Reassembly of a Bioluminescent Protein Renilla Luciferase Directed

Bioconjugate Chem. 2009, 20, 15–19

15

COMMUNICATIONS Reassembly of a Bioluminescent Protein Renilla Luciferase Directed through DNA Hybridization Kyle A. Cissell, Yasmeen Rahimi, Suresh Shrestha, and Sapna K. Deo* Department of Chemistry and Chemical Biology, Indiana University-Purdue University Indianapolis, Indiana 46202. Received July 18, 2008; Revised Manuscript Received November 10, 2008

Reassembly of split reporter proteins, also referred to as protein complementation, is utilized in the detection of protein-protein or protein-nucleic acid interactions. In this strategy, a reporter protein is fragmented into two inactive polypeptides to which interacting/binding partners are fused. The interaction between fused partners leads to the formation of a reassembled, active reporter. In this Communication, we have presented a proof-ofconcept for the detection of a target nucleic acid sequence based on the reassembly of the bioluminescent reporter Renilla luciferase (Rluc), which is driven by DNA hybridization. Although, reassembly of Rluc though protein interactions has been demonstrated by others, the Rluc reassembly through DNA hybridization has not been shown yet, which is the novelty of this work. It is well established that bioluminescence detection offers significant advantages due to the absence of any background signal. In our study, two rationally designed fragments of Rluc were conjugated to complementary oligonucleotide probes. Hybridization of the two probes with fused Rluc fragments resulted in the reassembly of the fragments, generating active Rluc, measurable by the intensity of light given off upon addition of coelenterazine. Our study also shows that the reassembly of Rluc can be inhibited by an oligonucleotide probe that competes to bind to the hybridized probe-Rluc fragment complex, indicating a potential strategy for the quantitative detection of target nucleic acid. We were able to achieve the reassembly of Rluc fused to oligonucleotide probes using femtomole amounts of the probe-fragment protein conjugate. This concentration is approximately 4 orders of magnitude less than that reported using green fluorescent protein (GFP) as the reporter. A DNA-driven Rluc reassembly study performed in a cellular matrix did not show any interference from the matrix.

The split protein reassembly method has found applications as a biochemical tool to detect specific protein-protein and protein-nucleic acid interactions (1-8). In this method, two rationally designed fragments of a reporter protein are fused to proteins that are capable of interacting when close to one another. The fragments in isolation are inactive; however, biomolecular interactions between fused proteins leads to the formation of an active reporter, allowing for the detection of protein-protein interactions. This strategy has been successfully used in the reassembly of reporter proteins, specifically, dihydrofolate reductase, β-lactamase, ubiquitin, aminoglycoside kinase, Renilla luciferase, firefly luciferase, red fluorescent protein, and green fluorescent protein (GFP) driven by interacting proteins (4, 5, 9-11). There is only one report on reporter protein reassembly achieved through nucleic acid hybridization using split GFP as the reporter (12). In this study, the reassembly of split GFP upon formation of a hybrid between two complementary oligonucleotides at a 200 nM concentration was accomplished. Design and development of fluorescent proteins such as GFP have revolutionized the biochemical field. However, these proteins have some limitations because of the requirement of high energy excitation and high background signal, reducing the sensitivity of the signal measurement. In that respect, bioluminescent proteins that emit light by a * To whom correspondence should be addressed. Phone: 317-2787532. E-mail: [email protected].

biochemical reaction do not require a high energy excitation source and do not suffer from background noise in signal measurement (13-17). In a recent paper by Porter et al., these advantages of the bioluminescent enzyme firefly luciferase (Fluc) were utilized to detect proteins, small molecules, RNA, and double stranded DNA, both methylated and nonmethylated (18). This was accomplished through the expression of the Fluc fragments flanked by expressed protein domains, which bound to the specific target, bringing split Fluc fragments within close proximity. This allows for protein reassembly to occur, resulting in a luminescent signal upon the addition of luciferin, the substrate for Fluc. However, this method is not a hybridizationbased assay that works for any target nucleic acid. This method will work only for targets for which peptide domains that bind to the target can be engineered. Taking this into consideration, here we evaluated in a proof-of-concept study of whether a bioluminescent protein, specifically Renilla luciferase (Rluc), is amenable to protein reassembly driven by DNA hybridization (Figure 1) similar to reassembly that can be achieved through protein complementation. We have also demonstrated the usefulness of this method in the detection of target nucleic acid. The reporter, Rluc, is a monomeric ∼38 kDa protein that catalyzes the oxidative decarboxylation of coelenterazine, the substrate that binds to Rluc, to coelenteramide, resulting in the production of light (19-21). One of the rationales for using Rluc in this study is that it is a small monomeric bioluminescent enzyme that is amenable to cellular imaging (16, 22-26). It

10.1021/bc8003099 CCC: $40.75  2009 American Chemical Society Published on Web 12/09/2008

16 Bioconjugate Chem., Vol. 20, No. 1, 2009

Figure 1. Schematic representation of the Rluc fragment reassembly driven through oligonucleotide probe hybridization.

Figure 2. X-ray crystal structure of Rluc8, a mutant of Rluc showing the catalytic triad. The N-terminal Rluc fragment region is shown in gray, and the C-terminal Rluc fragment region is shown in black.

can be detected at very low levels over a broad dynamic range using inexpensive luminometers. Rluc also does not require posttranslational modification for its activity. Several analogues of its substrate coelenterazine have been synthesized with different properties in terms of emission wavelength, quantum efficiency, and cell permeability, thus enhancing the applications of Rluc. Unlike fluorescence, there is no need for external excitation, and thus, problems with light scattering and source instability are avoided. The most critical issue in the success of protein reassembly from its split fragments is the selection of a cleavage site in the protein. Availability of the three-dimensional structure can aid in the design of the protein fragments. Other considerations in the design of successful protein reassembly include small protein size, monomeric structure, and simplicity of the assay system. Reassembly of split Renilla luciferase from its fragments, driven by the interactions between the proteins myoD and Id in vivo, has been demonstrated (27), although the fragments were not isolated or purified. On the basis of this work a cleavage site of Gly229-Lys230 (Figure 2) was chosen to dissect Rluc in our study. The X-ray crystal structure of a mutant of Rluc (RL8) was solved recently (Figure 2) (28); however, we decided to select the cleavage site on the basis of prior complementation work performed in detecting protein-protein interactions using the native Rluc. The X-ray crystal structure of RL8 mutant shows that it has a classic R/β-hydrolase fold at its core and contains a catalytic triad made from Asp120-Glu144-His285. The cleavage site selected in our study, Gly229-Lys230, would be such that the residue Asp120 and Glu144 of the catalytic triad will be in the N-terminal fragment, whereas His285 will be in the C-terminal fragment. This suggests that although the substrate binding may occur in the N-terminal fragment, the catalytic triad is not completed until the two fragments come close for the generation of the bioluminescence signal. In the future, other cleavage sites will be evaluated on the basis of the X-ray crystal structure of the Rluc mutant. It would be

Communications

Figure 3. Picture of 12.5% SDS-PAGE gel of purified Rluc fragments stained with Coomassie Blue. Lane 1: molecular weight protein marker. Lane 2: purified N-terminal Rluc fragment. Lane 3: purified C-terminal Rluc fragment. Lane 4: molecular weight protein marker.

interesting to cleave Rluc such that the triad is kept intact in one of the fragments while the substrate binding is disturbed in the absence of assembly of the two fragments. This type of fragmentation where the active site is maintained may reduce the overall time needed for the reassembly of protein as observed in the case of split GFP work (12). Using the gene for native Rluc from the plasmid phRL-CMV as the template, primers were designed to introduce a noncleavable 6-histidine tag to allow purification of the Rluc fragments as well as codons for a flexible amino acid linker and a cysteine residue for oligonucleotide probe conjugation (Supporting Information). The Rluc N-terminal fragment was constructed by introducing two stop codons after the amino acid Gly229 using site-directed mutagenesis (Supporting Information). Similarly, primers were designed to obtain the gene sequence of the C-terminal fragment of Rluc (amino acid 230-311) with a unique cysteine near its N-terminus (Supporting Information). Rluc fragments were expressed in E. coli cells and purified using a copper-immobilized Hi-Trap column (Supporting Information). The purity of the two fragments was verified using SDS-PAGE and Coomassie staining (Figure 3). The SDS-PAGE shows that the two fragments, the N-terminal fragment (∼30 kDa) and the C-terminal fragment (∼16 kDa), were purified efficiently. The concentrations of purified Rluc fragments were determined using a Bradford assay. The luminescence activity of the individual Rluc fragments (1 nmol) was monitored by adding the substrate coelenterazine (0.5 µL, 1 mg/mL). As expected, no luminescence signal was observed in the presence of substrate, indicating the inability of fragments to process the substrate (Figure 4). Next, the two fragments (1 nmol each) without the probe attached were mixed together followed by the addition of coelenterazine to check for any nonspecific binding between the fragments in the absence of oligonucleotide probes. The mixing of the two fragments did not result in the generation of any bioluminescence signal indicating that spontaneous reassembly between Rluc fragments does not occur (Figure 4). Thiol-modified oligonucleotide probes were conjugated to the split Rluc fragments using maleimide-sulfhydryl chemistry through the homobifunctional cross-linker BM(PEO)2 (1,8-bismaleimidotriethylene glycol). The Rluc N-terminal fragment contains three cysteine residues buried within the protein structure. Therefore, we hypothesize that a cysteine introduced at the C-terminus during the construction of the fragment will be the major target site for oligonucleotide conjuation through the thiol moiety. The Rluc C-terminal fragment does not contain a cysteine residue; therefore, the cysteine introduced near the N-terminus of the C-terminal fragment via site-directed mutagenesis was targeted for the oligonucleotide conjugation. In order to perform the conjugation, Rluc fragments were first reduced using tricholroethyl phospine (TCEP). Rluc fragments

Communications

Bioconjugate Chem., Vol. 20, No. 1, 2009 17

Figure 4. Bar graph showing the luminescence activity obtained in control and samples. The data are based on an average response of three samples ( the standard deviation. The inset shows a plot of luminescence activity obtained after mixing different moles of the probe1-Rluc N-terminal fragment conjugate and probe2-Rluc C-terminal fragment conjugates and adding coelenterazine. The data are based on an average response of three samples ( the standard deviation.

were then reacted with BM(PEO)2 using the manufacturer’s protocol in a 1:2 ratio of fragments to linker (Supporting Information). Both oligonucleotide probes (probe 2, 5′TCAACATCAGTCTGATAAGCTA3′-thiol and probe 1, thiol5′TAGCTTATCAGACTGATGTTGA3′) were reduced using 40 mM TCEP at 37 °C and were added at a 5 times mole excess to the Rluc-BM(PEO)2 conjugate. Probe 2 was conjugated to the N-terminal fragment, and probe 1 was fused to the C-terminal fragment. Since there is a unique cysteine introduced via site-directed mutagenesis in the C-terminal fragment, only one probe is conjugated to each fragment. Even though the N-terminal fragment contains three cysteine residues, the cysteine introduced via site-directed mutagenesis is more accessible for conjugation than the other three cysteine thiols. Unreacted probes and cross-linker were then separated using dialysis in sodium phosphate buffer (50 mM sodium phosphate, pH 7.2, containing 300 mM NaCl). In general, chemical conjugation techniques are not highly reproducible. However, in our studies we found about 70-80% conjugation efficiency monitored using SDS-PAGE analysis. This conjugation efficiency was found to be independent of the amount of mole excess of oligonucleotide probes used in the conjugation reaction. Furthermore, these conjugates were observed to be either mono-or bioligonucleotide conjugated as monitored using SDS-PAGE. Unreacted split fragments were not separated from the fragment-probe conjugates since the fragments alone are not active and thus will not contribute to any background signal. Initially, a control study was performed to determine any background signal from fragment-probe conjugates. The bioluminescence measurements were performed by adding coelenterazine to Rluc-N-terminal fragment-probe 2 and to RlucC-terminal fragment-probe 1 separately. The luminescence signal obtained using probe-fragment conjugates was comparable to that obtained using buffer mixed with coelenterazine. Next, 5 nmol of each of the two probe-fragment conjugates were mixed together at 37 °C for 30 min, and bioluminescence intensity was measured after the addition of coelenterazine. Figure 4 depicts a bar graph of luminescence intensity obtained upon mixing the two complementary probe-fragments and the intensity obtained in the control study. The control resulted in a minimal background luminescence representative of the luminescence of coelenterazine alone. In comparison, a strong bioluminescence signal was observed upon the reassembly of Rluc fragments driven through probe hybridization. In another study, we mixed various concentrations of probe-Rluc fragment conjugates and measured bioluminescence signal after incuba-

Figure 5. Luminescence emission profile obtained using a Cary Eclipse fluorometer of reassembled Rluc (solid line) and intact Rluc (dotted line) after adding coelenterazine.

tion (Figure 4 inset). A linear increase in luminescence with increasing concentrations of the two probe-fragment conjugates was obtained. We were able to detect Rluc with a detection limit of 5 fmoles (2.5 × 10-11 M) of probe-Rluc fragment conjugates. This detection limit was calculated corresponding to the signal obtained using the formula signal of the blank + 3 standard deviation of the blank. In a study performed using GFP as the reporter for fragment reassembly driven through DNA hybridization, the concentrations of probe-GFP fragment conjugates employed were 200 nM (12). Our results indicate that we were able to detect oligonucleotide probe hybridization using Rluc reassembly approximately 4 orders of magnitude lower than that with split GFP. A bioluminescence emission profile of the reassembled Rluc compared to the native Rluc (Figure 5) was obtained using a cuvette on a Cary Eclipse spectrophotometer (Varian, CA). The reassembled protein showed an emission maximum of 495 nm, while the native Rluc showed a 485 nm emission peak. This red shift of reassembled protein versus native protein has been observed previously with GFP (12). It is thought that the shift in emission wavelength may be explained by the change in proximity of the amino acids around the chromophore compared to the native Rluc, as well as the presence of the negatively charged DNA molecules driving the reassembly.

18 Bioconjugate Chem., Vol. 20, No. 1, 2009

Next, we studied the possibility of the potential application of reassembly of Rluc for the target nucleic acid detection. This was evaluated by examining whether the fragment reassembly driven through oligonucleotide hybridization can be competitively inhibited by a free complementary probe. When a 13.2 nmol of free oligonucleotide probe (5′-TAGCTTATCAGACTGATGTTGA-3′) was added to the reassembled Rluc fragmentprobe complex (1 nmol each), we observed a reduction in the bioluminescence signal. Approximately 45% reduction in bioluminescence intensity in the presence of free probe compared to that in the control without the free probe was observed. Although we have used excess free probe, complete quenching of the bioluminescence was not observed. We believe that the possible reason for this is when the free probe is added to the fragment-probe hybridized complex, a higher temperature than 37 °C may be needed to separate hybridized strands, which may be due to a tight assembly between the probe-fragment hybridized complex. A similar phenomenon was observed in the study performed using GFP split fragments in oligonucleotide hybridization (12). To further validate this result, when we added 3.4 nmol of the free probe to the probe-directed reassembled Rluc (1 nmol each), a 25% reduction in the bioluminescence intensity was observed. This result indicates that the free probe can prevent DNA-directed reassembly of Rluc in a concentration-dependent manner. Ideally, a competitive assay can be performed such that the free probe is allowed to hybridize with the complementary probe-Rluc fragment followed by the addition of the second probe-Rluc fragment. An alternative method would be to simultaneously mix probefragment conjugates and the free probe. Another strategy to perform target nucleic acid quantitation can be based on designing the probes such that they bind adjacently to the target. Here, the Rluc fragments can be conjugated to the probes and directly added to the target to be detected. This would allow reassembly of fragments and the generation of active Rluc in the presence of the target, yielding an increase in the signal rather than a decrease and hence a higher sensitivity of detection. All of these strategies and application of Rluc reassembly in nucleic acid quantitation are under evaluation in our laboratory. Finally, we performed the reassembly of probe-Rluc fragments in a cellular extract. This study was performed to evaluate whether the Rluc reassembly strategy will work in a more complex matrix. Another objective that was fulfilled by this experiment was to evaluate whether any nonspecific nucleic acid present in the sample will interfere with the hybridization of probes and hence the subsequent reassembly of Rluc fragments. For this study, we used the extract of E. coli strain ER 2566. We employed the sonication method to obtain the extract, which is also a method to get the genomic DNA (29). After sonicating E.coli strain ER 2566 cells in 50 mM sodium phosphate buffer, pH 7.2, containing 300 mM NaCl, 1 nmol of the probe 2-Nterminal Rluc fragment and probe 1-C-terminal Rluc fragment were mixed with the cell extract and hybridized. Bioluminescence emission intensity was measured after adding coelenterazine. The luminescence intensity obtained was comparable to that obtained in the in vitro study (Figure 6). A comparison of the light intensity in the buffer matrix and in the E. coli cell matrix showed an approximately 4.1% decrease from buffer matrix to cell matrix, which is within the allowed 10% limit. Therefore, reassembly of Rluc driven through DNA hybridization is not significantly affected by the sample matrix. In summary, our work demonstrates that Rluc reassembly can be performed in vitro driven through oligonucleotide hybridization. The detection limit for Rluc reassembly was 5 fmoles (2.5 × 10-11 M) of probe-Rluc fragment conjugates, which is approximately 4 orders of magnitude lower than that reported using probe-GFP fragment conjugates. Our study also

Communications

Figure 6. Bar graph showing the luminescence intensity obtained after adding coelenterazine to the samples containing probe-fragment conjugates in buffer and in E. coli cellular extract. The data are based on an average response of three samples ( the standard deviation.

shows that the reassembly can be inhibited in the presence of a complementary oligonucleotide indicating that this strategy can be useful in the quantitative detection of target nucleic acid. The bioluminescence emission, small size, and monomeric structure make Rluc an efficient label for the development of complementation assays. Furthermore, the method developed offers the advantage of parallel analysis in a 96-well microtiter plate and makes it suitable for high throughput nucleic acid quantitation applications.

ACKNOWLEDGMENT This work was supported by a grant from the National Institutes of Health (EB008231-01). Supporting Information Available: Methods for Rluc fragment genetic construction, expression and purification of protein fragments, conjugation of probe to Rluc fragments, and all hybridization studies, along with relevant figures. This material is available free of charge via the Internet at http:// pubs.acs.org.

LITERATURE CITED (1) Paulmurugan, R., and Gambhir, S. S. (2005) Firefly luciferase enzyme fragment complementation for imaging in cells and living animals. Anal. Chem. 77, 1295–1302. (2) Paulmurugan, R., and Gambhir, S. S. (2007) Combinatorial library screening for developing an improved split-firefly luciferase fragment-assisted complementation system for studying protein-protein interactions. Anal. Chem. 79, 2346–2353. (3) Paulmurugan, R., Umezawa, Y., and Gambhir, S. S. (2002) Noninvasive imaging of protein-protein interactions in living subjects by using reporter protein complementation and reconstitution strategies. Proc. Natl. Acad. Sci. U.S.A. 99, 15608– 15613. (4) Ooi, A. T., Stains, C. I., Ghosh, I., and Segal, D. J. (2006) Sequence-enabled reassembly of beta-lactamase (SEER-LAC): A sensitive method for the detection of double-stranded DNA. Biochemistry 45, 3620–3625. (5) Magliery, T. J., Wilson, C. G., Pan, W., Mishler, D., Ghosh, I., Hamilton, A. D., and Regan, L. (2005) Detecting proteinprotein interactions with a green fluorescent protein fragment reassembly trap: scope and mechanism. J. Am. Chem. Soc. 127, 146–157. (6) Stains, C. I., Porter, J. R., Ooi, A. T., Segal, D. J., and Ghosh, I. (2005) DNA sequence-enabled reassembly of the green fluorescent protein. J. Am. Chem. Soc. 127, 10782–10783. (7) Kim, S. B., Otani, Y., Umezawa, Y., and Tao, H. (2007) Bioluminescent indicator for determining protein-protein interactions using intramolecular complementation of split click beetle luciferase. Anal. Chem. 79, 4820–4826.

Communications (8) Deo, S. K. (2004) Exploring bioanalytical applications of assisted protein reassembly. Anal Bioanal Chem 379, 383–390. (9) Ozawa, T., Kaihara, A., Sato, M., Tachihara, K., and Umezawa, Y. (2001) Split luciferase as an optical probe for detecting protein-protein interactions in mammalian cells based on protein splicing. Anal. Chem. 73, 2516–2521. (10) Gangopadhyay, J. P., Jiang, S. Q., and Paulus, H. (2003) An in vitro screening system for protein splicing inhibitors based on green fluorescent protein as an indicator. Anal. Chem. 75, 2456–2462. (11) Galarneau, A., Primeau, M., Trudeau, L. E., and Michnick, S. W. (2002) Beta-lactamase protein fragment complementation assays as in vivo and in vitro sensors of protein protein interactions. Nat. Biotechnol. 20, 619–622. (12) Demidov, V. V., Dokholyan, N. V., Witte-Hoffmann, C., Chalasani, P., Yiu, H. W., Ding, F., Yu, Y., Cantor, C. R., and Broude, N. E. (2006) Fast complementation of split fluorescent protein triggered by DNA hybridization. Proc. Natl. Acad. Sci. U.S.A. 103, 2052–2056. (13) Whitehead, T. P., Kricka, L. J., Carter, T. J., and Thorpe, G. H. (1979) Analytical luminescence: its potential in the clinical laboratory. Clin. Chem. 25, 1531–1546. (14) Thore, A. (1979) Luminescence in clinical analysis. Ann. Clin. Biochem. 16, 359–369. (15) Shrestha, S., Paeng, I. R., Deo, S. K., and Daunert, S. (2002) Cysteine-free mutant of aequorin as a photolabel in immunoassay development. Bioconjugate Chem. 13, 269–275. (16) Contag, C. H., and Bachmann, M. H. (2002) Advances in in vivo bioluminescence imaging of gene expression. Annu. ReV. Biomed. Eng. 4, 235–260. (17) Kricka, L. J. (1991) Chemiluminescent and bioluminescent techniques. Clin. Chem. 37, 1472–1481. (18) Porter, J. R., Stains, C. I., Jester, B. W., and Ghosh, I. (2008) 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. 130, 6488–6497. (19) Matthews, J. C., Hori, K., and Cormier, M. J. (1977) Substrate and substrate analogue binding properties of Renilla luciferase.

Bioconjugate Chem., Vol. 20, No. 1, 2009 19 Biochemistry 16, 5217–5220. (20) Matthews, J. C., Hori, K., and Cormier, M. J. (1977) Purification and properties of Renilla reniformis luciferase. Biochemistry 16, 85–91. (21) Hart, R. C., Matthews, J. C., Hori, K., and Cormier, M. J. (1979) Renilla reniformis bioluminescence: luciferase-catalyzed production of nonradiating excited states from luciferin analogues and elucidation of the excited state species involved in energy transfer to Renilla green fluorescent protein. Biochemistry 18, 2204–2210. (22) Mirasoli, M., Deo, S. K., Lewis, J. C., Roda, A., and Daunert, S. (2002) Bioluminescence immunoassay for cortisol using recombinant aequorin as a label. Anal. Biochem. 306, 204–211. (23) Greer, L. F., III., and Szalay, A. A. (2002) Imaging of light emission from the expression of luciferases in living cells and organisms: a review. Luminescence 17, 43–74. (24) Gould, S. J., and Subramani, S. (1988) Firefly luciferase as a tool in molecular and cell biology. Anal. Biochem. 175, 5–13. (25) Desai, U. A., Deo, S. K., Hyland, K. V., Poon, M., and Daunert, S. (2002) Determination of prostacyclin in plasma through a bioluminescent immunoassay for 6-keto-prostaglandin F1alpha: implication of dosage in patients with primary pulmonary hypertension. Anal. Chem. 74, 3892–3898. (26) Deo, S. K., and Daunert, S. (2001) Luminescent proteins from Aequorea victoria: applications in drug discovery and in high throughput analysis. Fresenius’ J. Anal. Chem. 369, 258–266. (27) Paulmurugan, R., and Gambhir, S. S. (2003) Monitoring protein-protein interactions using split synthetic renilla luciferase protein-fragment-assisted complementation. Anal. Chem. 75, 1584–1589. (28) Loening, A. M., Fenn, T. D., and Gambhir, S. S. (2007) Crystal structures of the luciferase and green fluorescent protein from Renilla reniformis. J. Mol. Biol. 374, 1017–1028. (29) Zhang, L., Foxman, B., Gilsdorf, J. R., and Marrs, C. F. (2005) Bacterial genomic DNA isolation using sonication for microarray analysis. Biotechniques 39, 640–644. BC8003099