Data & Analytics

Single-Domain Protein A-Engineered Magnetic Nanoparticles: Toward a Universal Strategy to Site-Specific Labeling of Antibodies for Targeted Detection of Tumor Cells

Description
Single-Domain Protein A-Engineered Magnetic Nanoparticles: Toward a Universal Strategy to Site-Specific Labeling of Antibodies for Targeted Detection of Tumor Cells
Published
of 10
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  Single-Domain Protein A-EngineeredMagnetic Nanoparticles: Toward aUniversal Strategy to Site-SpecificLabeling of Antibodies for TargetedDetection of Tumor Cells Serena Mazzucchelli, †,‡,¶ Miriam Colombo, †,‡,¶ Clara De Palma, † Agnese Salvade`, †,‡ Paolo Verderio, †,‡ Maria D. Coghi, ‡ Emilio Clementi, †,§ Paolo Tortora, ‡ Fabio Corsi, † and Davide Prosperi ‡,  , * † Dipartimento di Scienze Cliniche “Luigi Sacco”, Universita` di Milano, Ospedale L. Sacco, via G.B. Grassi 74, 20157 Milano, Italy,  ‡ Dipartimento di Biotecnologie eBioscienze, Universita` di Milano-Bicocca, piazza della Scienza 2, 20126 Milano, Italy,  § Istituto Scientifico Eugenio Medea, 23842 Bosisio Parini, Italy, and  Istituto diScienze e Tecnologie Molecolari, CNR, via Fantoli 16/15, 20138 Milano, Italy.  ¶ These authors contributed equally to the research. N oninvasive medical imaging meth-ods currently represent a major is-sue in the prevention and treat-ment of malignant diseases. 1 Among them,magnetic resonance imaging (MRI) is rap-idly becoming the gold-standard techniquefor an in-depth tissue investigation withhigh-spatial resolution. 2,3 However, in sev-eral circumstances, MRI suffers from lowsensitivity, which is critical when accuratedetection and monitoring of localized ma-lignancies developing in the early stages of the disease are required. For this reason,great efforts have been made to develop ef-ficient contrast enhancing agents aimed atimproving the signal difference betweentheareaofinterest( e.g .,thebloodpool)andthe background. In clinical application, thesystemic injection of appropriate doses of contrast agents, mostly based on gadolin-ium complexes ( e.g ., Gd-DTPA), 4,5 results ina nonspecific enhancement of the MRI sig-nal. When a sharp tissue distribution is de-sired, targeted contrast agents are designedin such a way that they localize to specificcell type through active binding mecha-nisms, which exploit the conjugation of thesignal enhancer with suitable markers.Among them, most used are folate, smallpeptides, and antibodies, which stimulatespecific recognition with the respective tu-mor cell receptor. 6  8 Unfortunately, in mostcases, antibody-conjugated Gd-DTPAproved to be largely unsuccessful due tothe relatively low sensitivity of MRI and thelow density of cell target receptors, thus re-quiring administration of excessive gadolin-ium doses. This drawback can be partiallyovercome by using magnetic nanoparticlesbased on iron oxide (MNPs), which havebeen demonstrated to induce large incre-ments in transverse relaxation rate uponbinding with a 10 6 signal amplification overGd-DTPA. 9 While Gd-based agents enhancethesignalin T  1 -weightedimages,MNPspro-vide strong signal enhancement in  T  2 -weighted images, owing to a different con-trasting mechanism. 10,11 For this reason,MNPs have attracted much attention in bio-medical diagnostics in view of their useful-ness as contrast agents for magnetic *Address correspondence todavide.prosperi@unimib.it.Received for review June 10, 2010and accepted August 31, 2010. PublishedonlineSeptember8,2010. 10.1021/nn101307r © 2010 American Chemical Society ABSTRACT  Highly monodisperse magnetite nanocrystals (MNC) were synthesized in organic media andtransferred to the water phase by ultrasound-assisted ligand exchange with an iminodiacetic phosphonate. Theresulting biocompatible magnetic nanoparticles were characterized by transmission electron microscopy, dynamiclight scattering, and magnetorelaxometry, indicating that this method allowed us to obtain stable particledispersions with narrow size distribution and unusually high magnetic resonance  T  2  contrast power. Thesenanoparticles were conjugated to a newly designed recombinant monodomain protein A variant, which exhibiteda convincingly strong affinity for human and rabbit IgG molecules. Owing to the nature of antibody-protein Abinding, tight antibody immobilization occurred through the Fc fragment thus taking full advantage of thetargeting potential of bound IgGs. If necessary, monoclonal antibodies could be removed under controlledconditions regenerating the srcinal IgG-conjugatable MNC. As a proof of concept of the utility of our paramagneticlabeling system of human IgGs for biomedical applications, anti-HER-2 monoclonal antibody trastuzumab wasimmobilized on hybrid MNC (TMNC). TMNC were assessed by immunoprecipitation assay and confocal microscopyeffected on HER-2-overexpressing MCF-7 breast cancer cells, demonstrating excellent recognition capability andselectivity for the target membrane receptor. KEYWORDS:  magnetic nanoparticles · targeted MRI contrast agents · phasetransfer · biolabeling · protein A · breast cancer  A  R   T   I     C  L   E   www.acsnano.org VOL. 4  ▪  NO. 10  ▪  5693–5702  ▪  2010  5693  resonance imaging (MRI) 12  14 and biosensing. 15 More-over, the hydroxyl-rich surface of iron oxides promotesthe direct conjugation of organic and biological mol-ecules by surface chemical technology. 16  The designand construction of a successful target-oriented biohy-brid paramagnetic nanoprobe is a primary goal towardthe development of highly efficient contrast enhancerforMRI,standingattheinterfacebetweennanotechnol-ogy and molecular biology. The ultimate challenge isrepresented by the development of reliable strategiesfor the conjugation of targeting biomolecules, espe-cially monoclonal antibodies (IgGs), to MNPs. These in-clude passive/electrostatic physical adsorption of IgGs,tightimmobilizationexploitingthestronginteractionof biological counterparts, such as biotin-streptavidin,which requires prior IgG biotinylation, 17 or the forma-tion of covalent chemical connections, which is oftenconsidered the most practical choice. 18,19 Althoughthese approaches may offer different solutions andhave been successfully employed in several circum-stances, they all share the same basic limitation, that isa non-site-specific binding to IgG molecule, which af-fects the targeting efficiency of the antibody. As a mat-ter of fact, the actual conservation of the targeting bio-activity of immobilized IgGs is not obvious and remainsa crucial issue, which must be addressed in designing asuccessful targeted nanoprobe.An interesting option might be envisaged in theuse of a natural peptide linker endowed with high affin-ity for IgGs, such as protein A. Protein A is a cell-wall as-sociated protein exposed on the surface of the gram-positive bacterium  Staphilococcus aureus . 20  The primarystructure consists of a 42 kDa single polypeptide chain,folded into five highly homologous domains, named E,D, A, B, and C, each consisting of 56  61 residues. 21  The interest for this molecule in biotechnology residesmainly in three useful properties: (1) the protein struc-ture is stable over a broad range of pH (2  12) and inthe presence of various detergents; (2) it can bind re-versibly a large variety of IgGs  via  their Fc fragmentthrough its consensus sequence (Asn-Gln-Phe-Asn-Lys-Glu), 21 (3) IgG-protein A complex can be dissociated un-der controlled conditions (pH 3.5  4.5) without appar-ent loss of activity. 22  The affinity of protein A forimmunoglobulins is not conserved among the differ-ent classes and isotypes. In particular, it exhibits high af-finity to human, rabbit, and guinea pig IgGs. Recently,natural protein A has been used immobilized on mag-netic polyadsorbent for IgG separation and/or purifica-tion procedures. 23 As protein A recognizes the Fc por-tion of IgGs, it is expected to mediate an orderly Fc site-specific antibody immobilization on MNPs resulting ina target-directed Fab presentation. 24 In this paper, we present a multidisciplinary ap-proach to the design and synthesis of a universal mag-netic nanohybrid consisting of a high-quality iron oxidenanocrystal core conjugated to a suitably bioengi-neered small variant of protein A for the smart immobi-lization of IgGs. The resultant targeted nanoparticle isshown to exhibit high affinity and selectivity for the ap-propriate tumor markers. As a case study for our inves-tigations, we focused on the monoclonal antibody tras-tuzumab, which is commonly employed in clinicaltherapy of breast cancer, as a model for the develop-ment of our tumor-targeting magnetic nanoprobe. Trastuzumab (tz) is a humanized monoclonal antibodyconsisting of two antigen-specific sites that bind to the juxtamembrane portion of the extracellular domain of  Scheme 1. Synthesis of Pegylated Trastuzumab-Functionalized MNC (TMNC)       A      R      T      I      C      L      E VOL. 4  ▪  NO. 10  ▪  MAZZUCCHELLI  ET AL . www.acsnano.org 5694  the “Human Epidermal growth factor Receptor 2” (HER-2), which is found overexpressed in several metastasiz-ing breast cancer cells. 25 RESULTS AND DISCUSSION Our aim was to synthesize a model magnetic nano-particle hybrid system containing a specific IgG bind-ing functionality with the lowest molecular weight toreduce the nanoparticle overall size. We reasoned thata recombinant derivative of protein A consisting of onesingle IgG-binding domain, namely the B domain,would meet such criteria. This strategy displays severaladvantages: (1) as for entire protein A, all the antibod-ies are presented in the same orientation on the nano-particle, with the active Fab portions directed in an op-timal configuration for antigen binding; (2) the use of asingle B domain of protein A results in a very small bio-molecular support for the antibody, which is expectedtopartiallyreducetheimmunogenicityofthewholehy-brid nanoparticle, while conserving remarkable bind-ing affinity for IgGs; (3) the bioengineering approach tothe protein A variant allows for the artificial introduc-tion of selectively reactive functionalities in the peptidesequence, such as cysteine thiols, for site-specific conju-gation onto the nanoparticle surface. The promisingcombination of all these potential advantagesprompted us to explore the possibility of improvingthe current methods for the synthesis of IgG-functionalized nanohybrids. We chose an organic-phase approach for the synthesis of the magnetitecore nanostructure, providing highly uniform and crys-talline magnetic nanoparticles endowed with strong in-trinsic relaxivity and narrow size distribution. This ap-proach, while effective, had a main limitation, whichwas the poor solubility of the resulting surfactant-coated nanoparticles. Hence, an efficient ligand ex-change was required to transfer them into an aqueousenvironment avoiding particle aggregation. Scheme 1depicts the general strategy followed to synthesize tz-functionalized magnetic nanocrystals (TMNC), whichwere then assessed according to their capability to rec-ognize the HER-2 receptor antigen both in a whole cellextract and in living cultured breast cancer cells. Development of Water-Stable Profunctional Iron OxideNanocrystals.  High-quality, 8 nm hydrophobic iron oxidenanocrystals (MNC0, Figure 1, inset) were synthesizedby solvothermal decomposition from iron  oleate com-plex in a solution of octadecene in the presence of oleic acid as capping agent. 26  The synthesized uniformnanocrystals were finely dispersed in chloroform andtreated under continuous sonication with an excess of  N  -phosphonomethyl iminodiacetic acid phosphonate(PMIDA) dissolved in aqueous ammonia solution to pro-mote the ligand exchange on the MNC0 surface, lead-ing to the highly water dispersible, readily functionaliz-able MNC1. The phase transfer reaction occurredquickly thanks to the higher affinity of PMIDA phospho-nate group toward iron oxide compared to oleic car-boxylate (Figure 1a, inset). After the phase transfer,MNC1 maintained the srcinal average crystal size (8 Figure 1. (a)  T  2  relaxometry analysis of as-synthesized MNC1. The in-verse of experimental  T  2  values obtained at different MNC1 concen-trations is plotted  vs  iron concentration. The experimental data arefitted by a line. The line slope indicates the  T  2  relaxivity. Inset: MNCphase transfer from organic solvent (chloroform) to aqueous solu-tion. (b) Hydrodynamic size distribution histograms of oleic-coatedFe 3 O 4  nanoparticles (dashed line) and after ligand exchange withPMIDA (MNC1, continuous line). Diameters were measured by DLSin chloroform and water, respectively. Inset: TEM images of as-synthesized MNC0 in hexane (top) and of MNC1 in PBS (bottom).Scale bars  50 nm.Figure 2. Schematic representation of spaBC3 engineered se-quence. Glutathione  S -transferase (GST, blue), spaBC3 (red), histi-dine tag (green), and cysteine tripod (orange). PreScission pro-tease cleavage site between amino acids Gln 226  and Gly 227  isevidenced.  A  R   T   I     C  L   E   www.acsnano.org VOL. 4  ▪  NO. 10  ▪  5693–5702  ▪  2010  5695  nm, by TEM), the final nanoparticle shape was uni-formly spherical, while a slight increase in the hydrody-namic diameter from 21  2 nm (MNC0 in chloroform)to 32  3 nm (MNC1 in water) was determined by dy-namiclightscattering(DLS)(Figure1b),probablyduetothe change of solvent and solvation aptitude of thenew ligands. The resulting MNC1 surface, rich of acces-sible carboxyl groups, stabilized well the colloidal par-ticles in buffered solution by electrostatic repulsions,concomitantly providing a useful support for furtherfunctionalization. In addition, we performed relaxomet-ric measurements to determine the  T  2  enhancing capa-bility of synthesized nanocrystals. The water-solubleMNC1 exhibited a remarkably high transverse relaxiv-ity ( r  2  255 mM  1 s  1 at 20 mHz, B 0  0.47 T), deducedby the slope of the line obtained by plotting 1/ T  2  vs iron concentration. This value, as compared with the re-laxivity of commercially available  T  2  contrast agentsbased on polymer-coated iron oxides, such as Endorem(Guerbet, 160 mM  1 s  1 ), Ferumoxytol (Adv. Magnet-ics, 83 mM  1 s  1 ), and Resovist (Schering, 151 mM  1 s  1 ), suggests that high contrast power can be achievedwith MNC1, on the same order of the best  T  2  contrastagents reported so far. 27 Carboxylate functionalities were converted intoamine ends by reaction with the bifunctional diamino-linker 2,2-(ethylenedioxy)bis(ethylamine) (EDBE)  viaN  -hydroxysuccinimidyl ester (NHS) activation. EDBE-modified nanoparticles (MNC2) exhibited a hydrody-namic diameter of 57  3 nm, as determined by DLS. The average number of amino groups on MNC2 wasquantified by an adapted version of the Dunnill’s proto-col, 28 resulting in 150 amines per particle. Our two-step procedure involving hot organic synthesis of monocrystalline iron oxide followed by ligand ex-change and phase transfer allowed us to obtain stablewater-soluble nanoparticles with unique contrastpower and narrow size distribution, typical of the ther-mal synthesis in high-boiling organic solvent, yet de-void of the drawbacks associated to the biological fluidincompatibility. The high relaxivity value of our solubleMNC makes this nanoprobe a promising contrast agentapplicable in magnetic resonance imaging (MRI) forthe potential noninvasive diagnosis of malignances.MNC2 were stable for months with nondetectableprecipitation in several buffered media, including Dul-becco’s PBS, pH 7.4; borate buffer, pH 8.5; acetatebuffer, pH 5.0; and tris HCl, pH 6  8. The amino func-tionalities on the particle surface allowed the MNC2conjugation with  N  -succinimidyl-3-[2-pyridyldithio]-propionate (SPDP)  via  NHS ester resulting in the thiol-reactive pro-functional MNC3 (Scheme 1). MNC3 werethe ideal building block for bioconjugation with thiol-engineered protein A variant, as PDP functionality is un-affected by nucleophiles but very reactive toward sulf-hydryl ends of organic and biological thiol-containingmolecules under mild conditions by formation of stabledisulfide bridges. Design, Expression, and Purification of an Engineered Cys 3 -Ended Variant of a Single-Domain Fragment of Protein A from  Staphylococcus aureus.  The B domain sequence, previouslyreported by Abrahmsen  et al  ., 21 was modified inserting Bam HI and  Sma I restriction sites at the 5 =  and 3 =  posi-tions,respectively.Themodifiedgenewasclonedtoex-press the B domain in fusion with glutathione S -transferase (GST). However, proteins purified by GSTaffinity chromatography may contain small amounts of GST and GST-affine contaminants in eluted fractions.For this reason, a supplementary 6  His affinity tag wasintroduced at the C-terminal to completely remove theresidual GST-deriving impurities through an additionalaffinitypurificationstep. 29  Toachieveasite-specificcon- Figure 3. (a) SDS-PAGE of spaBC3 purified fractions. Molecularweight markers (MWM, 5  L) Precision Plus Unstained were used.Purified protein (4  g) was loaded after elution with PreScissionprotease (left). After purification with Ni-NTA agarose to removeGSTcontaminants,thesameamountofspaBC3wasloadedonSDS-PAGEwithorwithout0.1MDTTinrunningbuffer(rightandmiddle,respectively). (b) SpaBC3 binding assay. SpaBC3 (250 ng) was fil-tered through PVDF membrane and incubated with rabbit IgG ortz. IgG removal was obtained by Na-citrate incubation. GST protein(C-, 250 ng), as negative control, and rabbit IgG or tz (250 ng), aspositives, were used. The presence of rabbit IgG or tz was revealedby antirabbit or antihuman secondary antibodies conjugated toHRP, respectively.       A      R      T      I      C      L      E VOL. 4  ▪  NO. 10  ▪  MAZZUCCHELLI  ET AL . www.acsnano.org 5696   jugation of the protein A variant to MNC, the presenceof one selectively reactive group in the protein primarystructure was required. As the recombinant B domainwas devoid of cysteine residues, we envisaged that theincorporation of a cysteine tail would result in the intro-duction of a highly reactive thiol group, concomitantlyavoiding potentially competitive interferences fromother functional moieties in the peptide sequence. Inlight of our preliminary results showing an unexpectedweakness of the protein  nanoparticle linkage with onesingle Cys tagged behind the His 6  terminal (data notshown), we substituted the single Cys residue with aCys 3  tripod leading to an improved binding effective-ness of the thiol end (Figure 2). The engineered domain B of protein A from Staphilococcus aureus  containing a C-terminal His 6 Cys 3 tail (spaBC3) was cloned in fusion with GST tag in pGEX-6P-1vectorandexpressedinBL21(DE3) E. coli  strain.Sp-aBC3 was subsequently purified using a GST affinity col-umn and eluted with cleavage with PreScissionprotease, obtaining a 7.48 kDa protein with a good de-gree of purity showing the presence of two residualcontaminants at 15 and 25 kDa, respectively, whichwere removed by further purification through a Ni-NTAagarose column. SDS-PAGE performed under conven-tional reducing conditions showed a remarkably diffuseband at  ca.  10 kDa in correspondence of spaBC3, sug-gestingthatduringacrylamidegelmigrationtheforma-tion of disulfide bridges between spaBC3 moleculesleading to aggregates might occur. By addition of 0.1 Mdithiothreitol (DTT) in the running buffer, a more de-fined thin spaBC3 band was obtained. These experi-ments are summarized in Figure 3a. While a DTT con-centration of about 25 mM is usually estimated to besufficient to break intermolecular monodentate disul-fide bridges, 30 here the presence of a cysteine tripodforced us to use more strongly reducing conditions tomaintain spaBC3 in its monomer form until its conjuga-tion with nanoparticles. SpaBC3 Retains IgG Binding Activity. Recombinantsp-aBC3producedin E. coli  andfilteredonapolyvinylidenefluoride (PVDF) membrane was assessed in regard toits ability to bind IgG molecules. In Figure 3b, a dot blotexample of spaBC3 complex incubated with rabbit IgGis presented. The presence of bound IgG was revealedbyastrongsignalofantirabbitsecondaryantibodycon- jugated with horseradish peroxidase (HRP). To evalu-atethepossibilitytorecyclespaBC3formultipleusages,we tested the IgG removal by incubating the spaBC3-IgG complex in 50 mM Na-citrate buffer, pH 3 at roomtemperature (RT), observing a remarkable reduction of the signal from 30 min of treatment. The same spaBC3sample was then incubated with a solution containingtz, confirming the ability of spaBC3 to capture differentIgGs even after Na-citrate treatment. These results sug-gested to us that spaBC3 might be exploited for thepro-functionalization of nanoparticles leading to a uni-versal nanocarrier for the efficient, recyclable, andtarget-directed conjugation of a large variety of IgGs. Conjugation of spaBC3 to MNC through Sulfhydryl Groups. Sp-aBC3 was incubated in 0.1 M DTT to reduce the cys-teineresidues,andthentheexcessofDTTwasremovedby gel filtration. The purified monomer protein ob-tained by this procedure was immediately incubatedwith MNC3 and left overnight. Under these conditions,the thiol groups of cysteine residues displayed high re- Figure 4. (a) Assessment of SpaBC3-MNC conjugation. MNPA was treated with 0.1 M DTT, resulting in spaBC3 release. Thereduced MNPA was loaded on a PVDF membrane, together with untreated MNPA and the supernatant from the reductionmixture as controls. The presence of spaBC3 was detected by antihuman HRP-antibody only in untreated MNPA and in su-pernatant samples, confirming that SpaBC3 was anchored onto MNC essentially  via  disulfide bridges. C  and C  are GSTprotein (250 ng) and spaBC3 (250 ng), respectively. (b) SpaBC3 on MNPA retains its IgG binding activity. MNPA (500 ng) wasfiltered through PVDF membrane and incubated with rabbit IgG or tz. IgG removal was obtained by Na-citrate incubation.GST protein (C  , 250 ng) and spaBC3 (C  , 250 ng) were used as negative and positive controls, respectively. MNPA captureof rabbit IgG or tz was revealed by antirabbit or antihuman secondary HRP-antibodies, respectively.  A  R   T   I     C  L   E   www.acsnano.org VOL. 4  ▪  NO. 10  ▪  5693–5702  ▪  2010  5697
Search
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x