP repeats.8 Using these methods, we and others have previously engineered ELPs to respond to multiple triggers such as pH, light, and other extrinsic triggers.4,9,10 In this paper, we design an ELP with a new functionality by embedding multiple, periodic copies of a peptide motif that binds calcium between ELP repeats. This design was motivated by our goal of designing an ELP whose phase transition can be selectively triggered by calcium. Calcium is of interest as a trigger for biomedical applications because in vivo its compartmental concentration can vary by over 4-orders of magnitude. The extracellular concentration of calcium is maintained in the ?1?3 mM range, while the intracellular concentration of calcium is rigidly maintained between 50 and 200 nM.11,12 This large and spatially localized di?erence in calcium concentration could, in principle, provide a trigger that a?ects the phase transition of an ELP in vivo. Despite the importance of calcium as a potential trigger of stimulus responsive behavior, to our knowledge, there are no examples of calcium-sensitive thermally responsive polymers.
? 2013 American Chemical Society



Previously published work employed calcium for gelation or mineralization and did not systematically show the sensitivity or selectivity of the system to calcium. Alginate exhibits calciumsensitive gelation13 and has been incorporated into synthetic thermally responsive polymers,14,15 but alginate derived systems are intrinsically limited in their scope, as alginate is a naturally derived biopolymer that is isolated from sea weed,13 so that its conjugates cannot be engineered at the molecular level. Prieto et al. have engineered calcium phosphate binding sites into ELPs to create calcium phosphate-ELP precipitates as biomaterials for bone repair.8 This study was also motivated by a previous study, in which we synthesized a calmodulin-ELP fusion protein in which calmodulin acts as a calcium-binding allosteric actuator of the ELP transition.16 Calmodulin, one of the multitude of proteins that have evolved to respond to in vivo changes in calcium concentration, undergoes a large conformational change upon binding calcium.17 This large conformational change alters the surface properties of calmodulin, leading to a change in the Tt of a calmodulin-ELP fusion. To our knowledge, this is the ?rst example of the phase transition of a polymer that is triggered by a protein?ligand binding event. The transition is triggered by addition of CaCl2 at millimolar concentrations as opposed to the 103?104-fold greater concentrations required for other chloride salts to trigger the phase transition of ELPs.18 Furthermore, this behavior is highly selective for Ca2+ as compared to structurally similar divalent cations such as Mg2+. Building upon this study, herein we sought to examine whether
Received: April 2, 2013 Revised: May 21, 2013 Published: May 24, 2013
2347
dx.doi.org/10.1021/bm400464s | Biomacromolecules 2013, 14, 2347?2353

Biomacromolecules a minimal structural element, a calcium-binding loop of calmodulin, would be su?cient to impart calcium sensitivity to ELPs. To do so, we embedded multiple copies of a calciumbinding loop periodically into an ELP (herein termed CELP) and quanti?ed the calcium-triggered modulation of the ELP Tt. We also fused a CELP to another ELP block that is not calcium-sensitive, with the goal of tuning the amphiphilicity of the diblock ELP by the addition of calcium with the goal of triggering its self-assembly as a function of calcium concentration.

Article

ELP Nomenclature. ELPs are designated ELP[XaYbZc]-n where X, Y, and Z are the guest residues, a, b, and c are their relative ratios, and n is the number of total VPGXG repeats. For example, ELP[V1A8G7]96 consists of 96 repeating pentapeptides, 6 of which are VPGVG, 48 are VPGAG, and 42 are VPGGG. ELP[V]-90 consists of 90 repeating pentapeptides, all of which have the guest residue valine. CELPs are designated CELP[X-n]-m, where X-n denotes the guest residue and length of the ELP spacer that occurs after each calcium-binding sequence (for example, [V-10] means 10 repeats of VPGVG), and m denotes the number of times the calcium-binding sequence and ELP spacer monomer unit is repeated. For example, CELP[V-10]-8 has a monomer unit of one calcium binding sequence plus 10 repeats of VPGVG that is repeated 8 times. CELP block copolymers (CELPBCs) are designated CELP- n/m, where n donates the number of pentapeptide repeats in the hydrophilic ELP block with guest residue composition of V, A, and G at a ratio of 1:8:7, respectively, and m denotes the number of times the calcium-binding sequence and ELP[V-20] spacer monomer unit is repeated. All ELP and CELP sequences are listed in Table S1. Gene Synthesis. Four CELPs were synthesized using recursive directional ligation (RDL; Figure S1).3 The synthesis of CELP[V-10]8 by RDL is detailed in the Supporting Information. Block copolymer genes were obtained by the same procedure; ELP[V1A8G7] genes previously synthesized in our group were inserted into CELP vectors at the N-terminus of the CELP gene. Protein Expression and Puri?cation. Genes encoding CELPs or CELPBC were cloned into a modi?ed pET25b(+) expression vector, transformed into the BLR E. coli cell line, and grown in 1 L cultures using Terri?c Broth dry medium according to standard protocols. The CELPs were then puri?ed using inverse transition cycling (ITC) according to the method described by Meyer and Chilkoti19 with a few notable exceptions. While most ELPs are puri?ed in phosphatebu?ered saline (PBS), CELPs and CELPBC were puri?ed, stored, and characterized in a 10 mM HEPES bu?er with 140 mM NaCl at pH 7.4 to avoid the precipitation of calcium phosphate that occurs upon the addition of calcium to PBS. In addition, a pH-lowering step was used to help prevent impurities from binding to the acidic amino acids. Following the ?rst cold spin, the pH was lowered to 4.0 before the ?rst hot spin. The ELP pellets were resuspended in a 50 mM sodium succinate bu?er at pH 4.0 after the ?rst four hot spins. Before the fourth hot spin, the samples were ?ooded with a 20-fold excess of EDTA per calcium binding site to competitively bind any calcium ions. After the fourth hot spin, the pellets were resuspended in distilled deionized water to prepare for dialysis to remove all calcium ions. The samples were dialyzed against distilled, deionized water with a 10000 Da molecular weight cuto? to remove all traces of calcium and other salts. Purity was veri?ed by SDS-PAGE (Figure S2). Transition Temperature. The phase transition of CELPs and CELPBC was characterized by monitoring the absorbance of an aqueous solution at 350 nm (OD350) as a function of temperature (1 °C/min) on a UV?visible spectrophotometer equipped with a multicell thermoelectric temperature controller (Cary 300 Bio; Varian, Inc.). The Tt is de?ned as the temperature at the maximum of the derivative of OD350 with respect to temperature. Dynamic and Static Light Scattering. The temperaturetriggered self-assembly of CELPBC was monitored with dynamic
2348



EXPERIMENTAL SECTION

light scattering (DLS) using a DynaPro Plate Reader (Wyatt Technology Corporation, Santa Barbara, CA) to determine the CELPBC’s hydrodynamic radius (Rh). All CELPBC solutions were prepared in HEPES bu?er at a concentration of 10 μM and ?ltered through an Anotop 10 Whatman 20 nm ?lter. Samples were contained in a 384-well plate with a sample volume of 35 μL, and each sample was covered with a drop of mineral oil to prevent evaporation. DLS data were collected at a 90° angle (18 acquisitions, each 10 s) heated at 1 °C intervals from 20 to 60 °C. The autocorrelation function was analyzed using a cumulant and regularization algorithm for Rayleigh spheres provided by the manufacturer to determine Rh and mass fractions of the unimer, nanoparticles, and aggregate as the sample underwent its phase transition. Simultaneous SLS/DLS measurements were obtained for angles between 30° and 150° at 5° increments done in 3 runs each 15 s at each angle at 25 °C and 40 °C. Samples for the ALV/CGS-3 goniometer system were prepared by ?ltering solutions of 10 μM CELPBC with 500 mM CaCl2 through an Anotop 10 Whatman 100 nm ?lter into a 10 mm disposable borosilicate glass tube (Fischer). The glass tube was precleaned by washing three times with ?ltered ethanol (0.2 μm cellulose acetate ?lter).

RESULTS AND DISCUSSION CELP Design. The genetically encoded synthesis of ELPs provides precise control of their length, sequence and architecture. Herein, we sought to take advantage of these features to seamlessly embed calcium-binding peptide sequences into ELP sequences by recombinant DNA techniques to develop calcium-responsive ELPs. The CELP design incorporates a periodic calcium-binding peptide along the polypeptide backbone separated by ELP segments as illustrated in Figure 1.



Figure 1. Design of calcium-sensitive elastin-like polypeptides (CELPs). CELPs contain negatively charged calcium binding sequences spaced at regular intervals along the ELP backbone. The calcium ions bind to the calcium binding sequences, which partially neutralizes the charge, decreases the CELP’s transition temperature below solution temperature, and thereby isothermally triggers the phase transition of CELP.

The calcium-binding sequence was derived from the calciumbinding loop found in calmodulin (Figure S3). Of calmodulin’s four calcium-binding sites, sites I and II have dissociation constants (Kd) of ?10?5 M, while sites III and IV have slightly higher a?nity with a Kd of ?10?6 M.20,21 Among these four calcium-binding sites in calmodulin, site II, with the peptide sequence DADGNGTIDFPE, has the fewest overall charged residues and no positive charges. This peptide sequence was used in the CELPs because its charge pro?le potentially minimizes the contribution of electrostatics to the behavior of the CELPs. The CELPs consist of four or more copies of the site II calcium-binding sequence of calmodulin separated by the ELP spacer, roughly mimicking the primary sequence of calmodulin, and are listed in Table 1. The ELP chosen as a spacer for all
dx.doi.org/10.1021/bm400464s | Biomacromolecules 2013, 14, 2347?2353

Biomacromolecules Table 1. Sequences of the Three CELPsa
CELP CELP[V-10]-8 CELP[V-20]-4 CELP[V-20]-6
a

Article

amino acid sequence [VGDADGNGTIDFPEG?[VPGVG]10 ]8 [VGDADGNGTIDFPEG?[VPGVG]20]4 [VGDADGNGTIDFPEG?[VPGVG]20]6

molecular weight (kDa) 44.5 38.6 57.9

Underlined amino acids denote the calcium-binding sequence adapted from calmodulin.

Figure 2. (A) Transition temperature of 10 μM CELP[V-20]-4 and ELP[V]-90 at di?erent concentrations of CaCl2, MgCl2, and NaCl. (B) Transition temperatures of 10 μM CELP[V-20]-4 at salt concentrations between 0.1 and 10 mM, the physiological relevant range.

CELPs was VPGVG, because valine is a hydrophobic guest residue with a low Tt, so that the expected increase in the Tt due to the inclusion of charged residues from the calciumbinding peptides would still place the Tt of the CELP in a temperature range where it could be conveniently detected.3 Two variables were systematically varied in our design of CELPs: (1) the number of calcium-binding sequences within a ?xed CELP length, and (2) the overall CELP length. Two di?erent ELP spacer lengths were used, ELP[V-10] and ELP[V20] that consist of 10 repeats of VPGVG and 20 repeats of VPGVG between calcium binding sequences, respectively. CELP[V-10]-8 and CELP[V-20]-4 have the same total length of ELP (80 pentapeptides each), but CELP[V-10]-8 has twice as many calcium binding sequences (eight) as CELP[V-20]-4 (four). We varied the number of calcium-binding sequences, as we wished to examine its e?ect on the CELPs responsiveness to calcium. The e?ect of overall length of the CELP chain was examined for the ELP[V-20] spacer; four and six repeats of the calcium-binding sequence and ELP[V-20] spacer unit were synthesized and characterized with respect to their phase transition behavior. CELP Thermal Behavior. The transition temperatures of the CELPs were determined as a function of calcium concentration by measuring the optical density at 350 nm (OD350) and the Tt was determined to be the temperature at which the maximum of the derivative of OD350 with respect to temperature occurs (Figure S4). The Tt was also determined for a negative control, ELP[V]-90, which is an ELP with the same approximate amino acid length and the same guest residue composition as the ELP spacer in CELP, but is missing the calcium-binding sequence. CELP[V-20]-4 shows greatly increased sensitivity to calcium in the 1 to 100 mM range as indicated by the large decrease in Tt with increasing CaCl2 concentration as compared to the negative control ELP[V]-90 which shows no signi?cant change in Tt (Figure 2A). In addition, the transition temperatures of CELP[V-20]-4 are well above the Tt of ELP[V]-90. ELP[V]-90 has a Tt of 33 °C at an ELP concentration of 10 μM while CELP[V-20]-4 has a Tt of
2349

74 °C at a CELP concentration of 10 μM and no calcium. This 41 °C increase in Tt is caused by the negative charges contributed by the calcium-binding sequences; the four binding sequences in CELP[V-20]-4 contribute 16 amino acids that are negatively charged at pH 7.4. These negative charges greatly increase the overall hydrophilicity of the CELP molecule and cause the Tt to increase. With an increase in Ca2+ concentration, Cl? concentrations rise as well, necessitating deconvolution of the e?ect of calcium binding from the salting-out e?ect of Cl? as increasing the concentration of kosmotropic salt decreases the solubility of ELPs and enables their phase transition to be isothermally triggered.22 This behavior is qualitatively described by the Hofmeister series, which predicts the e?ect of di?erent cations and anions on protein, and ELP, solubility.23 The e?ect of salts on ELPs is similar to their e?ect on protein solubility, with the exception that ELPs are far more sensitive to salt than most proteins. For example, while it takes several M of (NH4)2SO4 to “salt out” proteins, ELPs will phase separate in aqueous solution and aggregate by the addition of only several hundred mM of (NH4)2SO4.16 Hence, to delineate the speci?c e?ect of calcium binding on ELPs from a nonspeci?c salting out e?ect, we compared the CELPs with equivalent ELPs that do not contain the calcium-binding sequence. In addition, we further determined the sensitivity and selectivity of CELPs by comparing their behavior in the presence of calcium versus other divalent and monovalent cations. The Tt’s of CELP[V-20]-4 and ELP[V]-90 were determined from their temperature-dependent turbidity pro?les for a range of MgCl2, CaCl2, and NaCl concentrations (Figure 2A). As mentioned previously, CELP[V-20]-4 is signi?cantly more sensitive to calcium at a concentration below 100 mM compared to ELP[V-90], as indicated by the large decrease in the CELP[V-20]-4 Tt as opposed to no change in the ELP[V]90 Tt with increasing CaCl2 concentrations below 100 mM. The observed e?ect of Ca2+ is therefore clearly due to the presence of calcium-binding sequences in CELP as opposed to a nonspeci?c salting out e?ect. However, above 100 mM, a
dx.doi.org/10.1021/bm400464s | Biomacromolecules 2013, 14, 2347?2353

Biomacromolecules decrease in ELP[V]-90 Tt is observed with increasing CaCl2 concentration, which is due to the e?ect of Cl?, as predicted by the Hofmeister series, thus, indicating that for CELP[V-20]-4 salting-out e?ects become relevant above 100 mM. ELP[V]-90 shows no selectivity for Na+, Mg2+, or Ca2+ in terms of their e?ect on its Tt, whereas each cation has a di?erent e?ect on the Tt of CELP[V-20]-4. At a salt concentration between 0.1 and 10 mM (Figure 2B), CELP[V-20]-4 is highly sensitive to divalent cations such as calcium and magnesium, but not to a monovalent cation, such as sodium. We compared the sensitivity of CELP[V-20]-4 to calcium and magnesium by the slope of Tt versus the log of salt concentration. The slopes for the two divalent ions are similar, suggesting little selectivity toward calcium as compared to magnesium. However, at all salt concentrations, the e?ect of Ca2+ is larger than Mg2+ in decreasing the Tt of CELP, so that calcium is somewhat more e?ective than magnesium in interacting with the charged residues within the calciumbinding sequences and thereby decreasing the Tt. The Tt of CELP[V-20]-4 at di?erent CELP concentrations and calcium concentrations is shown in Figure 3A. With no CaCl2 added and at a CELP concentration of 10 μM, the Tt of CELP[V-20]-4 is 74 °C. By adding just 1 mM CaCl2, the Tt drops to 62 °C, a 12 °C di?erence. At the increased CELP concentration of 100 μM, the Tt with no CaCl2 added drops to 60 °C and upon the addition of 1 mM CaCl2 drops to 54 °C, a

Article

Figure 3. (A) Transition temperature (Tt) as a function of CaCl2 concentration for CELP[V-20]-4 at concentrations of 10, 25, 50, and 100 μM in HEPES bu?er. (B) Transition temperature as a function of CaCl2 concentration for 10 μM CELP[V-10]-8, CELP[V-20]-4, and CELP[V-20]-6 in HEPES bu?er.
2350

6 °C di?erence. This indicates that the sensitivity to calcium decreases with increasing CELP concentrations for calcium concentrations below 1 mM. This decrease in sensitivity could be a simple e?ect of reducing the ratio of calcium to binding sites available, thereby shifting the equilibrium, leaving some calcium sites unbound in the CELP at the higher CELP concentrations. A comparison of the transition temperatures of three di?erent CELPs is shown in Figure 3B. An increase in the total length of the CELP by adding two additional repeat units to obtain CELP[V-20]-6 lowers the Tt by ?5 °C as compared to CELP[V-20]-4 which follows the expected trend for single segment ELPs, that is the ELP chain length inversely correlates with its Tt. CELP[V-10]-8 and CELP[V-20]-4 have the same total length of ELP (80 pentapeptides each), but CELP[V-10]8 has twice as many calcium binding sequences (eight) as CELP[V-20]-4 (four). The increased number of negative charges contributed by the additional calcium binding sequences in CELP[V-10]-8 cause a signi?cant increase in its Tt compared to that of CELP[V-20]-4. Because of the larger number of calcium binding sites, a calcium concentration of 10 mM must be added before the Tt of CELP[V-10]-8 is within the maximum temperature of ? 95 ° C that the UV spectrophotometer can reach, while for CELP[V-20]-4, the Tt is below the maximum temperature, even in the absence of calcium. Furthermore, the sensitivity of CELP[V-10]-8 to calcium is greater than that of CELP[V-20]-4 as indicated by the slope of Tt vs log of CaCl2 concentration for CELP[V-10]8. This e?ect is consistent with the larger number of calciumbinding sites within the CELP[V-10]-8 sequence, as compared to CELP[V-20]-4. CELP Diblock Copolymers. Several examples of diblocks formed from an ELP with a hydrophobic block with a low Tt (Tt1) and hydrophilic block with a high Tt (Tt2) have been shown to form spherical micelles, while other block copolymer designs have also been shown to form vesicles.24?27 Above a critical micellization temperature (CMT), ELP diblocks selfassemble into spherical micelles in which the hydrophobic block forms the core of the micelle while the hydrophilic block forms the corona, presumably because the more hydrophobic block (Tt1 block) preferentially desolvates with an increase in temperature. We synthesized an ELP diblock (CELPBC) with a hydrophobic: hydrophilic block length ratio of 1:1, with a CELP as the hydrophobic block to sensitize the ELP’s selfassembly to calcium. At low calcium concentrations, both blocks are hydrophilic and the CELPBC is soluble. We hypothesized that as the calcium concentration is increased, binding of Ca2+ to the negatively charged residues in the CELP block should shield the negative charges and increase the hydrophobicity of this block, so that at a critical calcium concentration, the newly acquired amphiphilicity of the diblock drives self-assembly of the CELPBC into micelles (or other selfassembled nanomeso scale structures). A schematic of a CELPBC and its self-assembly into spherical micelles triggered by calcium binding is shown in Figure 4. The hydrophobic segment in our design of the CELPBC is a CELP[V-20]-6 segment. To impart su?cient amphiphilicity to a CELPBC upon calcium-binding, the hydrophilic segment of the diblock was chosen to be ELP[V1A8G7]-128, a 128 pentapeptide ELP composed of guest residues valine, alanine, and glycine at a ratio of 1:8:7. This ELP has a high Tt; a concentration of 25 μM, this segment has a Tt of 78 °C. This CELPBC, composed of the ELP[V1A8G7]-128 block fused to CELP[V-20]-6, has a total
dx.doi.org/10.1021/bm400464s | Biomacromolecules 2013, 14, 2347?2353

Biomacromolecules

Article

Figure 4. Schematic illustrating micelle formation by CELP block copolymers consisting of one CELP block and one hydrophilic, high Tt ELP block. When the calcium concentration is increased, calcium binding to the CELP block depresses its Tt leading to self-assembly into spherical micelles.

molecular weight of 106 kDa, and will be referred to herein as CELP-128/6. Nanoparticle formation by CELPBC’s was successfully demonstrated by temperature-programmed turbidimetry measurments and DLS as a function of solution temperature.

Turbidity measurements for 10 μM CELP-128/6 are shown in Figure 5A with increasing concentration of calcium. In the absence of calcium, CELP-128/6 transitions from a unimer to a polydisperse, micrometer size aggregate at a Tt of 55 °C. With the addition of 10 mM CaCl2, the CELP-128/6 undergoes a transition from unimer to aggregates at a lower Tt of 47 °C, but no self-assembly into nanoparticles is observed. At calcium concentrations of 50 mM and greater, however, evidence of nanoparticle formation at intermediate temperatures can be seen in the turbidity pro?les, as in our experience, a region of low OD350 (?0.1?0.5) preceding aggregation is indicative of nanoparticle formation. At 50 mM CaCl2, the nanoparticle region extends from 41 to 45 °C. The temperature range over which nanoparticles form increases with increasing CaCl2 concentrations, as at 500 mM CaCl2 nanoparticles are observed from 32 to 49 °C. This extension of the micelle formation temperature range with increasing CaCl2 concentrations can be attributed to two e?ects: (1) an increase in the calcium bound to the CELP block thereby increasing the block’s hydrophobicity and (2) the salting-out e?ect of Cl?. To separate the speci?c e?ect of Ca2+ binding from that of a nonspeci?c salting out e?ect, we examined the thermal behavior of a control ELP diblock that lacks the Ca2+ binding sequence, as a function of CaCl2 (Figure S5). This ELP diblock is composed of a hydrophilic block of the same composition as CELPBCs, ELP[V1A8G7]-128, and an ELP[V]-60 hydrophobic block, and has been previously shown to self-assemble into micelles with an increase in temperature.25 The temperature at which

Figure 5. (A) Turbidity measurements for 10 μM CELP-128/6 as a function of temperature at increasing concentrations of CaCl2. Turbidity pro?les were obtained by monitoring optical density at 350 nm in HEPES bu?er as the solution was heated at a rate of 1 °C/min. The increase in optical density indicates formation of ELP nanoparticles larger than an ELP unimer. (B) Hydrodynamic radius (Rh) and turbidity as a function of temperature for 10 μM CELP-128/6 in HEPES bu?er with 500 mM CaCl2. (C) Turbidity measurements for 10 μM CELP-128/6 as a function of temperature in HEPES bu?er with increasing amounts of NaCl and (D) MgCl2.
2351
dx.doi.org/10.1021/bm400464s | Biomacromolecules 2013, 14, 2347?2353

Biomacromolecules the unimer self-assembles into micelles decreases with increasing salt concentrations (i.e., the temperature at which a small increase in OD350 is ?rst observed). Therefore, the extension of nanoparticle formation temperature range for CELPBCs could be a result of the e?ect of Cl? on Tt as well as calcium binding. While turbidity measurements suggest the formation of nanoparticles, DLS measurements are necessary to con?rm their presence. The hydrodynamic radius (Rh) of CELP-128/6 was determined by DLS as a function of temperature. Figure 5B shows a comparison of Rh with the turbidity measurements for 10 μM CELP-128/6 with 500 mM CaCl2 as a function of solution temperature. From 20 to 32 °C, the Rh is ?8 nm, which corresponds to the size expected for a soluble 106 kDa polypeptide with a largely random coil structure. From 33 to 46 °C, the Rh is between 45 and 49 nm. At temperatures above 47 °C, micrometer-sized aggregates form. These results correspond closely with the turbidity measurements. In addition, the di?usion coe?cient (Dapp) was measured at scattering angles from 30° to 150° (Figure S6A). The change in Dapp with scattering angle was found to be insigni?cant, suggesting the nanoparticles formed are monodisperse and spherical. The radius of gyration (Rg) was determined from the slope of the Zimm plot obtained from static light scattering and was found to be 23.8 nm giving a ratio of Rg/Rh of 0.5, indicating a spherical particle topology, consistent with the formation of spherical micelles (Figure S6B). The aggregation number was determined by dividing the y-intercept of the Zimm plot obtained for the unimer measured at 25 °C by that for the micelle measured at 40 °C. These micelles were found to have an aggregation number of 155 unimers/micelle. It is interesting to note that while large changes in Tt are observed in single block CELPs with the addition of small amounts of calcium, higher amounts of calcium must be added to a diblock in solution before nanoparticles form. The nanoparticles require such a high calcium level to form because in the absence of calcium, both blocks are quite, and similarly, hydrophilic, as seen by their Tts, the value of which correlates with the mean hydropathy of the ELP. The Tt of CELP[V-20]6 at 10 μM in HEPES bu?er is ?70 °C and is close to the Tt of ELP[V1A8G7]-128. To drive self-assembly of the CELP-128/6 diblock ELP formed by fusion of these two blocks, enough calcium must be added to selectively drive binding of Ca2+ to the CELP block to increase the hydrophobicity of this block relative to ELP[V1A8G7]-128 block. Micelles begin to form when the polymer attains a critical level of amphiphilicity that is imparted by the selective binding of Ca2+ to the CELP block and by the selective desolvation of this block with an increase in temperature. For example, for CELP-128/6 at a solution concentration of 10 μM, 100 mM CaCl2 has to be added and the solution temperature raised above 38 °C (the CMT of this diblock at 10 μM) to trigger its self-assembly into spherical micelles. These results also suggest that to create CELPBCs that are more sensitive to calcium, the hydrophobicity of the CELP block should be increased by incorporating more hydrophobic residues at the guest position, so that self-assembly can be triggered at a lower calcium concentration and lower temperature. Calcium Selectivity of CELPBC. We next investigated the selectivity and sensitivity of CELPBC to calcium relative to magnesium and sodium. Figure 5 shows turbidity pro?les for 10 μM CELP-128/6 with increasing amounts of three di?erent salts. NaCl, does not trigger nanoparticle formation even at a
2352

Article

concentration of 1.5 M at any temperature between 20 °C and 70 °C but instead transitions from unimer to aggregate at 24 °C (Figure 5C). MgCl2, in contrast, triggers nanoparticle formation, but only at a concentration of ≥100 mM (Figure 5D), while CaCl2 triggers nanoparticle formation at a lower 50 mM concentration (Figure 5A). These results demonstrate that the self-assembly of CELP-128/6 is divalent ion selective and that the self-assembly shows a 2-fold selectivity for calcium as compared to magnesium.

CONCLUSION Calcium-sensitive elastin-like polypeptides (CELPs) were designed as thermally sensitive calcium-binding calmodulin analogs with the goal of generating thermal and calciumresponsive peptide polymers. These arti?cial polypeptides consist of ELP spacer segments separated at regular intervals by the calcium-binding amino acid sequence found at the second calcium-binding site of calmodulin. CELPs are extremely sensitive to Ca2+ compared to monovalent cations. They do not, however, retain selectivity for calcium over other divalent cations such as Mg2+ that is displayed by calmodulin, indicating that the tertiary structure of calmodulin is critical to the selectivity displayed by calmodulin for Ca2+ as compared to its closest divalent homologue, Mg2+. The CELPs were further used to engineer calcium sensitive nanoparticles by synthesis of a diblock polypeptide composed of a hydrophilic noncalcium sensitive ELP block and a CELP block that becomes more hydrophobic upon binding calcium. We show that addition of calcium at concentrations between 50?500 mM imparts su?cient amphiphilicity to the diblock polypeptide between 33 and 46 °C to trigger its self-assembly into monodisperse spherical micelles with a hydrodynamic radius of ?50 nm. Future design of CELPs to enhance their calcium sensitivity will explore other calcium binding sequence with a higher calcium a?nity and lower net charge, as well as increasing the hydrophobicity of the ELP segments in the calcium binding domain.21,28?34



■ ■ ■ ■

ASSOCIATED CONTENT

S Supporting Information *

Additional data on temperature-programmed turbidimetry and light scattering data and analysis is available. This material is available free of charge via the Internet at http://pubs.acs.org.

AUTHOR INFORMATION

Corresponding Author Notes

*E-mail: chilkoti@duke.edu The authors declare no competing ?nancial interest.

ACKNOWLEDGMENTS The authors would like to acknowledge ?nancial support from NIH Grants GM61232 and EB001630 and support from NSF’s Triangle MRSEC DMR-1121107 REFERENCES
(1) Urry, D. W. Prog. Biophys. Mol. Biol. 1992, 57 (1), 23?57. (2) Urry, D. W.; Luan, C. H.; Parker, T. M.; Gowda, D. C.; Prasad, K. U.; Reid, M. C.; Safavy, A. J. Am. Chem. Soc. 1991, 113 (11), 4346? 4348. (3) Meyer, D. E.; Chilkoti, A. Biomacromolecules 2002, 3 (2), 357? 367.
dx.doi.org/10.1021/bm400464s | Biomacromolecules 2013, 14, 2347?2353

Biomacromolecules
(4) Callahan, D. J.; Liu, W. E.; Li, X. H.; Dreher, M. R.; Hassouneh, W.; Kim, M.; Marszalek, P.; Chilkoti, A. Nano Lett. 2012, 12 (4), 2165?2170. (5) Trabbic-Carlson, K.; Setton, L. A.; Chilkoti, A. Biomacromolecules 2003, 4 (3), 572?580. (6) Trabbic-Carlson, K.; Meyer, D. E.; Liu, L.; Piervincenzi, R.; Nath, N.; LaBean, T.; Chilkoti, A. Protein Eng. Des. Sel. 2004, 17 (1), 57?66. (7) Massodi, I.; Bidwell, G. L.; Raucher, D. J. Controlled Release 2005, 108 (2?3), 396?408. (8) Prieto, S.; Shkilnyy, A.; Rumplasch, C.; Ribeiro, A.; Arias, F. J.; Rodriguez-Cabello, J. C.; Taubert, A. Biomacromolecules 2011, 12 (5), 1480?1486. (9) Alonso, M.; Reboto, V.; Guiscardo, L.; Mate, V.; RodriguezCabello, J. C. Macromolecules 2001, 34 (23), 8072?8077. (10) Strzegowski, L. A.; Martinez, M. B.; Gowda, D. C.; Urry, D. W.; Tirrell, D. A. J. Am. Chem. Soc. 1994, 116 (2), 813?814. (11) Carafoli, E. Annu. Rev. Biochem. 1987, 56, 395?433. (12) Barry, W. H.; Bridge, J. H. B. Circulation 1993, 87 (6), 1806? 1815. (13) Kuo, C. K.; Ma, P. X. Biomaterials 2001, 22 (6), 511?521. (14) Shi, J.; Alves, N. M.; Mano, J. F. Macromol. Biosci. 2006, 6 (5), 358?363. (15) Wang, J.-Y.; Jin, Y.; Xie, R.; Liu, J.-Y.; Ju, X.-J.; Meng, T.; Chu, L.-Y. J. Colloid Interface Sci. 2011, 353 (1), 61?68. (16) Kim, B.; Chilkoti, A. J. Am. Chem. Soc. 2008, 130 (52), 17867? 17873. (17) Finn, B. E.; Evenas ? , J.; Drakenberg, T.; Waltho, J. P.; Thulin, E.; Forsen ? , S. Nat. Struct. Biol. 1995, 2 (9), 777?783. (18) Cho, Y. H.; Zhang, Y. J.; Christensen, T.; Sagle, L. B.; Chilkoti, A.; Cremer, P. S. J. Phys. Chem. B 2008, 112 (44), 13765?13771. (19) Meyer, D. E.; Chilkoti, A. Nat. Biotechnol. 1999, 17 (11), 1112? 1115. (20) Wang, C. L. A. Biochem. Biophys. Res. Commun. 1985, 130 (1), 426?430. (21) Strynadka, N. C. J.; James, M. N. G. Annu. Rev. Biochem. 1989, 58, 951?998. (22) Cho, Y.; Zhang, Y.; Christensen, T.; Sagle, L. B.; Chilkoti, A.; Cremer, P. S. J. Phys. Chem. B 2008, 112, 13765?13771. (23) Hofmeister, F. Arch. Exp. Pathol. Pharmakol. 1888, 25 (1), 1? 30. (24) Lee, T. A. T.; Cooper, A.; Apkarian, R. P.; Conticello, V. P. Adv. Mater. (Weinheim, Ger.) 2000, 12 (15), 1105?1110. (25) Dreher, M. R.; Simnick, A. J.; Fischer, K.; Smith, R. J.; Patel, A.; Schmidt, M.; Chilkoti, A. J. Am. Chem. Soc. 2008, 130 (2), 687?694. (26) Simnick, A. J.; Lim, D. W.; Chow, D.; Chilkoti, A. Polym. Rev. 2007, 47 (1), 121?154. (27) Martin, L.; Castro, E.; Ribeiro, A.; Alonso, M.; RodriguezCabello, J. C. Biomacromolecules 2012, 13 (2), 293?298. (28) Bertini, I.; Sigel, A.; Sigel, H. Handbook on Metalloproteins; Marcel Dekker: New York, 2001. (29) Stuart, D. I.; Acharya, K. R.; Walker, N. P. C.; Smith, S. G.; Lewis, M.; Phillips, D. C. Nature 1986, 324 (6092), 84?87. (30) Kim, S. B.; Lim, J. W. Asian-Australas. J. Anim. Sci. 2004, 17 (10), 1459?1464. (31) Kroger, N.; Bergsdorf, C.; Sumper, M. EMBO J. 1994, 13 (19), 4676?4683. (32) Hofmann, S. L.; Goldstein, J. L.; Orth, K.; Moomaw, C. R.; Slaughter, C. A.; Brown, M. S. J. Biol. Chem. 1989, 264 (30), 18083? 18090. (33) Fliegel, L.; Ohnishi, M.; Carpenter, M. R.; Khanna, V. K.; Reithmeier, R. A. F.; Maclennan, D. H. Proc. Natl. Acad. Sci. U.S.A. 1987, 84 (5), 1167?1171. (34) Zhao, H.; Waite, J. H. J. Biol. Chem. 2006, 281 (36), 26150? 26158.

Article

2353

dx.doi.org/10.1021/bm400464s | Biomacromolecules 2013, 14, 2347?2353


相关文章:
更多相关标签: