A Chemically Functionalized Magnetic Nanoplatform for Rapid and

Dec 3, 2012 - Daily Rodriguez-Padrón , Alain R. Puente-Santiago , Alina M. Balu , Antonio A. ... Raluca M. Fratila , María Moros , Jesús M. de la F...
0 downloads 0 Views 4MB Size
Download PDF
Article pubs.acs.org/Biomac

A Chemically Functionalized Magnetic Nanoplatform for Rapid and Specific Biomolecular Recognition and Separation Po-Chiao Lin,*,† Ching-Ching Yu,‡ Huan-Ting Wu,‡ Ying-Wei Lu,‡,§ Chia-Li Han,§ An-Kai Su,§ Yu-Ju Chen,*,§ and Chun-Cheng Lin*,‡ †

Department of Chemistry, National Sun Yat-sen University, 70, Lienhai Road, Kaohsiung 80424, Taiwan Department of Chemistry, Tsing Hua University, 101, Sec. 2, Kuang Fu Road, Hsinchu 30013, Taiwan § Institute of Chemistry, Academia Sinica, 128, Academia Road Sec. 2, Taipei 115, Taiwan ‡

S Supporting Information *

ABSTRACT: We have developed a target-molecule-functionalized magnetic nanoparticle (MNP)-based method to facilitate the study of biomolecular recognition and separation. The superparamagnetic property of MNPs allows the corresponding biomolecules to be rapidly separated from crude biofluids with a significant improvement in recovery yield and specificity. Various MNPs functionalized with tag molecules (chitin, heparin, and amylose) were synthesized for recombinant protein purification, and several probe-functionalized MNPs, such as nitrilotriacetic acid (NTA)@MNP and Pk@MNP, exhibited excellent extraction efficiency for proteins. In a cell recognition study, mannose-functionalized MNPs allowed specific purification of Escherichia coli with FimH adhesin on the surface. In an immunoprecipitation assay, the antibody-conjugated MNPs reduced the incubation time from 12 to 1 h while maintaining a comparable efficiency. The functionalized MNPs were also used in a membrane proteomic study that utilized the interaction between streptavidin-functionalized MNPs and biotinylated cell membrane proteins. Overall, the functionalized MNPs were demonstrated to be promising probes for the specific separation of targets from proteins to cells and proteomics.

■

INTRODUCTION Because of emerging biothreats from microorganisms and secreted biomolecules, the development of rapid, sensitive, and accurate methods for the separation and identification of macro-biomolecules is important.1 Modern instruments and technologies for analyzing biomolecules, especially proteins, enable the large-scale analysis of a proteome in a single run. However, direct analysis of a diluted molecular target from a complicated biosample, such as human serum and crude cell lysates, remains difficult. Therefore, sample preparation and cleanup are critical steps for enhancing the detection sensitivity and providing reliable analyses. Furthermore, some pathogenic microorganisms are highly active in small numbers, which makes early detection difficult. Conventionally, a cell-culture process is necessary to satisfy the sample concentration requirement. However, the prolonged cell-culture time required for concentrating samples considerably increases the operation time and delays the diagnosis.2 To circumvent these problems, materials such as magnetic microbeads (MMBs) or resins have been widely used as solid supports for the separation and purification of biomolecules in past decades.3 However, although the biomolecular targets are in an aqueous medium, the rather large size of microscale materials results in a heterogeneous interface, which diminishes the extraction efficiency. Consequently, there is a need for new solid carriers with reduced size and improved biological tolerance to improve © 2012 American Chemical Society

the efficiency of the isolation and enrichment of target molecules. In the past decade, nanotechnology has been used for the study of extremely weak biomolecular interactions, such as protein−protein and protein−ligand interactions.4 The unique quantum properties of nanoscale materials (e.g., metal, insulator, and semiconductor) offer excellent prospects for designing highly sensitive and innovative biosensing devices and machines.5,6 Through appropriate surface modifications, functionalized nanomaterials can be useful for studying various biomolecular recognition events, especially characterized by a very weak binding affinity.7 Because of the high surface area-tovolume ratio of nanomaterials, the surface-bound ligand density should substantially increase and result in a “multivalent effect” that should enhance the binding affinity.8 For example, the mannose-functionalized gold nanoparticle has been demonstrated to be a good multivalent ligand for the study of carbohydrate−protein interactions, which generally exhibit a low binding affinity (Ka = 103−104 M−1).9 Furthermore, the rather high miscibility of nanoparticles in aqueous media provides a tremendous driving force for diffusion, which enhances the interaction between surface-bound ligands and Received: October 9, 2012 Revised: November 28, 2012 Published: December 3, 2012 160

dx.doi.org/10.1021/bm301567w | Biomacromolecules 2013, 14, 160−168

Biomacromolecules

Article

promising probes for the specific separation of various targets including proteins and cells.

the corresponding biomolecules in a nearly pseudohomogeneous phase. This property remarkably bridges the gap between aqueous media and solid materials. Among the available nanoparticles, magnetic nanoparticles (MNPs)10,11 have been used in many applications, such as protein sorting,12,13 contrast agents for magnetic resonance imaging (MRI), drug targeting and delivery, and hyperthermia therapy.14−17 On the basis of their unique magnetic properties, MNPs can be rapidly isolated from aqueous fluids without centrifugation or chromatography, thereby effectively preventing contamination and degradation during the purification process.18,19 Recently, we have successfully exploited MNPbased platforms in diverse applications, such as affinity extractions,20−22 enzymatic reactions,23,24 carbohydrate−lectin interactions,9 epitope mapping,25 small-molecule screening,26 and a phosphoproteomic study.22 However, the difficulty in controlling monodispersed MNPs in aqueous solution and the lack of robust chemistry for surface functionalization severely limit the applications of MNPs in biomedical studies. In addition, the nonspecific adsorption of undesired proteins on the MNPs results in ambiguous analyses. A chemical strategy for the modification of MNPs to provide low background interference is crucial to provide a useful probe for bioseparations. Therefore, in this report, amide bonds and noncovalent adsorption were used to assemble a diverse array of biomolecules, including carbohydrate antigens, polysaccharides, and proteins, onto the surface of magnetic iron oxide (Fe3O4) nanoparticles. In applications of these functionalized MNPs to biomolecular purifications (as illustrated in Figure 1),

■

MATERIALS AND METHODS

Materials. Tetraethyl orthosilicate (TEOS, Sigma-Aldrich), 3aminopropyltrimethoxysilane (APTMS, Sigma-Aldrich), 1-propanol (Acros), 25% ammonia solution (Acros), sinapinic acid (SA, SigmaAldrich), suberic acid bis-N-hydroxysuccinimide ester (DSS, SigmaAldrich), amylose (Sigma-Aldrich), and chitin (Sigma-Aldrich) were used as received. The monomeric acrylamide/bisacrylamide solution (40%, 29:1) was purchased from Bio-Rad. Trypsin (modified, sequencing grade) was obtained from Promega (Madison, WI). The BCA and Bradford protein assay reagent kits as well as the EZ-Link Sulfo-NHS-SS-Biotin were obtained from Pierce. Sodium dodecyl sulfate (SDS) was purchased from GE Healthcare. Ammonium persulfate (APS), iodoacetamide, and tetramethylethylenediamine (TEMED) were purchased from Amersham Biosciences. EDTA was purchased from Merck. Tris(2-carboxyethyl)phosphine hydrochloride (TCEP), triethylammonium bicarbonate (TEABC), phosphatebuffered saline (PBS), sodium carbonate (Na2CO3), sucrose, potassium chloride (KCl), magnesium chloride hexahydrate (MgCl2), HEPES, methyl methanethiosulfonate (MMTS), trifluoroacetic acid (TFA), and high-performance liquid chromatography (HPLC)-grade acetyl nitrile (ACN) were purchased from SigmaAldrich. Formic acid (FA) was purchased from Riedel de Haen (Seelze, Germany). Water was obtained from a Milli-Q Ultrapure Water Purification System (Millipore, Billerica, MA). Transmission electron microscopy (TEM) images were obtained using a JEM-2100 electron microscope (JEOL Co. 200 KV). Scanning electron microscopy (SEM) images were obtained using a Hitachi S-4200 scanning microscope. General Procedure for the Preparation of Small-MoleculeConjugated MNPs. NH2@MNP (1 mg) was prepared as reported27 and dispersed into 100 μL of dimethyl sulfoxide (DMSO). After sonication for 30 min, suberic acid DSS (5 mg, 0.015 mmol) was added to the solution, which was then stirred at room temperature for 1 h. The resulting nanoparticles (OSu@MNP) were washed with DMSO (100 μL) three times to remove excess DSS. Fifty microliters of a solution containing the amine-functionalized small molecule, i.e., Pk antigen25 or NTA,22 at a concentration of 100 mM was added to the above activated OSu@MNP and then shaken at 4 °C for 12 h. Then, 50 μL of 100 mM aqueous ethanolamine solution was added to the mixture, followed by shaking for an additional 6 h at room temperature. The resulting nanoparticles were washed with PBS (100 μL, pH 7.4, 0.1 M) five times to give small-molecule-conjugated MNPs as a black powder. Basic Hydroylsis of Chitin. One gram of chitin powder (from crab shells, practical grade) was hydrolyzed using a 50% NaOH (10 mL) solution at 25 °C for 7 days. The resulting viscid mixture was diluted with 90 mL of deionized H2O (10 x dilution). Then, 100 mL of propanol was added to the mixture, which was centrifuged at 8500 rpm for 15 min. The isolated precipitation was redissolved in 100 mL of deionized H2O, and 100 mL of propanol was then added to the mixture. After centrifugation at 500 rpm for 2 min, the large chitin fragments were removed, and the supernatant was subsequently centrifuged at 8500 rpm for 15 min. After the above steps were repeated two times, the water-soluble chitin (∼500 mg) was harvested as a white powder after lyophilization. Preparation of Polysaccharide@MNP. The water-soluble polysaccharides (chitin or starch, 10 mg) were dissolved in deionized H2O (3 mL). Then, 5 mg of the Fe3O4 MNPs were added, and the solution was heated to reflux at 100 °C for 2 h. After being washed three times with deionized H2O (1 mL), the polysaccharide@MNPs were obtained as a black powder. Preparation of Protein-Conjugated MNPs. A 25 μL volume of a PBS buffer solution (pH 7.4) containing the desired protein (≥5 μg/ μL) was added to the OSu@MNP (0.1 mL of 10 mg/mL solution), and the resulting mixture was incubated at 4 °C for 30 min. Then, 25 μL of an aqueous solution containing the surface blocking reagent,

Figure 1. General concept of the diverse applications of biomoleculefunctionalized MNPs.

several probe-functionalized MNPs, such as nitrilotriacetic acid(NTA) and trisaccharide Pk antigen-functionalized MNPs (NTA@MNP and Pk@MNP, respectively) exhibited excellent extraction efficiency of target proteins. In a cell recognition study, mannose-functionalized MNPs permitted specific purification of Escherichia coli with FimH adhesin on the surface. In an immunoprecipitation (IP) assay, antibodyconjugated MNPs reduced the incubation time from 12 to 1 h while maintaining comparable extraction efficiency. The functionalized MNPs were also used in a study of membrane proteomics that utilized the interaction between streptavidinfunctionalized MNPs and biotinylated cell membrane proteins. Overall, the functionalized MNPs were demonstrated to be 161

dx.doi.org/10.1021/bm301567w | Biomacromolecules 2013, 14, 160−168

Biomacromolecules

Article

methoxyethylene glycol (MEG),27 at a concentration of 100 mM was added to the mixture, followed by incubation at 4 °C for 12 h. After being separated by a magnet, the protein-conjugated MNPs were washed 5 times with PBS (100 μL, pH 7.4, 0.1 M) to yield the protein@MNPs as a black powder. Purification of Shiga-like Toxin B Subunit by Pk@MNPs. E. coli cells that expressed the Shiga-like toxin B subunit (Slt-B) were lysed in lysis buffer (0.1 mg/mL polymyxin B and 5 mM EDTA in PBS buffer) at 37 °C for 1 h. The Pk@MNP (500 μg) was added to the cell lysate solution, and the resulting mixture was vigorously mixed for 30 min. The supernatant was subsequently removed by magnetic separation, and the resulting MNPs were washed three times with a PBS buffer solution. The MNP complex was then dispersed in an eluting buffer (4.8 M MgCl2), and the solution was stirred for 10 min. After being separated by a magnet, the eluted Slt-B was collected and analyzed using a BCA protein quantitation assay. The extraction efficiency of Pk@MNPs was determined to be 31 ± 1.2 μg of protein per mg of Pk@MNPs. Purification of Chitin-Binding Domain (CBD)-Fused Cytidine Monophosphate (CMP)−Sialic Acid Synthetase by Chitin@ MNPs. E. coli BL21 (DE3) cells that expressed CBD-fused CMP− sialic acid synthetase (CSS)23 were lysed by ultrasonication in the column buffer solution (20 mM Tris-HCl, pH 8.5, 500 mM NaCl, 0.1 mM EDTA). The chitin@MNPs (5 mg) were incubated with 1 mL of the cell lysate, and the resulting mixture was vigorously mixed for 30 min. The MNP complexes were then separated using a magnet, and the isolated MNP complex was washed three times with the column buffer solution (500 μL). The MNP complex was then dispersed in the eluting buffer (80 mM 1,4-dithiothreitol (DTT) in column buffer) and stirred for 10 min. After being separated with a magnet, the eluted CBD-fused CSS was collected and analyzed using the BCA protein quantitation assay. The extraction efficiency of the chitin@MNPs was determined to be 9.36 ± 5.5 μg of protein per milligram of chitin@ MNPs. His-Tagged Protein Purification by NTA@MNPs. The E. coli cells that expressed the His-tagged protein were lysed in a buffer (300 mM NaCl, 10 mM imidazole, and 50 mM sodium phosphate buffer, pH 8.0) for 15 min. The Ni2+ chelated NTA@MNPs (200 mg) were incubated with 2 mL of cell lysate, and the resulting mixture was vigorously mixed for 1 h at 4 °C. The supernatant was then removed by magnetic separation, and the resulting MNPs were washed 2 times with Tris-HCl buffer (20 mM) and then 2 times with a 10 mM imidazole solution. The MNP complex was then dispersed in the elution buffer (50−500 mM imidazole solution) and stirred for 10 min. After being separated with a magnet, the eluted His tag fusion protein was collected and analyzed using the Bradford protein quantitation assay. The extraction efficiency of NTA@MNP was determined to be 7.38 ± 0.1 μg of protein per mg of NTA @MNPs. Purification of Human IgG Antibody by Protein G@MNPs. Protein G@MNPs (5 mg) were incubated with 1 mL of human serum, and the resulting solution was vigorously mixed for 30 min. The MNP complex was then separated using a magnet, and the isolated MNP complex was washed 3 times with a TTBS buffer (500 μL). The MNP complex was then dispersed in the elution buffer (0.1 M Gly, pH 3.0) and incubated for 10 min. After being separated with a magnet, the eluted human IgG was collected and analyzed using the BCA protein quantitation assay. The extraction efficiency of human IgG antibody was determined to be 125.68 ± 10.9 μg of protein per mg of protein G@MNPs. Rapid IP by Anti-β-catenin@MNPs. Nanoprobe based immunoextraction was performed by first precleaning the sample solution (100 μL) with 200 μg of blank MNPs at room temperature for 2 min to reduce nonspecific binding. After removal of the blank MNPs with a magnet, 50 μg of anti-β-catenin polyclonal antibody-conjugated MNPs (pAb@MNPs) were added, and the mixture was gently rotated for 1 h. After affinity extraction of the target antigens on the MNPs, the nonspecifically adsorbed molecules were removed by washing the MNPs with 100 μL of an incubation buffer (25 mM Tris-HCl, 200 mM NaCl, 1 mM EDTA, 0.5 mM DTT, 1% NP-40, and 10% isopropanol solution, pH 7.4) and 100 μL ammonium bicarbonate (50

mM), and each washing was repeated two times. The incubated MNPs were resuspended in 20 μL of 1× sample buffer (125 mM Tris-HCl, pH 6.8, containing 2% SDS, 10% glycerol, 0.001% bromophenol blue and 5% 2-mercaptoethanol). Boiled aliquots were subjected to SDS polyacrylamide gel electrophoresis (SDS-PAGE) in a 10% acrylamide gel using a mini-gel apparatus (BioRad, Germany) and blotted onto nitrocellulose membranes (Millipore) at a current density of 1 mA/ cm2 using a semidry blot system (BioRad, Germany). The membranes were blocked with 5% skim-milk powder in TTBS (50 mM Tris-HCl, pH 7.5, containing 150 mM NaCl). The rabbit polyclonal anti-βcatenin antibody was diluted to 1:1000 using TTBS buffer. Horseradish peroxidase-conjugated antirabbit polyvalent rabbit immunoglobulin, which served as a secondary antibody, was used at a dilution of 1:5000−20000 (Amersham Pharmacia Biotech). The detection of antibodies was performed using the ECL system (Amersham) based on the manufacturer’s instructions. Separation of Bacteria by Mannose@MNPs. E. coli ORN178 (transformed with pDsRed2 for red fluorescence) and ORN208 (transformed with pGFP for green fluorescence) strains were used as previously reported.28 E. coli cells were grown overnight in LB medium at 37 °C to a density of approximately 107 cells/mL. Mannose@MNPs were added to a 1.5 mL centrifuge tube that contained an E. coli solution (107 cells/mL, 1 mL) in LB medium. The resulting mixture was gently shaken at room temperature for 30 min. After incubation, the samples were placed on a magnet for 1 min. The supernatant was carefully removed by pipetting, and the resulting nanoparticles were washed three times and then resuspended with LB to the original volume. The presence of E. coli was detected using fluorescent microscopy. Biotinylation of Membrane Proteins. HeLa cells were grown at 37 °C in RPMI 1640 medium with 10% FBS. The dishes (10 cm) that contained the cells were washed three times with prewarmed (37 °C) PBS, and then 10 mL of EZ-Link Sulfo-NHS-SS-Biotin solution (0.1 mg/mL, freshly prepared with PBS) was added. After the cells were incubated at 37 °C for 15 min, the reaction solution was removed by decanting the supernatant, and then 10 mL of lysine solution (0.1 mg/ mL) was added to quench the reaction. Affinity Isolation of Membrane Proteins of HeLa Cells. Biotinylated cells were washed twice with PBS and transferred into 1 mL of cold PBS. The cells were recollected by centrifugation at 1000g at 4 °C for 5 min and then resuspended with a hypotonic buffer (10 mM HEPES at pH 7.5, 1.5 mM MgCl2, 10 mM KCl, and 1× protease inhibitor mixture (Calbiochem)). The solution was kept on ice for 15 min and then homogenized with 50 strokes of a Dounce homogenizer. The nuclei were removed by centrifugation at 1000g at 4 °C for 10 min. The concentration of KCl in the supernatant was adjusted to 150 mM through the addition of a stock KCl solution (1 M). A 1-mL aliquot of suspended streptavidin-conjugated magnetic nanoparticles (streptavidin@MNPs, 10 mg/mL, prewashed four times with ice-cold hypotonic buffer before use) was added to the supernatant, and the resulting suspension was rotated at 4 °C overnight. The streptavidin@ MNPs were collected using a magnetic plate and sequentially washed three times with 1 mL of ice-cold 1 M KCl, three times with 1 mL of ice-cold 0.1 M Na2CO3 (pH 11.5), and then once with 50 mM TEABC. The streptavidin@MNPs were resuspended in a solution containing 6 M urea, 5 mM EDTA, and 2% SDS in 0.1 M TEABC to obtain the plasma membrane fraction. Gel-Assisted Digestion of Plasma Membrane Proteins. To completely elute the plasma membrane proteins from the streptavidin@MNPs, the proteins were chemically reduced by 5 mM TCEP at 37 °C for 30 min followed by reaction with an alkylation reagent, MMTS (2 mM), at 37 °C for 30 min. The streptavidin@ MNPs were removed using a magnetic plate, and the supernatant was subjected to gel-assisted digestion.29 To incorporate the proteins into a gel directly in the microcentrifuge tube, 18.5 μL of acrylamide/ bisacrylamide solution (40%, 29:1), 2.5 μL of 10% APS, and 1 μL of 100% TEMED were then applied to the membrane protein-containing tube without electrophoresis. The gel was cut into small pieces and then washed several times with 1 mL of TEABC that contained 50% ACN. The gel samples were further dehydrated with 100% ACN and 162

dx.doi.org/10.1021/bm301567w | Biomacromolecules 2013, 14, 160−168

Biomacromolecules

Article

Scheme 1. Preparation of Biomolecule-Functionalized MNPsa

(a) TEOS, APTMS, PrOH, 55 °C, 16 h; (b) DSS, DMSO, r.t., 1 h; (c) R-NH2, DMF, 4 °C, 12 h, then capping reagent; (d) protein/antibody, 4 °C, 30 min, then capping reagent, 12 h; (e) polysaccharides, H2O, 110 °C, 2 h. TEOS, tetraethyl orthosilicate; APTMS, 3-aminopropyltrimethoxysilane; DSS, suberic acid bis-N-hydroxysuccinimide ester. a

then completely dried using a SpeedVac. Proteolytic digestion was then performed using trypsin (protein:trypsin = 10:1, w/w) in 25 mM TEABC followed by incubation at 37 °C overnight. The peptides were extracted from the gel using sequential extraction with 200 μL of 25 mM TEABC, 200 μL of 0.1% TFA in water, 200 μL of 0.1% TFA in ACN, and 200 μL of 100% ACN. The solutions were combined and concentrated using a SpeedVac. LC-ESI-MS/MS Analysis. The peptide samples were reconstituted in 0.1% (v/v) FA in H2O and analyzed using a Waters Q-TOF Premier (Waters Corp., Milford, MA). The samples were injected into a 20 mm × 180 μm trap column, separated by a 200 mm × 75 mm Waters1 ACQUITY 1.7 μm BEH C18 column using a nanoACQUITY Ultra Performance LC system (Waters Corp., Milford, MA), and finally eluted with a linear gradient of 0−80% of 0.1% FA in ACN for 120 min at 300 nL/min. The MS was operated in the electrospray ionization positive V mode with a resolving power of 10 000. A NanoLockSprayTM source (Waters Corp., Milford, MA) was used for accurate mass measurements, and the lock mass channel was sampled every 30 s. The mass spectrometer was calibrated with a synthetic human [Glu1]-fibrinopeptide B solution (1 pmol/μL, Sigma-Aldrich) delivered through the NanoLockSpray source. Data acquisition was operated in the data-directed analysis mode to automatically switch between a full MS scan (m/z 400−1600, 0.6 s) and three MS/MS scans (m/z 100−1990, 1.2 s for each scan) sequentially for the three most intense ions present in the full MS scan. Data Processing and Analysis. The peak list resulting from MS/ MS spectra was exported to mgf format by Mascot Distiller v2.1.1.0. The combination of multiple data sets was batch-searched by Mascot v2.2 (Matrix science, London, United Kingdom) against the International Protein Index (IPI) human database30 (v3.64, 84,032 sequences) from the European Bioinformatics Institute using the following constraints: only tryptic peptides with up to two missed cleavage sites, and 0.1-Da mass tolerance for MS and 0.1-Da mass

tolerance for MS/MS fragment ions. Methylation (Cys), deamidation (Asn, Gln), and oxidation (Met) were specified as variable modifications. Only unique peptides with a Mascot individual ion score higher than the Mascot identity scores (p < 0.05) were confidently assigned. In each MS/MS spectrum, at least four b- and yions were observed. To evaluate the false discovery rate in protein identification, we performed a decoy database search against a randomized decoy database created by Mascot using identical search parameters and validation criteria. Annotations. For subcellular localization and molecular function annotations, all of the proteins identified in this study were analyzed against the Gene Ontology consortium using the GoMiner software package.

■

RESULTS AND DISCUSSION Preparation of Functionalized MNPs. The Fe 3 O 4 nanoparticles were prepared as previously reported,27 and they were used as the core structure for the preparation of diversely functionalized MNPs. For conjugation of smallmolecule probes and proteins, the surface of the core MNP was further modified to have amino-functionality using APTMS through the sol−gel process.27 The resulting NH2@MNPs were then transformed to activated ester OSu@MNPs by reacting with the suberic acid DSS cross-linker. The activated OSu@MNPs were respectively conjugated with aminecontaining molecules, such as NTA, trisaccharide Pk antigen31,32 (αGal(1→4)βGal(1→4)βGlc), mannose, streptavidin, antibody, and protein A/protein G, through the formation of amide bonds, as shown in Scheme 1 (for characterization of the functionalized MNPs, see Figure S1 A−E). For the polysaccharide-functionalized MNPs, the polysaccharide, such 163

dx.doi.org/10.1021/bm301567w | Biomacromolecules 2013, 14, 160−168

Biomacromolecules

Article

Figure 2. Purification of Slt-B subunit by Pk@MNPs. (A) workflow of Pk@MNP-based purification; (B) directed MALDI-TOF measurement for MNPs and SDS electrophoresis analysis with Coomassie blue staining.

lectin was completely separated from the medium by Pk@ MNPs after incubation for 30 min (Figure S1 (F)). Shiga and Shiga-like toxins (SLT) belong to the family of classic AB5-type bacterial toxins, which contain a single catalytically active A-subunit and five copies of a B-subunit, in which the B-subunit pentamer has been reported to specifically recognize the globotriose (Pk) blood group antigen.33−35 In this study, the SLT B-subunit (Slt-B) was cloned into the high-level expression vector pJLB28. After overexpression, the E. coli cells were collected and lysed using polymyxin B, which is an antibiotic that permeates cell membranes. The crude cell lysate was incubated with Pk@MNP for 30 min and then separated using a magnet. The resulting Slt−B-Pk@MNP complex was washed three times with PBS followed by direct detection with MALDI-TOF MS, which did not require tedious desalting and elution processes to produce a peak at 7672.8 Da that indicated the presence of the Slt-B monomer (Figure 2). Slt-B was obtained by eluting the Slt-B−Pk@MNP complex with 4.8 M MgCl2. Additionally, the efficiency of the isolation of Slt-B was characterized by SDS-PAGE. Only a single protein band, without any contaminating protein, was observed on the gel. Furthermore, through an appropriate washing process, the Pk@ MNPs were regenerated and reused up to six times without losing their affinity for Slt-B (Figure 2). The extraction efficiency was determined to be 31 ± 1.2 μg of Slt-B per milligram of Pk@MNPs (Table 1). Compared with our previously reported Pk-AuNP-based probe,25 Pk@MNP was able to purify proteins faster and more easily. For the recognition and isolation of large macrobiomolecules, such as those on the cell surface, MNPs are advantageous because (i) the high surface/volume ratio offers greater surface area for attaching carbohydrates and for capturing pathogens and (ii) the size of the NP is generally approximately 2 orders of magnitude less than that of a bacterium, which allows multiple NPs to be attached to a bacterial cell to facilitate magnetic separation. To evaluate the applicability of the MNP-based platform in the study of cellular recognition, FimH adhesin, which is a mannose adhesion protein expressed by E. coli, was adopted to demonstrate the

as amylose, heparin, and chitin, was hydrolyzed and subsequently coated on the core Fe3O4 nanoparticles through noncovalent electrostatic interactions and/or hydrogen bonding between the hydrolyzed polysaccharide and the MNP (Scheme 1). The polar ligands on the surface facilitated the dispersion of the functionalized MNPs in aqueous solution. Applications of Small-Molecule Modified MNPs for Bioseparation. Utilizing the robust surface chemistry of the MNPs, different affinity-based molecular tags were synthesized and assembled onto the surface of the MNPs. The resulting functionalized MNPs were used to bind to the corresponding proteins that were expressed in crude cell lysates. We first applied NTA@MNPs, which are commonly used in the Ni2+/ His tag protein purification system. The NTA-attached MNPs (NTA@MNP) were charged with Ni2+ followed by incubation with E. coli cell lysate that contained 6x His-tag fused recombinant Chi-A. After simple incubation and magnetic separation procedures, the purified protein was obtained. As shown in Figure S2, analysis using denatured SDS-PAGE with Coomassie blue staining revealed an intense band at 90 kDa that indicated the overexpression of Chi-A. To extend the applications of the nanoprobe-based protein purification system and demonstrate the advantage of using MNPs, we then used carbohydrate-conjugated MNPs to purify target lectins and bacteria. To assess the efficiency of MNP separation in biofluids, the unique affinity interactions of trisaccharide Pk antigen/PA-IL lectin (a galactose-binding protein) and Pk antigen/Shiga-like toxin B subunit were chosen as models to evaluate the efficiency of the MNP-based protein purification system. In a preliminary study, Pk antigen-conjugated MNPs (Pk@MNPs) were used to recognize and isolate the PA-IL lectin from an aqueous medium. After incubation at room temperature, the solid phase was separated from the aqueous medium within 1 min using a conventional magnet, and the remaining PA-IL in the supernatant was also analyzed using the matrix-assisted laser desorption ionization-time-of-flight mass spectrometry technique (MALDI-TOF MS). The results indicate that the PA-IL 164

dx.doi.org/10.1021/bm301567w | Biomacromolecules 2013, 14, 160−168

Biomacromolecules

Article

obtained aggregates exhibited a red fluorescent signal during microscopic observation, which indicates the specific separation of ORN178 (Figure 3F). Therefore, the advantages of using the nanoprobes for cell separation are that (i) only a short incubation time (30 min) is required, and (ii) a simple washing process with a PBS buffer is sufficient. The high sensitivity and specificity make this nanoprobe an ideal purification tool for cell recognition and sorting. Separation of Target Proteins by PolysaccharideFunctionalized MNPs. Because of the low water solubility of high-molecular-weight polysaccharides, basic hydrolysis is required to produce low-molecular-weight oligosaccharide fragments that show aqueous miscibility. Thus, through noncovalent electrostatic interactions and/or hydrogen bonding between the hydrolyzed polysaccharides and MNPs, amylose, heparin, and chitin were successfully coated on the MNPs. To demonstrate the diversity of the functionalized MNPs, we examined the interactions between the polysaccharide-functionalized MNPs and their corresponding binding proteins, including heparin@MNPs and a heparin-binding protein (A27L, Figure S4), amylose@MNPs and starch-binding protein (SBP, Figure S5) fused to enhanced green fluorescent protein (EGFP), and chitin@MNPs and CBD-fused CSS (Figure S6). The high density of surface carbohydrate ligands and the homogeneous colloid interface resulted in impressive extraction efficiency, as shown in Table 1. Among these polysaccharide binding proteins, SBP and CBD are popular affinity tags in protein engineering and are widely used during recombinant protein purification. Amylose@MNP and chitin@MNP were able to respectively isolate SBP-fused EGFP and CBD-fused CSS with a yield of 16.39 ± 1.2 and 9.36 ± 5.5 μg per milligram of nanoparticles, which reflects differences in binding affinity and in the individual size of the proteins. Purification of IgG Antibody by Protein A/Protein G @ MNPs and Rapid IP by Antibody@MNPs. To expand the

Table 1. Efficiency of Purification of Expressed Proteins by Functionalized MNPs probe

target protein

Carbohydrate Type Pk antigen Shiga-like toxin B subunit heparin heparin binding protein amylose starch-binding domain (SBD) fusion EGFP chitin CBD fused CSS General Probe NTA His tag fusion Chi-A protein protein G human IgG antibody protein A human IgG antibody

extract efficiency (extract protein μg/mg MNP) 31.04 ± 1.2 10.67 ± 0.6 16.39 ± 1.2 9.36 ± 5.5 7.42 ± 0.1 125.68 ± 10.9 97.59 ± 3.3

specific interaction of functionalized nanoprobes with targeted cells. To visualize the formation of mannose@MNP−E. coli aggregates using fluorescent microscopy, two E. coli strains, ORN178 and ORN208, were respectively transformed with pDsRed2 and pGFP genes to produce red and green fluorescent signals. ORN 178 expresses FimH on its pili, whereas ORN 208 has no expression of FimH.28,36 After incubation of the mannose@MNPs with the respective bacterial strains for 10 min, the magnetically separated MNPs complexed with pDsRed2-expressing ORN178 produced clear red fluorescence, whereas no distinguishable signal was observed when ORN208 was used (Figure 3A vs C). Using optical phase imaging, we confirmed red fluorescence in ORN178 bacteria captured on the surface of the MNPs (Figure 3B and Figure S3), which exhibited bacteria-induced aggregation. In contrast, no notable aggregation occurred after treatment of the mannose@MNPs with ORN208 (Figure 3C,D). To demonstrate the efficiency of the mannose@MNPs for the separation of specific bacterial strains, the mannose@MNPs were mixed with a pool containing both ORN178 and ORN208 cells (Figure 3E) for 30 min. After magnetic separation, the

Figure 3. Images of ORN 178 cells captured by mannose@MNPs (A) under fluorescent microscopy and (B) optical microscopy; images of ORN 208 cells captured by mannose@MNPs (C) under fluorescent microscopy and (D) optical microscopy; (E) mixing pool of ORN 178 and ORN 208 cells; (F) fluorescent microscope image of bacteria extracted by mannose@MNPs from a mixed pool. 165

dx.doi.org/10.1021/bm301567w | Biomacromolecules 2013, 14, 160−168

Biomacromolecules

Article

conjugated MMBs were also applied to isolate the antibody-βcatenin complex from the cell lysate as a conventional control. As shown in lane 1 of Figure 4, after 16 h of IP with MMBs,

application of probe molecules from small organic molecules to biomacromolecules, two specific antibody-binding proteins, protein G and protein A, were assembled onto the surface of MNPs to give nanoextractors that can be used for the purification of human IgG antibodies from human serum. To avoid the nonspecific binding of highly abundant human serum albumin (HSA) and other highly abundant proteins, such as apolipoproteins and transferrin, the surface blocking spacer MEG27 was used to suppress the adsorption of undesired proteins. The loading and extraction capacities of protein Gand protein A-conjugated MNPs (G@MNP and A@MNP) were carefully studied using bicinchoninic acid (BCA) protein quantitation assays and SDS-PAGE analyses (Figures S7−S10). The protein G@MNPs were used to successfully purify human IgG from human serum (125.67 ± 10.86 μg protein/mg MNPs) without contamination by undesired proteins, whereas the purification efficiency from human serum of protein A@ MNPs was 97.59 ± 3.43 μg of IgG per milligram of MNPs (Table 1). IP is a critical technique for the separation of specific proteins from complex biofluids. The IP technique uses an antibody to interact with a specific protein to form a protein complex. In the IP technique, protein G- or protein A-conjugated beads are used to efficiently separate the protein complex from an aqueous medium, which is followed by identification of the isolated proteins.37,38 However, the heterogeneous interface between microbeads and the aqueous medium considerably reduces the comprehensive interaction and subsequently diminishes the efficiency of the IP technique. Furthermore, the nonspecific binding of undesired contaminants is a primary limiting factor in IP techniques. To suppress nonspecific interactions, an additional step known as precleaning is introduced before the IP process. During the precleaning step, the target sample pool is treated with an excess amount of protein G-conjugated beads before the addition of the antibody. The precleaning process is expected to remove the nonspecifically interacting proteins. To apply protein-conjugated MNPs to the widely used IP technique, antibody-functionalized MNPs (Ab@MNPs) were prepared as previously reported27 and used as a new IP extractor to overcome the aforementioned problems. In contrast to the use of immobilized protein G during the conventional IP process, Ab@MNPs are directly separated using an external magnet. Furthermore, the Ab@MNPs can be recycled if necessary, which significantly reduces the cost of the antibody. In the developed MNP-IP process, a 4-fold excess (compared to Ab@MNPs) of blank MEG@MNPs was used as a “pre-cleaning” probe to remove the nonspecific binding molecules, and this process only required 2 min of incubation. After the precleaning step, the blank MNPs were removed and Ab@MNPs were added to the supernatant, which was followed by gentle rotation at 25 °C for 1 h. Subsequently, the nanoparticles were magnetically separated, and the resulting complex was washed twice with incubation buffer (25 mM TrisHCl, 200 mM NaCl, 1 mM EDTA, 0.5 mM DTT, 1% NP-40, and 10% isopropanol solution at pH 7.4) followed by ammonium bicarbonate buffer. To examine the applicability of the MNP-IP process, β-catenin, which is a component of the Wnt/wingless signaling cascade,39 was used as a target protein that is present at low concentrations in crude cell lysates. Polyclonal anti-β-catenin Ab@MNP was used as a probe in the MNP-IP process to purify β-catenin from SW480 cell lysates. To compare separation efficiencies, commonly used protein G-

Figure 4. SDS-PAGE of IP of β-catenin by antibody-conjugated microbeads and MNPs. Lane 1: silver staining of microbead-based IP; lane 2: silver staining of MNP-based IP; lane 3: Western blotting of microbead-based IP; lane 4: Western blotting of MNP-based IP.

silver staining showed no distinguishable bands around 98 kDa (corresponding to β-catenin). In addition, the Western blot result (lane 3) only exhibited a weak β-catenin signal. These results indicate that MMB is not suitable as a solid extractor for such diluted target proteins. By contrast, MNP-IP resulted in a remarkable β-catenin band (lane 2, silver stain), which was further confirmed by Western blotting (lane 4) and in-gel digestion followed by LC-MS/MS analysis. The identified proteins also contained several known proteins associated with β-catenin, which indicates the potential of MNP-IP for rapid and sensitive protein separation and protein−protein interaction studies. Furthermore, the use of MNP-IP substantially shortened the incubation time from 16−24 h (conventional IP) to less than 2 h. Affinity Isolation of Plasma Membrane Proteins of HeLa Cells by Streptavidin@MNPs. Plasma membrane proteins play important roles in many cellular functions, such as ion and solute transport, sensing of chemical and physical signals from external stimuli, and adaptation to environmental and cell-matrix interactions. Therefore, membrane proteins, especially plasma membrane proteins, have the greatest potential to be candidates for disease biomarkers.40 To demonstrate the applicability of functionalized MNPs in the discovery of potential biomarkers, a streptavidin@MNP24 was designed to be an affinity probe for plasma membrane proteins labeled with biotin. In this study, the pilot experiments were performed on HeLa cells. The prewashed HeLa cells were labeled with EZ-Link Sulfo-NHS-SS-Biotin solution, washed with PBS and then incubated with lysine solution to quench the reaction. After being washed with PBS, the cells were disrupted in a hypotonic buffer using a Dounce homogenizer, and then the nuclei were removed by centrifugation; consequently, a postnucleus supernatant was generated. The supernatant was incubated with streptavidin@MNPs overnight. Subsequently, the streptavidin@MNPs were collected using a magnet to yield a biotinylated-strepavidin@MNPs complex. Through reduction and alkylation reactions, the biotinylated plasma membrane proteins were eluted from the streptavidin@MNPs and subjected to gel-assisted digestion, as described previously.41 The extracted peptides were analyzed using the liquid 166

dx.doi.org/10.1021/bm301567w | Biomacromolecules 2013, 14, 160−168

Biomacromolecules

Article

Finally, using streptavidin@MNPs, we successfully isolated and identified 125 proteins located on or associated with the cell surface and organelle membranes of HeLa cells. The cooperation of this functionalized MNP platform with various analytical tools such as MS, molecular spectral analysis, and biomedical imaging is believed to have an important role for further studies of important bimolecular interactions, especially those with extremely weak affinity.

chromatography electrospray ionization tandem mass spectrometry technique (LC-ESI-MS/MS), and the proteins were identified using the Mascot search engine. In this study, 203 proteins were identified. Among the proteins with known subcellular localizations according to the Gene Ontology annotations, 100 membrane proteins (69 plasma membrane and 31 organelle membrane proteins) and 25 membraneassociated cytoskeletal proteins were identified (Figure 5).

■

ASSOCIATED CONTENT

S Supporting Information *

Experimental details of functionalized MNP synthesis and characterization, protein purification procedure, SEM image, and SDS-PAGE results. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected] (C.-C.L.); [email protected]. edu.tw (P.-C.L.); [email protected] (Y.-J.C.). Notes

The authors declare no competing financial interest.

■

Figure 5. Subcellular localization of affinity purified proteins using streptavidin@MNPs. The annotations were obtained from the Gene Ontology database.

ACKNOWLEDGMENTS This work was financially supported by the National Sun-Yatsen University, National Tsing Hua University, Academia Sinica, and the National Science Council, Taiwan.

■

Excluding the unknown proteins, 70.6% of the identified proteins were annotated as membrane proteins. A similar experiment was also performed using avidin@MNPs (data not shown), and both experiments resulted in consistent and specific purification of membrane proteins from a complex cell lysate.

REFERENCES

(1) Lazcka, O.; Del Campo, F. J.; Muñoz, F. X. Biosens. Bioelectron. 2007, 22, 1205−1217. (2) Leoni, E.; Legnani, P. P. J. Appl. Microbiol. 2001, 90, 27−33. (3) Svec, F.; Fréchet, J. M. Science 1996, 273, 205−211. (4) Katz, E.; Willner, I. Angew. Chem., Int. Ed. 2004, 43, 6042−6108. (5) Niemeyer, C. M. Angew. Chem., Int. Ed. 2001, 40, 4128−4158. (6) Alivisatos, P. Nat. Biotechnol. 2004, 22, 47−52. (7) Rosi, N. L.; Mirkin, C. A. Chem. Rev. 2005, 105, 1547−1562. (8) Gestwicki, J. E.; Cairo, C. W.; Strong, L. E.; Oetjen, K. A.; Kiessling, L. L. J. Am. Chem. Soc. 2002, 124, 14922−14933. (9) Lin, C.-C.; Yeh, Y.-C.; Yang, C.-Y.; Chen, G.-F.; Chen, Y.-C.; Wu, Y.-C.; Chen, C.-C. Chem. Commun. 2003, 23, 2920−2921. (10) Lu, A.-H.; Salabas, E. L.; Schüth, F. Angew. Chem., Int. Ed. 2007, 46, 1222−1244. (11) Beveridge, J. S.; Stephens, J. R.; Williams, M. E. Annu. Rev. Anal. Chem. 2011, 4, 251−273. (12) Pan, Y.; Long, M. J. C.; Lin, H.-C.; Hedstrom, L.; Xu, B. Chem. Sci. 2012, 3, 3495−3499. (13) Pan, Y.; Du, X.; Zhao, F.; Xu, B. Chem. Soc. Rev. 2012, 41, 2912−2942. (14) Mornet, S.; Vasseur, S.; Grasset, F.; Duguet, E. J. Mater. Chem. 2004, 14, 2161−2175. (15) Gupta, A. K.; Gupta, M. Biomaterials 2005, 26, 3995−4021. (16) Weissleder, R.; Kelly, K.; Sun, E. Y.; Shtatland, T.; Josephson, L. Nat. Biotechnol. 2005, 23, 1418−1423. (17) Won, J.; Kim, M.; Yi, Y.-W.; Kim, Y. H.; Jung, N.; Kim, T. K. Science 2005, 309, 121−125. (18) Gu, H.; Xu, K.; Xu, C.; Xu, B. Chem. Commun. 2006, 9, 941− 949. (19) Liu, Z.; Li, M.; Yang, X.; Yin, M.; Ren, J.; Qu, X. Biomaterials 2011, 32, 4683−4690. (20) Chou, P.-H.; Chen, S.-H.; Liao, H.-K.; Lin, P.-C.; Her, G.-R.; Lai, A.-C.; Chen, J.-H.; Lin, C.-C.; Chen, Y.-J. Anal. Chem. 2005, 77, 5990−5997. (21) Wang, K.-Y.; Chuang, S.-A.; Lin, P.-C.; Huang, L.-S.; Chen, S.H.; Ouarda, S.; Pan, W.-H.; Lee, P.-Y.; Lin, C.-C.; Chen, Y.-J. Anal. Chem. 2008, 80, 6159−6167.

■

CONCLUSIONS Progress in studying protein−protein interactions and proteomic research largely depends on the sensitivity and specificity of the affinity probe used to enrich the target proteins over a wide concentration range, from 10−3 M to a single biomolecule.40 The introduction of nanotechnology promises the development of platforms that possess high sensitivity and specificity for the enrichment of desired proteins from complex biofluids. However, the inconvenience of material preparation always hampers biological applications. The design of simple and uniform MNP surface modifications, as demonstrated in this study, largely reduces the gap between material preparation and practical application without tedious synthetic design and optimization. We have demonstrated that several probe-functionalized MNPs exhibit excellent extraction efficiency for various targets from proteins to cells. For example, Pk@MNPs were used to specifically isolate PA-IL and Slt-B from an aqueous medium and cell lysate, respectively, with a short incubation as well as easy and rapid magnetic separation. Mannose@MNPs were used to specifically isolate the FimH-containing E. coli strain ORN178 without contamination by the FimH knockout strain ORN208. Subsequently, to suppress the nonspecific interactions caused by the MNP surface, an innovative MNP-based IP method was established and demonstrated to exhibit several-fold higher sensitivity than the traditional protein G protein-functionalized microbeads in conventional IP. The short duration (2 h) and the dispensability of protein G and protein A substantially economized the operation time and the cost of materials. 167

dx.doi.org/10.1021/bm301567w | Biomacromolecules 2013, 14, 160−168

Biomacromolecules

Article

(22) Wu, H.-T.; Hsu, C.-C.; Tsai, C.-F.; Lin, P.-C.; Lin, C.-C.; Chen, Y.-J. Proteomics 2011, 11, 2639−2653. (23) Yu, C.-C.; Lin, P.-C.; Lin, C.-C. Chem. Commun. 2008, 11, 1308−1310. (24) Yu, C.-C.; Kuo, Y.-Y.; Liang, C.-F.; Chien, W.-T.; Wu, H.-T.; Chang, T.-C.; Jan, F.-D.; Lin, C.-C. Bioconjugate Chem. 2012, 23, 714− 724. (25) Chien, Y.-Y.; Jan, M.-D.; Adak, A. K.; Tzeng, H.-C.; Lin, Y.-P.; Chen, Y.-J.; Wang, K.-T.; Chen, C.-T.; Chen, C.-C.; Lin, C.-C. ChemBioChem 2008, 9, 1100−1109. (26) Lin, P.-C.; Tseng, M.-C.; Su, A.-K.; Chen, Y.-J.; Lin, C.-C. Anal. Chem. 2007, 79, 3401−3408. (27) Lin, P.-C.; Chou, P.-H.; Chen, S.-H.; Liao, H.-K.; Wang, K.-Y.; Chen, Y.-J.; Lin, C.-C. Small 2006, 2, 485−489. (28) Shih, P.-H.; Shiu, J.-Y.; Lin, P.-C.; Lin, C.-C.; Veres, T.; Chen, P. J. Appl. Phys. 2008, 103, 07A316−1. (29) Han, C.-L.; Chien, C.-W.; Chen, W.-C.; Chen, Y.-R.; Wu, C.-P.; Li, H.; Chen, Y.-J. Mol. Cell Proteomics 2008, 7, 1983−1997. (30) Kersey, P. J.; Duarte, J.; Williams, A.; Karavidopoulou, Y.; Birney, E.; Apweiler, R. Proteomics 2004, 4, 1985−1988. (31) Hsieh, S.-Y.; Jan, M.-D.; Patkar, L. N.; Chen, C.-T.; Lin, C.-C. Carbohydr. Res. 2005, 340, 49−57. (32) Chen, Y.-J.; Chen, S.-H.; Chien, Y.-Y.; Chang, Y.-W.; Liao, H.K.; Chang, C.-Y.; Jan, M.-D.; Wang, K.-T.; Lin, C.-C. ChemBioChem 2005, 6, 1169−1173. (33) Karmali, M. A.; Petric, M.; Lim, C.; Fleming, P. C.; Arbus, G. S.; Lior, H. J. Infect. Dis. 1985, 151, 775−782. (34) Naiki, M.; Marcus, D. M. Biochem. Biophys. Res. Commun. 1974, 60, 1105−1111. (35) Lindberg, A. A.; Brown, J. E.; Strömberg, N.; Westling-Ryd, M.; Schultz, J. E.; Karlsson, K. A. J. Biol. Chem. 1987, 262, 1779−1785. (36) Lin, C.-C.; Yeh, Y.-C.; Yang, C.-Y.; Chen, C.-L.; Chen, G.-F.; Chen, C.-C.; Wu, Y.-C. J. Am. Chem. Soc. 2002, 124, 3508−3509. (37) Barro, M.; Patton, J. T. Proc. Natl. Accd. Sci. U.S.A. 2005, 102, 4114−4119. (38) Levy, F.; Muehlethaler, K.; Salvi, S.; Peitrequin, A. L.; Lindholm, C. K.; Cerottini, J. C.; Rimoldi., D. Mol. Biol. Cell 2005, 16, 1777− 1787. (39) Park, C. S.; Lee, M. S.; Oh, H.; Choi, K. Y.; Yeo, M. G.; Chun, J.; Song, W. K. Eur. J. Cell Biol. 2007, 86, 111−123. (40) Yildirim, M. A.; Goh, K. I.; Cusick, M. E.; Barabási, A. L.; Vidal, M. Nat. Biotechnol. 2007, 25, 1119−1126. (41) Anderson, N. L.; Anderson, N. G. Mol. Cell Proteomics 2002, 1, 845−867.

168

dx.doi.org/10.1021/bm301567w | Biomacromolecules 2013, 14, 160−168