Demonstration of protein-fragment complementation assay using purified firefly luciferase fragments
© Ohmuro-Matsuyama et al.; licensee BioMed Central Ltd. 2013
Received: 3 October 2012
Accepted: 15 March 2013
Published: 28 March 2013
Human interactome is predicted to contain 150,000 to 300,000 protein-protein interactions, (PPIs). Protein-fragment complementation assay (PCA) is one of the most widely used methods to detect PPI, as well as Förster resonance energy transfer (FRET). To date, successful applications of firefly luciferase (Fluc)-based PCA have been reported in vivo, in cultured cells and in cell-free lysate, owing to its high sensitivity, high signal-to-background (S/B) ratio, and reversible response. Here we show the assay also works with purified proteins with unexpectedly rapid kinetics.
Split Fluc fragments both fused with a rapamycin-dependently interacting protein pair were made and expressed in E. coli system, and purified to homogeneity. When the proteins were used for PCA to detect rapamycin-dependent PPI, they enabled a rapid detection (~1 s) of PPI with high S/B ratio. When Fn7-8 domains (7 nm in length) that was shown to abrogate GFP mutant-based FRET was inserted between split Fluc and FKBP12 as a rigid linker, it still showed some response, suggesting less limitation in interacting partner’s size. Finally, the stability of the probe was investigated. Preincubation of the probes at 37 degreeC up to 1 h showed marked decrease of the luminescent signal to 1.5%, showing the limited stability of this system.
Fluc PCA using purified components will enable a rapid and handy detection of PPIs with high S/B ratio, avoiding the effects of concomitant components. Although the system might not be suitable for large-scale screening due to its limited stability, it can detect an interaction over larger distance than by FRET. This would be the first demonstration of Fluc PCA in vitro, which has a distinct advantage over other PPI assays. Our system enables detection of direct PPIs without risk of perturbation by PPI mediators in the complex cellular milieu.
KeywordsProtein-protein interaction Firefly luciferase Bioluminescence Protein fragment complementation assay Thermostability In vitro diagnostics
While PPI is conveniently assayed by PCA in vivo and in cultured cells, it is often desirable to be performed also in vitro, to know whether the interaction is direct or not, since many PPI mediators, inhibitors and enhancers may exist in the complex cellular milieu. To date, Porter et al. successfully utilized a cell-free transcription/translation system for performing firefly luciferase (Fluc)-based PCA in vitro. Their probes can be prepared in shorter time period than the probes expressed in E. coli, yeast or mammalian cells. However, the assay needs expensive cell-free lysate, and whose components always have a risk to affect PPI. Here we report a PCA using purified interacting proteins fused with split Fluc proteins, to detect PPI in a defined solution. We also examined the stability of the probes, and the detectable distance between the interacting partners. To the best of our knowledge, this will be the first investigation of pure in vitro PPI, based on Fluc PCA.
For Fluc PCA, several successful split sites are reported to date. To investigate on this issue, another pair of N-terminal domain (1–398) together with the same C-terminal domain (394–547) was employed according to a previous attempt that gave good response in vivo (Additional file 1: Figure S1) . We could express and purify the new N-domain similarly (Additional file 1: Figure S1A), and obtained clear rapamycin-dependent signals at several concentrations of FRB/N(1–398) – FKBP/C pair (Additional file 1: Figure S1, B-E). However, compared with the FRB/N(1–437) – FKBP/C(394–547) pair, the signal intensity was lower, and the resultant S/B ratio was rather unstable, albeit not low (Additional file 1: Figure S1F). This was probably because of the instability of the N-domain (1–398), which is reflected to its degradation products observed in Additional file 1: Figure S1A.
Due to the nature of split enzyme, we reasoned that the instability of the purified probes could be a possible demerit of Fluc-based PCA. To examine the stability of the probes, the mixtures of 100 nM FKBP/C and 100 nM FRB/N with or without equimolar rapamycin were incubated at 37°C for 0–60 min, and applied to the luminescence assay (Additional file 2: Figure S2). As a result, the incubation resulted in a time-dependent decrease in luminescent intensity of the pair added with rapamycin (Additional file 2: Figure S2A). As short as 15 min incubation resulted in the activity loss to less than one-third of that before incubation. The longer incubation for 60 min resulted in virtually no signal difference between the pairs with and without rapamycin (Additional file 2: Figure S2, B and C). These results clearly indicate that the Fluc PCA probes have a limited stability in this purified milieu. However, the results might also reflect their limited stability in vivo or in lysate, which is probably sequestered by the newly synthesized probes.
On the other hand, the PCA pair of FKBP/Fn7-8/C and FRB/N at 750 nM displayed a moderate but significant S/B ratio, clearly showing its superiority over FRET, although the observed rapamycin dependency was not so clear when reduced amounts (100 nM) of the probes were used (Figure 6C). Nevertheless, the Fluc PCA system was shown to work for detecting PPI over longer distance, which was not detectable by FRET.
Previously, Fluc-based PCA has been utilized as a sensitive tool to investigate PPI in vivo and in cultured cells. However, considering possible disturbance by the co-existing components acting as PPI mediators, enhancers and inhibitors, PPI assay in vitro will also give us invaluable information. Furthermore, the in vitro assay has the possibility to study PPIs between pathogen-derived cytotoxic proteins and host proteins, which are normally impossible to perform in vivo. In this study, we showed that Fluc PCA also works with purified elements. The system displayed high S/B ratio and unexpectedly rapid response (~1 s). In PCA, the background signal due to nonspecific activation of the reporter can be a problem. As in Figure 2C, luminescent intensities of non-interacting pairs were slightly higher than those of single probe alone, probably reflecting the nonspecific reconstitution of split Fluc. However, when the specific PPI was induced by rapamycin, it resulted in markedly higher signal, giving high S/B ratio. Therefore, in practice the nonspecific signals hardly interfere with the PPI detection (Figure 2).
The only but not easily avoidable drawback of this system is the instability of probe as shown in Additional file 2: Figure S2. Apparently, the system is not very suitable for large-scale screening that needs longer sample preparation time. However, proper cooling of the reagents and samples during the preparation period will surely reduce such inactivation. Therefore, we think that the system has a potential for low-cost and high-throughput drug screenings. On the other hand, a distinct merit of this system is robustness to the interaction detection over longer distance. In fact, insertion of either long helical linker or a rigid large protein domain did not deteriorate PPI detection, while FRET signal was barely detectable when Fn7-8 was inserted (Figure 6B). These results show a lesser limitation in distance between interacting partners in PCA than in FRET. Therefore, this system will be available for the interactions between large proteins that cannot be assayed by FRET. While the reason for observed nonspecific PPI signal remains unclear, it may be due to the interaction between the inserted sequence and the other probe. In addition, we could observe the higher S/B ratio with the higher concentration of a probe set (Figure 6C). Use of higher probe concentration is desirable to assay larger interacting partners. The nearly straight shape of the kinetic curve observed might reflect the large dimension of Fn7-8, which would lower the reconstitution efficiency and subsequent luminescent reaction rate.
Collectively, considering the demerits and the merits, Fluc PCA using purified elements enables us a rapid and convenient detection of direct PPI with high confidence.
The feasibility of Fluc PCA in vitro using purified elements was demonstrated for the first time. The assay was found to have high S/B ratio, and allow the detection of larger size of interacting partner than detectable by FRET. In spite of relatively weak thermostability, the assay will be applicable to handy detection of various PPIs in vitro including in vitro diagnostics and drug screening.
ATP and D-luciferin (LH2) were from Sigma, St. Louis, MO. MOPS (3-(N-morpholino)propanesulfonic acid) was from Dojindo, Kumamoto, Japan. Rapamycin was from Wako Pure Chemical Industries, Osaka, Japan or LKT Laboratories, St. Paul, MN. Synthetic genes for E. coli codon-optimized human FKBP12, appended with Nco I / Sfi I and NotI sites at the 5’ and 3′ ends, respectively, were from Mr Gene GmbH, Regensburg, Germany. The plasmid pFH154 encoding human fibronectin cDNA was from Health Science Research Resources Bank (HSRRB), Osaka, Japan. Other reagents in the highest grade available were from Wako Pure Chemical Industries unless otherwise indicated.
Construction of FKBP/FRB fused Fluc fragments
The DNA fragment encoding Fluc was obtained by PCR using pGEX-Ppy vector  as a template, and primers LucNotG4SB (5′- gg cgc gcc GCG GCC GCC GGT GGT GGT GGT AGC ATG GAA GAC GCC AAA AAC ATA AAG-3′) encoding a G4S linker and NotI site (underlined) and LucXhoF (5′- g gcg cgc CTC GAG CTT TCC GCC CTT CTT GGC CT- 3′) containing XhoI site (underlined). Similarly, N- and C-terminal domain genes were amplified with primers LucNotG4SB and Luc437XhoF (5′-g gcg cgc CTC GAG GCG GTC AAC TAT GAA GAA GTG- 3′), and Luc394NotG4SB (5′- gg cgc gcc GCG GCC GCC GGT GGT GGT GGT AGC GGA CCT ATG ATT ATG TCC GG-3′) and LucXhoF, respectively. The amplified fragments were cloned into pET32b (Merck, Darmstadt, Germany) between the NotI and XhoI sites, to give pET32/Fluc, pET32/FlucN and pET32/FlucC, respectively. The synthetic genes encoding FKBP12 and FRB cDNAs were digested with NcoI and NotI, and the digested fragments inserted in pET32/FlucN and pET32/FlucC digested with the same enzymes each other, to give pET32/FKBP/FlucN, pET32/FKBP/FlucC, pET32/FRB/FlucN and pET32/FRB/FlucC.
Construction of FRB fused FlucN(1–398) fragments
The DNA fragment encoding Fluc N-terminal domain (1–398) was obtained by PCR using pET32/FRB/FlucN as a template, and primers LucNotG4SB and Luc398XhoF (5′- t gtt tac ata CTC GAG cat aat cat agg tcc tct tac- 3′) containing XhoI site (underlined). The amplified fragments were cloned into pET32/FRB/FlucN between the NotI and XhoI sites, to give pET32/FRB/FlucN (1–398).
Construction of p53/Mdm2 fused Fluc fragments
The DNA fragment encoding transactivation domain of p53 (residue 15–29) and two restriction sites (underlined) was obtained by thermal cycling using following oligonucleotides: p53NcoBack (5′-gg aat tCC ATG GCT AGT CAG GAA ACA TTT TCA GAC CTA TGG AAA C-3′) and p53NotFor (5′- g gga ttc tGC GGC CGC GTT TTC AGG AAG TAG TTT CCA TAG GTC TG-3′). Mdm2 gene (residue 17–125) was obtained by PCR with human Mdm2 gene as a template and mdm2NocBack (5′-gg aat tCC ATG GCT TCG GAA CAA GAG ACC C-3′) and mdm2NotFor (5′-g gaa ttc tGC GGC CGC CTG CTG ATT GAC TAC TAC C-3′) as primers. The amplified fragments were digested with NcoI and NotI, and inserted to pET32/FKBP/FlucC and pET32/FRB/FlucN digested with the same enzymes, to give pET32/p53/FlucN, pET32/p53/FlucC, pET32/Mdm2/FlucN and pET32/Mdm2/FlucC.
Insertion of 4 × DDAKK between FKBP12 and FlucC
The two oligonucleotides DDAKK4_NotBack2 (5′ –g gaa ttc GCG GCC GCA GAT GAT GCT AAA AAA GAT GCT AAA AAA GAT GAT GCC AAG AAG GAC GAC GC- 3′) containing NotI site (underlined), and DDAKK4_EagFor2 (5′ –g gaa ttC GGC CGA TTT TTT AGC ATC ATC TTT TTT CGC GTC GTC CTT CTT GGC- 3′) containing EagI site (underlined) were annealed and extended using a thermal cycler. The fragment was digested with EagI, and inserted into pET32/FKBP/FlucC digested with NotI, to give pET32/FKBP/4 × DDAKK/FlucC.
Insertion of 7×DDAKK between FKBP12 and FlucC
pET32/FKBP/4 × DDAKK/FlucC was amplified with the primers, DDAKK_VectorNotB (5′-GA CGA CGC CAA AAA AGA TGA TGC CAA GAA GG-3′), and DDAKK_VectorNotF (5′-CT TTT TTA GCA TCA TCT GCG G-3′). The 4 × DDAKK fragment inserted in pET32/FKBP/4 × DDAKK/FlucC was amplified with the primers DDAKK_LinkerB (5′ –GA TGA TGC TAA AAA AGA TG- 3′) and DDAKK_LinkerF (5′-TT TTT TGG CGT CGT CTT TTT TCG CGT CGT C- 3′). The two amplified fragments were connected using In-Fusion HD cloning kit (Takara-Bio, Shiga, Japan), to obtain pET32/FKBP/8 × DDAKK/FlucC, which resulted in unexpected acquisition of pET32/FKBP/7 × DDAKK/FlucC.
Construction of FKBP/FRB fused with fluorescent proteins
The cDNAs for Ypet on pYpet-His (kindly provided by Dr. PS Daugherty)  and Cerulean  made from pEBFP-N1 plasmid (Clontech, Takara-Bio) were amplified using specific primers with 5′-terminal NotI and 3′-terminal XhoI sites, digested with NotI and XhoI, and inserted into pET32/FKBP/FlucN and pET32/FRB/FlucN digested with the same enzymes, to give pET32/FKBP/Cerulean and pET32/FRB/Ypet.
Insertion of Fn7-8 as a rigid linker
The sequence for Fn7-8 was amplified from human fibronectin cDNA with the primers Fn7EagBack (5′-g gaa ttC GGC CGC ACC ATT GTC TCC ACC AAC AAA C- 3′) and Fn8EagFor (5′-g gaa ttC GGC CGA TGT TTT CTG TCT TCC TCT AAG- 3′) each containing an EagI site. The amplified fragment was digested with EagI, and was inserted into pET32/FKBP/FlucC and pET32/FKBP/Cerulean digested with same enzyme, respectively.
Expression and purification of probe proteins
All the fusion proteins were expressed in E. coli BL21 (pLysS, DE3) (Novagen) as a thioredoxin and hexahistidine-tagged protein. To purify the expressed fusion proteins, Talon metal affinity resin (Clontech) was used according to the manufacturer’s instruction. Concentration of the purified protein was determined by CBB-stained SDS-PAGE co-loaded with various concentrations of BSA as a concentration standard. The protein added with final 15 % glycerol was stored at −80°C before use.
Detection of PCA
The purified probe proteins with or without rapamycin were suspended in 100 mM MOPS, 10 mM MgSO4, pH 7.3. The mixture (50 μl each) was dispensed to a well of 96-well half well white plate (Corning-Costar, NY, USA). The light intensity was measured immediately after injection of 50 μl 2 × substrate solution (40 mM ATP and 150 μM LH2 in 100 mM MOPS, 10 mM MgSO4, pH 7.3) with a periodical integration for 0.1 s using a luminometer Phelios AB-2350 (ATTO, Tokyo, Japan).
Fluorescence spectra were measured by F-2500 fluorescence spectrophotometer (Hitachi High-Technologies, Tokyo, Japan). Samples were diluted in 250 μl PBS, pH 7.4. The mixture of FKBP/cerulean or FKBP/Fn/cerulean and FRB/Ypet (40 nM each) was excited at 433 nm, and the fluorescent spectra at 455–600 nm were recorded in the presence and absence of 40 nM rapamycin.
FK506 binding protein
7–8 domains of fibronectin type III
fluorescence resonance energy transfer
green fluorescent protein
protein-fragment complementation assay
We thank Keiichi Ayabe and Aoi Kimura for preparing Fluc-N and Fluc-C genes, and Yukiko Gotoh for human mdm2 cDNA. YOM was supported by SENTAN, Japan Science and Technology Agency, Japan.
- Venkatesan K, Rual JF, Vazquez A, Stelzl U, Lemmens I, Hirozane-Kishikawa T, Hao T, Zenkner M, Xin X, Goh KI: An empirical framework for binary interactome mapping. Nat Methods. 2009, 6 (1): 83-90. 10.1038/nmeth.1280.View ArticleGoogle Scholar
- Zhang QC, Petrey D, Deng L, Qiang L, Shi Y, Thu CA, Bisikirska B, Lefebvre C, Accili D, Hunter T: Structure-based prediction of protein-protein interactions on a genome-wide scale. Nature. 2012, 490 (7421): 556-560. 10.1038/nature11503.View ArticleGoogle Scholar
- Shekhawat SS, Ghosh I: Split-protein systems: beyond binary protein-protein interactions. Curr Opin Chem Biol. 2011, 15 (6): 789-797. 10.1016/j.cbpa.2011.10.014.View ArticleGoogle Scholar
- Hu CD, Kerppola TK: Simultaneous visualization of multiple protein interactions in living cells using multicolor fluorescence complementation analysis. Nat Biotechnol. 2003, 21 (5): 539-545. 10.1038/nbt816.View ArticleGoogle Scholar
- Magliery TJ, Wilson CG, Pan W, Mishler D, Ghosh I, Hamilton AD, Regan L: Detecting protein-protein interactions with a green fluorescent protein fragment reassembly trap: scope and mechanism. J Am Chem Soc. 2005, 127 (1): 146-157. 10.1021/ja046699g.View ArticleGoogle Scholar
- Luker KE, Smith MC, Luker GD, Gammon ST, Piwnica-Worms H, Piwnica-Worms D: Kinetics of regulated protein-protein interactions revealed with firefly luciferase complementation imaging in cells and living animals. Proc Natl Acad Sci U S A. 2004, 101 (33): 12288-12293. 10.1073/pnas.0404041101.View ArticleGoogle Scholar
- Paulmurugan R, Massoud TF, Huang J, Gambhir SS: Molecular imaging of drug-modulated protein-protein interactions in living subjects. Cancer Res. 2004, 64 (6): 2113-2119. 10.1158/0008-5472.CAN-03-2972.View ArticleGoogle Scholar
- Paulmurugan R, Umezawa Y, Gambhir SS: Noninvasive imaging of protein-protein interactions in living subjects by using reporter protein complementation and reconstitution strategies. Proc Natl Acad Sci U S A. 2002, 99 (24): 15608-15613. 10.1073/pnas.242594299.View ArticleGoogle Scholar
- Remy I, Michnick SW: A highly sensitive protein-protein interaction assay based on Gaussia luciferase. Nat Methods. 2006, 3 (12): 977-979. 10.1038/nmeth979.View ArticleGoogle Scholar
- Kim SB, Ozawa T, Watanabe S, Umezawa Y: High-throughput sensing and noninvasive imaging of protein nuclear transport by using reconstitution of split Renilla luciferase. Proc Natl Acad Sci U S A. 2004, 101 (32): 11542-11547. 10.1073/pnas.0401722101.View ArticleGoogle Scholar
- Porter JR, Stains CI, Jester BW, Ghosh I: A general and rapid cell-free approach for the interrogation of protein-protein, protein-DNA, and protein-RNA interactions and their antagonists utilizing split-protein reporters. J Am Chem Soc. 2008, 130 (20): 6488-6497. 10.1021/ja7114579.View ArticleGoogle Scholar
- Chen J, Zheng XF, Brown EJ, Schreiber SL: Identification of an 11-kDa FKBP12-rapamycin-binding domain within the 289-kDa FKBP12-rapamycin-associated protein and characterization of a critical serine residue. Proc Natl Acad Sci U S A. 1995, 92 (11): 4947-4951. 10.1073/pnas.92.11.4947.View ArticleGoogle Scholar
- Chiu MI, Katz H, Berlin V: RAPT1, a mammalian homolog of yeast Tor, interacts with the FKBP12/rapamycin complex. Proc Natl Acad Sci U S A. 1994, 91 (26): 12574-12578. 10.1073/pnas.91.26.12574.View ArticleGoogle Scholar
- Paulmurugan R, Gambhir SS: Combinatorial library screening for developing an improved split-firefly luciferase fragment-assisted complementation system for studying protein-protein interactions. Anal Chem. 2007, 79 (6): 2346-2353. 10.1021/ac062053q.View ArticleGoogle Scholar
- Banaszynski LA, Liu CW, Wandless TJ: Characterization of the FKBP.rapamycin.FRB ternary complex. J Am Chem Soc. 2005, 127 (13): 4715-4721. 10.1021/ja043277y.View ArticleGoogle Scholar
- Kussie PH, Gorina S, Marechal V, Elenbaas B, Moreau J, Levine AJ, Pavletich NP: Structure of the MDM2 oncoprotein bound to the p53 tumor suppressor transactivation domain. Science. 1996, 274 (5289): 948-953. 10.1126/science.274.5289.948.View ArticleGoogle Scholar
- Vassilev LT, Vu BT, Graves B, Carvajal D, Podlaski F, Filipovic Z, Kong N, Kammlott U, Lukacs C, Klein C: In vivo activation of the p53 pathway by small-molecule antagonists of MDM2. Science. 2004, 303: 844-848. 10.1126/science.1092472.View ArticleGoogle Scholar
- Zhang J, Campbell RE, Ting AY, Tsien RY: Creating new fluorescent probes for cell biology. Nat Rev Mol Cell Biol. 2002, 3 (12): 906-918. 10.1038/nrm976.View ArticleGoogle Scholar
- Tsien RY: The green fluorescent protein. Annu Rev Biochem. 1998, 67: 509-544. 10.1146/annurev.biochem.67.1.509.View ArticleGoogle Scholar
- Wu P, Brand L: Resonance energy transfer: methods and applications. Anal Biochem. 1994, 218 (1): 1-13. 10.1006/abio.1994.1134.View ArticleGoogle Scholar
- Arai R, Ueda H, Kitayama A, Kamiya N, Nagamune T: Design of the linkers which effectively separate domains of a bifunctional fusion protein. Protein Eng. 2001, 14 (8): 529-532. 10.1093/protein/14.8.529.View ArticleGoogle Scholar
- Ohashi T, Galiacy SD, Briscoe G, Erickson HP: An experimental study of GFP-based FRET, with application to intrinsically unstructured proteins. Protein Sci. 2007, 16 (7): 1429-1438. 10.1110/ps.072845607.View ArticleGoogle Scholar
- Kitayama A, Yoshizaki H, Ohmiya Y, Ueda H, Nagamune T: Creation of a thermostable firefly luciferase with pH-insensitive luminescent color. Photochem Photobiol. 2003, 77 (3): 333-338. 10.1562/0031-8655(2003)077<0333:COATFL>2.0.CO;2.View ArticleGoogle Scholar
- Nguyen AW, Daugherty PS: Evolutionary optimization of fluorescent proteins for intracellular FRET. Nat Biotechnol. 2005, 23 (3): 355-360. 10.1038/nbt1066.View ArticleGoogle Scholar
- Rizzo MA, Springer GH, Granada B, Piston DW: An improved cyan fluorescent protein variant useful for FRET. Nat Biotechnol. 2004, 22 (4): 445-449. 10.1038/nbt945.View ArticleGoogle Scholar
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.