Promoting Self-Assembly of Collagen-Related Peptides into Various

Article pubs.acs.org/Langmuir

Promoting Self-Assembly of Collagen-Related Peptides into Various Higher-Order Structures by Metal−Histidine Coordination Wei Hsu, Yi-Lun Chen, and Jia-Cherng Horng* Department of Chemistry, National Tsing Hua University, Hsinchu, Taiwan 30013, R.O.C. Frontier Research Center on Fundamental and Applied Sciences of Matters, National Tsing Hua University, Hsinchu, Taiwan 30013, R.O.C. S Supporting Information *

ABSTRACT: Collagen is an important and widely used biomaterial and therapeutic. The construction of large-scale collagen structures via the self-assembly of small collagenrelated peptides has been extensively studied in the past decade. Here, we report a highly effective and simple means to assemble small synthetic collagen-related peptides into various higher-order structures by utilizing metal−histidine coordination. In this work, two short collagen-related peptides in which histidine residues were incorporated as metal binding sites were designed and chemically synthesized: HG(PPG)9GH (X9) and HG(PPG)4(PHG)(PPG)4GH (PHG). Circular dichroism measurements indicated that these two peptides form only marginally stable collagen triple helices but that their stability can be increased upon the addition of metal ions. Dynamic light scattering analyses, turbidity measurements, TEM, and SEM results demonstrated the metal ion-dependent selfassembly of X9 and PHG into supramolecular structures ranging from various nanofibrils to microscale spherical, laminated, and granulated assemblies. The topology and size of these higher-order structures depends both on the metal ion identity and the location of the binding sites. Most intriguingly, the assembled fibrils show similar D-periodicity to that of natural collagen. Our results demonstrate that metal−histidine coordination can serve as an effective force to induce the self-assembly of unstable collagen-related peptides into higher-order structures.

■

INTRODUCTION Biologically derived and synthetic materials that mimic or are compatible with natural extracellular matrices (ECMs) have been extensively studied in regenerative medicine and tissue engineering.1 In particular, collagen has increasingly drawn attention as an important biomaterial because collagen exists as the most abundant protein in animals and the major component of ECMs.2−4 Although natural collagen is widely used in biomedical therapy, there are still some problems and limitations. For example, it is relatively difficult to introduce specific sequence modifications on natural collagen. Therefore, the search for an effective way to assemble synthetic and short collagen-related peptides (CRPs) into higher-order structures has emerged as an important topic in the field of biomaterials research.5,6 In the past decade, a variety of approaches using covalent and noncovalent interactions have been used to aid the assembly of CRPs. Most of these strategies used noncovalent forces including electrostatic interactions,7 cation−π interactions,8 aromatic π−π stacking,9,10 aromatic−proline interactions,11 and metal−ligand coordination,12−17 whereas some of them involved covalent interactions, such as cysteine knots,18−21 native chemical ligation,22 and star PEG polymer conjugation.23 In addition, recent work demonstrated the © 2012 American Chemical Society

efficacy of salt-bridge hydrogen bonds to propagate the selfassembly of CRPs into nanofibres and a hydrogel.24 Of the assembly strategies mentioned above, metal−ligand coordination typically induces much larger structures than other methods.5 A recent study also showed that His-tagged proteins can be decorated on the surface or incorporated into the core of metal-assembled collagen structures.25 However, previously reported metal−ligand coordination-assisted CRP self-assembly methodologies require the incorporation of organic ligands in peptides, which may introduce additional synthetic steps and complicate peptide preparation.12−17 Thus, an approach to metal-assisted CRP self-assembly that does not require extra ligand modifications would be beneficial. As a first step to achieve this goal, we designed and synthesized a short Pro-Pro-Gly (PPG) repeat CRP, HG(PPG)7GH, in which histidine residues are incorporated at the termini. Circular dichroism (CD) measurements showed that metal ions induce the folding of HG(PPG)7GH into a triple-helical conformation, indicating that metal−histidine coordination can drive collagen triple helix formation. We further prepared two PPG repeat Received: November 5, 2011 Revised: January 11, 2012 Published: January 14, 2012 3194

dx.doi.org/10.1021/la204351w | Langmuir 2012, 28, 3194−3199

Langmuir

Article

Technologies) by solid phase methods. The peptides were synthesized on a 0.1 mmol scale using the Fmoc-Pro-Pro-GlyOH tripeptide and Fmoc-protected amino acids, HBTUmediated coupling, and standard reaction cycles. Use of a Rink amide resin with a MBHA linker generated an amidated Cterminus following the cleavage. Upon the completion of peptide synthesis, the N-terminus of each peptide was acetylated using acetic anhydride. Cleavage of the peptides from the resin and removal of side-chain protecting groups was achieved with a 3.0 mL solution of 95% trifluoroacetic acid (TFA), 2.5% triisopropylsilane (TIS), and 2.5% H2O (v/v). The crude peptides were precipitated by adding cold methyl tert-butyl ether and purified by reversed phase (RP) HPLC using a Thermo BioBasic semipreparative C18 column (250 × 10 mm, 5 μm particle size) with a linear gradient of acetonitrile and water containing 0.1% (v/v) TFA. The gradient was 0 to 60% acetonitrile over 60 min and the flow rate was 3.0 mL/min. The peptides were more than 90% pure according to the HPLC analysis using a Thermo BioBasic analytical C18 column (250 × 4.6 mm, 5 μm particle size). The calculated and observed molecular masses were HG(PPG)7GH [M + H]+: 2206.09 (Calcd), 2206.13 (observed), X9 [M + Na]+: 2730.33 (Calcd), 2729.62 (observed), and PHG [M + Na]+: 2770.34 (Calcd), 2769.63 (observed). Circular Dichroism Spectroscopy. Circular dichroism (CD) measurements were performed on an AVIV Model 410 spectrometer using quartz cuvettes with a path length of 1 mm. For HG(PPG)7GH, the peptide−metal solutions were prepared in pH 8.0 buffer containing 20 mM sodium phosphate and 1 mM Cu2+ or Ni2+ to obtain a peptide concentration of 230 μM. For X9 and PHG, the peptides (400 μM) were dissolved in pH 7.4 and 20 mM sodium phosphate buffer and mixed with various metal ions (160 μM). All peptide solutions were incubated at 4 °C for at least 24 h before CD measurements to permit the formation of collagen triple helices. Thermal unfolding curves were obtained by monitoring the CD signals at 227 nm for HG(PPG)7GH, and at 225 or 226 nm for X9 and PHG in a 1 mm path length cuvette with a bandwidth of 1 nm. The CD signal was recorded every 2 °C and the heating rate was 0.4 °C/min. The Tm value was obtained from the first derivative of the ellipticity versus temperature curve (Supporting Information for details). Turbidity Measurements. Turbidity experiments were conducted on a JASCO V-630 spectrophotometer by measuring the absorbance at 313 nm and 4 °C. The samples were prepared in 20 mM phosphate buffer at pH 7.4 with a peptide concentration of 400 μM and a metal concentration of 160 μM. For chelation competition experiments, the turbidity was measured at ambient temperature after incubating the peptide/ Zn2+ solution at 4 °C for more than 24 h and followed by adding 1.6 mM ethylenediaminetetraacetic acid (EDTA). Dynamic Light Scattering Spectroscopy. Dynamic light scattering (DLS) experiments were performed on a 90-Plus nanoparticle size analyzer (Brookhaven Instruments) equipped with a 660 nm laser at a 90° scattering angle and a temperature controller. The samples containing 400 μM peptide and 160 μM metal ions were measured using 50 μL plastic cuvettes. All samples were incubated at 4 °C for at least 2 days before DLS analysis. The measurements were conducted at 26 °C in pH 7.4 20 mM phosphate buffer. The hydrodynamic diameters were obtained by analyzing intensity autocorrelation functions with 90-Plus particle sizing software (Brookhaven Instruments). For data analysis, a viscosity value (η) of 0.890 cP and a refractive index (n) of 1.330 were used for the buffer.

Figure 1. Illustration of a PPG triplet and the self-assembly of PPG repeat collagen-related peptides (X9 and PHG) via metal−histidine coordination. Histidine residues are shown in ball-and-stick format.

CRPs, HG(PPG)9GH (X9), and HG(PPG)4(PHG)(PPG)4GH (PHG), to study if we can assemble CRPs into fibrils and even higher-order structures by simply implanting metal−histidine coordination. As shown in Figure 1, we anticipated that metal− histidine coordination could promote both end-to-end and radial assembly of these two CRPs. Dynamic light scattering (DLS) analysis and electron microscopy showed that these two peptides do indeed self-assemble into large-scale structures upon the addition of metal ions. Furthermore, the metal− histidine coordination-assisted assembled fibrils from these peptides exhibit very clear D-periodicity. In summary, we report a metal−histidine coordination strategy for CRP self-assembly into nanofibrils and microscale structures.

■

EXPERIMENTAL METHODS General Materials and Instrumentation. All Fmocprotected peptides were purchased from Advanced ChemTech or Chem-Impex International. Rink amide MBHA resin was obtained from Novabiochem. Piperidine, triisopropylsilane (TIS) and trifluoroacetic acid (TFA) were purchased from Alfa Aesar. 4-Methylmorpholine (NMM) and acetic anhydride (Ac2O) were purchased from Sigma-Aldrich. N-Methyl-2pyrrolidone (NMP), dimethylformamide (DMF), acetonitrile, and methyl tert-butyl ether were purchased from Mallinckrodt Baker. Cobalt(II) sulfate heptahydrate, copper(II) chloride dihydrate, nickel(II) chloride hexahydrate, and zinc(II) chloride were the metal ion source and were purchased from SigmaAldrich. These materials were directly used without further purification. The Fmoc-Pro-Pro-Gly-OH tripeptide was synthesized by the method described previously.26 NMR spectroscopy was conducted with a Varian INOVA-500 spectrometer at National Tsing Hua University Instrumentation Center. ESIMS mass spectra were obtained using a Q-TOF LC/MS/MS (Micromass Inc.) spectrometer at National Chiao Tung University Precious Instrument Center. MALDI-TOF mass spectra of the peptides were obtained on an Autoflex III Smartbeam LRF200-CID spectrometer (Bruker Daltonics). TEM images were acquired using a JEOL JEM-1400 TEM at National Chung Hsing University Instrumentation Center. Peptide Synthesis and Purification. The peptide sequences used in this study are listed in Table S1 of the Supporting Information. All of the peptides were synthesized on a PS3TM 3-channel serial peptide synthesizer (Protein 3195

dx.doi.org/10.1021/la204351w | Langmuir 2012, 28, 3194−3199

Langmuir

Article

Table 1. Melting Temperatures (Tm) for Collagen-Related Peptides and Their Self-Assembly Times in the Presence of Specified Metal Ions peptide

metal ions added

Tm (°C)a

t1/2 of self-assemblyb

X9

none Co2+ Ni2+ Cu2+ Zn2+ none Co2+ Ni2+ Cu2+ Zn2+

36 36 40 42 37, 67 24 26 30 34, 71c 55c

>24 h >24 h >24 h ∼6.6 h ∼13.4 h >1.5 h >1.5 h >1.5 h ∼7.0 min ∼6.5 min

PHG

The deviation of Tm values was ±3 °C for X9/Zn2+, PHG/Cu2+, and PHG/Zn2+ solutions, whereas that was ±2 °C for others. bt1/2 values for the self-assembly process are the times taken to reach 50% of the maximal turbidity. cTm values were estimated since only small fraction of triple helices existed in solution after extensive aggregates appeared. a

Figure 3. Turbidity curves for (A) X9 and (B) PHG solutions (400 μM) collected in the presence of different metal ions (160 μM) at 4 °C by monitoring their absorbance at 313 nm.

Scanning Electron Microscopy. SEM images of the collagen aggregates were obtained using a JSM-7000F JEOL FESEM with an operating voltage of 5.0 kV. The peptide solution (400 μM) was incubated with metal ions (80, 160, 240, 320, or 400 μM) at 4 °C for at least 4 days until the white peptide precipitates were observed. The sample was then centrifugated and the supernatant was carefully removed. The remaining residue was resuspended in DI water and a droplet of the resuspended solution was air-dried on a silica wafer. The dried silica wafer was coated with Pt before taking SEM images.

■

RESULTS AND DISCUSSION To test the hypothesis that metal−histidine coordination can serve as a driving force for the folding of CRPs into triple helices, a Pro-Pro-Gly repeat short peptide HG(PPG)7GH was synthesized by solid phase methods. Since (Pro-Pro-Gly)n peptides form much less stable collagen triple helices than (Pro-Hyp-Gly)n peptides,4,27 here we chose Pro-Pro-Gly (PPG) instead of Pro-Hyp-Gly (POG) as the repeat unit to investigate how significant an effect metal−histidine coordination could have on the folding and self-assembly of unstable CRPs. Thus, histidine residues were incorporated in the peptide for metal binding sites. CD measurements indicated that this peptide does not form a triple helix in solution at the concentration of 230 μM but that it does fold into a triple helix with a melting temperature (Tm) around 20 °C upon the addition of Cu2+ or Ni2+ (Figure S1 of the Supporting Information). The results clearly show that metal−histidine coordination can assist the folding of CRPs and that the effect is relatively profound suggesting the possibility that the self-assembly of CRPs might be promoted by simply

Figure 2. Hydrodynamic diameters (d) measured by dynamic light scattering experiments for (A) X9 and (B) PHG solutions in the presence of different metal ions. All measurements were conducted at pH 7.4 in 20 mM phosphate buffer at 26 °C. The solutions (400 μM peptides +160 μM metal ions) were incubated at 4 °C for at least 24 h prior to measurements.

Transmission Electron Microscopy. TEM images of the collagen fibrils were acquired using a JEOL JEM-1400 TEM with an operating voltage of 120 kV. The mixture of peptide (400 μM) and metal ions (160 μM) was incubated at 4 °C for more than 2 days. After incubation, the mixed solution was deposited on a 200-mesh carbon coated copper grid and the excess solution was carefully removed. After droplets of the samples were air-dried, the resulting copper grid was stained with 0.3% uranyl acetate and air-dried. 3196

dx.doi.org/10.1021/la204351w | Langmuir 2012, 28, 3194−3199

Langmuir

Article

Figure 4. TEM images for the assembled structures of (A) X9 and (B) PHG after incubating with Cu2+ or Zn2+. (C) The assembled collagen fibers show D-periodicity with a D value of ∼9.5 nm. Before TEM images taken, 400 μM peptides and 160 μM metal ions were mixed in pH 7.4 20 mM phosphate buffer and incubated at 4 °C for at least 2 d. The scale bar is 100 nm.

monitoring the absorbance at 313 nm (Figure 3) showed that only Cu2+−CRPs and Zn2+−CRPs mixtures exhibited a turbidity increase within 24 h. This finding indicates that Cu2+ and Zn2+ more profoundly accelerate CRP self-assembly than do the other metal ions, in agreement with our DLS analyses and CD measurements. As shown in Table 1, the PHG peptide assembles much faster than the X9 peptide upon the addition of Cu2+ and Zn2+ suggesting that increasing the number of metal−histidine coordination sites can likewise hasten the self-assembly process. TEM was used to assess the morphology of the assembled structures formed from X9 and PHG. As shown in Figure 4, TEM images show that both X9 and PHG can form fibrils even in the absence of metal ions and that the fibrils exhibit weak Dperiodicity. The D-periodic pattern of the assembled fibers (Figure 4 and Figure S4 of the Supporting Information) becomes more pronounced and the periodic distance (D) remains the same in the presence of metal ions, indicating that metal ions significantly assist self-assembly and enhance the D-periodic pattern. As shown in part C of Figure 4, D is ∼9.5 nm − which is shorter than that of native collagen (∼67 nm) and the previous reported POG repeat CRPs (∼18 nm)7 but similar to the recent result reported by Pires et al.16 On the basis of the average distance of 0.31 nm for each residue in a polyproline II helix, the length of the 31 residues in X9 and PHG can be estimated to be ∼9.6 nm, in good agreement with the periodic distance found in their assembled fibers. This D-periodic pattern corresponds with the assembly mechanism that is delineated in Figure 1. It is quite surprising to observe such D-periodicity in PPG repeat CRPs, implying that even an unstable CRP can be modified to form and mimic the structural pattern found in natural collagen fibers.

implanting metal−histidine coordination. To explore whether CRPs can be induced to self-assemble exclusively by metal− histidine coordination, we further designed and synthesized two small CRPs, X9, and PHG. These two CRPs contain only natural amino acids, and histidine residues are incorporated in the peptides for metal binding sites. Our CD measurements indicated that both X9 and PHG can fold into marginally stable collagen triple helices in solution with Tm values of 36 and 24 °C respectively and that their thermal stability is increased by the addition of metal ions (Co2+, Ni2+, Cu2+, Zn2+) (Table 1 and Figure S3 of the Supporting Information). Of the metal ions tested, Cu2+ and Zn2+ ions have the most significant triple-helix stabilizing effects. Notably, the CD signals of some Cu2+ or Zn2+containing peptide solutions are weaker than others because aggregates appear in the solution after the incubation with Cu2+ or Zn2+. The thermal transition of PHG is particularly unclear in the presence of Cu2+ or Zn2+, and the Tm value can only be estimated. This enhanced aggregation may imply that the interactions of Cu2+−histidine and Zn2+−histidine are stronger than those of Co2+−histidine and Ni2+−histidine. DLS analysis (Figure 2) revealed that the hydrodynamic diameters of metal−CRPs are larger than those of CRPs in solution, and that the Cu2+ and Zn2+−CRPs’ hydrodynamic diameters are the largest. The increase in hydrodynamic diameters for the peptide solutions in the presence of metal ions indicates that metal ions induce the formation of higher-order structures. The observations that Cu2+−PHG and Zn2+−PHG solutions have the largest hydrodynamic diameters conform to that the CD signals of these two samples are weak due to the appearance of aggregates in solution. Turbidity measurements performed by 3197

dx.doi.org/10.1021/la204351w | Langmuir 2012, 28, 3194−3199

Langmuir

Article

Figure 5. SEM images for the large aggregates of (A) X9 and (B) PHG after incubating with Cu2+ or Zn2+. Before SEM images were taken, 400 μM peptides and 400 μM metal ions were mixed in pH 7.4, 20 mM phosphate buffer and incubated at 4 °C for at least 4 d. The scale bar is 10 μm in the panel of X9/Zn2+ and 1 μm in other panels.

To the best of our knowledge, this is the first demonstration that CRPs composed of PPG repeats can assemble into D-periodic collagen fibers. TEM images also show that metal−X9 mixtures form more linear fibers, whereas the fibers formed by metal− PHG mixtures are wider and prone to congregate. These observations indicate that metal−X9 assembles in an end-to-end manner, whereas metal−PHG assembles both end-to-end and radially. Our findings strongly indicate that the nature of the assembled structures can be modulated by the location of metal− histidine coordination sites, and are consistent with our a priori predictions. After incubation with Cu2+ or Zn2+, the aggregates observed for both X9 and PHG peptide solutions were further examined by SEM. From SEM images (Figure 5 and Figure S9 of the Supporting Information), we found that both X9 and PHG can self-assemble into micrometer scale structures in the presence of Cu2+ or Zn2+ but with different morphology. The finding that X9 and PHG can form microscale assemblies after incubation with Cu2+ or Zn2+ is consistent with DLS analyses that Cu2+−CRPs and Zn2+−CRPs have larger hydrodynamic diameters than Ni2+−CRPs and Co2+−CRPs in solution (Figure 2). We also performed the peptide assembly with different molar ratios of peptide/Cu2+ or peptide/Zn2+ and assessed their structures by SEM. At the lower concentration of metal ions, X9 assembles into laminate-like structures, whereas PHG forms granular structures (Figures S5−S8 of the Supporting Information). Upon increasing the metal ion concentration, the

structures formed by PHG are not altered (Figures S7 and S8 of the Supporting Information, and Figure 5), but X9 can further assemble into ultralarge structures (Figures S5 and S6 of the Supporting Information, and Figure 5). The effect of peptide− metal stoichiometry on X9 seems more significant than that on PHG. When the molar ratio of peptide to metal is 1:1, Zn2+, in particular induced X9 to form spherical structures with a diameter of 70−80 nm. The spherical structures formed by the X9/Zn2+ mixture are larger than those formed by a POG repeat peptide that contains both histidine residues and organic ligands,12 suggesting that mere metal−histidine coordination is sufficient to drive the self-assembly of even small and unstable CRPs into ultralarge collagen structures. Compared to PHG, X9 has a much slower assembly rate but forms a much larger and more organized collagen structure. For X9, Cu2+ induces assembly twice as fast as Zn2+ (t1/2 is ∼6.6 versus ∼13.4 h), but the assembled structures induced by Cu2+ are smaller and less organized than those induced by Zn2+. The slower assembly rate in the presence of Zn2+ may facilitate the production of an ultralarge and well-organized collagen structure. Furthermore, we also studied the reversibility of the metal-induced assembly process. We assembled the peptides in the presence of Zn2+ and then incubated the assemblies with EDTA to deplete Zn2+. After depleting the Zn2+, we then conducted turbidity measurements and DLS analysis. The results (Table S2 and Figure S10 of the Supporting Information) clearly show that the self-assembly process is reversible for PHG because the turbidity disappears and the hydrodynamic diameters are 3198

dx.doi.org/10.1021/la204351w | Langmuir 2012, 28, 3194−3199

Langmuir

Article

(7) Rele, S.; Song, Y.; Apkarian, R. P.; Qu, Z.; Conticello, V. P.; Chaikof, E. L. J. Am. Chem. Soc. 2007, 129, 14780−14787. (8) Chen, C.-C.; Hsu, W.; Kao, T.-C.; Horng, J.-C. Biochemistry 2011, 50, 2381−2383. (9) Cejas, M. A.; Kinney, W. A.; Chen, C.; Leo, G. C.; Tounge, B. A.; Vinter, J. G.; Joshi, P. P.; Maryanoff, B. E. J. Am. Chem. Soc. 2007, 129, 2202−2203. (10) Cejas, M. A.; Kinney, W. A.; Chen, C.; Vinter, J. G.; Harold R. Almond, J.; Balss, K. M.; Maryanoff, C. A.; Schmidt, U.; Breslav, M.; Mahan, A.; Lacy, E.; Maryanoff, B. E. Proc. Natl. Acad. Sci., U.S.A. 2008, 24, 8513−8518. (11) Kar, K.; Ibrar, S.; Nanda, V.; Getz, T. M.; Kunapuli, S. P.; Brodsky, B. Biochemistry 2009, 48, 7959−7968. (12) Pires, M. M.; Chmielewski, J. J. Am. Chem. Soc. 2009, 131, 2706−2712. (13) Pires, M. M.; Przybyla, D. E.; Chmielewski, J. Angew. Chem., Int. Ed. 2009, 48, 7813−7817. (14) Przybyla, D. E.; Chmielewski, J. J. Am. Chem. Soc. 2008, 130, 12610−12611. (15) Przybyla, D. E.; Chmielewski, J. J. Am. Chem. Soc. 2010, 132, 7866−7867. (16) Pires, M. M.; Przybyla, D. E.; Rubert Pérez, C. M.; Chmielewski, J. J. Am. Chem. Soc. 2011, 133, 14469−14471. (17) Pires, M. M.; Lee, J.; Ernenwein, D.; Chmielewski, J. Langmuir 2012, ASAP, DOI: 10.1021/la203848r. (18) Koide, T.; Homma, D. L.; Asada, S.; Kitagawa, K. Bioorg. Med. Chem. Lett. 2005, 15, 5230−5233. (19) Kotch, F. W.; Raines, R. T. Proc. Natl. Acad. Sci., U.S.A. 2006, 103, 3028−3033. (20) Yamazaki, C. M.; Asada, S.; Kitagawa, K.; Koide, T. J. Pept. Sci. 2008, 14, 186−186. (21) Yamazaki, C. M.; Asada, S.; Kitagawa, K.; Koide, T. Biopolymers 2008, 90, 816−823. (22) Paramonov, S. E.; Gauba, V.; Hartgerink, J. D. Macromolecules 2005, 38, 7555−7561. (23) Rubert Pérez, C. M.; Panitch, A.; Chmielewski, J. Macromol. Biosci. 2011, 11, 1426−1431. (24) O’Leary, L. E. R.; Fallas, J. A.; Bakota, E. L.; Kang, M. K.; Hartgerink, J. D. Nat. Chem. 2011, 3, 821−828. (25) Pires, M. M.; Ernenwein, D.; Chmielewski, J. Biomacromolecules 2011, 12, 2429−2433. (26) Jenkins, C. L.; Vasbinder, M. M.; Miller, S. J.; Raines, R. T. Org. Lett. 2005, 7, 2619−2622. (27) Engel, J.; Bachinger, H. P. Top. Curr. Chem. 2005, 247, 7−33.

reduced upon the addition of EDTA. In contrast, the reversibility of X9 assembly is weaker than for PHG because some aggregates could still be detected by DLS after the addition of EDTA, indicating that the microscale Zn2+−X9 assemblies formed via slow growth are relatively compact and more stable.

■

CONCLUSIONS In this work, we have used small, histidine-containing PPG repeat CRPs to demonstrate that metal−histidine coordination can serve as an effective force to induce the self-assembly of short synthetic and unstable CRPs into various nano- to microscale collagen structures. Because the designed CRPs only use natural amino acids and do not require any organic ligand modifications, they can be easily prepared via well-established solid phase methods and should be highly biocompatible. Thus, our results provide another simple and efficient means to create large scale collagen-related biomaterials from small peptides. Notably, the collagen fibers assembled by these two peptides in the presence of metal ions exhibit the D-periodicity found in natural collagen suggesting that these CRPs could mimic natural collagen and have potential applications in biological systems. Moreover, the size and topology of the assembled collagen structures depends on the location and number of metal−histidine coordination sites in a CRP, the metal ions added, and the assembly rates. Controlling and optimizing these variables will enable the straightforward production of collagen-related biomaterials with a wide range of structures and properties.

■

ASSOCIATED CONTENT

S Supporting Information *

Peptide sequences, CD-monitored thermal unfolding curves, DLS analysis and turbidity measurements for EDTA competition experiments, additional TEM and SEM images. This material is available free of charge via the Internet at http:// pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*Phone: +886-3-5715131, ext 35635; fax: +886-3-5711082; e-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We are thankful to the Taiwan National Science Council (NSC 99-2113-M-007-016-MY2 and NSC 98-2119-M-007-011) and National Tsing Hua University (100N2011E1) for support of this work, and to Prof. Chia-Min Yang for providing the DLS spectrometer. We are particularly grateful to Dr. Matthew Shoulders (The Scripps Research Institute) for reading our manuscript and valuable suggestions.

■

REFERENCES

(1) Lutolf, M. P.; Hubbell, J. A. Nat. Biotechnol. 2005, 23, 47−55. (2) Brodsky, B.; Persikov, A. V. Adv. Protein Chem. 2005, 70, 301−339. (3) Brodsky, B.; Thiagarajan, G.; Madhan, B.; Kar, K. Biopolymers 2008, 89, 345−353. (4) Shoulders, M. D.; Raines, R. T. Annu. Rev. Biochem. 2009, 78, 929−958. (5) Przybyla, D. E.; Chmielewski, J. Biochemistry 2010, 49, 4411− 4419. (6) Yu, S. M.; Li, Y.; Kim, D. Soft Matter 2011, 7, 7927−7938. 3199

dx.doi.org/10.1021/la204351w | Langmuir 2012, 28, 3194−3199