Microbially Synthesized Repeats of Mussel Foot Protein Display

Research Article Cite This: ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

www.acsami.org

Microbially Synthesized Repeats of Mussel Foot Protein Display Enhanced Underwater Adhesion Eugene Kim,† Bin Dai,† James B. Qiao,† Wenlu Li,† John D. Fortner,†,‡ and Fuzhong Zhang*,†,‡,§ †

Department of Energy, Environmental & Chemical Engineering, ‡Institute of Materials Science and Engineering, and §Division of Biological & Biomedical Sciences, Washington University in St. Louis, One Brookings Drive, St. Louis, Missouri 63130, United States

ACS Appl. Mater. Interfaces Downloaded from pubs.acs.org by KAOHSIUNG MEDICAL UNIV on 11/28/18. For personal use only.

S Supporting Information *

ABSTRACT: Mussels strongly adhere to a variety of surfaces by secreting byssal threads that contain mussel foot proteins (Mfps). Recombinant production of Mfps presents an attractive route for preparing advanced adhesive materials. Using synthetic biology strategies, we synthesized Mfp5 together with Mfp5 oligomers containing two or three consecutive, covalentlylinked Mfp5 sequences named Mfp5(2) and Mfp5(3). The force and work of adhesion of these proteins were measured underwater with a colloidal probe mounted on an atomic force microscope and the adsorption was measured with a quartz crystal microbalance. We found positive correlations between Mfp5 molecular weight and underwater adhesive properties, including force of adhesion, work of adhesion, protein layer thickness, and recovery distance. DOPA-modified Mfp5(3) displayed a high force of adhesion (201 ± 36 nN μm−1) and a high work of adhesion (68 ± 21 fJ μm−1) for a cure time of 200 s, which are higher than those of previously reported Mfp-mimetic adhesives. Results presented in this study highlight the power of synthetic biology in producing biocompatible and highly adhesive Mfp-based materials. KEYWORDS: mussel foot proteins, bioadhesive, biomimetic, polymers, underwater, split intein, force of adhesion, DOPA

■

INTRODUCTION Nature has evolved a wide array of protein-based materials and composites (e.g., silk, elastin, collagen) with mechanical properties and functions exceeding the performance of many synthetic polymers and even metal alloys. These remarkable properties arise from the multiscale assembly of a narrow set of basic, repetitive peptide sequences.1−3 Engineering microbial cell factories for heterologous production of such proteins offers the opportunity to provide sufficient supplies of natural materials from cheap and renewable feedstocks (e.g., cellulosic biomass and simple sugars), presenting a sustainable and costeffective approach for advanced material manufacturing. Further, editing protein sequences and engineering material assembly processes allow for tunable material properties and enable novel applications. In this study, we engineer a microbial process to produce strong adhesive mussel foot proteins (Mfps) and characterize their properties. Mfps secreted in the byssal threads of mussels have an extraordinary ability to adhere to various surfaces underwater, a feat which is unachievable by most chemical or synthetic adhesives.4,5 To date, seven main types of Mfps have been © XXXX American Chemical Society

identified, and different Mfp types spatially organized at specific locations of the mussel byssal plaque have been found to serve different functions.6,7 Although Mfp2 and Mfp4, which are localized at the core of the mussel byssal thread plaque, are mostly responsible for cohesive protein−protein interactions, the intrinsically disordered Mfp3 and Mfp5 are localized at the distal end of the plaque and play major roles in surface adhesion (Figure 1a).6 Mfps have varying levels of 3,4dihydroxyphenylalanine (DOPA), which arise from posttranslational modification of tyrosine residues by tyrosinases, a group of natively expressed hydroxylating enzymes.7 Extensive studies have illustrated the essential roles of DOPA in forming both adhesive protein−surface interactions via bidentate hydrogen bonding, metal complexation, and hydrophobic interactions and in cohesive protein−protein interactions via bis- and tris-DOPA-Fe3+ complexation, DOPA− DOPA hydrogen bonding, aryloxyl radicalization, π−π stacking Received: August 29, 2018 Accepted: November 12, 2018

A

DOI: 10.1021/acsami.8b14890 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 1. (a) Schematic representation of the design and production of strong underwater adhesives using engineered microbes. The distal end of the mussel plaque is abundant in Mfp5, whose DNA sequence is recoded for E. coli production using standard genetic parts. (b) Mfp5 chains can adhere to various surfaces via multiple types of interactions, e.g., bidentate hydrogen bonding and metal complexation. Mfp5 can also cohesively interact with neighboring chains via bisDOPA hydrogen bonding, aryloxyl radicalization crosslinking, and physical chain entanglements. Compared to low-molecular-weight (MW) proteins, high MW proteins are expected to chemically interact and entangle into a more robust network of interactions.

Figure 2. (a) Schematic of gene constructs and the split intein-mediated ligation processes of Mfp5(2)-IntN and IntC-Mfp5(1) for the production of Mfp5(3). (b) Sodium dodecyl sulfate poly acrylamide gel electrophoresis (SDS-PAGE) with Coomassie blue staining of split intein-fused Mfp5 reactants (lanes 2 and 3) and purified Mfp5(3) product (lane 4) with a protein marker standard (lane M). Lane 1 contains the whole cell lysate of wild type E. coli that does not express target protein.

interactions, and cation−π interactions (Figure 1b).8−11 Among the different types of Mfps, Mfp5 has the highest known level of DOPA (∼26 mol %) and displays the highest adhesion strength (2.3−7 mJ m−2).6 These findings have inspired the design of various DOPA-based synthetic biomimetic materials.2,4,12−17 Moreover, recent studies indicate that other residues and sequence features may also play important roles in surface adhesion. For example, over half of the tyrosine residues in Mfp5 have neighboring lysine residues, which are believed to assist in priming surfaces for facile interaction with DOPA.18 These results suggest that Mfps

employ a more complex mechanism than just using DOPAbased interactions. Fully understanding the sequence− structure−function relationship is ultimately the key for designing better biomimetics, and at present,recombinant Mfp5 is an attractive adhesive material for various applications, such as in surgical glues and for underwater repair.19,20 Recombinant Mfp5 has been previously expressed in E. coli.21 After enzymatic conversion of tyrosine residues to DOPA, the resulting protein displayed a force of adhesion of 49 nN μm−1 (force normalized to the radius of a contacting probe).21 A recent innovative approach fused Mfp5 to the B


Research Article


Figure 3. (a) Schematic of the colloidal probe atomic force microscopy experimental setup. The cantilever with a colloidal probe (1) approaches the surface of the sample deposited on mica and (2) interacts with oligomer chains with varying cure times. The probe then begins to (3) retract from the sample, and protein−surface and protein−probe interactions start to break. Applying longer separation distances (4) causes protein− protein interactions to break further until the probe is completely separated from the mica surface. Representative approach−retract curves of (b) Mfp5(1)DOPA, (c) Mfp5(2)DOPA, and (d) Mfp5(3)DOPA with a cure time of 10 s were collected on multiple days at multiple sample locations. Triplicate measurements were taken at individual locations shown in the same color. Zrel represents the separation distance between surfaces relative to the points of maximum contact (positive force of adhesion).

to ligate separately expressed Mfp5(1) and Mfp5(2) to yield the Mfp5(3) protein. SIs are autocatalytic peptides that catalyze the spontaneous splicing−ligation reactions between two SI-fused target proteins, assembling the two target proteins covalently in trans (Figure 2a).23,24 For this study, a recently engineered Cfa SI was employed due to its rapid protein splicing rate in denaturing conditions (e.g., 8 M urea), which we used to efficiently extract and solubilize the target Mfp5 proteins.25 The N-terminal split intein (IntN) and C-terminal split intein (IntC) were genetically fused to Mfp5(2) and Mfp5(1), respectively, resulting in two fusion proteins, Mfp5(2)-IntN and IntC-Mfp5(1). These fusion proteins in whole cell lysate mixtures were then mixed, yielding Mfp5(3), which was further purified from the nonspecific proteins, the spliced SI complex, and the unreacted low-molecular-weight proteins to a final yield of about 20 mg L−1 of total culture with 96.4% purity (Figure 2b,c). Purified proteins were then reacted with tyrosinase to convert tyrosine to DOPA, yielding modified proteins, named Mfp5(1)DOPA, Mfp5(2)DOPA, and Mfp5(3)DOPA. Matrix-assisted laser desorption/ionization-time of flight mass spectrometry (MALDI-TOF MS) analysis showed increased molecular weights with an average tyrosine modification rate of 65% under our conditions (Figures S5 and S6), consistent with previous modification efficiencies from similar methods.2,4 Further quantification by the acid-borate difference spectroscopy analysis showed similar DOPA content among all three Mfp samples and slightly lower conversion efficiencies (48− 55%) compared to those from MALDI-TOF analysis (Figure S7).26 Discrepancies between these quantification methods have been observed previously.2,4 Underestimation of DOPA content by spectrophotometric assays could be caused by steric

major subunit of curli protein, CsgA, which exposed the Mfp5 domain on the exterior of self-assembled curli fibers.2 The βsheet structure of the curli fiber contains intermolecular hydrogen bonds, enabling interactions between multiple fused Mfp5 domains. The resulting material displayed underwater force of adhesion values of up to 120 nN μm−1. We hypothesized that covalent linking of multiple Mfp5 proteins at the proteins’ termini may also lead to strong underwater adhesives due to the increased probability of protein−surface interactions and the enhanced density of interprotein interactions and entanglements, effectively reducing adhesion failure. Here, we developed a microbial-based approach to produce covalently linked Mfp5 oligomers and evaluated their underwater adhesion properties.

■

RESULTS AND DISCUSSION Using synthetic DNA with codons optimized for E. coli expression, we designed the Mytilus galloprovincialis Mfp5 together with Mfp5 oligomers containing two or three consecutive Mfp5 sequences named Mfp5(2) and Mfp5(3) and cloned them into standardized expression vectors (Figures S1 and S2).22 Initial experiments showed that both Mfp5(1) and Mfp5(2) could be produced in E. coli and purified using affinity chromatography to final yields of about 100 and 200 mg of purified proteins per liter of cell culture, respectively, with purities of 99.9 and 92.2% (Figure S3), respectively. However, Mfp5(3) designed using the same gene optimization algorithm cannot be expressed under identical conditions. The lack of Mfp5(3) expression is likely caused by a combination of the repetitive nature of the coding sequence, translation inhibition by complex mRNA secondary structures, and a high demand for tyrosyl tRNA. To solve this issue, we used split inteins (SI) C


Research Article


Figure 4. (a) Positive approach curves for Mfp5(1)DOPA (yellow), Mfp5(2)DOPA (blue), and Mfp5(3)DOPA (green) showing repulsive interactions during probe approaching. (b) Fitting of representative approach curves to the AdG model (red solid lines). (c) Average equilibrium thicknesses of the three modified Mfps. The average thickness of each protein was obtained from at least 18 fittings. Error bars represent standard deviations (n ≥ 18). Statistically significant differences (one-way ANOVA with Tukey’s HSD post hoc analysis, p < 0.01) are indicated by asterisks.

Figure 5. (a) Force of adhesion and (b) work of adhesion versus probe cure time for BSA (black), Mfp5(1) (gray), Mfp5(1)DOPA (yellow), Mfp5(2)DOPA (blue), and Mfp5(3)DOPA (green). Error bars represent standard deviations (n ≥ 21).

hindrances that prevent assay reagents from reacting with internally buried DOPA residues. Colloidal probe atomic force microscopy (AFM) was used to examine asymmetric adhesive and cohesive characteristics of the Mfp5 proteins between a colloidal probe and a proteinadsorbed mica surface under aqueous conditions (Figure 3a). The colloidal probe was moved toward the protein layer (probe approach) and then incubated for various times before the probe was retracted from the surface (probe retraction). We refer to this incubation period as a “curing” process during which physical entanglement between protein chains and chemical interactions between proteins and the colloidal probe surface, as well as between protein chains, are established. For all tested protein samples, approach−retract curves were

collected on several days from multiple sample locations and for multiple cure times at each sample location (Figure 3b−d). Figure 4 shows approach curves of DOPA-modified Mfps. For all three Mfp5 proteins, no negative force of adhesion values were observed during probe approach, indicating that the interaction was dominated by repulsive interactions under our experimental conditions (Figure 4a).27 The approach curves were fitted to the Alexander-de Gennes (AdG) model (see Experimental Section), which was employed to describe the physisorption of intrinsically disordered gelatin coils, polymer layers, and other proteins to colloidal probes.27,28 Our fitted results suggest apparent average protein layer equilibrium thicknesses (L) of approximately 8.5, 16, and 29 nm for Mfp5(1)DOPA, Mfp5(2)DOPA, and Mfp5(3)DOPA, respectively (Figure 4b,c). These values are greater than the estimated D


Research Article


adhesion energies that were 10.5 and 19.2 fold higher than that of Mfp5(1)DOPA, respectively, also exhibiting a chain-lengthdependent increase. The longest curing time that allowed a reliable approach− retract curve measurement for Mfp5(3)DOPA was 200 s because longer cure times produced adhesion strengths that were too high and prevented recovery of retract curves, even at the maximum AFM probe separation of our equipment (1 μm). At a cure time of 200 s, Mfp5(3)DOPA exhibited a force of adhesion of ∼201 nN μm−1 and a work of adhesion of ∼68 fJ μm−1, values which are higher than previously reported for Mfpmimetic adhesives (Figure 6).2,13 Fitting the time course of

radii of gyration of each of these proteins (3.1, 4.4, and 5.5 nm for Mfp5(1)DOPA, Mfp5(2)DOPA, and Mfp5(3)DOPA, respectively) when adopting random-coil structures.29 These greater lengths suggest that a protein multilayer is present and that with increasing molecular weight, there is an increase in the amount of protein adsorbed on the surface. This observation is consistent with our hypothesis that high-molecular-weight increases interprotein interactions and/or entanglements, leading to thicker protein layers (Figure 4c). The force of adhesion of our proteins was calculated from the maximum attraction force of the retract curves and was normalized by the radius of the colloidal probe.9,13,21,27 Similar to previous reports, all proteins displayed a cure timedependent adhesion, with increasing forces of adhesion after longer curing times (Figure 5a).30,31 For all cure times, bovine serum albumin (BSA) and unmodified Mfp5(1) exhibited little adhesion ( 0.1). system takes the following conditions into account: (a) split intein genes, ribosome binding sites (RBS), 5′-untranslated regions (5′UTR), antibiotic resistance markers, promoters, and origins of replication, which are flanked with appropriate restriction sites that allow them to be easily swapped with other corresponding genes of interest, (b) restriction sites that exist within open reading frames introduce amino acids that are small, flexible, and are not expected to change the protein behavior and dynamics, (c) restriction sites are all distinct with respect to one another to facilitate the assembly of complete plasmids in one pot and in one step. All restriction sites used are schematically mapped out in Figures S1 and S2. E. coli strain MDS42pdu was used as a host strain for cloning of all genes and plasmids in this study.40 The amino acid sequences of Mytilus galloprovincialis Mfp5(1), Mfp5(2), and Mfp5(3) were codonoptimized for E. coli expression using the Gene Designer 2.0 software package (DNA 2.0 Inc.). All designed DNA sequences were chemically synthesized by Integrated DNA Technologies Inc. (San Jose, CA) (Table S1). These synthetic genes were then amplified using polymerase chain reaction (PCR) with corresponding forward and reverse primers, as listed in Table S2. All mfp5 genes were amplified with BglII and BamHI sites on the 5′ and 3′ ends, respectively, for insertion into the pE7a-AKTK-H6 backbone containing the same sites, which was previously PCR amplified from plasmid pE7a-GFP22,41 with the addition of short coding sequences, 5′-ATGGCTAAGACTAAACATCATCACCATCATCAC-3′, translating to N′-MAKTK-H6-C′ (Figures S1 and S2). The AKTK expression tag, immediately following the start codon, has been shown to increase translation initiation rates,42−44 and the Nterminal hexahistidine-tag (His6) was used for downstream protein purification. This backbone was named pE7a-AKTK-H6. The amplified gene and backbone were digested with BglII and BamHI. The backbone was further treated with alkaline phosphatase to minimize unwanted recircularization. The digested backbone and gene inserts were then ligated with T4 ligase. The proper orientation of each gene within the isolated plasmid was confirmed with restriction digestion using BglII and BamHI followed by Sanger sequencing (Eurofins Genomics, Louisville, KY). To construct plasmids containing split intein-fused Mfp5 proteins, the amino acid sequences of Cfa N- and C-inteins (CfaN and CfaC, respectively)25 were first codon optimized and chemically synthesized using the same method as described above. The synthesized genes containing the CfaC domain, which is flanked by NdeI and KpnI sites, and the CfaN domain, which is flanked by Kpn2I and XhoI sites, were digested directly from the synthesized DNA. The mfp5(2) and mfp5(1) genes were amplified with corresponding primers (Table S2) containing KpnI and Kpn2I restriction sites. Amplified mfp5(2) and mfp5(1) fragments were digested (Figure S1) and ligated with CfaN and CfaC into pE7a-N′-AKTK and pE7a-H10-C′ backbones,

a single protein layer is formed and fully occupies the surface of the gold-coated quartz crystal surface, the moles of proteins adsorbed per area are expected to decrease as molecular weight increases. Our AFM results suggest that protein layers increased with Mfp molecular weight. These two opposing trends may cancel each other out leading to similar moles of protein absorbed per unit area.

■

CONCLUSIONS In summary, we engineered E. coli to successfully produce Mfp5-based adhesive proteins with strong underwater adhesion capabilities. Under our experimental conditions, Mfp5(3)DOPA displayed high force and work of adhesion values that were comparable to or higher than those of previously reported Mfp-mimetic adhesives.2,13 With a longer curing time, the force of adhesion for Mfp5(3) could be even higher than what was measured, but was beyond the detection limit of our method. Such adhesive proteins can potentially be used to replace natural Mfps in a variety of applications, with even better performance due to their stronger underwater adhesion. Our results demonstrated that protein−protein cohesive interactions play an important role in determining both force and work of adhesion. We observed positive correlations between Mfp5’s molecular weight and its measured force and work of adhesion. These correlations indicate that producing high-molecular-weight proteins might be a natural strategy to obtain extensive cohesive interactions, as evidenced in Mfp2 and Mfp4, which have the highest molecular weights within the byssal plaque core (45 and 90 kDa, respectively) among all Mfps and play cohesive roles in mussel plaques.16,17,39 The observed molecular-weight-dependent adhesion can be used in designing rules to guide future engineering efforts, creating even stronger underwater adhesives.

■

EXPERIMENTAL SECTION

Chemicals and Reagents. Unless otherwise noted, all chemicals and reagents were obtained from Millipore Sigma (Saint Louis, MO). Plasmid purification and gel extraction kits were purchased from iNtRON Biotechnology (Seoul, South Korea). FastDigest restriction enzymes and T4 DNA ligase were purchased from Thermo Fisher Scientific (Austin, TX) and were used according to the suggested protocols from the manufacturer. Plasmid Construction. A BioBrick system was used to facilitate cloning of two-part split intein-fused material protein domains.22 This F


Research Article

ACS Applied Materials & Interfaces respectively. The corresponding backbones were both PCR amplified and contain distinct restriction sites for ligation of the appropriate genes in the correct locations (Figure S1). Expression of Recombinant Proteins. E. coli strain BL21(DE3) (Thermo Fisher Scientific, Waltham, MA) was used as a host strain for expression of all Mfps in this study. E. coli strains containing the plasmids listed in Table S3 were cultured in shake flasks with the Luria−Bertani (LB) broth containing 10 g/L tryptone, 10 g/L sodium chloride, and 5 g/L yeast extract with the appropriate antibiotic (100 μg/mL ampicillin). Fresh transformants were cultivated overnight in 50 mL of LB medium at 37 °C. Overnight cultures were then used to inoculate 1 L of fresh LB medium in Erlenmeyer flasks at an initial OD600 = 0.08. Cultures were grown at 37 °C with shaking to OD600 = 0.6, then induced by addition of 500 μM (for Mfp5(1) and CfaCMfp5(1)) or 50 μM (Mfp5(2) and Mfp5(2)-CfaN) IPTG. The culture was further cultivated at 37 °C at 250 rpm for another 5−7 h. Cells were harvested by centrifugation at 4500g for 20 min at 4 °C. Centrifuged cell pellets were either directly extracted or stored at −80 °C until needed. Protein Purification. For Mfp5(1) and Mfp5(2), cell pellets were resuspended in 10 mL of guanidine lysis buffer (6 M guanidine hydrochloride (BioBasic Inc., Amherst, NY), 50 mM potassium phosphates, and 300 mM sodium chloride at pH 7.4) per gram of wet cells and lysed by agitation at 250 rpm. The lysates were centrifuged at 20 000g for 20 min at 18 °C. To reduce the viscosity, collected lysates were further sonicated on ice (to maintain a roughly ambient temperature) for 30 min with a QSonica probe sonicator using 5 s on/5 s off cycles. The lysates were filtered through 0.2 μm filter membranes. Both proteins were purified using an AktaPure Fast Protein Liquid Chromatograph (FPLC, GE Healthcare Inc., Chicago, IL) equipped with a 5 mL nickel affinity chromatography column (GE Healthcare). The column was pre-equilibrated with guanidine lysis buffer followed by sample loading. After washing with 5−10 column volumes (CVs) of guanidine wash buffer (6 M guanidine hydrochloride, 50 mM potassium phosphates, 300 mM sodium chloride, and 50 mM imidazole at pH 7.4), proteins were eluted and fractionated with 5−10 CVs of guanidine elution buffer (6 M guanidine hydrochloride, 50 mM potassium phosphates, 300 mM sodium chloride, and 250 mM imidazole at pH 7.4). Purified Mfp5(1) and Mfp5(2) proteins were examined by SDS-PAGE, as shown in Figures S3 and 2. Split Intein-Mediated Ligation and Purification of Mfp5(3). Cell pellets containing CfaC-Mfp5(1) and Mfp5(2)-CfaN-His10 fusion proteins were separately resuspended in 10 mL of urea lysis buffer (8 M urea, 100 mM sodium phosphates, and 300 mM sodium chloride at pH 7.4) per gram of wet cells and lysed by agitation at 250 rpm overnight. The lysates were centrifuged at 20 000g for 20 min at 18 °C. Clarified cell lysate was then mixed at a final reactant ratio of 4:1 (CfaC-Mfp5(1)/Mfp5(2)-CfaN-His10) based on densitometric analysis of Coomassie Blue-stained SDS-PAGE gels. The excess IntC-Mfp5(1) does not contain His-tag, thus can be easily separated from the ligated product. The lysate mixture was stirred at 30 °C for 8 h. The mixed lysate was then filtered through a 0.2 μm filter membrane and purified by nickel affinity chromatography as described above. Post-Translational Modification with Tyrosinase. Purified protein solutions were first dialyzed against 100 mM sodium acetate buffer at pH 5.5 using a 10 kDa molecular weight cut off (MWCO) dialysis membrane (Thermo Fisher Scientific). Dialyzed proteins were then diluted to a concentration of 4 mg/mL in 100 mM sodium acetate buffer at pH 5.5 with 100 mM ascorbic acid and filtered. Tyrosinase was added to a final concentration of 250 U mL−1, and the mixture was incubated at 37 °C with agitation at 250 rpm for 30 min. After the reaction, the solution was filtered, and the enzyme activity in the flow-through was quenched by adding 0.2 mL of 6 N HCl per mL of reaction. The solution was filtered a final time and then was dialyzed extensively in 5% acetic acid at 4 °C and lyophilized. Matrix-Assisted Laser Desorption/Ionization-Time of Flight (MALDI-TOF) Mass Spectrometry (MS) Analysis. Purified proteins at a final concentration of ∼10 μM in 0.1% v/v. trifluoroacetic acid (TFA) solution were mixed with dihydroxyaceto-

phenone (DHAP) and spotted on a stainless steel MALDI target plate. Samples were analyzed using a Shimadzu AXIMA Resonance MALDI-TOF Mass Spectrometer (Shimadzu, Columbia, MD) at the Saint Louis University Protein Core Facility. Positive-ion mass spectra analysis was conducted in standard linear mode with a laser power/ acceleration voltage of 60−100 V. For Mfp5(1)DOPA with an expected molecular weight of 10 978.02 Da, the single- and double-charged state peaks could be assigned at m/z values of 5489.01 and 3659.34, respectively (Figure S5). The quadruple ion trap available for MS analysis conducted in this study allowed for a high peak detection resolution of 1 Da, but was limited to a maximum m/z ratio of 15 000 Da. Consequently, for Mfp5(2)DOPA with an expected molecular weight of 19 793.86 Da, only the double-charged state peak could be assigned at an m/z value of 9896.93 (Figure S6). For Mfp5(3)DOPA with an expected molecular weight of 29 849.88 Da, a triple-charged state was necessary to identify a signal within the measurable m/z range, however, this species was not clearly discernible from the noise and minor contaminants existing in this sample (data not shown). Acid-Borate Difference Spectroscopy for Quantification of DOPA in Modified Mfp5’s. To further confirm and quantify the conversion from tyrosine to DOPA in the Mfp5 proteins tested in this study, acid-borate difference was used.26 In aqueous acid solutions, DOPA maximally absorbs light at a wavelength of 280 nm. Acidborate difference spectroscopy relies on using borate to react with DOPA, which creates diol-borate complexes, which shifts the maximum absorptivity to a wavelength of approximately 292 nm. This complexation reaction is specific to DOPA and is not expected to react with other aromatic residues that lack diol functional groups, such as tyrosine, tryptophan, or phenylalanine. When borate is introduced, DOPA-containing proteins have been shown to have a maximum absorptivity at a slightly higher wavelength of approximately 294 nm, which has been shown in a previous study.2 A stock solution of L-DOPA was freshly prepared at a concentration of 100 mg L−1 with the addition of 0.2 mM hydrochloric acid. Diluted solutions were made from this stock solution at concentrations of 20, 30, 40, 50, 60, 70, 80, and 90 mg L−1 DOPA, with an additional acid solution control with no DOPA present. A UV-transparent 96-well plate was preread over a wavelength range of 230−330 nm. For each standard solution, 100 μL was added to a preread well, followed by 100 μL of 0.1 N HCl. In a separate well, 100 μL of standard solution was added to a preread well, followed by 100 μL of 0.1 M sodium tetraborate. The spectra within the measured range for each standard solution are shown in Figure S7a, with the maximum absorptions at 292 nm, plotted as a linear standard curve in Figure S7b. The same steps were carried out with known concentrations of each protein to be measured from known masses of protein and known volumes of solvent (0.2 mM hydrochloric acid), and the spectra for these samples are shown in Figure S7c. A summary of the known concentrations of proteins analyzed, their molecular weights, and amount of DOPA in each protein is tabulated in Figure S7d. AFM Colloidal Probe Analysis. Colloidal probe atomic force microscopy was used to characterize our protein samples by measuring force and work of adhesion using a colloidal probe AFM cantilever (Figure S8). This type of measurement is known as asymmetric adhesion where protein is bound firmly on the mica surface upon sample preparation and binds temporarily to the glass probe upon measurement, as outlined in previous studies.2,6 PFT mode was performed using a Bruker Multimode 8 High Resolution AFM (Bruker Inc., Billerica, MA) with calibrated glass colloidal probe cantilevers (radii of 5 μm and spring constants between 5.4 and 16 N m−1, Novascan, Ames, IA). Force curves were measured at a frequency of 1.0 Hz. Analysis of force curves was performed using the Nanoscope Analysis 1.8 software (Bruker Inc.). For each protein sample, we collected force curves in triplicate in at least seven regions to comprehensively assess each sample. For sample preparation, 10 μL of protein solution in PBS buffer was pipetted on a mica surface and set quiescently for 30 min. After extensive washing, the mica surface was mounted on the AFM stage, and the adsorbed protein was probed under a buffered condition (100 mM sodium phosphates pH 7.4). Bovine serum albumin (BSA) and G


Research Article

ACS Applied Materials & Interfaces unmodified Mfp5(1) were used as controls and measured under the same condition. Approach curves were fitted to the Alexander-de Gennes (AdG) model ÄÅ ÉÑ 7/4 ÑÑ 16πkBTRL ÅÅÅÅ ij 2L yz5/4 ij Z + 2h yz ÑÑ j z j z 7 5 12 F(Z) = + − Å ÑÑ j z j z Å ÑÑ 35s 3 ÅÅÅÇ k Z + 2h { k 2L { ÑÖ (1) where F is the measured force of adhesion, Z is the measured separation distance, kB is the Boltzmann constant, T is the absolute ambient temperature (∼298 K), R is the contact probe radius (5 μm). L, h, and s are fitted parameters that represent equilibrium film thickness, the offset distance that considers material compressibility upon maximum probe contact, and the average distance between occupied sorption sites, respectively.27 This model has been used to describe the interaction of random-coil polymers and proteins on surfaces.27,45 The data were fit using MATLAB. Data far beyond the fitted equilibrium thicknesses were not included to prevent misfitting at larger separation distances. All fitted data had correlation coefficients (r2) of at least 99%. The fitted equilibrium thickness (L) parameters were further compared with both the radii of gyration assuming random-coil configurations, ⟨R⟩, and the end-to-end distances when the proteins are fully stretched out as rigid rods, Lr. These distances were calculated using the formulas ⟨R ⟩ =

N ⟨l⟩2

unbound protein off the quartz sensors until the frequency shift was stable (Figure S13). The adsorbed mass was calculated using the Sauerbrey equation.

Δm =

(6)

where C (=17.7 ng cm−2 Hz−1) is a constant that combines all constants relating to using gold-coated quartz sensors for deposition with the QCM instrument used in this study.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.8b14890.

■

(2)

Sequence information; additional SDS-PAGE; MALDITOF MS data; acid-borate difference spectroscopy data; and additional AFM and QCM adhesion data (PDF)

AUTHOR INFORMATION

Corresponding Author

(3)

A 1 + BQ log t

(5)

Δm = − C·Δf

*E-mail: [email protected]. Phone: (314) 935-7671. Fax: (314) 935-7211.

where N is the number of amino acids in the protein and l is the length of one amino acid (3.5 Å).29 Following probe approaching, the probe was cured on the protein sample for a delay time varying from 0 to 50 s for most measurements. The force of adhesion was taken at the point of maximum force (most negative force). The work of adhesion was calculated as the area “under” the retraction curve, relative to the zero nN baseline (Figures 3, 5, S9, and S10). For the strongest oligomer tested in this study, Mfp5(3)DOPA, force of adhesion values with each respective cure time tested were averaged and fit to a logistic model of the form log F =

2 2 Fquartz

where Δm is change in mass, Δf is change in crystal resonance frequency, A is the active area of the crystal between electrodes, μquartz is the shear modulus of the quartz crystal, ρquartz is the density of the quartz crystal, and Fquartz is the reference frequency. This equation can be simplified to

and Lr = Nl

−Δf · A · μquartz ρquartz

ORCID

Wenlu Li: 0000-0003-3938-1624 Fuzhong Zhang: 0000-0001-6979-7909 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the Office of Naval Research under the award number N000141512515 awarded to F.Z. We would also like to thank Dr. Elijah Thimsen for providing plasma etchers for cleaning AFM colloidal probe tips, Dr. Daniel Giammar for providing a lyophilizer for freeze-drying protein, Dr. David Wood at the Saint Louis University Protein Core Facility for the MALDI-TOF Mass Spectrometry data and analysis, and Dr. Janie Brennan for productive discussions and insight.

(4)

where F is the measured force of adhesion, t is the cure time, and A, B, and Q are fitted constants. This model was appropriate given the assumption that an infinitesimally small cure time would result in zero force of adhesion, whereas an infinite cure time would result in a theoretical maximum force of adhesion (represented by the fit constant “A”) upon complete underwater curing of the material. The log−log scaling of the force of adhesion and cure time axes allowed for the fitted model to give rise to an unpatterned residual, which indicates a high quality of fitting (Figure S12b). Quartz Crystal Microbalance with Dissipation Monitoring (QCM-D) Analysis. Gold-coated quartz crystal sensors (QSX 301, Biolin Scientific, Gothenburg, Sweden) were used for QCM analysis. The sensors were cleaned by UV irradiation for 15 min, heated in a mixture of ammonia (25%) and hydrogen peroxide (30%) at 75 °C for 5 min, then thoroughly rinsed with distilled water, dried with N2, and subjected to 10 min of UV irradiation before being mounted in the QCM flow modules (Biolin Scientific). PBS buffer carrier solution was flowed through an injector valve into the flow cell modules containing the quartz sensors at a flow rate of 0.1 mL min−1 until stable baselines were achieved. Protein solutions were prepared to a final concentration of 10 μM (or 2 mg mL−1 in the case of BSA) in PBS buffer (100 mM sodium phosphates, 300 mM sodium chloride pH 7.4) and flowed through the flow cell at a flow rate of 10 μL min−1 for 60 min. PBS buffer carrier solution was flowed through to wash

■

REFERENCES

(1) So, C. R.; Fears, K. P.; Lear, D. H.; Scancella, J. M.; Wang, Z.; Liu, J. L.; Orihuela, B.; Rittschof, D.; Spillmann, C. M.; Wahl, K. J. Sequence Basis of Barnacle Cement Nanostructure Is Defined by Proteins with Silk Homology. Sci. Rep. 2016, 6, No. 36219. (2) Zhong, C.; Gurry, T.; Cheng, A. A.; Downey, J.; Deng, Z.; Stultz, C. M.; Lu, T. K. Strong Underwater Adhesives Made by SelfAssembling Multi-Protein Nanofibres. Nat. Nanotechnol. 2014, 9, 858−866. (3) Yuvienco, C.; More, H. T.; Haghpanah, J. S.; Tu, R. S.; Montclare, J. K. Modulating Supramolecular Assemblies and Mechanical Properties of Engineered Protein Materials by Fluorinated Amino Acids. Biomacromolecules 2012, 13, 2273−2278. (4) Brennan, M. J.; Kilbride, B. F.; Wilker, J. J.; Liu, J. C. A Bioinspired Elastin-Based Protein for a Cytocompatible Underwater Adhesive. Biomaterials 2017, 124, 116−125.

H


Research Article

ACS Applied Materials & Interfaces (5) Wheat, J. C.; Wolf, J. S. Advances in Bioadhesives, Tissue Sealants, and Hemostatic Agents. Urol. Clin. 2009, 265−275. (6) Lee, B. P.; Messersmith, P. B.; Israelachvili, J. N.; Waite, J. H. Mussel-Inspired Adhesives and Coatings. Annu. Rev. Mater. Res. 2011, 41, 99−132. (7) Waite, J. H. Mussel Adhesion − Essential Footwork. J. Exp. Biol. 2017, 220, 517−530. (8) Ooka, A. A.; Garrell, R. L. Surface-Enhanced Raman Spectroscopy of DOPA- Containing Peptides Related to Adhesive Protein of Marine Mussel, Mytilus edulis. Biopolymers 1999, 57, 92−102. (9) Cha, H. J.; Hwang, D. S.; Lim, S.; White, J. D.; Matos-Perez, C. R.; Wilker, J. J. Bulk Adhesive Strength of Recombinant Hybrid Mussel Adhesive Protein. Biofouling 2009, 25, 99−107. (10) Wei, W.; Petrone, L.; Tan, Y.; Cai, H.; Israelachvili, J. N.; Miserez, A.; Waite, J. H. An Underwater Surface-Drying Peptide Inspired by a Mussel Adhesive Protein. Adv. Funct. Mater. 2016, 26, 3496−3507. (11) Gebbie, M. A.; Wei, W.; Schrader, A. M.; Cristiani, T. R.; Dobbs, H. A.; Idso, M.; Chmelka, B. F.; Herbert Waite, J.; Israelachvili, J. N. Tuning Underwater Adhesion with Cation-φ Interactions. Nat. Chem. 2017, 9, 473−479. (12) Wei, Q.; Achazi, K.; Liebe, H.; Schulz, A.; Noeske, P. L. M.; Grunwald, I.; Haag, R. Mussel-Inspired Dendritic Polymers as Universal Multifunctional Coatings. Angew. Chem., Int. Ed. 2014, 53, 11650−11655. (13) Ahn, B. K.; Das, S.; Linstadt, R.; Kaufman, Y.; MartinezRodriguez, N. R.; Mirshafian, R.; Kesselman, E.; Talmon, Y.; Lipshutz, B. H.; Israelachvili, J. N.; Waite, J. H. High-Performance MusselInspired Adhesives of Reduced Complexity. Nat. Commun. 2015, 6, No. 8663. (14) North, M. A.; Del Grosso, C. A.; Wilker, J. J. High Strength Underwater Bonding with Polymer Mimics of Mussel Adhesive Proteins. ACS Appl. Mater. Interfaces 2017, 9, 7866−7872. (15) Jenkins, C. L.; Siebert, H. M.; Wilker, J. J. Integrating Mussel Chemistry into a Bio-Based Polymer to Create Degradable Adhesives. Macromolecules 2017, 50, 561−568. (16) Jenkins, C. L.; Meredith, H. J.; Wilker, J. J. Molecular Weight Effects upon the Adhesive Bonding of a Mussel Mimetic Polymer. ACS Appl. Mater. Interfaces 2013, 5, 5091−5096. (17) Wei, W.; Yu, J.; Gebbie, M. A.; Tan, Y.; Martinez Rodriguez, N. R.; Israelachvili, J. N.; Waite, J. H. Bridging Adhesion of MusselInspired Peptides: Role of Charge, Chain Length, and Surface Type. Langmuir 2015, 31, 1105−1112. (18) Maier, G. P.; Rapp, M. V.; Waite, J. H.; Israelachvili, J. N.; Butler, A. Adaptive Synergy between Catechol and Lysine Promotes Wet Adhesion by Surface Salt Displacement. Science 2015, 349, 628− 632. (19) Kim, H. J.; Hwang, B. H.; Lim, S.; Choi, B. H.; Kang, S. H.; Cha, H. J. Mussel Adhesion-Employed Water-Immiscible Fluid Bioadhesive for Urinary Fistula Sealing. Biomaterials 2015, 72, 104−111. (20) Shin, J.; Lee, J. S.; Lee, C.; Park, H. J.; Yang, K.; Jin, Y.; Ryu, J. H.; Hong, K. S.; Moon, S. H.; Chung, H. M.; Yang, H. S.; Um, S. H.; Oh, J. W.; Kim, D. I.; Lee, H.; Cho, S. W. Tissue Adhesive CatecholModified Hyaluronic Acid Hydrogel for Effective, Minimally Invasive Cell Therapy. Adv. Funct. Mater. 2015, 25, 3814−3824. (21) Hwang, D. S.; Yoo, H. J.; Jun, J. H.; Moon, W. K.; Cha, H. J. Expression of Functional Recombinant Mussel Adhesive Protein Mgfp-5 in Escherichia Coli. Appl. Environ. Microbiol. 2004, 70, 3352. (22) Lee, T. S.; Krupa, R. A.; Zhang, F.; Hajimorad, M.; Holtz, W. J.; Prasad, N.; Lee, S. K.; Keasling, J. D. BglBrick Vectors and Datasheets: A Synthetic Biology Platform for Gene Expression. J. Biol. Eng. 2011, 5, 15−17. (23) Carvajal-Vallejos, P.; Pallisse, R.; Mootz, H. D.; Schmidt, S. R. Unprecedented Rates and Efficiencies Revealed for New Natural Split Inteins from Metagenomic Sources. J. Biol. Chem. 2012, 287, 28686− 28696.

(24) Shah, N. H.; Eryilmaz, E.; Cowburn, D.; Muir, T. W. Naturally Split Inteins Assemble through a “Capture and Collapse” Mechanism. J. Am. Chem. Soc. 2013, 135, 18673−18681. (25) Stevens, A. J.; Brown, Z. Z.; Shah, N. H.; Sekar, G.; Cowburn, D.; Muir, T. W. Design of a Split Intein with Exceptional Protein Splicing Activity. J. Am. Chem. Soc. 2016, 138, 2162−2165. (26) Waite, J. H. Determination of (Catecholato)Borate Complexes Using Difference Spectrophotometry. Anal. Chem. 1984, 56, 1935− 1939. (27) Chang, D. P.; Abu-Lail, N. I.; Guilak, F.; Jay, G. D.; Zauscher, S. Conformational Mechanics, Adsorption, and Normal Force Interactions of Lubricin and Hyaluronic Acid on Model Surfaces. Langmuir 2008, 24, 1183−1193. (28) Micklavzina, B. L.; Longo, M. L. Characterization of Repulsive Forces and Surface Deformation in Thin Micellar Films via AFM. Langmuir 2017, 33, 10483−10491. (29) Fitzkee, N. C.; Rose, G. D. Reassessing Random-Coil Statistics in Unfolded Proteins. Proc. Natl. Acad. Sci. USA 2004, 101, 12497− 12502. (30) Xu, L. C.; Siedlecki, C. A. Effects of Surface Wettability and Contact Time on Protein Adhesion to Biomaterial Surfaces. Biomaterials 2007, 28, 3273−3283. (31) Mondon, M.; et al. Scanning-Force Techniques to Monitor Time-Dependent Changes in Topography and Adhesion Force of Proteins on Surfaces. Anal. Bioanal. Chem. 2003, 375, 849−855. (32) Rastogi, S. C.; Rastogi, P.; Mendiratta, N. Bioinformatics Methods and Applications: Genomics Proteomics and Drug Discovery, 3rd ed.; Prentice-Hall of India Pvt. Limited, 2008. (33) Janshoff, A.; Neitzert, M.; Oberdörfer, Y.; Fuchs, H. Force Spectroscopy of Molecular SystemsSingle Molecule Spectroscopy of Polymers and Biomolecules. Angew. Chem., Int. Ed. 2000, 39, 3212−3237. (34) Baek, D.; Hemthavy, P.; Saito, S.; Takahashi, K. Evaluation of Energy Dissipation Involving Adhesion Hysteresis in Spherical Contact between a Glass Lens and a PDMS Block. Appl. Adhes. Sci. 2017, 5, 4. (35) Li, J.; Celiz, A. D.; Yang, J.; Yang, Q.; Wamala, I.; Whyte, W.; Seo, B. R.; Vasilyev, N. V.; Vlassak, J. J.; Suo, Z.; Mooney, D. J. Tough Adhesives for Diverse Wet Surfaces. Science 2017, 357, 378−381. (36) Donahoe, C. D.; Cohen, T. L.; Li, W.; Nguyen, P. K.; Fortner, J. D.; Mitra, R. D.; Elbert, D. L. Ultralow Protein Adsorbing Coatings from Clickable PEG Nanogel Solutions: Benefits of Attachment under Salt-Induced Phase Separation Conditions and Comparison with PEG/Albumin Nanogel Coatings. Langmuir 2013, 29, 4128−4139. (37) Fant, C.; Elwing, H.; Höök, F. The Influence of Cross-Linking on Protein-Protein Interactions in a Marine Adhesive: The Case of Two Byssus Plaque Proteins from the Blue Mussel. Biomacromolecules 2002, 3, 732−741. (38) Choi, Y. S.; Kang, D. G.; Lim, S.; Yang, Y. J.; Kim, C. S.; Cha, H. J. Recombinant Mussel Adhesive Protein Fp-5 (MAP Fp-5) as a Bulk Bioadhesive and Surface Coating Material. Biofouling 2011, 27, 729−737. (39) Lin, Q.; Gourdon, D.; Sun, C.; Holten-Andersen, N.; Anderson, T. H.; Waite, J. H.; Israelachvili, J. N. Adhesion Mechanisms of the Mussel Foot Proteins Mfp-1 and Mfp-3. Proc. Natl. Acad. Sci. USA 2007, 104, 3782−3786. (40) Csörg, B.; Fehér, T.; Tímár, E.; Blattner, F. R.; Pósfai, G. LowMutation-Rate, Reduced-Genome Escherichia coli: An Improved Host for Faithful Maintenance of Engineered Genetic Constructs. Microb. Cell Fact. 2012, 11, No. 11. (41) Anderson, J. C.; Dueber, J. E.; Leguia, M.; Wu, G. C.; Goler, J. A.; Arkin, A. P.; Keasling, J. D. BglBricks: A Flexible Standard for Biological Part Assembly. J. Biol. Eng. 2010, 4, 1−12. (42) Tats, A.; Remm, M.; Tenson, T. Highly Expressed Proteins Have an Increased Frequency of Alanine in the Second Amino Acid Position. BMC Genomics 2006, 7, 1−13. (43) Bivona, L.; Zou, Z.; Stutzman, N.; Sun, P. D. Influence of the Second Amino Acid on Recombinant Protein Expression. Protein Expression Purif. 2010, 74, 248−256. I


Research Article

ACS Applied Materials & Interfaces (44) Tuller, T.; Carmi, A.; Vestsigian, K.; Navon, S.; Dorfan, Y.; Zaborske, J.; Pan, T.; Dahan, O.; Furman, I.; Pilpel, Y. Theory An Evolutionarily Conserved Mechanism for Controlling the Efficiency of Protein Translation. Cell 2010, 141, 344−354. (45) Leckband, D.; Israelachvili, J. Intermolecular Forces in Biology. Q. Rev. Biophys. 2001, 34, 105−267.

J


Microbially Synthesized Repeats of Mussel Foot Protein Display

Recommend Documents