Binary polypeptide system for permanent and oriented protein immobilization
© Ferrari et al. 2010
Received: 11 January 2010
Accepted: 12 May 2010
Published: 12 May 2010
Skip to main content
© Ferrari et al. 2010
Received: 11 January 2010
Accepted: 12 May 2010
Published: 12 May 2010
Many techniques in molecular biology, clinical diagnostics and biotechnology rely on binary affinity tags. The existing tags are based on either small molecules (e.g., biotin/streptavidin or glutathione/GST) or peptide tags (FLAG, Myc, HA, Strep-tag and His-tag). Among these, the biotin-streptavidin system is most popular due to the nearly irreversible interaction of biotin with the tetrameric protein, streptavidin. The major drawback of the stable biotin-streptavidin system, however, is that neither of the two tags can be added to a protein of interest via recombinant means (except for the Strep-tag case) leading to the requirement for chemical coupling.
Here we report a new immobilization system which utilizes two monomeric polypeptides which self-assemble to produce non-covalent yet nearly irreversible complex which is stable in strong detergents, chaotropic agents, as well as in acids and alkali. Our system is based on the core region of the tetra-helical bundle known as the SNARE (soluble N-ethylmaleimide-sensitive factor attachment protein receptor) complex. This irreversible protein attachment system (IPAS) uses either a shortened syntaxin helix and fused SNAP25-synaptobrevin or a fused syntaxin-synaptobrevin and SNAP25 allowing a two-component system suitable for recombinant protein tagging, capture and immobilization. We also show that IPAS is suitable for use with traditional beads and chromatography, planar surfaces and Biacore, gold nanoparticles and for protein-protein interaction in solution.
IPAS offers an alternative to chemical cross-linking, streptavidin-biotin system and to traditional peptide affinity tags and can be used for a wide range of applications in nanotechnology and molecular sciences.
Two-component affinity-based tools underlie basic molecular research and are invaluable for the development of drugs and diagnostics . Applications include affinity chromatography, microarray technologies, microplate-based screens and many biotechnological processes . The main factor underlying a successful outcome often relies on firm, irreversible immobilization of a protein in a defined orientation either on a solid surface or in a 3-dimensional matrix. Existing immobilization technologies suffer from a number of disadvantages. For example, in the case of chemical protein coupling , one can achieve irreversible surface immobilization, but the product may be in a non-functional state due to orientation issues and chemical modifications. Chemical crosslinking through reactive amino acid side chains of proteins often results in a range of products due to the availability of large number of such groups on a single protein molecule and limited specificity of reactions. The outcome of chemical labelling will depend strongly on reaction conditions such as pH, temperature, etc., and the efficiency of chemical derivatization would often vary from batch to batch. Other chemoselective methods, independent of the reactive terminal amino acids, such as Staudinger ligation , require the presence of groups which do not occur in natural or recombinantly produced proteins such as triaryl phosphines and azides. Thus, none of the chemical modification techniques when applied to proteins can achieve the same specificity and selectivity of labelling as affinity-based systems. The most popular binary affinity system utilizes a uniquely strong biotin-streptavidin interaction, however attachment of either biotin or streptavidin (normally tetrameric) to a target protein still requires chemical conjugation and is therefore less site-specific. Recombinant technologies for protein expression, on the other hand, allow a convenient encoding, in the expression vector, of polypeptide affinity tags allowing immobilization on a specific binding substrate. Examples of such polypeptide tag systems include: His-tag binding to metal, glutathione-S-transferase binding to glutathione, maltose-binding protein binding to maltose, strep-tag peptide binding to streptavidin, myc-tag peptide binding to anti-myc antibody-containing surfaces [4–8]. Although it is possible to immobilize a protein in a site-selective way using these polypeptide tags, in all these cases immobilization is either non-permanent or too expensive (antibody-based affinity surfaces). Clearly, the ideal immobilization technique should be capable of both an irreversible coupling as with chemical modifications and selective labelling as affinity based systems. Such system should also allow for a site-specific orientation of the target protein, and be simple, robust and affordable (unlike antibody-based systems, which are prone to degradation, denaturation and are expensive to produce).
Most current affinity tags can only operate in mild conditions, i.e. neutral pH, low ionic strength and physiological temperatures. In the emerging field of nanobiotechnology, conjugation which can resist harsh conditions may be required during fabrication of micro- or nano-arrays, micro-fluidic devices or bio-conjugation to quantum dots or other nanoparticles. Furthermore, enzymes resistant to denaturants, acidic or alkaline conditions are catching attention due to their ability to accelerate reactions in the food and paper industry and in toxic waste removal. Clearly, to better exploit the potential of recombinant proteins for nanobiotechnology, new robust affinity system(s) capable of irreversible capture and immobilization in harsh environments need to be developed. We and others shown previously that three neuronal SNARE proteins, syntaxin, SNAP25 and synaptobrevin, form a very tight tetra-helical bundle commonly known as the SNARE complex [9–12]. In this complex, both syntaxin and synaptobrevin contribute a single α-helix, whereas SNAP25 contributes two α-helices. One fascinating feature of the neuronal SNARE complex is its stability and resistance to harsh treatments, including urea and sodium dodecyl sulphate (SDS) . Only boiling in SDS can break the SNARE complex in vitro; in vivo the complex is dissociated by an intracellular ATPase . Previously, Rothman and colleagues demonstrated that SNARE proteins expressed on the cell surface can fuse cells . The unique properties of the SNARE coiled-coil bundle, however, have not been considered for other applications. Here we report a binary SNARE-based affinity system for protein capture and immobilization, which is permanent and irreversible under physiological buffer conditions.
When the two proteins were separately mixed with the syntaxin beads we detected binding of each protein (Fig. 3C). To confirm that binding of syntaxin to either B-S or S-B results in the conventional SNARE complex, we tested whether the syntaxin beads with immobilized B-S or S-B can also pull-down complexin, which is known to bind selectively to the neuronal SNARE complex . Indeed, the pull-down in Fig. 3D shows that complexin could specifically bind to syntaxin beads only after addition of B-S or S-B. The complexin binding suggests that the four helices bundle is parallel. Furthermore, the melting temperature of the B-S and S-B complexes, measured by heating in presence of 2% SDS at different temperatures, is 50°C (data not shown), and suggests a tight assembly of SNARE helices .
Similarly to what we did for the syntaxin/three-helical IPAS, we then immobilized GST-SNAP25 on a Biacore chip to prove the possibility of capturing the NL on the chip surface. Fig. 6C shows the effective immobilization of NL on top of SNAP25 and the strong resistance of the complex to a series of harsh washes.
Here we described a novel binary affinity system for protein capture that can withstand very harsh conditions. The irreversible protein attachment system (IPAS) utilizes 3 SNARE proteins which were converted into two tags. Our affinity system is based on the neuronal SNARE complex, a bundle of four α-helices interacting through strong hydrophobic forces . It is believed that the SNARE complex formation happens by a 'zippering' mechanism starting at the N-termini of four SNARE motifs. The complex has an extremely slow dissociation rate with a half-life estimated to be a billion years under non-denaturating conditions in vitro but can be dissociated inside cells by an ATPase [14, 18]. Generally, SNARE proteins play a key role in fusion of intracellular vesicles with their target membranes. To date, more than 100 SNARE proteins have been discovered which carry highly conserved ~70 aa heptad repeat motifs responsible for tight SNARE interactions . It, thus, will be of interest to evaluate usefulness of other SNARE proteins for affinity systems. Tandem fusion of SNARE proteins is a practical invention which has not been considered previously, but as shown here allows production of high-affinity reagents. Naturally, the most attractive feature of the SNARE-based protein capture is the potential of the IPAS tags to be fused to proteins of interest via recombinant means. The resulting fusion products can then be nearly permanently immobilized to a solid support via a simple mixing with the corresponding immobilization support (i.e., syntaxin beads, syntaxin or GST-SNAP25 Biacore chips, GST-SNAP25 or GST-NL gold nanoparticles). When necessary, either of the tags in our binary system can be chemically linked to surfaces of beads, chips and microarray plates, or modified by chemical or recombinant introduction of functional groups. Our tested SNARE-based bimolecular affinity system affords an inexpensive, nearly irreversible linking of required protein modules or firm capture of tagged molecules on surfaces. The irreversible nature of the SNARE complex makes the conventional thermodynamic analysis difficult; under normal buffer conditions the dissociation of the IPAS peptides is not detectable with either of the methods we tested (beads pull down, Biacore) and was impossible to estimate even for naturally occurring SNARE complexes . The use of α-helical bundles as affinity tags has been attempted before based on heterodimerization of coiled-coils ~40 aa peptides [21–23]. However, in contrast to the de novo engineering, we chose a biomimetic strategy focusing on a known tight interaction that was perfected by evolution to drive fusion of cellular membranes . Our work presents the first evidence that an affinity system based on SNARE proteins can work, maintaining the unique property of the SNARE complex - extremely stable interaction that can withstand harsh conditions. Although here we presented two IPAS systems that are based on a single helix (syntaxin) interacting with a three-helical fusion (S-B or B-S) and an alternative IPAS based on two double helices (NL and SNAP25), we anticipate that other SNARE configurations would be also possible.
As a practical application in the field of nanobiotechnology we have reported the assembly of the tetra-helical complex on the surface of gold nanoparticles, detected by measuring the change in the colloidal gold surface plasmon resonance peak. Red shift in the SPR peak of gold nanoparticles depends on and changes linearly with the refractive index of the surrounding medium . The red shift due to the immobilization of protein is also well documented [25, 26] and results from the apparent increase in the overall size of the gold nanoparticles. We have observed slight blue shift following the assembly of the tetra-helical "NanoLock" complex. No change in optical properties was detected when any of the non-interacting proteins were incubated with the derivatized gold sol. The blue shift indicates that the assembly is likely to result in the increased density of protein packing on the surface of the gold, which is expected, because of the nature of the binary peptides, based on the virtually irreversible binding of SNARE proteins. The addition of GST protein to the NL peptide apparently makes no difference for the tetra-helical self-assembly of GST-NL or NL with GNP-GST-SNAP25. And neither the addition of GST affects self-assembly of SNAP25 with GNP-GST-NL. This is significant because it means that our self-assembling system is not affected by the protein "load" added to either of the binary peptides (SNAP25 or NL). Our results also show that the self-assembly of SNAP25 and NL peptides may be easily controlled irrespective of the protein "load" used. We have also shown that our system is sensitive to the orientation of proteins on the gold surface. This is consistent with the previously reported ability of GNP based methods to distinguish chiral differences [27, 28]. Thus, our results indicate that gold nanoparticles uses are not limited to the detection of protein-protein interactions but may also be used for monitoring protein folding. Previously reported applications of gold nanoparticles for protein conformational changes were limited to detecting pH changes [29, 30], thermodynamic stability, unfolding or to aggregation assays. However, unlike previous reports, where protein folding was detected only through nanoparticle aggregation [31–33], the NanoLock binary peptides assembly does not result in the loss of gold nanoparticles, which remain in the sol and could therefore be used for downstream applications.
The emerging field of nanotechnology increases the demand for tailored conjugation methods for the development of nanochips, microarrays and also for nanodevices for drug delivery [34–37]. Biomaterial and tissue engineering can also benefit from the presented conjugation method for decoration of inert fibrous scaffolds with biologically active molecules . Finally, industrial processes involving immobilized enzymes could require non-covalent yet stable conjugation specifically designed to be resistant to harsh treatments .
We designed three pairs of self assembling polypeptides mimicking the neuronal SNARE complex: the first is made by a 6 kDa sytaxin peptide and the 32 kDa fusion of synaptobrevin and SNAP25 (B-S), the second is made by the same syntaxin peptide and the 32 kDa fusion of SNAP25 and synaptobrevin (S-B) and the third pair is represented by the SNAP25 protein and a 17 kDa fusion of syntaxin and brevin. The affinity systems presented here provides a novel concept that can be utilized for tailored applications in many different technologies.
GST fusions with the full-length rat SNAP25B (aa 1-206) with cysteine to alanine mutations, rat synaptobrevin2 (aa 1-96), complexin II and GST alone were cloned in pGEX-KG vector. His-tag rat SNAP25B (aa 1-206) with cysteine to alanine mutation was cloned on pET vector. Plasmids encoding S-B and B-S fusion proteins were made by attaching optimized SNAP25B DNA (commercially obtained from ATG Biosynthetics) on the N-terminus and C-terminus of synaptobrevin2 (aa 1-84) in the pGEX-KG vector. The plasmid encoding the NL fusion protein was made by attaching the DNA sequence of rat syntaxin3 (195-253) to rat synaptobrevin2 (1-84) in the pGEX-KG vector. The amino acid sequences of S-B, B-S and NL are:
GSADESLESTRRMLQLVEESKDAGIRTLVMLDEQGEQLERIEEGMDQINKDM KEAEKNLTDLGKFAGLAVAPANKLKSSDAYKKAWGNNQDGVVASQPARVV DEREQMAISGGFIRRVTNDARENEMDENLEQVSGIIGNLRHMALDMGNEIDT QNRQIDRIMEKADSNKTRIDEANQRATKMLGSGSGSSGASGEQKLISEEDLSG GSAGSGSSAGMSATAATVPPAAPAGEGGPPAPPPNLTSNRRLQQTQAQVDEV VDIMRVNVDKVLERDQKLSELDDRADALQAGASQFETSAAKL,
GSMSATAATVPPAAPAGEGGPPAPPPNLTSNRRLQQTQAQVDEVVDIMRVN VDKVLERDQKLSELDDRADALQAGASQWETSAAKLSGAGSGAGSAGSGSAE DADMRNELEEMQRRADQLADESLESTRRMLQLVEESKDAGIRTLVMLDEQG EQLERIEEGMDQINKDMKEAEKNLTDLGKFAGLAVAPANKLKSSDAYKKAA GNNQDGVVASQPARVVDEREQMAISGGFIRRVTNDARENEMDENLEQVSGII GNLRHMALDMGNEIDTQNRQIDRIMEKADSNKTRIDEANQRATKMLGSG,
GSEGRHKDIVRLESSIKELHDMFMDIAMLVENQGEMLDNIELNVMHTVDHV EKARDEAKRAGILDSMGRLELKLMSATAATVPPAAPAGEGGPPAPPPNLTSN RRLQQTQAQVDEVVDIMRVNVDKVLERDQKLSELDDRADALQAGASQFETS
GST fusion proteins were produced in BL21 Escherichia coli and purified on glutathione-sepharose beads (Amersham Biosciences), followed either by elution with glutathione (GST-tag proteins) or by thrombin cleavage (SNARE part only). His-tag SNAP25 was purified using Ni-NTA agarose beads (Qiagen) and eluted with Imidazole. Eluted proteins were further purified by gel filtration on a Superdex 200 column (Amersham Biosciences) equilibrated in buffer A (20 mM HEPES, 100 mM NaCl, pH 7.2). The 47 aa syntaxin peptide corresponding to the SNARE interaction part of the syntaxin sequence 201-248 was commercially obtained from Peptide Synthetics. Biotinilation of GST-SNAP25 have been obtained using biotin-maleimide from Sigma.
0.75 mg of syntaxin peptide was cross-linked to 0.28 g (dry weight) of CNBr-activated Sepharose 4B beads (Amersham Biosciences) which were pre-washed in 1 mM HCl and pre-hydrolysed for 4 h at room temperature in the coupling buffer (0.1 M NaHCO3, 0.5 M NaCl, pH 8.3). Coupling reaction have been carried in the same buffer for 2 h at 20°C followed by an overnight blocking of the active groups at 4°C with 1 M ethanolamine, pH 8.5. Beads were washed with 0.1 M acetate buffer, 0.5 M NaCl, pH 4.0, followed by 0.1 M Tris-HCl, 0.5 M NaCl, pH 8.0, and finally buffer A. Fluorescence was visualized on a Bio-Rad confocal microscope (Fig. 1B). Control beads were prepared following the same protocol but in absence of the syntaxin peptide.
To analyze formation of the SDS-resistant SNARE complex (Fig. 1C and Fig. 6B), proteins were incubated (final concentration 1 μM) in buffer B (20 mM HEPES, 100 mM NaCl, 0.8% (w/v) n-octylglucoside, pH 7.2) for 30 min at 20°C in a total volume of 20 μl. We noticed that n-octylglucoside aids the formation of the SNARE bundle. The reactions were stopped by the addition of SDS-containing sample buffer, and proteins were separated by SDS-PAGE and visualised by Coomassie staining. Note that SNARE complex, likely due to its closed conformation, migrates faster than the apparent sum of the monomers sizes.
Syntaxin and control beads (see preparation above, Fig. 2A, B, 3C, 3D, 4A, 5A, B, C, D), streptavidin-sepharose (Sigma, Fig. 4C), Ni-NTA-agarose (Qiagen, Fig. 4C) and glutathione-sepharose beads (Amersham Biosciences, Fig. 4C, 5E) in buffer B were incubated in the presence of an excess of proteins for 30 min at 20°C in a reaction volume of 50 μl with constant shaking. In the experiments shown in Fig. 5D and 5E buffer B has been replaced by calf serum. The beads were washed three times with 20 mM HEPES, pH 7.0, 1 M NaCl, 1 mM EDTA, 0.1% Triton X-100 and 1 mM DTT by low-speed centrifugation followed by two additional washes with buffer A. When using Ni-NTA beads EDTA was omitted to avoid elution by chelation. Bound protein was eluted into SDS containing sample buffer, heated at 100°C for 3 min and analysed by SDS-PAGE and Coomassie staining. When testing disassembly of the binary affinity system, various solutions indicated in the figures were applied to the beads for 10 min at 20°C followed by standard washes.
GST activity assay of the immobilized enzyme was performed with 10 μl beads containing 2 μg of GST-S-B or control beads, according to the manufactures instructions (Sigma). Absorbance at 340 nm was measured in a Tecan plate reader and presented after subtraction of the background signal from control beads.
Experiments shown in Fig. 4B and 6C were performed using a Biacore 2000 system (GE Healthcare). Following the initial wash of the CM5 chip with 1% SDS (1 min), 100 mM Phosphoric acid (1 min) and 100 mM NaOH (2 min), the chip was used to covalently immobilise either the syntaxin peptide or GST-SNAP25 (0.05 mg/ml in 0.1 M acetate buffer, pH 5, containing 0.5% DMSO and 0.8% n-octylglucoside. Following blocking of the chip surface with 0.1 M ethanolamine and a wash with 1% SDS (1 min), 100 mM phosphoric acid (1 min) and 100 mM NaOH (2 min), the chip surface was loaded with 0.13 mg/ml S-B protein in buffer B for 5 min (Fig. 4B) and 0.10 mg/ml NL in buffer B for 45 min (Fig. 6C). To check stability of the formed complex, the loaded chip was washed consecutively with a selection of washing or denaturing reagents for 1 min each. To control for any background drifts and for background subtraction the data were compared to the values obtained for the unloaded channel. All measurements were performed at 25°C.
Gold sols were prepared by reducing Tetrachloroauric acid hydrate with sodium citrate. 40 ml of 0.02% w/w solution of HAuCl4 (Alfa Aesar 36400) in deionised water (equivalent to 0.01% w/w gold) was heated to boiling point under constant stirring. 4 ml of 1% sodium citrate was added under rapid stirring, which continued for another 15 minutes. Sodium azide was added to cooled gold sols to final concentration of 0.05% (w/v). To determine protein binding capacity of gold nanoparticles (GNP), series of bovine serum albumin (BSA) dilutions ranging from 1 μg/ml to 10 mg/ml were made. 100 μl of each of BSA dilutions was added to individual 500 μl aliquots of colloidal gold and the samples were incubated at 25°C with constant, gentle agitation. Following the 30 min incubation 600 μl aliquots of 20% NaCl were added to individual GNP-BSA samples. The GNP-BSA sample with the lowest BSA content (100 μl or 1 mg/ml BSA per 500 μl of the GNP preparation) where no colour change was observed contained sufficient protein for total coverage of the colloidal gold present. The same ratio of protein to GNPs was used in all following experiments.
Three groups of GNPs were produced by derivatization with GST, GST-S25 and GST-NL proteins. Each protein was diluted to 0.02% w/v with buffer A and a number of individual GNP samples for each protein were made by mixing 50 μl of 0.02% w/w GNP sol with 300 μl of GNP buffer (0.1% sodium citrate, 0.02% NaN3) and 50 μl of 0.02% GST, GST-S25 or GST-NL to make GNP-GST, GNP-GST-S25 and GNP-GST-NL respectively. Following 70 min incubation at 25°C with gentle agitation, n-octylglucoside was added to each sample to a final concentration of 0.8%. Following 10 min incubation at 25°C a second set of proteins was added to the fully derivatized GNP-protein sols. Protein concentrations were 0.02% w/v for GST, GST-S25 and GST-NL, and 0.01% for S25 and NL proteins. 50 ul of each protein was added, followed by 30 min incubation at 25°C. Absorption spectra were taken using Helios Alpha UV-Vis spectrophotometer, wavelength resolution 1 nm. All samples were prepared individually at least in duplicate and the experiment repeated twice.
EF and BD acknowledge the British Council which supported the British-Italian Partnership Programme. The MRC Technology partly supported this work via the Development Gap Funding.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.