Circularly Permutated Bioluminescent Probes for Illuminating Ligand

Dec 2, 2008 - Engineered bioluminescent enzymes provide a creative approach for illuminating the intracellular molecular events. We demonstrate a new ...
0 downloads 0 Views 1MB Size
Download PDF
2480

Bioconjugate Chem. 2008, 19, 2480–2486

Circularly Permutated Bioluminescent Probes for Illuminating Ligand-Activated Protein Dynamics Sung Bae Kim, Moritoshi Sato, and Hiroaki Tao* Research Institute for Environmental Management Technology, National Institute of Advanced Industrial Science and Technology (AIST), 16-1 Onogawa, Tsukuba 305-8569, Japan, and Graduate School of Arts and Sciences, The University of Tokyo, 3-8-1 Komaba, Meguro-ku, Tokyo 153-8902, Japan. Received September 5, 2008; Revised Manuscript Received November 3, 2008

Engineered bioluminescent enzymes provide a creative approach for illuminating the intracellular molecular events. We demonstrate a new strategy for molecular imaging of bioactive small molecules using circularly permutated luciferases (cpLuc), derived from Gaussia princeps (GLuc), Photinus pyralis (FLuc), and Pyrearinus termitilluminans (CBLuc): (i) the luciferases were first dissected into two fragments, (ii) the original N- and C-termini were linked with a GS linker and the new termini were created in an appropriate site, and (iii) the new ends were correspondingly linked with proteins of interest, e.g., a ligand-binding domain (LBD) and an LBD-recognition protein. When the ends of the cpCBLuc were linked with LBD of estrogen receptor (ER) and Src homology 2 domain of Src (SH2 of Src), the estrogen can trigger an intramolecular ER LBD-Src SH2 binding. This assembly subsequently provokes an approximation of the adjacent fragments of cpCBLuc recovering the enzyme activity. These probes were surprisingly stable in the mammalian cells and largely exhibited a decreased background luminescence (e.g., 1000 times in case of cpGLuc) and improved signal-to-noise ratios, compared with the nonCP indicators. This study offers a new strategy for luciferase-aided probing systems. Our study is the first to fabricate circular permutation (CP) in luciferases for tracing the molecular dynamics of proteins.

INTRODUCTION Engineered functional proteins have provided major contributions to advancements in the bioindustry and have brought a breakthrough in the treatment of human diseases. Manipulation of enzymes as functional proteins is largely governed by the knowledge of the principles that direct enzyme catalysis and enzyme structures, elucidated by X-ray crystallography (1). Recent revolutionary advances in enzyme manipulation technologies now allow researchers to carry out quantitative examination of the molecular dynamics and cell signaling in living cells (2). While the green fluorescent protein (GFP) that acts as an excellent functional protein has been widely used, enzymes catalyzing light emission are currently proving their distinguished advantages in the quantitative, specific, signal-enhanced, noninvasive, and real-time investigation of intracellular molecular events (3-9). As a creative approach for probing molecular events, a circular permutation (CP) of bioluminescent enzymes can be an effective technique. Many of the beetle-derived bioluminescent enzymes are members of a superfamily of acyl-adenylate forming enzymes and consist with larger N-terminal and smaller C-terminal domains (10, 11). When the active-site region between the domains is dissected into two fragments, the activity of the enzyme may be temporally lost. A dissected active site can be placed in the opposite side of the other dissected site in the probe backbone, by CP. Upon stimulation of the ligand, the fragmented luciferase recovers its activity via an intramolecular complementation. As the possibility of encounter between the fragmented active sites by CP is rare, we can expect a dramatic decrease in the background activity of the enzyme and an enhancement of the signal-to-background ratios. The molecular structure is illustrated with the firefly luciferase (FLuc), as shown in Figure 1. * Corresponding author. [email protected].

On the basis of this speculation, the feasibility of an indicator with circularly permutated luciferases was explored with the luciferases derived from Gaussia princeps (Gaussia luciferase (GLuc); Nanolight), Photinus pyralis (firefly luciferase (FLuc); Promega), and Pyrearinus termitilluminans (click beetle luciferase (CBLuc); Promega). We first examined the circularly permutated CBLuc (cpCBLuc) for illuminating the phosphorylation of estrogen receptor (ER), and cpFLuc for imaging the ligand-stimulated protein-protein interactions. The original N- and C-termini of CBLuc were linked with a GS linker, and the new N- and C-termini were constructed between I439 and K440 of CBLuc. Each end of the cpCBLuc was extensively linked with the ligand binding domain of the estrogen receptor (ER LBD) and the Src homology 2 domain of Src (Src SH2). This fusion protein was surprisingly expressed stably in mammalian cells and tolerated the insertion of ER LBD and Src SH2. The probe was sensitive to 4-hydroxytamoxifen (OHT) and emitted the characteristic bioluminescence. Similarly, we synthesized a bioluminescent probe that traces the real-time dynamics of Ca2+ as a representative second messenger in living mammalian cells. In Figure 1, the current mechanism of the CP of enzymes is illustrated with the structure of FLuc, as the crystal structures of CBLuc and GLuc were not presented. It should be noted that Ca2+ and estrogen are a typical second messenger and a major steroid hormone, respectively. Thus, the above-mentioned examples of GLuc and CBLuc demonstrate the general applicability of the current CP concept in illuminating the protein dynamics in living mammalian cells.

EXPERIMENTAL PROCEDURES Plasmid Construction. The respective cDNA fragments consisting of circularly permutated luciferase probes were generated by the polymerase chain reaction (PCR) to introduce each unique restriction site at each end of the fragments, using adequate primers and templates. For example, a plasmid carrying Yellow

10.1021/bc800378a CCC: $40.75  2008 American Chemical Society Published on Web 12/02/2008

Circularly Permutated Bioluminescent Probes

Bioconjugate Chem., Vol. 19, No. 12, 2008 2481

Figure 1. (A) Schematic diagram of circularly permutated luciferase probes. The original N- and C-terminals of CBLuc, GLuc, and FLuc were fused with a 10 GS linker. On the other hand, their new terminal ends were created at I439, Q105, and D415, respectively. Abbreviations: Src SH2, SH2 domain of Src; CBLuc-N and -C, N- and C-terminal fragment of click beetle luciferase; ER LBD, ligand binding domain of estrogen receptor; M13, a 26-residue calmodulin-binding peptide of myosin light-chain kinase; CaM, calmodulin. (B) The crystal structure of firefly luciferase (FLuc). N- and C-terminal fragments are colored gray and red, respectively. The scissors mark shows the dissected point (D415) in the structure. (C) Comprehensive illustration of the molecular structure of CPF as an example. (D) Comparison of the ligand sensitivities of the present CP probes. The luminescence intensities from COS-7 cells carrying each probe were determined in the presence or absence of the specific stimulators, i.e., OHT for CPC, DHT for CPF, histamine for CPG.

Cameleon-3.1 (YC3.1), donated by Dr Miyawaki (12), was utilized as a template for amplifying Xenopus laeVis CaM, by using PCR. A cDNA encoding humanized Gaussia princeps luciferase (GLuc; Nanolight) was consumed as a template of PCR for generating the N- and C-terminal fragments, where the known N-terminal secretion signal (1-17 AA) of GLuc was excluded for confining GLuc to the cytoplasm as determined before (6, 13). The synthesized constructs are illustrated in Figure 1A. The enzyme-restricted cDNA fragments were tandemly ligated, as shown in Figure 1A, and subcloned into pcDNA 3.1(+) (Invitrogen). The plasmids were named pCPC (cpCBLuc probe), pCPG (cpGLuc probe), and pCPF (cpFLuc probe), according to the type of luciferases that were circularly permutated in the single-chain probes. The expressed fusion probes may be called CPC, CPG, and CPF, respectively. In addition, the cDNA of ER LBD in pCPC was pointmutated at Y537 from TAT to TTT, i.e., Y537F. The construct was subcloned into the pcDNA 3.1(+) backbone. This plasmid was named pCPC-mut. As another reference to our cpCPC probe, a probe with straight-ordered fragments of luciferase was similarly constructed, as shown in Figure 3B. The N- and C-terminal fragments were tandemly fused in the straight order, where a 10 GS linker was intervened between the fragments. Then, the N- and C-terminal ends of chimera cDNA were flanked with cDNAs of Src SH2 and ER LBD, respectively. The plasmid with this construct was named pCPC-ctrl. In addition, an integrated-molecule-format (IMF) probe was constructed in parallel, as illustrated in Figure 3D. The plasmid was named pSMC. As a reference to CPG, an IMF probe was constructed, as shown in Figure 5B. The plasmid encoding this probe was named pCPG-ctrl. All the plasmids constructed for this study were sequenced to ensure fidelity, using a BigDye Terminator Cycle Sequencing kit and a genetic analyzer ABI Prism310 (Applied Biosystems). Comparison of Relative Luminescence Intensities by CPC, CPF, or CPG. The relative luminescence intensities by CPC, CPF, or CPG were compared in the presence or absence of each ligand (Figure 1(D)). COS-7 cells derived from African

green monkey kidney were cultured in a 24-well plate with Dulbecco’s modified Eagle’s medium (DMEM; Sigma), supplemented with 10% steroid-free fetal bovine serum (FBS) and 1% penicillin-streptomycin (P/S) at 37 °C in a 5% CO2 cell incubator (Sanyo). The COS-7 cells were transiently transfected with pCPC, pCPF, or pCPG (0.2 µg per well) using a transfection reagent, TransIT-LT1 (Mirus). The cells were extensively incubated in the CO2 incubator for 16 h. The cells carrying pCPC or pCPF were stimulated with 10-6 M of OHT or dihydrotestosterone (DHT) (final concentration). The luminescence intensities were developed with a BrightGlo substrate solution (Promega) and integrated for 15 s using a luminometer (Minilumat LB9506; Berthold). The brief procedure for the use of the Bright-Glo substrate solution is as follows: The mammalian cells were washed once with PBS, 20 min after stimulation with a ligand. 40 µL of the substrate (D-leuciferin) solution was added to each well of the plates. Three minutes after substrate addition, the plate was tapped gently, and the subsequent cell lysates were transferred to a test tube for determination of the luminescence intensities. One the other hand, the cells carrying pCPG were harvested by trypsinization and centrifugation. The cells were gently mixed with a 40 µL substrate (coelenterazine) solution containing 1 mM histamine or PBS. The luminescence intensities were then determined with the luminometer (Minilumat LB9506). Comparison of the Ligand Sensitivity of CPC and CPCmut. We examined whether the phosphorylation of ER LBD at Y537 motivated the intramolecular interaction between Src SH2 and ER LBD in the CPC and CPC-mut, as illustrated in Figure 2B. COS-7 cells cultured in a 24-well plate were transiently transfected with pCPC or pCPC-mut. Sixteen hours after transfection, the cells were stimulated with various ligands (10-6 M) for 20 min. The ligands used were as follows: vehicle (0.1% DMSO; final concentration), OHT, estrone, E2, DHT, and cortisol. The subsequent luminescence intensities were estimated with the luminometer.

2482 Bioconjugate Chem., Vol. 19, No. 12, 2008

Kim et al.

Figure 2. (A) Mutagenesis study for the association of ER LBD with Src SH2 domain. Tyrosine 537 in ER LBD is replaced with lysine, and the plasmid is named pCPC-mut. The luminescence intensities from COS-7 cells carrying pCPC or pCPC-mut are compared in response to various ligands (n ) 3). (B) Comprehensive illustration of ligand-sensing mechanisms of CPC and CPC-mut. OHT-ER LBD binding exerts phosphorylation of ER LBD at Y537, which subsequently triggers Src SH2-ER LBD interactions. On the other hand, mutation at Y537 demolishes this interaction. (C) Relative sensitivity of CPC to various antagonists. E2 competed with the antagonists for binding ER LBD. (D) Spectra from the COS-7 cells carrying CPC before and after stimulation of OHT.

Comparison of Inhibitory Effects of E2 to AntagonistBound CPC. On the basis of the result that CPC is sensitive to ER antagonists, and not to the agonists, the inhibitory effects of E2 to the antagonized CPC were examined (Figure 2C). COS-7 cells cultured in 24-well plates were transiently transfected with pCPC. Sixteen hours after transfection, the cells were prestimulated with 10-5 M E2 for 5 min. The cells were additionally incubated with one of the following nuclear receptor antagonists (10-6 M: final concentration): ciglitazone, ICI182780, genistein, or OHT. The resulting luminescence intensities were compared with those from the cells stimulated with an antagonist alone. Comparison of the Ligand Sensitivity of CPC and CPCctrl. The ligand sensitivity of CPC was compared with the CPCctrl (Figure 3A). COS-7 cells cultured in 24-well plates were transiently transfected with pCPC or pCPC-ctrl. Sixteen hours after transfection, the cells were stimulated with varying concentrations of E2 or OHT for 20 min. The resulting luminescence intensities were recorded with the luminometer. Dose-Response Curves of CPC to Ligands and the Comparison with CPC-ctrl. Ligand sensitivity of CPC in COS-7 cells was examined with varying concentrations of steroids (Figure 3C). COS-7 cells cultured in the 24-well plates were transiently transfected with pCPC. Sixteen hours after transfection, the cells were stimulated with varying concentrations of ligands. The developed luminescence intensities were recorded with the luminometer (Minilumat LB9506). As another reference to CPC, the sensitivity of SMC was examined in parallel with the same experimental condition as CPC. COS-7 cells carrying pSMC were stimulated with varying concentrations of OHT, and the resulting luminescence intensities were determined with the luminometer (Minilumat LB9506). Kinetics of Ligand-Probe Binding. The kinetic aspects of ligand-probe binding were estimated with COS-7 cells carrying

pCPC (Figure 4). COS-7 cells cultured in 12-well plates were transiently transfected with pCPC and incubated for another 16 h. The luminescence intensities were recorded at 2, 5, 10, 20, and 30 min after the addition of 10-6 M OHT or vehicle (0.1% DMSO). In the meantime, relaxation of the OHT-CPC binding was examined by refreshing the culture medium after OHT stimulation (Figure 4B). The cells carrying pCPC were stimulated with 10-6 M OHT or vehicle (0.1% DMSO) for 20 min, and the culture media were then replaced with fresh, ligand-free media. The luminescence intensities at 10, 20, and 60 min after medium replacement were developed with a Bright-Glo substrate solution. The intensities at each time period were compared with those from the cells mock-stimulated with the vehicle (0.1% DMSO). Comparison of Relative Luminescence Intensities after Being Screened by Bandpass Filters. The wavelengths of light emitted from COS-7 cells were examined by using a series of bandpass filters (Figure 4C). COS-7 cells cultured in a 24-well, glass-bottom plate were transfected with pCPC and incubated for 16 h. The cells were saturated with a 500 µL Hank’s balanced salt solution (HBSS) buffer containing coelenterazine and one of the following stimulators: vehicle (0.1% DMSO), 10-6 M E2, or 10-6 M OHT. Twenty minutes after stimulation with the ligand, the relative luminescence intensities from the cells were monitored with a series of bandpass filters, whose wavelengths were as follows: 510 ( 10 nm, 535 ( 10 nm, 540 ( 10 nm, 560 ( 10 nm, and 610 ( 10 nm. Determination of the Ca2+ Dynamics in COS-7 Cells with CPG. The dynamics of the cytosolic Ca2+ in COS-7 cells were monitored with CPG and CPG-ctrl (Figure 5A). COS-7 cells cultured in a black, glass-bottom plate (24-well) were transfected with pCPG or pCPG-ctrl, and incubated in a cell incubator for 16 h. The cells were saturated with 300 µL HBSS buffer

Circularly Permutated Bioluminescent Probes

Bioconjugate Chem., Vol. 19, No. 12, 2008 2483

Figure 3. Ligand sensitivity of CPC. (A) Comparison of ligand sensitivity of CPC and CPC-ctrl. Inset shows the absolute luminescence intensities from the cells carrying CPC and CPC-ctrl in response to OHT and E2. The left X-axis shows the luminescence intensities by CPC (open bars), whereas the right X-axis indicates the intensities by CPC-ctrl (closed bars). (B) Comprehensive illustration of the molecular structures of CPC and CPC-ctrl. Abbreviations: OHT, 4-hydroxytamoxifen; E2, 17β-estradiol; DHT, dihydrotestosterone; Src SH2, Src homology 2 (SH2) domain of Src; ER LBD, ligand binding domain of estrogen receptor; CBLuc-N and -C, N- and C-terminal domains of click beetle luciferase. (C) Dose-response curves of CPC in response to varying concentrations of ligands. (D) Illustrative comparison of the molecular structures of CPC and SMC.

containing coelenterazine. The luminescence variances were monitored with a bioluminescence plate reader (Mithras LB 940; Berthold) every 30 s before and after the addition of histamine in the range 0.1-1 mM.

RESULTS Structural Basis of CPC and Its Phosphorylation Recognition. The Y537 in ER LBD was phosphorylated upon the stimulation of ER ligands, which is sensitively recognized by the Src SH2 domain (14). We utilized this typical non-genomic interaction of ER with Src in constructing our CP probe. This phosphorylation-mediated binding was compared with a reference probe carrying point-mutated ER LBD at Y537F (CPCmut) (Figure 2A,B). CPC recognized various steroids and was particularly sensitive to an estrogen antagonist, OHT, whereas CPC-mut was insensitive to all the ligands (Figure 2A). The results show that (i) the intramolecular interactions occur via phosphorylation of ER LBD at Y537, (ii) Y537 in ER is phosphorylated ligand-dependently, (iii) the luminescence intensities by CPC are capable of indexing the antagonistic activities of ligands, and (iv) from an analytical point of view, signal-to-noise ratios are large enough to discriminate among the non-genomic activities of the ligands. It was previously debated on whether 17β-estradiol (E2) exerts phosphorylation of ER at Y537. Our experiment reveals that not only E2 but also OHT could induce phosphorylation of ER at Y537. Although both E2 and OHT phosphorylate ER LBD, OHT may induce more favorable conformational changes of ER LBD in the recovery of CBLuc activity than E2. This observation is analogous to an earlier luciferase study on ER antagonists (5). We also examined the influence of E2 on CPC-antagonist binding (Figure 2C). E2 negatively contributed to the luminescence intensities developed by the antagonists. The results show that the (i) antagonists compete with E2 in binding ER LBD,

and (ii) OHT is the most efficient antagonist barely influenced by 10× excess of E2. Luminescence spectra were obtained for examining whether the recovery of the red-orange light is selective to OHT (Figure 2D). The spectra showed that red-orange light was enhanced by OHT and the λmax was the same as that of intact CBLuc. This result was in concordance with the results with bandpass filters in Figure 4C and also with the result of our previous study (9). Comparison of Ligand Sensitivity of CPC with That of Its Non-CP Controls. The ligand sensitivities of CPC were examined with varying doses of steroids (Figure 3). CPC selectively sensed OHT and even responded to 10-9 M of OHT. The half-maximal effective concentration (EC50) was approximately 5 × 10-8 M. On the other hand, CPC-ctrl did not show any luminescence variance in response to OHT or E2 (Figure 3A). The reason was found to be in the extremely high background luminescence, even in the absence of stimulators. The background luminescences by CPC-ctrl were ca. 1000 times higher than those by CPC. This resulted in poor signal-to-noise ratios, as shown in Figure 3A. CPC showed a 10-fold enhanced detection limit, when compared with a CP-free probe (SMC) (Figure 3C). The structural difference between CPC and SMC is illustrated in Figure 3D. Even in the absence of a stimulator, SMC emitted a background luminescence that was seven times stronger than CPC, i.e., 38 500 vs 5500 RLU (n ) 3). The reason for the improved detection limit is explained by the fact that CP of CBLuc decreases the basal interactions between the ER LBD and SH2 in the absence of a ligand. Kinetics of CPC-Ligand Binding. The ligand-sensing kinetics of CPC was monitored with COS-7 cells carrying pCPC (Figure 4). The luminescence intensities by CPC were greatly enhanced from 5 min after OHT addition and reached a plateau

2484 Bioconjugate Chem., Vol. 19, No. 12, 2008

Kim et al.

Figure 4. Ligand-binding kinetics of CPC. (A) Time course of OHT-CPC binding was examined. At 2, 5, 10, 20, and 30 min after OHT-CPC binding, the luminescence variances were monitored. (B) Cancellation of OHT-CPC binding by medium replacement was determined. The luminescence intensities were retained until 60 min after medium replacement. (C) Relative luminescence intensities after being screened by bandpass filters. The light from the cells carrying CPC passed well through a 610 nm bandpass filter, indicating that the signal is red-orange light.

Figure 5. (A) Comparison of ligand-binding kinetics of CPG (solid lines) with that of SMG (dot lines). Inset shows the increased luminescence intensities due to histamine (n ) 3). (B) A cartoon diagram showing the Ca2+ recognition and conformation changes of CPG and CPG-ctrl. Abbreviations: GLuc-N and -C, N- and C-terminal fragments of Gaussia luciferase; ∆RLU, change in the amount of luminescence intensity in the presence or absence of histamine; M13, calmodulin-binding peptide of myosin light-chain kinase; CaM, calmodulin.

in 20 min. The total response time, 20 min, comprised the total time for (i) penetration of OHT across the plasma membrane, (ii) OHT-ER LBD binding and conformation change of ER LBD, (iii) subsequent binding of ER LBD with Src SH2, and (iv) intramolecular complementation between the fragments of CBLuc. Using cell-free assays, it was previously proven that the net ligand-ER binding and conformation change of ER were completed within 1 min (15). Therefore, it can be considered that the large portion of the total response time, 20 min, was

consumed during the penetration of OHT into the cytosol. This response time is considerably slower than those shown by our previous probes (10, 16). Hydrophilicity of OHT, when compared with other steroids, may cause the late plasma membrane penetration. Dissociation kinetics of CPC-ligand binding was examined with COS-7 cells carrying pCPC (Figure 4B). First, the cells were stimulated with OHT for 20 min. The culture medium was then replaced with a fresh, OHT-free medium. The

Circularly Permutated Bioluminescent Probes

luminescence intensities within 1 h after the change of the medium were not decreased considerably. Even washout for 60 min did not weaken the luminescence intensities from the cells. These results show that OHT-triggered Src SH2-ER binding can be endured for quite a long time. Wavelengths of Light Emitted from COS-7 Cells Carrying pCPC. The wavelengths and intensities of light from COS-7 cells carrying pCPC were determined with a series of bandpass filters after the stimulation of ligands (Figure 4C). The light from COS-7 cells efficiently passed a 610 ( 10 nm filter, which indicates that the probe emits red-orange light. The luminescence intensities by OHT were five times stronger than those by the vehicle (0.1% DMSO), whereas the intensities by E2 were merely 1.8 times stronger. These results show that OHT selectively elevates the red-orange light, and thus, the light can be a specific index of antagonistic activities of ligands to ER. Ligand-Sensing Chemistry of CPG. As a representative of second messengers, cytosolic Ca2+ levels in living mammalian cells were estimated with our CPG (Figure 5). Gaussia princepsderived luciferase was dissected at Q105, and the fragments were circularly permutated with a 10 GS linker. The outer terminals were fused with M13 and CaM, respectively. A control probe, CPG-ctrl, was synthesized in parallel with the same components but in different order (CP-free; Figure 5B). In response to 1 mM histamine, CPG quickly increased its luminescence intensities, which reached a plateau in approximately 10 min. The inset in Figure 5A shows the dose-response curve (n ) 3). The probe was even sensitive to 0.25 mM of histamine, and the linear range was found between 0.25 and 0.75 mM of histamine. On the other hand, CPG-ctrl was unstable in emitting light and exhibited approximately 100× higher background intensity than the CPG, even in the absence of histamine. Stimulation with 1 mM histamine seldom increased the luminescence intensities. Only the substrate coelenterazine was saturated in the COS-7 cells within 8.5 min (Supporting Information Figure S1(C)). These results conclude that the CP of GLuc favors the decrease in background intensities, as shown in the case of CPC and CPC-ctrl. The decreased basal luminescence enabled us to firmly determine the dynamics of free Ca2+ levels triggered by histamine.

DISCUSSION The concepts of CP were previously introduced with green fluorescent protein (GFP) variants, where the hydrophobicity of the chromophore varied according to a stimulator (17, 18). It is interesting to compare the ligand-sensing mechanism of our CP-based probes with that of the conventional CP-based fluorescent probes. The on/off system of fluorescent probes with cpGFP is considered to depend on the altered quantum yield of fluorescence by solvent penetration into the chromophore (19, 20). Particularly, an insertion of peptides at the β-sheet linkers temporarily disrupts fluorescence, owing to solvent penetration within the protein core, which interferes with the fluorophore-βsheet interactions. The inhibitory action is relieved by condensive interaction between the inserted peptides. On the other hand, the recovery of luminescence in our probes is based on the physical approximation after dissociation between the completely separated fragments of a split luciferase, and not on the hydrophobicity variance like cpGFP. Conventionally, real-time imaging of dynamics of second messengers relied heavily on the fluorescence resonance energy transfer (FRET) (12, 21, 22). The FRET between GFP variants can be detectable with a fluorescence microscope. However, only few cells can be observed by this method with autofluorescence. Our cpGLuc probe (CPG) provides a simple wholecell assay and real-time imaging, and is useful for tracing the

Bioconjugate Chem., Vol. 19, No. 12, 2008 2485

molecular dynamics in mammalian cells. Our study was the first example to determine the ligand-triggered dynamics of a second messenger with a bioluminescent probe in the living cells. The probe with cpCBLuc (CPC) was sensitive to antagonists and exhibited a 1000× lower background luminescence when compared with a non-CP probe (CPC-ctrl). This dramatic decrease of basal luminescence intensity can be explained as follows: the CBLuc is considered to consist of two major domains, between which the active site may probably exist (10, 11). In our study, the potential active site region was divided into two fragments, causing a temporary loss of activity. A dissected active site of CBLuc was located in the opposite side of the other dissected site in the single-chain probe backbone, with respect to CP. In this conformation, the encountering chance between the fragmented active sites should be extremely decreased in the basal conditions. This structural reason may contribute to (i) the dramatic decrease of the basal activity of the probe and (ii) the improved signal-to-background ratios in the study. It was previously debated whether the 17β-estradiol (E2) exerts phosphorylation of estrogen receptor (ER) at Y537 (23, 24). Our experiment revealed that not only E2 but also 4-hydroxytamoxifen (OHT) could induce phosphorylation of ER at Y537. Although both E2 and OHT phosphorylate ER LBD, OHT modulates a more appropriate conformation change on ER LBD for recruiting the adjacent SH2 domain than E2. Previously, an intramolecular folding sensor was developed, where a singlecomponent protein (i.e., ER LBD), and not a two-component protein like ours, was sandwiched between the fragments of split-Renilla luciferase (RLuc) (5). It is interesting that their probe was similarly sensitive only to antagonists and was believed to be controlled by the conformational variances of ER LBD. On the whole, we synthesized efficient bioluminescent probes with circularly permutated fragments of luciferases. Our probes were surprisingly stable in the mammalian cells after permutation, and the permutated positions were tolerant to the insertion of the detector proteins. The probes were sensitive to their respective ligands and exhibited a dramatically decreased background luminescence and improved signal-to-background ratios, when compared with the CPCctrl. This study offers a new strategy for the enzyme-aided probing systems that overcomes the limitations of the conventional probes. Our study was the first to fabricate CP in the enzymes. Any protein-protein interactions and molecular dynamics may be examined with these probes by simply replacing the detector proteins. Supporting Information Available: Washout effects of ligand-elevated Ca2+ levels in COS-7 cells carrying pCPG. were specified in Figure S1(A). Kinetics of coelenterazine saturation in COS-7 cells were examined with the cells carrying CPG-ctrl (Figure S1(C)). This material is available free of charge via the Internet at http://pubs.acs.org.

LITERATURE CITED (1) Brannigan, J. A., and Wilkinson, A. J. (2002) Protein engineering 20 years on. Nat. ReV. Mol. Cell Biol. 3, 964–970. (2) Weissleder, R., and Ntziachristos, V. (2003) Shedding light onto live molecular targets. Nat. Med. 9, 123–128. (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) Paulmurugan, R., and Gambhir, S. S. (2005) Firefly luciferase enzyme fragment complementation for imaging in cells and living animals. Anal. Chem. 77, 1295–1302.

2486 Bioconjugate Chem., Vol. 19, No. 12, 2008 (5) Paulmurugan, R., and Gambhir, S. S. (2006) An intramolecular folding sensor for imaging estrogen receptor-ligand interactions. Proc. Natl. Acad. Sci. U.S.A. 103, 15883–15888. (6) Remy, I., and Michnick, S. W. (2006) A highly sensitive protein-protein interaction assay based on Gaussia luciferase. Nat. Meth. 3, 977–979. (7) Kim, S. B., Ozawa, T., Watanabe, S., and Umezawa, Y. (2004) High-throughput sensing and noninvasive imaging of protein nuclear transport by using reconstitution of split Renilla luciferase. Proc. Natl. Acad. Sci. U.S.A. 101, 11542– 11547. (8) Kim, S. B., Kanno, A., Ozawa, T., Tao, H., and Umezawa, Y. (2007) Nongenomic activity of ligands in the association of androgen receptor with Src. ACS Chem. Biol. 2, 484– 492. (9) Kim, S. B., Umezawa, Y., Kanno, K. A., and Tao, H. (2008) An integrated-molecule-format multicolor probe for monitoring multiple activities of a bioactive small molecule. ACS Chem. Biol. 3, 359–372. (10) 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. (11) Viviani, V. R., Uchida, A., Viviani, W., and Ohmiya, Y. (2002) The influence of Ala243 (Gly247), Arg215 and Thr226 (Asn230) on the bioluminescence spectra and pH-sensitivity of railroad worm, click beetle and firefly luciferases. Photochem. Photobiol. 76, 538–544. (12) Miyawaki, A., Llopis, J., Heim, R., McCaffery, J. M., Adams, J. A., Ikura, M., and Tsien, R. Y. (1997) Fluorescent indicators for Ca2+ based on green fluorescent proteins and calmodulin. Nature 388, 882–887. (13) Tannous, B. A., Kim, D. E., Fernandez, J. L., Weissleder, R., and Breakefield, X. O. (2005) Codon-optimized Gaussia luciferase cDNA for mammalian gene expression in culture and in vivo. Mol. Ther. 11, 435–443. (14) Varricchio, L., Migliaccio, A., Castoria, G., Yamaguchi, H., de Falco, A., Di Domenico, M., Giovannelli, P., Farrar, W., Appella, E., and Auricchio, F. (2007) Inhibition of estradiol receptor/Src association and cell growth by an estradiol receptor alpha tyrosine-phosphorylated peptide. Mol. Cancer Res. 5, 1213–1221.

Kim et al. (15) Rich, R. L., Hoth, L. R., Geoghegan, K. F., Brown, T. A., LeMotte, P. K., Simons, S. P., Hensley, P., and Myszka, D. G. (2002) Kinetic analysis of estrogen receptor/ligand interactions. Proc. Natl. Acad. Sci. U.S.A. 99, 8562–8567. (16) Kim, S. B., Awais, M., Sato, M., Umezawa, Y., and Tao, H. (2007) Integrated molecule-format bioluminescent probe for visualizing androgenicity of ligands based on the intramolecular association of androgen receptor with its recognition peptide. Anal. Chem. 79, 1874–1880. (17) Baird, G. S., Zacharias, D. A., and Tsien, R. Y. (1999) Circular permutation and receptor insertion within green fluorescent proteins. Proc. Natl. Acad. Sci. U.S.A. 96, 11241–11246. (18) Kawai, Y., Sato, M., and Umezawa, Y. (2004) Single color fluorescent indicators of protein phosphorylation for multicolor imaging of intracellular signal flow dynamics. Anal. Chem. 76, 6144–9. (19) Kotlikoff, M. I. (2007) Genetically encoded Ca2+ indicators: using genetics and molecular design to understand complex physiology. J. Physiol. 578, 55–67. (20) Souslova, E. A., and Chudakov, D. M. (2007) Genetically encoded intracellular sensors based on fluorescent proteins. Biochemistry (Moscow, Russ. Fed.) 72, 683–697. (21) Truong, K., Sawano, A., Mizuno, H., Hama, H., Tong, K. I., Mal, T. K., Miyawaki, A., and Ikura, M. (2001) FRET-based in vivo Ca2+ imaging by a new calmodulin-GFP fusion molecule. Nat. Struct. Biol. 8, 1069–1073. (22) Awais, M., Sato, M., Lee, X. F., and Umezawa, Y. (2006) A fluorescent indicator to visualize activities of the androgen receptor ligands in single living cells. Angew. Chem., Int. Ed. 45, 2707–2712. (23) Migliaccio, A., Castoria, G., Di Domenico, M., de Falco, A., Bilancio, A., Lombardi, M., Bottero, D., Varricchio, L., Nanayakkara, M., Rotondi, A., and Auricchio, F. (2002) Sex steroid hormones act as growth factors. J. Steroid Biochem. Mol. Biol. 83, 31–35. (24) Arnold, S. F., Melamed, M., Vorojeikina, D. P., Notides, A. C., and Sasson, S. (1997) Estradiol-binding mechanism and binding capacity of the human estrogen receptor is regulated by tyrosine phosphorylation. Mol. Endocrinol. 11, 48–53. BC800378A