Ribosome Display and Photo-Cross-Linking Techniques for In Vitro

Dec 31, 2013 - The identification of target proteins of bioactive small molecules as bioprobe candidates or drug seeds is indispensable for elucidatin...
0 downloads 0 Views 1MB Size
Download PDF
Technical Note pubs.acs.org/ac

Ribosome Display and Photo-Cross-Linking Techniques for In Vitro Identification of Target Proteins of Bioactive Small Molecules Akira Wada,*,†,‡,∥ Shuta Hara,†,§ and Hiroyuki Osada*,†,§,∥ †

Antibiotics Laboratory, RIKEN, 2-1 Hirosawa, Wako, Saitama 351-0198, Japan Precursory Research for Embryonic Science and Technology (PRESTO), Japan Science and Technology Agency (JST), 4-1-8 Honcho, Kawaguchi, Saitama 332-0012, Japan § Biomedical Science PhD Program, Tokyo Medical and Dental University, 1-5-45 Yushima, Bunkyo-ku, Tokyo 113-8510, Japan ∥ Chemical Biology Research Group, RIKEN Center for Sustainable Resource Science, 2-1 Hirosawa, Wako, Saitama 351-0198, Japan ‡

S Supporting Information *

ABSTRACT: The identification of target proteins of bioactive small molecules as bioprobe candidates or drug seeds is indispensable for elucidating their actions and predicting their side effects. To meet the current need, we developed a scheme for detection and identification of target proteins by using ribosome display and photo-cross-linking techniques, and demonstrated the feasibility. The mRNAs encoding full-length human proteins (FHPs) were constructed and translated in vitro to prepare pools of FHP−ribosome−mRNA complexes used for ribosome display selection. Expression levels of the FHPs were confirmed by Western blot analysis, and photo-cross-linked smallmolecule beads were assessed through cell-free synthesized FHP binding assay. After ribosome display selection against photocross-linked small-molecule beads, RT-PCR using mRNAs recovered from the selected ternary complexes and electrophoresis of the PCR products allowed specific detection of the target proteins binding to the beads. In addition, a repeat of ribosome display selection enabled us to identify the target proteins even if the molar quantity was one ten-thousandth of that of the other proteins in a cell-free synthesized FHP pool. Therefore, these results showed that ribosome display using photo-cross-linked smallmolecule beads and further extended FHP pool could be one of the powerful techniques for identification of unknown target proteins of bioactive small molecules.

R

ment of small molecules as novel bioprobes or therapeutic agents, an innovative approach to precisely and comprehensively identify target proteins is required. Other techniques such as phage display,4 mRNA display,5 and ribosome display6 have been developed in order to link peptides/proteins of interest as phenotypes with their corresponding genes as genotypes. These techniques are powerful tools used to select peptides/proteins bound specifically to a target molecule from libraries that comprise a wide variety of peptides/proteins. Recently, phage display has been adopted for the identification of target proteins of bioactive small molecules. Various target proteins have been successfully identified from pools that were prepared from cDNA encoding human proteins.4d−g However, phage display has several problems related to the use of cDNA isolated from human cells, the influence of codon usage bias, cell toxicity of peptides/proteins against E. coli as host cells, and lack of expression of proteins whose coding sequences are not in-frame with those of phage coat proteins. These challenges decrease

ecent studies in the boundary area of chemistry and biology have focused on the exploration and deconvolution of bioactive small molecules that specifically bind to target proteins or enzymes associated with the development of cancer and a variety of diseases.1 Small molecules can be used as bioprobes or drugs to inhibit intracellular signal transduction systems or to induce particular cellular functions. With this background, drug discovery research with chemical libraries consisting of natural products and synthetic compounds2 has actively expanded in academia as well as the pharmaceutical industry. However, in the past decade, whenever small molecules have been developed as pharmaceuticals, there have been difficulties associated with “identifying target proteins to which a small molecule binds”.3 Thus, the mechanisms and side effects of such small molecules cannot be predicted easily. Pull-down and yeast-3-hybrid techniques3e,f have been widely and routinely used to identify target proteins of bioactive small molecules. The pull-down technique enables the isolation of proteins from cell extracts that bind to smallmolecule-immobilized beads. The isolated proteins are analyzed by mass spectrometry to determine their amino acid sequences. However, it is often quite difficult to distinguish between target proteins and nontarget proteins that bind nonspecifically to the bead surfaces, as well as in identifying target proteins with low expression levels in cells. Therefore, for the further develop© 2013 American Chemical Society

Received: September 2, 2013 Accepted: December 31, 2013 Published: December 31, 2013 6768

dx.doi.org/10.1021/ac4030208 | Anal. Chem. 2014, 86, 6768−6773

Analytical Chemistry

Technical Note

complexes were purified with an RNeasy Mini Kit (Qiagen). Fifty microliters of the recovered mRNA was reverse transcribed at 50 °C for 60 min in a reaction mixture (100 μL) containing 4 μL of PrimeSTAR Reverse Transcriptase (TaKaRa), 5 μL of dNTP Mix (Promega), and 20 pmol of primer (rp-NS). The obtained cDNAs were applied to PCR for identifying target molecules or to synthesize DNA constructs for initiating the next round of ribosome display selection. The other experimental procedures and their conditions are described in the Supporting Information [SI].

the diversity of proteins that can be displayed on the surface of phage particles. In contrast to phage display, mRNA display and ribosome display, which are employed by cell-free protein synthesis systems, are unaffected by the limitations of using live cells. Therefore, cell-free display techniques are considered to be much more useful in the identification of target proteins of bioactive small molecules. McPherson and colleagues constructed a pool of proteins derived from cDNA of human tissues using PCR with random primers and successfully selected a target protein against FK506−biotin−streptavidin beads using mRNA display.5g Additionally, Yanagawa and colleagues identified the target protein of a phthalimide derivative by mRNA display selection on a microfluidic chip.5h However, mRNA display selection has not been often used for exploration of unknown target proteins. The reasons for this are likely related to mutations within stop codons and truncations of native cDNA due to random priming in PCR or to difficulties in the synthesis of a puromycin-attached DNA linker for construction of various protein−mRNA complexes. Therefore, in order to establish a reliable approach for detection and identification of target proteins, it would be essential to construct a pool consisting of full-length human proteins (FHPs) and adapt it to an adequate display technique. With this background, we focused on potentials of ribosome display suitable for performing desired selections at a faster speed and photo-cross-linking reactions that enable covalent immobilization of small molecules on bead surfaces. Thus, in this study, we constructed a scheme for detection and identification of target proteins by using ribosome display selection with photo-cross-linked small-molecule beads, and then demonstrated the feasibility and utility of these techniques.

■

RESULTS AND DISCUSSION Validation of the Scheme for Ribosome Display Selection. An overview of ribosome display selection of target proteins against photo-cross-linked small-molecule beads is described in Figure 1. In contrast to mRNA display selection

■

Figure 1. Schematic representation of in vitro identification of target proteins using the ribosome display technique and photo-cross-linked small-molecule beads. (I) In vitro transcription of DNA constructs encoding full-length human proteins (FHPs) for synthesizing mRNAs. (II) In vitro translation of the synthesized mRNA pool for constructing ternary complexes that consisted of FHP, ribosomes, and mRNA. (III) In vitro selection of ternary complexes bound to photo-cross-linked small-molecule beads. (IV) Recovery of mRNAs by dissociating the selected complexes through the addition of EDTA. (V) Reverse transcription (RT)-PCR for the reconstruction of DNAs used to identify target peptides/proteins or initiate the next round of selection.

EXPERIMENTAL SECTION Ribosome Display Selection. The DNA constructs DNA(FLAG-Ps-NS) and DNA(FHP-Ps-NS) (Figure 2) were transcribed in vitro using a RiboMAX Large Scale RNA Production System (Promega). The resulting mRNAs were purified with an RNeasy Mini Kit (Qiagen), and their concentrations were determined by measurement of the absorbance at 260 nm. The pools consisting of mRNAs encoding a FLAG peptide or FHPs were designed for each ribosome display selection and prepared as described in the text. These mRNA pools were separately translated using a PURESYSTEM7 classic II (BioComber) at 37 °C for 30 min in a 50 μL reaction mixture that contained 30 pmol of mRNA, and 1 μL of RNasin (Promega). After the translation of each mRNA pool was stopped by cooling on ice, solutions containing various ribosomal complexes were diluted 5-fold with Buffer S [60 mM Tris-AcO (pH 7.5), 180 mM NaCl, 60 mM Mg(AcO)2, 0.05% Tween 20] and mixed with 15 μL of ANTI-FLAG M2 Affinity Gel (Sigma), 15 μL of photo-crosslinked FK506 beads, or 15 μL of photo-cross-linked CsA beads. The resulting mixtures were gently rotated at 4 °C for 60 min. After washing the beads 5 times with 400 μL of Buffer W [50 mM Tris-AcO (pH 7.5), 150 mM NaCl, 50 mM Mg(AcO)2, 0.05% Tween 20], 100 μL of Buffer E [50 mM Tris-AcO (pH 7.5), 150 mM NaCl, 60 mM ethylenediamine-N,N,N′,N′tetraacetic acid (EDTA), 0.05% Tween 20] was added, and the mixture was gently shaken for 30 min to dissociate ribosomal complexes bound to the surfaces. The supernatants were then carefully collected, and mRNAs released from the selected

using wheat germ extract or rabbit reticulocyte lysate,5g,h we used an E. coli cell-free protein synthesis system for ribosome display selection in this study. This characteristic excluded the possibility that any orthologues of human proteins involved in the eukaryotic translational system behaved like target molecules and bound to the small molecules. As shown in Figure 1-II, a protein domain as a spacer at the C-terminus needs to be added to break various proteins of interest free from the ribosome tunnel and to reduce steric hindrances between the ribosome and proteins or between the ribosome and an intended molecule to which the target protein binds. Thus, we attached a protein spacer that was longer than a conventional one to the C-termini of FHPs (FKBP12, SOD1 (G86R), CypA, BID, KRas (G12V), DHFR, and p38α), which are associated with various diseases and cancers, as well as a FLAG peptide, which is a target epitope for anti-FLAG antibodies (SI). Subsequently, we confirmed whether these fusion proteins could be expressed using an E. coli cell-free 6769

dx.doi.org/10.1021/ac4030208 | Anal. Chem. 2014, 86, 6768−6773

Analytical Chemistry

Technical Note

DNA(FHP-Ps-NS)] were freshly constructed (Figure 2) and transcribed to synthesize the corresponding mRNAs. Then, the initial mRNA pool containing equimolar amounts of each mRNA was prepared, and a portion of it was reversetranscribed to synthesize cDNA for evaluating the quality of the mRNA pool. The synthetic cDNA was used as a template for performing PCR with primers that recognized the individual sequences encoding the peptide or FHPs. The resulting PCR products were evaluated by electrophoretic analysis. As shown in Figure 4A, all of the bands were observed to be uniform,

protein synthesis system. As depicted in Figure 2, the DNA constructs, DNA(FLAG-Ps-S) and DNA(FHP-Ps-FS) that

Figure 2. DNA constructs that enable synthesis of mRNAs encoding various fusion proteins used for the comparison of protein expression levels or performance of ribosome display selection. The T7 promoter (T7) and ribosome binding site (RBS) are necessary for in vitro transcription and translation, respectively. The coding sequences for a FLAG peptide or FHPs (FKBP12, SOD1 (G86R), CypA, BID, KRas (G12V), DHFR, and p38α) are inserted between original Sf iI restriction sites. ATG: start codon, Ps: the coding sequence for a flexible protein spacer, S: stop codon, NS: no stop, and FS: FLAG and stop codon.

have the coding sequences for FLAG fusion protein and FHP fusion proteins, respectively, were transcribed to synthesize the corresponding mRNAs. After translation of the mRNAs using this system, the expression level of each protein was assessed by Western blot analysis (Figure 3). The results indicated that all

Figure 4. Electrophoretic gel images of RT-PCR products derived from (A) the initial mRNA pool before selection and (B) total mRNAs recovered after selection with anti-FLAG antibody-immobilized beads. The initial mRNA pool contained equimolar amounts of each mRNA encoding a FLAG peptide and FHPs (F: FLAG-Ps-NS, 1: FKBP12-PsNS, 2: SOD1 (G86R)-Ps-NS, 3: CypA-Ps-NS, 4: BID-Ps-NS, 5: KRas (G12V)-Ps-NS, 6: DHFR-Ps-NS, and 7: p38α-Ps-NS). The band with weak fluorescence intensity observed in lane 2 of (B) was produced nonspecifically. (C) Enrichment of mRNA of the target peptide and exclusion of mRNAs of nontarget proteins through ribosome display selection. The amount of each RT-PCR product was quantified by analysis of fluorescence band intensities observed in (A) and (B).

Figure 3. Western blot analysis for relative comparison of the expression levels of a FLAG fusion protein (F: FLAG-Ps-S) and FHP fusion proteins (1: FKBP12-Ps-FS, 2: SOD1 (G86R)-Ps-FS, 3: CypAPs-FS, 4: BID-Ps-FS, 5: KRas (G12V)-Ps-FS, 6: DHFR-Ps-FS, and 7: p38α-Ps-FS). (A) The highest expression level of FLAG-Ps-S could be compared with the smallest expression level of FKBP12-Ps-FS, when the amount of FLAG-Ps-S corresponded to one-tenth of that of FKBP12-Ps-FS. (B) The expression levels of all of the fusion proteins except for FLAG-Ps-S could be compared by using an equal amount of each solution.

which indicated that the initial mRNA pool contained equimolar amounts of each type of the mRNA. Next, a pool of the peptide/protein−ribosome−mRNA complexes was prepared by translation of the initial mRNA pool and was screened against anti-FLAG antibody-immobilized beads. After RT-PCR for mRNAs recovered through ribosome display selection, gel electrophoresis for the PCR products revealed that only the band derived from the FLAG peptide exhibited high fluorescence intensity (Figure 4B). Furthermore, to assess changes in the abundance ratios of the peptide and proteins after selection, the fluorescence band intensities observed in A and B of Figure 4 were quantitatively compared (Table S1 in SI). A calculated positive value indicates an increase in the amount of the peptide or proteins after selection, whereas a calculated negative value indicates a decrease after selection. As shown in Figure 4C, the observation of a higher positive value signified that the FLAG peptide could be identified as a target molecule by the ribosome display selection according to the designed scheme.

of the fusion proteins could be expressed by E. coli translational machinery. On the basis of empirical data, it was also found that the expression levels were sufficient for performing ribosome display selection. Moreover, FLAG fusion protein showed the highest expression level (Figure 3A). On the other hand, the FKBP12 and p38α fusion proteins exhibited relatively lower expression levels (Figure 3B). To optimize and verify the scheme for identification of target molecules from a pool of protein−ribosome−mRNA complexes as described in Figure 1, we tried to perform ribosome display selection of a FLAG peptide that was known to interact with anti-FLAG antibody-immobilized beads. For construction of the ternary complexes displaying nascent proteins on the ribosomes idling at the 3′ terminus of mRNA, it is necessary to remove any stop codons in the mRNAs. Therefore, on the basis of the sequences of DNA(FLAG-Ps-S) and DNA(FHP-Ps-FS), DNAs without stop codons [i.e., DNA(FLAG-Ps-NS) and 6770

dx.doi.org/10.1021/ac4030208 | Anal. Chem. 2014, 86, 6768−6773

Analytical Chemistry

Technical Note

Synthesis and Assessment of Photo-cross-linked Small-Molecule Beads. Diazirines are able to form covalent bonds with all molecules, regardless of the absence of a functional group, because they are activated by photoirradiation to generate “carbenes”,8 which react readily with various sites of small molecules. Therefore, the photo-cross-linking technique using diazirines enables us to immobilize small molecules covalently on the bead surface in a state where the site that is needed to bind to their target proteins is exposed. To ascertain whether target proteins expressed by the cellfree protein synthesis system could specifically bind to photocross-linked small-molecule beads, cyclosporine A (CsA), FK506, and TPh A were immobilized on the beads using the photo-cross-linking technique (SI). Subsequently, the translation solutions of FHP fusion proteins having FLAG tags at their C-termini (SOD1 (G86R)-Ps-FS, CypA-Ps-FS, DHFR-PsFS, FKBP12-Ps-FS, and Pirin-Ps-FS whose cell-free expression was confirmed by Western blot analysis in Figure S1(A) in SI) were individually mixed with the photo-cross-linked CsA, FK506, or TPh A beads. After the addition of anti-FLAG antibody−HRP conjugates to the solutions, chemiluminescence from the reaction of HRP with light-emitting substrates was measured. Predictably, substantial chemiluminescence was not detected at the surfaces of beads mixed with the solution in the absence of mRNA (A-I, B-I, and C-I of Figure 5), the solution of SOD1 (G86R) fusion protein (Figures 5A-II), the solution of DHFR fusion protein (Figures 5B-II), and the solution of FKBP12 fusion protein (Figures 5C-II). On the other hand, strong light emission was observed at the surfaces of beads mixed with the solution of CypA fusion protein as a target protein of CsA (Figure 5A-III), the solution of FKBP12 fusion protein as a target protein of FK506 (Figure 5B-III), and the solution of Pirin fusion protein as a target protein of TPh A (Figure 5C-III). These results clearly demonstrated that photocross-linked small molecules on bead surfaces exposed sites needed to bind specifically to the cell-free synthesized target proteins. In addition, according to X-ray crystal structure analysis and isothermal titration calorimetry reported previously,2c TPh A bound to the iron active site of Pirin, and the dissociation constant (Kd) was 0.6 μM, which is larger than those of CsA (40 nM) and FK506 (0.4 nM). Therefore, the chromogenic assay was also carried out to assess the interaction between cellfree synthesized Pirin and photo-cross-linked TPh A beads in the absence/presence of Fe ions. As shown in Figure S1(B) in SI, a significantly increased change in the absorbance was observed only in the copresence of Pirin-Ps-FS and Fe ions. The result indicated that photo-cross-linked TPh A bound specifically to the iron active site of cell-free synthesized Pirin. One of the advantages of ribosome display selection is that the ribosomal complexes displaying target proteins bound to small molecules can be easily dissociated by only adding EDTA. This enables us to recover mRNAs involved in the selected complexes without any treatment. In general, most small molecules are very poorly soluble in aqueous solutions. Thus, it is difficult to prepare highly concentrated solutions for competitively eluting out target proteins that bind to smallmolecule-immobilized beads. In addition, since bioactive small molecules bind very tightly to target proteins, they could not be competitively eluted from the bead surfaces even if a solution of free small molecules was added in excess. Thus, to prove the above advantage experimentally, ribosomal complexes displaying FKBP12 were independently prepared and mixed with

Figure 5. Chemiluminescent assay for the binding of cell-free synthesized proteins to (A) photo-cross-linked CsA beads (I: no protein, II: SOD1 (G86R)-Ps-FS, and III: CypA-Ps-FS), (B) photocross-linked FK506 beads (I: no protein, II: DHFR-Ps-FS, and III: FKBP12-Ps-FS), and (C) photo-cross-linked TPh A beads (I: no protein, II: FKBP12-Ps-FS, and III: Pirin-Ps-FS). BI and CI denote bright-field images and chemiluminescence images, respectively. (D) Recovery of mRNAs from FKBP12-displaying ribosomal complexes bound to photo-cross-linked FK506 beads. Electrophoretic gel images of RT-PCR products derived from the mRNA of FKBP12 after recovery by elution with free FK506 (EF) or EDTA (EE). The concentrations of FK506 and EDTA were 10 μM and 60 mM, respectively. M denotes a DNA marker.

photo-cross-linked FK506 beads, and complexes binding to the bead surfaces were eluted by the addition of FK506 or EDTA. Then, RT-PCR was carried out using the recovered mRNA for each PCR cycle, and gel electrophoresis of the RT-PCR products was analyzed. As shown in Figure 5D, in the case of elution with FK506, the bands of interest were not observed even after 32 cycles of PCR. This result indicated the difficulty of achieving effective recovery of desirable ribosomal complexes by the addition of FK506, which has a high affinity for the target protein. On the other hand, in the case of elution with EDTA, the bands of interest appeared at the position of the arrow in Figure 5D, beginning at 23 cycles of PCR. Moreover, as the number of cycles increased, the band intensity increased. These results evidently indicated that the ribosomal complexes bound to FK506 on the bead surfaces could be sufficiently eluted by the addition of EDTA. Therefore, elution with EDTA enabled the dissociation of the selected complexes without the influences of target affinity and insolubility of small molecules, as well as the recovery of mRNAs for identifying target proteins. Identification of Highly Diluted Target Protein by Repeating Ribosome Display Selection. To verify whether 6771

dx.doi.org/10.1021/ac4030208 | Anal. Chem. 2014, 86, 6768−6773

Analytical Chemistry

Technical Note

highly diluted FKBP12 could be identified by repeating ribosome display selection with photo-cross-linked FK506 beads, an initial mRNA pool was freshly constructed to contain equimolar amounts of mRNAs encoding SOD1 (G86R), CypA, BID, KRas (G12V), DHFR, and p38α, and an extremely small amount of mRNA encoding FKBP12 (3 fmol), which corresponded to one ten-thousandth (1/10,000th) of the total mRNA (30 pmol). The quantity of the mRNA encoding FKBP12 was almost identical to that of the mRNA encoding one protein in a pool that would contain mRNAs encoding 10,000 species of FHPs. At this time, according to the scheme shown in Figure 1, ribosome display selection against photocross-linked FK506 beads was repeated twice. Then, after synthesis of cDNA by reverse transcription of the portions of mRNA before selection, mRNA after the first round of selection, and mRNA after the second round of selection, PCR was carried out separately at the same cycle number using each cDNA as a template. Gel electrophoresis of the PCR products revealed that the fluorescence band intensity increased dramatically with increasing selection round (Figure 6A). This tendency indicated that the mRNA of FKBP12 was recovered and amplified through selection of FKBP12−ribosome−mRNA complexes bound to photo-cross-linked FK506 beads. Furthermore, to evaluate the effectiveness of ribosome display selection, mRNAs before selection and after the second round of selection were analyzed by gel electrophoresis of the RT-PCR products. As shown in Figure 6B, the fluorescence band intensity representing FKBP12 expression was present only at trace level due to its extremely low concentration, while the fluorescence band intensities representing the expression level of the other proteins were uniform. After the second round of selection, a band derived from FKBP12 appeared at the position of the arrow in Figure 6C, indicating effective amplification of the corresponding mRNA. On the other hand, the fluorescence band intensity representing p38α expression decreased markedly, indicating that p38α was excluded as a nontarget protein during selection. Similar to p38α, we expected to observe decreases in the fluorescence intensity of bands derived from SOD1 (G86R), CypA, BID, KRas (G12V), and DHFR. However, fluorescence band intensities representing the expression level of these proteins did not change significantly after selection. Therefore, we quantified the fluorescence band intensities of each protein before selection and after the second round of selection (Table S2 in SI). As depicted in Figure 6D, only FKBP12 exhibited a high positive value compared with the other proteins. These data clearly demonstrated that FKBP12 was specifically selected as the target protein of FK506 and that the other proteins were treated as nontarget proteins during selection. Considering the protein expression levels (Figure 3) and the very low concentration of mRNA encoding FKBP12 in the initial mRNA pool, the abundance of FHP−ribosome−mRNA complexes was expected to decrease in the following order: SOD1 (G86R), CypA, BID, KRas (G12V), and DHFR > p38α ≫ FKBP12. However, as shown by the data in Figure 6D, FKBP12 was preferentially selected through specific binding to FK506. On the other hand, SOD1 (G86R), CypA, BID, KRas (G12V), and DHFR, which were expressed at almost the same levels, bound nonspecifically and equally to bead surfaces, regardless of their amino acid sequences and structures. Therefore, notable changes in the fluorescence intensities of bands derived from these proteins were not observed in the electrophoretic analysis. Following this principle, p38α was also

Figure 6. (A) Electrophoretic gel image of RT-PCR products derived from the mRNA of FKBP12 before selection (BS) and after the first round (1R) and the second round (2R) of ribosome display selection with photo-cross-linked FK506 beads (upper panel). The PCR products at each round were obtained by performing 17 cycles of PCR. The increase in the fluorescence band intensity indicated enrichment of the mRNA of FKBP12 (lower panel). Electrophoretic gel images of RT-PCR products derived from (B) the initial mRNA pool before selection and (C) total mRNAs recovered after the second round of selection (1: FKBP12-Ps-NS (1/10,000th), 2: SOD1 (G86R)-Ps-NS, 3: CypA-Ps-NS, 4: BID-Ps-NS, 5: KRas (G12V)-PsNS, 6: DHFR-Ps-NS, and 7: p38α-Ps-NS). The PCR products were obtained by performing 10 cycles of PCR. (D) Enrichment of mRNA of the target protein and exclusion of mRNAs of nontarget proteins through two rounds of selection. The amount of each RT-PCR product was quantified by analysis of fluorescence band intensities observed in (B) and (C).

thought to bind nonspecifically to bead surfaces. However, p38α was at a disadvantage in terms of competitive binding to the bead surfaces compared to SOD1 (G86R), CypA, BID, KRas (G12V), and DHFR because the expression of p38α was much lower than that of the other proteins. Thus, p38α was gradually excluded during selection. According to the consideration, if the expression level of FKBP12 was enhanced by performing further rounds of selection, specific binding of FKBP12 to photo-cross-linked FK506 beads would become dominant over nonspecific binding of the other proteins onto bead surfaces. Thus, a one-time ribosome display selection against photo-cross-linked FK506 beads using an mRNA pool consisting of equimolar amounts of each type of mRNA derived from the proteins (FKBP12, SOD1 (G86R), CypA, BID, KRas (G12V), DHFR, and p38α) was performed (SI). This selection was almost comparable to the selection of FKBP12 under conditions in which FKBP12-displaying ribosomal complexes became abundant through repeated selection. Predictably, electrophoretic analysis after this selection allowed us to observe only a single band derived from FKBP12 as a target protein (Figure S2 and Table S3 in SI ), supporting the above hypothesis. 6772

dx.doi.org/10.1021/ac4030208 | Anal. Chem. 2014, 86, 6768−6773

Analytical Chemistry

Technical Note

Martinez, N. E.; Görmer, K.; Rosin-Steiner, S.; Shimizu, T.; Ozaki, S.; Mikoshiba, K.; Watanabe, N.; Hall, D.; Vetter, I. R.; Osada, H.; Hedberg, C.; Waldmann, H. ChemBioChem. 2013, 14, 115−122. (3) (a) Lamb, J.; Crawford, E. D.; Peck, D.; Modell, J. W.; Blat, I. C.; Wrobel, M. J.; Lerner, J.; Brunet, J. P.; Subramanian, A.; Ross, K. N.; Reich, M.; Hieronymus, H.; Wei, G.; Armstrong, S. A.; Haggarty, S. J.; Clemons, P. A.; Wei, R.; Carr, S. A.; Lander, E. S.; Golub, T. R. Science 2006, 313, 1929−1935. (b) Young, D. W.; Bender, A.; Hoyt, J.; McWhinnie, E.; Chirn, G. W.; Tao, C. Y.; Tallarico, J. A.; Labow, M.; Jenkins, J. L.; Mitchison, T. J.; Feng, Y. Nat. Chem. Biol. 2008, 4, 59− 68. (c) Campillos, M.; Kuhn, M.; Gavin, A. C.; Jensen, L. J.; Bork, P. Science 2008, 321, 263−266. (d) Lomenick, B.; Olsen, R. W.; Huang, J. ACS Chem. Biol. 2011, 21, 34−46. (e) Ziegler, S.; Pries, V.; Hedberg, C.; Waldmann, H. Angew. Chem., Int. Ed. 2013, 52, 2744−2792. (f) Chidley, C.; Haruki, H.; Pedersen, M. G.; Muller, E.; Johnsson, K. Nat. Chem. Biol. 2011, 7, 375−383. (4) (a) Hoess, R. H. Chem. Rev. 2001, 101, 3205−3218. (b) Eisenhardt, S. U.; Schwarz, M.; Bassler, N.; Peter, K. Nat. Protoc. 2007, 2, 3063−3073. (c) Heinis, C.; Rutherford, T.; Freund, S.; Winter, G. Nat. Chem. Biol. 2009, 5, 502−507. (d) Sche, P. P.; McKenzie, K. M.; White, J. D.; Austin, D. J. Chem. Biol. 1999, 6, 707− 716. (e) Jin, Y.; Yu, J.; Yu, Y. G. Chem. Biol. 2002, 9, 157−162. (f) Shim, J. S.; Lee, J.; Park, H.-J.; Park, S.-J.; Kwon, H. J. Chem. Biol. 2004, 11, 1455−1463. (g) Kuroiwa, Y.; Takakusagi, Y.; Kusayanagi, T.; Kuramochi, K.; Imai, T.; Hirayama, T.; Ito, I.; Yoshida, M.; Sakaguchi, K.; Sugawara, F. PLoS One 2013, 8, e63073. (5) (a) Li, S.; Roberts, R. W. Chem. Biol. 2003, 10, 233−239. (b) Horisawa, K.; Tateyama, S.; Ishizaka, M.; Matsumura, N.; Takashima, H.; Miyamoto-Sato, E.; Doi, N.; Yanagawa, H. Nucleic Acids Res. 2004, 32, e169. (c) Cotton, S. W.; Zou, J.; Valencia, C. A.; Liu, R. Nat. Protoc. 2011, 6, 1163−1182. (d) Schlippe, Y. V.; Hartman, M. C.; Josephson, K.; Szostak, J. W. J. Am. Chem. Soc. 2012, 134, 10469−10477. (e) Hipolito, C. J.; Suga, H. Curr. Opin. Chem. Biol. 2012, 16, 196−203. (f) Ueno, S.; Yoshida, S.; Mondal, A.; Nishina, K.; Koyama, M.; Sakata, I.; Miura, K.; Hayashi, Y.; Nemoto, N.; Nishigaki, K.; Sakai, T. Proc. Natl. Acad. Sci. U.S.A. 2012, 109, 11121−11126. (g) McPherson, M.; Yang, Y.; Hammond, P. W.; Kreider, B. L. Chem. Biol. 2002, 9, 691−698. (h) Shiheido, H.; Terada, F.; Tabata, N.; Hayakawa, I.; Matsumura, N.; Takashima, H.; Ogawa, Y.; Du, W.; Yamada, T.; Shoji, M.; Sugai, T.; Doi, N.; Iijima, S.; Hattori, Y.; Yanagawa, H. PLoS One 2012, 7, e38878. (6) (a) Mattheakis, L. C.; Bhatt, R. R.; Dower, W. J. Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 9022−9026. (b) Lamla, T.; Erdmann, V. A. J. Mol. Biol. 2003, 329, 381−388. (c) Zahnd, C.; Amstutz, P.; Plückthun, A. Nat. Methods 2007, 4, 269−279. (d) He, M.; Taussig, M. J. Nat. Methods 2007, 4, 281−288. (e) Wada, A.; Sawata, S. Y.; Ito, Y. Biotechnol. Bioeng. 2008, 101, 1102−1107. (f) Mie, M.; Shimizu, S.; Takahashi, F.; Kobatake, E. Biochem. Biophys. Res. Commun. 2008, 373, 48−52. (g) Douthwaite, J. A. Methods Mol. Biol. 2012, 805, 45−57. (h) Ohashi, H.; Kanamori, T.; Osada, E.; Akbar, B. K.; Ueda, T. Methods Mol. Biol. 2012, 805, 251−259. (i) Wada, A. Front. Immunol. 2013, 4, 224. (7) Shimizu, Y.; Inoue, A.; Tomari, Y.; Suzuki, T.; Yokogawa, T.; Nishikawa, K.; Ueda, T. Nat. Biotechnol. 2001, 19, 751−755. (8) (a) Bayley, H. In Photogenerated Reagents in Biochemistry and Molecular Biology; Work, T., Burdon, R., Eds. Amsterdam: Elsevier, 1983; Vol. 12, p 187. (b) Brunner, J. Annu. Rev. Biochem. 1993, 62, 483−514. (c) Kanoh, N.; Honda, K.; Simizu, S.; Muroi, M.; Osada, H. Angew. Chem., Int. Ed. 2005, 44, 3559−3562.

Furthermore, for generalizing this technique, ribosome display selection of highly diluted target protein was carried out by using photo-cross-linked CsA beads (SI). By repeating the selection twice, CypA as a target protein of CsA could be preferentially identified from an initial mRNA pool that contained equimolar amounts of mRNAs encoding FKBP12, SOD1 (G86R), BID, KRas (G12V), DHFR, and p38α, and an extremely small amount of mRNA encoding CypA, which corresponded to one ten-thousandth (1/10,000th) of the total mRNA (Figure S3 and Table S4 in SI). On the basis of these results, it is clear that repeated ribosome display selection with photo-cross-linked smallmolecule beads and electrophoretic analysis of RT-PCR products could identify a target protein even if the molar quantity was one ten-thousandth of that of the other proteins in a cell-free synthesized FHP pool. Therefore, after further extension of the FHP pool, the ribosome display and photocross-linking techniques presented here could be powerful tools for the identification of unknown target proteins of various small molecules, which will contribute to a better understanding of their actions and mechanisms.

■

ASSOCIATED CONTENT

S Supporting Information *

Experimental procedures and supplementary data as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected] *E-mail: [email protected] Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This research was supported in part by the “Precursory Research for Embryonic Science and Technology (PRESTO)” from the Japan Science and Technology Agency (JST) (to A.W.) and the “Grant-in-Aid for Challenging Exploratory Research” from the Japan Society for the Promotion of Science (JSPS) (to A.W.). We thank Dr. Yasumitsu Kondoh and Ms. Kaori Honda for the synthesis of photo-cross-linked smallmolecule beads and also thank Ms. Shinobu Yamaguchi for technical assistance.

■

REFERENCES

(1) (a) Huang, J.; Zhu, H.; Haggarty, S. J.; Spring, D. R.; Hwang, H.; Jin, F.; Snyder, M.; Schreiber, S. L. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 16594−16599. (b) Bradner, J. E.; McPherson, O. M.; Koehler, A. N. Nat. Protoc. 2006, 1, 2344−2352. (c) Palchaudhuri, R.; Hergenrother, P. J. ACS Chem. Biol. 2011, 6, 21−33. (d) Kato, N.; Takahashi, S.; Nogawa, T.; Saito, T.; Osada, H. Curr. Opin. Chem. Biol. 2012, 16, 101−108. (2) (a) Stockwell, B. R. Nature 2004, 432, 846−854. (b) Bradner, J. E.; McPherson, O. M.; Mazitschek, R.; Barnes-Seeman, D.; Shen, J. P.; Dhaliwal, J.; Stevenson, K. E.; Duffner, J. L.; Park, S. B.; Neuberg, D. S.; Nghiem, P.; Schreiber, S. L.; Koehler, A. N. Chem. Biol. 2006, 13, 493−504. (c) Miyazaki, I.; Simizu, S.; Okumura, H.; Takagi, S.; Osada, H. Nat. Chem. Biol. 2010, 6, 667−673. (d) Vrijens, K.; Lin, W.; Cui, J.; Farmer, D.; Low, J.; Pronier, E.; Zeng, F. Y.; Shelat, A. A.; Guy, K.; Taylor, M. R.; Chen, T.; Roussel, M. F. PLoS One 2013, 8, e59045. Herrero Martín, D.; Boro, A.; Schäfer, B. W. PLoS One 2013, 8, e55072. (e) Zimmermann, T. J.; Bürger, M.; Tashiro, E.; Kondoh, Y.; 6773

dx.doi.org/10.1021/ac4030208 | Anal. Chem. 2014, 86, 6768−6773