Article pubs.acs.org/accounts
Site-Selective Conjugation of Native Proteins with DNA Julie B. Trads, Thomas Tørring, and Kurt V. Gothelf* Center for DNA Nanotechnology at the Interdisciplinary Nanoscience Center and Department of Chemistry, Aarhus University, Gustav Wieds Vej 14, 8000 Aarhus C, Denmark CONSPECTUS: Conjugation of DNA to proteins is increasingly used in academia and industry to provide proteins with tags for identification or handles for hybridization to other DNA strands. Assay technologies such as immuno-PCR and proximity ligation and the imaging technology DNA-PAINT require DNA−protein conjugates. In DNA nanotechnology, the DNA handle is exploited to precisely position proteins by self-assembly. For these applications, site-selective conjugation is almost always desired because fully functional proteins are required to maintain the specificity of antibodies and the activity of enzymes. The introduction of a bioorthogonal handle at a specific position of a protein by recombinant techniques provides an excellent approach to site-specific conjugation, but for many laboratories and for applications where several proteins are to be labeled, the expression of recombinant proteins may be cumbersome. In recent years, a number of chemical methods that target conjugation to specific sites at native proteins have become available, and an overview of these methods is provided in this Account. Our laboratory has investigated DNA-templated protein conjugation (DTPC), which offers an alternative approach to siteselective conjugation of DNA to proteins. The method is inspired by the concept of DNA-templated synthesis where functional groups conjugated to DNA strands are preorganized by DNA hybridization to dramatically increase the reaction rate. In DPTC, we target metal binding sites in proteins to template selective covalent conjugation reactions. By chelation of a DNA−metal complex with a metal binding site of the protein, an electrophile on a second DNA strand is aligned for reaction with a lysine side chain on the protein in the proximity of the metal binding site. The method is quite general because approximately one-third of all wild-type proteins contain metal-binding sites, including many IgG antibodies, and it is also applicable to His-tagged proteins. This emerging field provides direct access to site-selective conjugates of DNA to commercially available proteins. In this Account, we introduce these methods to the reader and describe current developments and future aspects.
■
INTRODUCTION DNA carries the genetic information in all known living organisms. The precise recognition and hybridization of complementary DNA sequences are crucial to this function. This inherent ability for encoded molecular recognition also makes DNA a unique tool in bionanotechnology. DNA− protein conjugates are increasingly exploited because they combine the programmability of DNA with the diverse functionalities of proteins. Designed hybridization of DNA− protein conjugates allows construction of enzyme cascades,1 proximity ligation assays,2 assembly of multiple proteins in complex DNA nanostructures,3 and functionalization of solid supports for use in biosensors and biochips.4 Combining the proteins’ function with the encoding and recognition ability of DNA is particularly important for immuno-PCR5 and the super-resolution imaging technique DNA-PAINT.6 Our laboratory has conducted research in DNA nanotechnology for more than a decade and several of our planned projects have required DNA−protein conjugates, but we have often been frustrated by the lack of efficient chemical methods to produce high quality conjugates from commercially available proteins. Therefore, we engaged in the development of new methods for bioconjugation, which eventually lead to the development of DNA-templated protein conjugation (DPTC).7 A number of limitations must be considered when designing DNA−protein conjugation methods. The most important are © 2017 American Chemical Society
kinetics, biocompatibility, conservation of protein function, and applicability. Both DNA and protein must have accessible functional groups to enable conjugation. The relatively inert nature of DNA limits the chemical reactions in which it can take part, but automated solid-phase DNA synthesis both allows control over sequence and enables the introduction of a large variety of functional groups required for DNA−protein conjugation. Proteins, on the other hand, contain multiple reactive groups on their surfaces that often result in nonspecific (global) labeling and can lead to heterogeneous mixtures with a wide distribution in stoichiometry. However, in the absence of better solutions, this is often accepted. Strategies for the conjugation of DNA to proteins typically require high reaction rates, because they are often performed at low concentrations. The reactive groups should therefore enable fast and selective chemical reaction. In addition, the reaction must be biocompatible to preserve protein function, which usually implies neutral aqueous conditions at temperatures between 4 and 37 °C. The reactants have to be sufficiently stable to participate in the conjugation before they are degraded in the aqueous media or by the functional groups present in proteins. Received: December 12, 2016 Published: May 9, 2017 1367
DOI: 10.1021/acs.accounts.6b00618 Acc. Chem. Res. 2017, 50, 1367−1374
Article
Accounts of Chemical Research
Figure 1. Conjugation using (a) homo-bifunctional NHS linker, (b) hetero-bifunctional linker, (c) bifunctional linker targeting tryptophan, (d) homo-bifunctional linker targeting the N-terminus, and (e) extraction of a nondiffusible organic cofactor followed by reconstitution with a DNAmodified cofactor.
DNA−protein conjugates from commercially available native proteins.
Common for almost all DNA−protein conjugates is the importance of conserved protein function and activity. Nonspecific labeling can result in modification in close proximity to or at the protein active site and thereby inactivate the conjugate. Nonspecifically labeled antibodies show increased binding to off-targets,8 and heavily modified antibodies have shown lowered maximum tolerated IgG dose and rapid clearance in rats.9 Because homogeneous conjugates provide better control of protein function, it is desirable to form DNA−protein conjugates in a site-selective manner. Advances in genetic engineering have allowed the development of site-specific methods for expressing proteins fused with bioorthogonal handles, recognition sequences, or self-labeling protein tags. These methods have been reviewed elsewhere.10,11 They allow precise control over stoichiometry and positioning of modifications, but they are also associated with several drawbacks. For example, expression of engineered proteins can be cumbersome and can require individual customization and optimization for each expressed protein; the methods can require special equipment and often a subsequent chemical reaction still has to be performed; and larger modifications such as expression as a fusion protein can significantly alter protein function. Site-selective labeling provides a beneficial middle ground with control over stoichiometry and position of the modification without the need for protein engineering. A great advantage of selective labeling is the opportunity to form
■
HISTORY A traditional and widely used method in DNA−protein conjugation is the application of bifunctional cross-linkers. A range of homo- and hetero-bifunctional linkers functionalized with reactive esters or Michael acceptors for reactions with amino- or thiol-groups are commercially available (Figure 1a,b). In a seminal paper by Niemeyer and co-workers, such a linker (sulfo-SMPB) was used to create DNA−steptavidin conjugates that could be used to assemble DNA−protein complexes (Figure 1b).12 Lysine residues are abundant on most protein surfaces, so the general targeting of primary amines on lysine side chains often results in nonspecific labeling. In contrast, cysteine occurs with low abundance (1−2%) and its side chain thiol has a high nucleophilicity. However, cysteine thiols are often locked in structurally important disulfide bridges in proteins or serve as catalytic residues in enzyme active sites. Several interesting methods have been developed to target other low-frequency amino acids such as tyrosine or tryptophan. These residues have bulky aromatic side groups and therefore tend to orient toward the interior of the protein, resulting in low occurrence at protein surfaces.13 Tyrosine can be modified by cyclic diazodicarboxamides,14 and this was exploited by Fruk and co-workers in the design of 1368
DOI: 10.1021/acs.accounts.6b00618 Acc. Chem. Res. 2017, 50, 1367−1374
Article
Accounts of Chemical Research
Figure 2. DNA−protein conjugation using DNA-templated protein conjugation (DTPC). (a) Antibody−DNA conjugation directed by a metal ion chelating tris(NTA) modified DNA strand. (b) Introduction of a cleavable linker by DPTC followed by oxidative cleavage and introduction of an aldehyde handle able to further react with alkoxyamine-PEG5000 (blue). (c) Modification of a DNA origami structure by Tf-DNA conjugates prepared by DTPC for investigation of the effect on cellular uptake.
amino group of lysine residues (pKaH 10.5) and the N-terminal amino group (pKaH 7.7).21 This strategy was used to modify the enzymes pseudoazurin (Paz) and nitrite reductase (NiR) by an NHS bifunctional linker at pH 7.5 (Figure 1d).22 The redox enzyme NiR is able to catalyze the reduction of nitrite to nitric oxide, and in nature the enzyme forms a complex with its electron-transfer partner protein, Paz.4 The proposed Nterminal modification eliminated problems of nonreproducible conjugation products and provided control over the protein orientation, which was desired in electron-transfer studies. Even though Tepper did not provide data to verify the site-selectivity of the proposed N-terminal modification, the Francis lab has previously demonstrated that such N-terminal selectivity can in fact be achieved.23 However, the method requires that the Nterminus is unmodified and solvent exposed. Conjugation of DNA to organic cofactor-dependent proteins can be achieved by extracting the natural cofactor and replacing it with a DNA-modified synthetic cofactor (Figure 1e). One of
three bifunctional linkers for modification of proteins such as streptavidin (STV) and myoglobin (Mb) (Figure 1c).15 The three linkers all contained a cyclic urazole moiety and either a maleimide, an azide, or a cyclooctyne group at the other end. The linkers were preactivated by oxidation with N-bromosuccinimide (NBS) and pyridine (Pyr) in dimethylformamide to form the cyclic diazodicarboxamide and subsequently reacted with the protein. DNA strands with a thiol, an alkyne, or an azide modification, respectively, were then used to prepare the DNA−protein conjugates. Several bioconjugation methods have been developed to target tyrosine16,17 and tryptophan.18−20 By taking advantage of these strategies, other bifunctional linkers could be developed in the future allowing more proteins with rarely occurring surface-exposed amino acids to be site-selectively labeled with DNA. The N-terminal amino group of proteins can be selectively targeted by exploiting the differences in pKa values of the ε1369
DOI: 10.1021/acs.accounts.6b00618 Acc. Chem. Res. 2017, 50, 1367−1374
Article
Accounts of Chemical Research
conjugate and transported it into the cell interior. The two metalloenzymes alkaline phosphatase and carboxypeptidase B were also labeled by the method, and both enzymes retained their catalytic activity after conjugation.7 IgG antibodies possess a histidine-rich cluster on each side of the constant Fc domain, which has been reported to bind metal ions.34 This metal-binding site was exploited to form DNA− antibody conjugates by DPTC. Yields were improved when changing the metal-ion from nickel(II) to copper(II), which is known to form more stable complexes.35 DNA-conjugates of anti-c-Myc, anti-FLAG, anti-EGFR, and anti-β-tubulin were formed. It was shown by SDS-PAGE that the DNA strand was conjugated to the heavy-chain that contains the histidine-rich cluster. For anti-β-tubulin, MS/MS revealed that the DNA strands were conjugated to one of the two lysine residues closest to the histidine cluster. For the anti-FLAG and anti-βtubulin conjugates, it was demonstrated that the target binding ability of both antibodies was retained.7 An advantage of DPTC is that only equal stoichiometry of the protein and the DNA strands are required. In traditional protein conjugation, one of the reaction partners is usually used in large excess, which may be costly and lead to multiple conjugation products. More recently, the method was improved by changing the reacting group from a NHS ester to an aldehyde, thereby enabling conjugation by reductive amination.36 The change alleviates problems of hydrolysis of the highly reactive NHS ester during storage and handling. Furthermore, the decrease in reactivity of the aldehyde decreases nontemplated conjugation and allows the labeling to be performed at higher concentrations. The improved method was also applied to introduce an aldehyde handle to the protein by insertion of a cleavable linker between the two macromolecules. The DNA− protein conjugates were first separated from nonlabeled protein by nondenaturing PAGE. A linker between DNA and protein containing a 1,2-diol was then cleaved oxidatively by NaIO4. The resulting aldehyde handle on the protein reacted quantitatively with alkoxyamine-PEG5000 as shown by SDSPAGE and nondenaturing PAGE gel analysis (Figure 2b).36 When the same approach was applied to the antibody trastuzumab, the negative controls also showed conjugation. This was attributed to the carbohydrates on the glycosylated antibody, as the vicinal diols in the carbohydrates are also cleaved by NaIO4 to form aldehydes at the protein. This was, however, circumvented by removal of the glycan chains of the antibody by pretreatment with PNGase F, followed by DTPC and oxidative formation of the aldehyde handle. Double labeling of the antibody by site-selective conjugation proximate to the metal-binding site on both of the heavy chains was also demonstrated.36 DNA−Tf conjugates prepared by DTPC were applied to modify a DNA origami structure to study the effect on cellular uptake of DNA nanostructures decorated with Tf (Figure 2c).37 DNA origami is a self-assembly technique that folds a long single-stranded DNA template into desired structures by annealing with hundreds of designed synthetic DNA “staple” strands.38 In that study, the DNA origami structure contained single-stranded DNA extensions complementary to the DNA on the Tf−DNA conjugates enabling self-assembly of the proteins along the edge of the structure. By confocal laser scanning microscopy analysis, it was shown that the Tf modified DNA structures were taken up by KB carcinoma cells in a dose-dependent manner correlated with increasing
the advantages of this approach is that the enzyme is only active once the conjugate is formed. The technique was pioneered by Niemeyer and co-workers by forming fully active conjugates of myoglobin (Mb) and horseradish peroxidase (HRP).24 Compared to the native cofactor, DNA-conjugation did alter the enzyme activity,25 and in addition, the association rate constant for the uptake of the DNA-modified heme was found to be several orders of magnitude lower than for native heme.26 Conjugates have also been formed by linkers containing a modified heme propionate and a DNA-binding moiety such as the DNA binding peptide GCN4,27 [Pt(bpy)(en)]2+, or the DNA intercalators acridine and ethidium.28−30
■
DNA-DIRECTED PROTEIN CONJUGATION The specificity of post-translational modifications of proteins in cells arises from specific interactions between the responsible enzyme and target protein before the covalent bond is formed. In the same way, researchers have explored affinity, directing groups, and proximity as mechanisms to obtain site-selectivity in protein modifications. The concept known as DNA templated synthesis (DTS) was pioneered by Liu and co-workers. The method exploits the fact that most covalent chemical reactions do not proceed at nanomolar or low micromolar concentrations, whereas DNA hybridization is fast even at picomolar concentrations.31 When DNA strands modified with small molecules containing reactive groups are mixed in nanomolar to low micromolar concentrations, encoded DNA hybridization can preorganize the reactive groups and facilitate efficient reactions by locally simulating high millimolar concentration.31,32 We developed the method DTPC by combining DTS with a metal affinity probe.7 In DTPC, a template DNA strand is used to direct the conjugation between a DNA strand and a protein. The template strand carries a tris(NTA) moiety that chelates metal ions and enables coordination to metal binding sites of proteins. This interaction is used to direct a complementary DNA strand modified with an activated N-hydroxysuccinimide (NHS) ester to a lysine ε-amine in the vicinity of the metalbinding site and thereby facilitate the formation of a covalent DNA−protein product (Figure 2a).7 We used recombinant green fluorescent protein (GFP) containing an N-terminal His6-tag as a model protein, but the method was also used to form DNA−protein conjugates with several other proteins containing either an N- or a C-terminal His6-tag. The reaction was performed at a slightly basic pH of 7.7 to suppress NHS ester hydrolysis and render a sufficient fraction of lysine amines as free base for the reaction to proceed. Furthermore, this avoided the destabilization of NTA−metal complex at lower pH. The concentration of macromolecules and salt were found to be crucial for avoiding nontemplated, random conjugation. Coupling yields varied somewhat from protein to protein, an observation we attributed to variations in lysine accessibility and reactivity or the affinity to the His6-tag. Metalloproteins, which are estimated to represent more than one-third of all native proteins found in Nature, were also labeled by DTPC.33 An example is serotransferrin (Tf), which is responsible for transporting iron(III) across the cell membrane. By coordination to one of the two metal binding sites in Tf, DTPC was used to modify it with a DNA strand in a site-selective manner. Analysis by MS/MS showed that mainly two lysines close to the metal-binding site were modified. After conjugation, the Tf receptor still accepted the DNA−protein 1370
DOI: 10.1021/acs.accounts.6b00618 Acc. Chem. Res. 2017, 50, 1367−1374
Article
Accounts of Chemical Research
Figure 3. DNA-programmed photoaffinity labeling (DPAL) directed by a ligand modified DNA strand followed by photolysis and labeling of the protein.
Figure 4. Binding of the universal linker to the Fc region of an IgG antibody.
was conjugated to a DNA strand to form a binding probe (BP). The photoreactive group diazirine and a tag were conjugated to opposite ends of a complementary DNA strand to form a capture probe (CP). Upon photolysis, a reactive carbene species was formed and reacted with the protein. This approach made it easy to change one part of the system without changing the others and to vary the spacing between components. To demonstrate the applicability of the system in a complex mixture of proteins, the immunophilin FKBP12 in cell lysate was successfully labeled with fluorescein or biotin by using a specific inhibitor as the ligand. To show the multiplexity of the system, three proteins were labeled with different CPs either individually or in combination by using BPs with three different ligands and DNA sequences.40 Through testing of different photoreactive groups and incorporation of multiple groups, it was possible to increase the cross-linking efficiency 7-fold.41 As an extension, Li and co-workers applied the method to enrich and select small molecule binders from a DNA-encoded library.42 Besides being a method for directed DNA−protein conjugation DPAL has also been used in the identification of transcriptions factors and proteins binding to post-translational modifications on proteins.43,44 The DPAL method demonstrates obvious potential for producing site-selective DNA− protein conjugates, if acceptable isolation yields can be obtained. Staphylococcal protein A is a surface protein located in the cell wall of the bacteria and capable of binding the Fc region of most types of IgG antibodies. The engineered variant EZZ protein lacking the membrane anchor domain of protein A is often used in biochemical research. Luo and co-workers used this to develop a universal adapter for IgG antibodies by fusing the EZZ protein with the self-labeling protein SNAP-tag and subsequently reacting it with a benzyl guanine modified DNA strand.45 The universal adapter was used for site-specific formation of DNA−antibody conjugates (Figure 4). The
amounts of Tf. Quantitative PCR (qPCR) was used to quantify the scaffold strand isolated from cell lysates. To avoid false positive signals from extracellular bound DNA structures, the cell samples were pretreated with benzonase nuclease. Currently, we are investigating DNA nanostructures as potential vehicles for targeted drug delivery. As this study shows increased cellular uptake of DNA structures, it provides auspicious potential for further studies of receptor-mediated endocytosis and improved cell delivery systems.
■
PERSPECTIVES In recent years, other methods exploiting various affinity sites of proteins and diverse chemistries have been reported, but in most of these examples, the purpose of the DNA−protein conjugation is analytical such as identifying a protein in cell lysate through a DNA-tag. Hence, the conjugates are not purified or analyzed for site-selectivity or are only produced in very minute amounts. In this section, we describe and evaluate the potential of these methods for making conjugates to assess the potential of the field. Photoaffinity labeling utilizes protein−ligand binding in combination with a photoreactive group to covalently label proteins. It is typically used to identify unknown proteins that bind to a specific ligand. The ligand is tethered to a photoreactive cross-linking group, for example, phenylazide or diazirine, that after incubation with a complex biological sample can be selectively activated by photolysis and induced to form a covalent coupling between the probe and the protein.39 In small molecule photoaffinity labeling, the probe typically consists of three interlinked components; the ligand, a crosslinking group, and a tag for identification or isolation of the labeled proteins. To increase modularity and the capacity for multiplexing, Li and co-workers developed a DNA-programmed photoaffinity labeling (DPAL) method (Figure 3).40 In DPAL, a dual-probe system was used. The ligand 1371
DOI: 10.1021/acs.accounts.6b00618 Acc. Chem. Res. 2017, 50, 1367−1374
Article
Accounts of Chemical Research
Figure 5. Protein−aptamer conjugation by photolysis and conjugation of a diazirine phosphoramidite probe.
modification was thus able to provide the proper balance between reactivity and stability to allow site-selective DNA− protein conjugation.48 Instead of forming a covalent coupling between aptamer and protein, an extended version of the aptamer could potentially be used as a guiding strand as in DTPC. This would allow subsequent removal of the aptamer by strand displacement and liberated the large area of the protein surface otherwise covered by the aptamer binding-site. Hence, these are elegant methods for protein conjugation. Drawbacks are that a new aptamer has to be obtained for each protein to be labeled and that the sequence cannot be designed a priori.
conjugates were tested by dot blot, gel electrophoresis separation, Western blot, and with microbeads by hybridizing a DNA strand modified with nanobarcodes, quantum dots, and enzymes with the attached DNA strand. In a dot blot assay, no exchange reaction between antibodies and DNA labels was observed. The conjugates were tested in situ, and they were able to simultaneously visualize insulin and glucose proteins in diabetic mouse pancreas tissue at their expected positions.45 The method does have the disadvantage of a rather large adaptor but is highly selective for conjugation of DNA to antibodies. However, it should be kept in mind that the interaction is noncovalent, which may result in exchange of DNA sequences when different IgG antibodies and DNA sequences are applied. Modified aptamers have been used to form covalent conjugates with proteins, where the covalent bond is established in proximity to the aptamer affinity region. The technique, called Aptamer-Based Affinity Labeling (ABAL), was developed by Famulok and co-workers.46 In this method, the potential disturbance of the aptamer protein binding by the reactive group was avoided by placing the reactive group distant from the aptamer binding site. An ABAL-linker containing the photoreactive phenyl azide and a biotin-tag was attached to the 5′-end of the aptamer. Three aptamers, targeting three proteins, in three different cellular environments (membrane, cytoplasm, and blood) were chosen. In all cases conjugation products were obtained, but small amounts of byproducts were observed in the negative controls. The technique was applied to conjugate aptamers to their target in cell membrane and in cell lysate.46 In a related study, Yang and co-workers developed a diazirine phosphoramidite probe that could be introduced directly by automated DNA synthesis into an aptamer. This allowed more flexibility in the insertion position of the linker to be tested, and it was found that the location affected the aptamer binding affinity and the cross-linking efficiency (Figure 5).47 Aptamer-directed conjugation to proteins was also used by Tan and co-workers to form DNA conjugates of platelet derived growth factor-BB (PDGF-BB).48 A DNA strand containing a 3′-ribonucleotide was activated by NaIO4 oxidation to form a dialdehyde-modified strand. The aldehyde was brought into proximity of a lysine residue on the protein target in the vicinity of the aptamer binding-region. By addition of NaBH3CN, a secondary amine conjugate was formed by reductive amination. However, some conjugation products were also observed in the negative control. The aptamer modification was therefore changed to an α,α-gem-difluoromethyl carboxylic acid ethyl ester group (F-carboxyl). Common aliphatic esters normally do not react spontaneously with lysine residues, but the introduction of the electron withdrawing fluorides increases the reactivity of the ester. The modified aptamer was incubated with PDGF-BB and the crosslinking product was obtained in a yield comparable to that of the dialdehyde-modified aptamer; nonspecific conjugation was avoided in controls. The reactivity of the F-carboxyl
■
OUTLOOK AND CONCLUSION The site-selective conjugation of DNA to native proteins has developed remarkably in recent years, and several new methods have become available. However, the range of proteins that can be targeted and the level of selectivity remain a challenge. The suitability of the methods varies significantly from protein to protein, and the opportunities for DNA conjugation have to be validated for each protein. If the protein of interest has a single reactive and available functional group in amino acids such as cysteine, tyrosine, and tryptophan, the targeting of such groups may be the most straightforward route to site-selective conjugation with a DNA strand. Lysines are most often too abundant to offer sideselective conjugation; however if the amino group at the Nterminus is available and not acetylated, it may offer an opportunity for selective conjugation.31 A range of specific binding sites in proteins may be addressed. If the protein binds strongly to a cofactor (prosthetic group), conjugation of DNA to the prosthetic group (in a way that does not inhibit the binding) has proven to be an efficient way to conjugate DNA to apoenzymes. It should, however, be kept in mind that this approach may alter the activity of the enzyme.27 Alternatively, the DNA−cofactor conjugate could be used to direct the coupling to a complementary DNA strand in a system similar to DTPC. Weaker and noncovalent interactions between a binding-site on a protein and a ligand for the protein may be used to template the formation of a covalent bond in the vicinity of the binding-site. In the DPTC method developed in our laboratories, a metal complex−DNA conjugate interacting with a metal binding site on the protein was used to guide a second DNA strand to react with a lysine in proximity to the binding site.7,36 Due to the template effect, the reaction is relatively fast and only requires stoichiometric amounts of the reagents, even at low concentrations. This method is probably the most general site-selective method and also works well for most proteins containing a His6-tag. However, the method requires a lysine in proximity to the metal binding site; yields and specificity may vary from protein to protein. Analogous to the metal affinity DTPC method, DNA strands modified with a small molecule ligand such as an inhibitor that 1372
DOI: 10.1021/acs.accounts.6b00618 Acc. Chem. Res. 2017, 50, 1367−1374
Accounts of Chemical Research binds noncovalently at a specific site of the protein can also be used to obtain site-selective DNA−protein conjugates. This has been used in the DNA-programmed photoaffinity labeling (DPAL) method, where the reactive group linking the DNA to the protein is a diazirine.40 The small molecule ligand−DNA templated conjugation may offer high site-selectivity. A similar approach for affinity labeling of proteins with small molecules to proteins has been widely used, and many of these methods can potentially be extended to DNA conjugation.49−53 Modified aptamers have in many setups been used to form covalent DNA−protein conjugates, and this method can be used for site-selective conjugation to proteins that contain no particular small molecule or metal binding sites, as long as an aptamer for the protein is available.46−48 Aptamers containing an extended region may potentially be used to template a second DNA strand to react with proteins. The topic of this Account has been site-selective conjugation of DNA to native proteins, with particular focus on affinitydirected reactions. Whereas these reactions may at first sight seem difficult due to the dual labeling of the DNA strand or the employment of more modified DNA strands, the method offers significant advantages. In addition to the site-selectivity, the noncovalent template effect increases the rate of the reaction dramatically, and only stoichiometric amounts of DNA and protein may be required. With the increasing demand for highquality DNA−protein conjugates, we believe that these and future methods will provide attractive solutions to the selective preparation of conjugates with native proteins.
■
Article
■
ACKNOWLEDGMENTS
■
REFERENCES
The authors are grateful to A. M. T. Thygesen for assistance in the making of the illustrations and to Prof. Thomas H. LaBean for proofreading the manuscript.
(1) Siu, K.-H.; Chen, R. P.; Sun, Q.; Chen, L.; Tsai, S.-L.; Chen, W. Synthetic scaffolds for pathway enhancement. Curr. Opin. Biotechnol. 2015, 36, 98−106. (2) Fredriksson, S.; Gullberg, M.; Jarvius, J.; Olsson, C.; Pietras, K.; Gustafsdottir, S. M.; Ostman, A.; Landegren, U. Protein detection using proximity-dependent DNA ligation assays. Nat. Biotechnol. 2002, 20, 473−477. (3) Linko, V.; Nummelin, S.; Aarnos, L.; Tapio, K.; Toppari, J.; Kostiainen, M. A. DNA-Based Enzyme Reactors and Systems. Nanomaterials 2016, 6, 139−155. (4) Meyer, R.; Giselbrecht, S.; Rapp, B. E.; Hirtz, M.; Niemeyer, C. M. Advances in DNA-directed immobilization. Curr. Opin. Chem. Biol. 2014, 18, 8−15. (5) Adler, M.; Wacker, R.; Niemeyer, C. M. Sensitivity by combination: immuno-PCR and related technologies. Analyst 2008, 133, 702−718. (6) Jungmann, R.; Avendano, M. S.; Woehrstein, J. B.; Dai, M.; Shih, W. M.; Yin, P. Multiplexed 3D cellular super-resolution imaging with DNA-PAINT and Exchange-PAINT. Nat. Methods 2014, 11, 313− 318. (7) Rosen, C. B.; Kodal, A. L.; Nielsen, J. S.; Schaffert, D. H.; Scavenius, C.; Okholm, A. H.; Voigt, N. V.; Enghild, J. J.; Kjems, J.; Torring, T.; Gothelf, K. V. Template-directed covalent conjugation of DNA to native antibodies, transferrin and other metal-binding proteins. Nat. Chem. 2014, 6, 804−809. (8) Kazane, S. A.; Sok, D.; Cho, E. H.; Uson, M. L.; Kuhn, P.; Schultz, P. G.; Smider, V. V. Site-specific DNA-antibody conjugates for specific and sensitive immuno-PCR. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 3731−3736. (9) Junutula, J. R.; Raab, H.; Clark, S.; Bhakta, S.; Leipold, D. D.; Weir, S.; Chen, Y.; Simpson, M.; Tsai, S. P.; Dennis, M. S.; Lu, Y. M.; Meng, Y. G.; Ng, C.; Yang, J. H.; Lee, C. C.; Duenas, E.; Gorrell, J.; Katta, V.; Kim, A.; McDorman, K.; Flagella, K.; Venook, R.; Ross, S.; Spencer, S. D.; Wong, W. L.; Lowman, H. B.; Vandlen, R.; Sliwkowski, M. X.; Scheller, R. H.; Polakis, P.; Mallet, W. Site-specific conjugation of a cytotoxic drug to an antibody improves the therapeutic index. Nat. Biotechnol. 2008, 26, 925−932. (10) Lotze, J.; Reinhardt, U.; Seitz, O.; Beck-Sickinger, A. G. Peptidetags for site-specific protein labelling in vitro and in vivo. Mol. BioSyst. 2016, 12, 1731−1745. (11) Chen, X.; Wu, Y. W. Selective chemical labeling of proteins. Org. Biomol. Chem. 2016, 14, 5417−5439. (12) Niemeyer, C. M.; Sano, T.; Smith, C. L.; Cantor, C. R. Oligonucleotide-directed self-assembly of proteins: semisynthetic DNA–streptavidin hybrid molecules as connectors for the generation of macroscopic arrays and the construction of supramolecular bioconjugates. Nucleic Acids Res. 1994, 22, 5530−5539. (13) Lodish, H. F.; Berk, A.; Kaiser, C. A.; Krieger, M.; Bretscher, A.; Ploegh, H.; Amon, A.; Scott, M. P. Molecular cell biology, 7th ed.; W.H. Freeman and Co.: New York, 2013; pp 33−34. (14) Ban, H.; Gavrilyuk, J.; Barbas, C. F. Tyrosine Bioconjugation through Aqueous Ene-Type Reactions: A Click-Like Reaction for Tyrosine. J. Am. Chem. Soc. 2010, 132, 1523−1525. (15) Bauer, D. M.; Ahmed, I.; Vigovskaya, A.; Fruk, L. Clickable tyrosine binding bifunctional linkers for preparation of DNA-protein conjugates. Bioconjugate Chem. 2013, 24, 1094−1101. (16) Joshi, N. S.; Whitaker, L. R.; Francis, M. B. A Three-Component Mannich-Type Reaction for Selective Tyrosine Bioconjugation. J. Am. Chem. Soc. 2004, 126, 15942−15943. (17) Gavrilyuk, J.; Ban, H.; Nagano, M.; Hakamata, W.; Barbas, C. F. Formylbenzene Diazonium Hexafluorophosphate Reagent for Tyro-
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. ORCID
Kurt V. Gothelf: 0000-0003-2399-3757 Funding
This work was supported financially by the Danish National Research Foundation (Grant Number DNRF81) and Aarhus University, Faculty of Science and Technology. Notes
The authors declare no competing financial interest. Biographies Julie B. Trads was born in 1987 in Odder, Denmark. She obtained her M.S. in 2014 and is currently a Ph.D. student in the laboratory of Prof. Kurt V. Gothelf working with DNA−protein conjugation. Thomas Tørring was born in 1983 in Aalborg, Denmark. He performed his Ph.D. in the laboratory of Prof. Kurt V. Gothelf. He worked as a Carlsberg postdoctoral fellow with Prof. Jason Crawford at Yale University before returning to Aarhus University as an Assistant Professor. His research interests include chemical biology, biosynthesis, and natural product chemistry. Kurt Gothelf graduated in 1995 from the group of Professor K. A. Jørgensen at Aarhus University, Denmark, after studies in organic chemistry and asymmetric catalysis. Then followed a post doctoral stay in Professor M. C. Pirrung’s group at Duke University, USA, after which he joined the faculty at Aarhus University in 2002 as an Associate Professor working with surface chemistry and DNA nanotechnology. Since 2007, he has been Full Professor and director of Center for DNA Nanotechnology at Aarhus University. 1373
DOI: 10.1021/acs.accounts.6b00618 Acc. Chem. Res. 2017, 50, 1367−1374
Article
Accounts of Chemical Research
(38) Rothemund, P. W. K. Folding DNA to create nanoscale shapes and patterns. Nature 2006, 440, 297−302. (39) Ruoho, A. E.; Kiefer, H.; Roeder, P. E.; Singer, S. J. The mechanism of photoaffinity labeling. Proc. Natl. Acad. Sci. U. S. A. 1973, 70, 2567−2571. (40) Li, G.; Liu, Y.; Liu, Y.; Chen, L.; Wu, S.; Liu, Y.; Li, X. Photoaffinity labeling of small-molecule-binding proteins by DNAtemplated chemistry. Angew. Chem., Int. Ed. 2013, 52, 9544−9549. (41) Li, G.; Liu, Y.; Yu, X.; Li, X. Multivalent photoaffinity probe for labeling small molecule binding proteins. Bioconjugate Chem. 2014, 25, 1172−1180. (42) Zhao, P.; Chen, Z. T.; Li, Y. Z.; Sun, D. W.; Gao, Y.; Huang, Y. Y.; Li, X. Y. Selection of DNA-Encoded Small Molecule Libraries Against Unmodified and Non-Immobilized Protein Targets. Angew. Chem., Int. Ed. 2014, 53, 10056−10059. (43) Liu, Y.; Zheng, W. L.; Zhang, W.; Chen, N.; Liu, Y.; Chen, L.; Zhou, X. Z.; Chen, X. S.; Zheng, H. F.; Li, X. Y. Photoaffinity labeling of transcription factors by DNA-templated crosslinking. Chem. Sci. 2015, 6, 745−751. (44) Bai, X.; Lu, C.; Jin, J.; Tian, S.; Guo, Z.; Chen, P.; Zhai, G.; Zheng, S.; He, X.; Fan, E.; Zhang, Y.; Zhang, K. Development of a DNA-Templated Peptide Probe for Photoaffinity Labeling and Enrichment of the Histone Modification Reader Proteins. Angew. Chem., Int. Ed. 2016, 55, 7993−7997. (45) Tran, T. N. N.; Cui, J. H.; Hartman, M. R.; Peng, S. M.; Funabashi, H.; Duan, F. P.; Yang, D. Y.; March, J. C.; Lis, J. T.; Cui, H. X.; Luo, D. A Universal DNA-Based Protein Detection System. J. Am. Chem. Soc. 2013, 135, 14008−14011. (46) Vinkenborg, J. L.; Mayer, G.; Famulok, M. Aptamer-based affinity labeling of proteins. Angew. Chem., Int. Ed. 2012, 51, 9176− 9180. (47) Zhang, H.; Song, Y.; Zou, Y.; Ge, Y.; An, Y.; Ma, Y.; Zhu, Z.; Yang, C. J. A diazirine-based photoaffinity probe for facile and efficient aptamer-protein covalent conjugation. Chem. Commun. 2014, 50, 4891−4894. (48) Wang, R. W.; Lu, D. Q.; Bai, H. R.; Jin, C.; Yan, G. B.; Ye, M.; Qiu, L. P.; Chang, R. S.; Cui, C.; Liang, H.; Tan, W. H. Using modified aptamers for site specific protein-aptamer conjugations. Chem. Sci. 2016, 7, 2157−2161. (49) Tamura, T.; Hamachi, I. Recent Progress in Design of ProteinBased Fluorescent Biosensors and Their Cellular Applications. ACS Chem. Biol. 2014, 9, 2708−2717. (50) Tsukiji, S.; Hamachi, I. Ligand-directed tosyl chemistry for in situ native protein labeling and engineering in living systems: from basic properties to applications. Curr. Opin. Chem. Biol. 2014, 21, 136− 143. (51) Foley, T. L.; Burkart, M. D. Site-specific protein modification: advances and applications. Curr. Opin. Chem. Biol. 2007, 11, 12−19. (52) Koniev, O.; Wagner, A. Developments and recent advancements in the field of endogenous amino acid selective bond forming reactions for bioconjugation. Chem. Soc. Rev. 2015, 44, 5495−5551. (53) Takaoka, Y.; Ojida, A.; Hamachi, I. Protein Organic Chemistry and Applications for Labeling and Engineering in Live-Cell Systems. Angew. Chem., Int. Ed. 2013, 52, 4088−4106.
sine-Selective Modification of Proteins and the Introduction of a Bioorthogonal Aldehyde. Bioconjugate Chem. 2012, 23, 2321−2328. (18) Antos, J. M.; McFarland, J. M.; Iavarone, A. T.; Francis, M. B. Chemoselective Tryptophan Labeling with Rhodium Carbenoids at Mild pH. J. Am. Chem. Soc. 2009, 131, 6301−6308. (19) Antos, J. M.; Francis, M. B. Selective Tryptophan Modification with Rhodium Carbenoids in Aqueous Solution. J. Am. Chem. Soc. 2004, 126, 10256−10257. (20) Seki, Y.; Ishiyama, T.; Sasaki, D.; Abe, J.; Sohma, Y.; Oisaki, K.; Kanai, M. Transition Metal-Free Tryptophan-Selective Bioconjugation of Proteins. J. Am. Chem. Soc. 2016, 138, 10798−10801. (21) Grimsley, G. R.; Scholtz, J. M.; Pace, C. N. A summary of the measured pK values of the ionizable groups in folded proteins. Protein Sci. 2009, 18, 247−251. (22) Tepper, A. W. J. W. Electrical Contacting of an Assembly of Pseudoazurin and Nitrite Reductase Using DNA-Directed Immobilization. J. Am. Chem. Soc. 2010, 132, 6550−6557. (23) MacDonald, J. I.; Munch, H. K.; Moore, T.; Francis, M. B. Onestep site-specific modification of native proteins with 2-pyridinecarboxyaldehydes. Nat. Chem. Biol. 2015, 11, 326−331. (24) Fruk, L.; Niemeyer, C. M. Covalent Hemin−DNA Adducts for Generating a Novel Class of Artificial Heme Enzymes. Angew. Chem., Int. Ed. 2005, 44, 2603−2606. (25) Fruk, L.; Müller, J.; Niemeyer, C. M. Kinetic Analysis of Semisynthetic Peroxidase Enzymes Containing a Covalent DNA− Heme Adduct as the Cofactor. Chem. - Eur. J. 2006, 12, 7448−7457. (26) Fruk, L.; Kuhlmann, J.; Niemeyer, C. M. Analysis of hemereconstitution of apoenzymes by means of surface plasmon resonance. Chem. Commun. 2009, 230−232. (27) Sakamoto, S.; Kudo, K. Design and Synthesis of Semi-Artificial Myoglobin possessing DNA-Binding Peptides on Heme Propionates. Bull. Chem. Soc. Jpn. 2005, 78, 1749−1756. (28) Takashima, H.; Fujimoto, E.; Hirai, C.; Tsukahara, K. Synthesis and spectroscopic properties of reconstituted zinc-myoglobin appending a DNA-binding platinum(II) complex. Chem. Biodiversity 2008, 5, 2101−2112. (29) Takashima, H.; Matsushima, Y.; Araki, Y.; Ito, O.; Tsukahara, K. Synthesis and photophysical properties of zinc myoglobin appending an ethidium ion as a DNA intercalator. J. Biol. Inorg. Chem. 2008, 13, 171−181. (30) Takashima, H.; Tara, C.; Namikawa, S.; Kato, T.; Araki, Y.; Ito, O.; Tsukahara, K. Photoinduced intramolecular electron-transfer reactions of reconstituted met- and zinc-myoglobins appending acridine and methylacridinium ion as DNA-binders. J. Phys. Chem. B 2006, 110, 26413−26423. (31) Li, X.; Liu, D. R. DNA-templated organic synthesis: nature’s strategy for controlling chemical reactivity applied to synthetic molecules. Angew. Chem., Int. Ed. 2004, 43, 4848−4870. (32) Li, G.; Zheng, W.; Liu, Y.; Li, X. Novel encoding methods for DNA-templated chemical libraries. Curr. Opin. Chem. Biol. 2015, 26, 25−33. (33) Andreini, C.; Bertini, I.; Cavallaro, G.; Holliday, G. L.; Thornton, J. M. Metal ions in biological catalysis: from enzyme databases to general principles. J. Biol. Inorg. Chem. 2008, 13, 1205− 1218. (34) Hale, J. E.; Beidler, D. E. Purification of Humanized Murine and Murine Monoclonal-Antibodies Using Immobilized Metal-Affinity Chromatography. Anal. Biochem. 1994, 222, 29−33. (35) Irving, H.; Williams, R. J. P. The stability of transition-metal complexes. J. Chem. Soc. 1953, 3192−3210. (36) Kodal, A. L.; Rosen, C. B.; Mortensen, M. R.; Torring, T.; Gothelf, K. V. DNA-Templated Introduction of an Aldehyde Handle in Proteins. ChemBioChem 2016, 17, 1338−1342. (37) Schaffert, D. H.; Okholm, A. H.; Sørensen, R. S.; Nielsen, J. S.; Tørring, T.; Rosen, C. B.; Kodal, A. L. B.; Mortensen, M. R.; Gothelf, K. V.; Kjems, J. Intracellular Delivery of a Planar DNA Origami Structure by the Transferrin-Receptor Internalization Pathway. Small 2016, 12, 2634−2640. 1374
DOI: 10.1021/acs.accounts.6b00618 Acc. Chem. Res. 2017, 50, 1367−1374
