Measuring CREB Activation Using Bioluminescent Probes That Detect

Apr 16, 2012 - Measuring CREB Activation Using Bioluminescent Probes That Detect KID–KIX Interaction in Living Cells. Tetsuya Ishimoto†, Hiroki Ma...
0 downloads 0 Views 3MB Size
Download PDF
Article pubs.acs.org/bc

Measuring CREB Activation Using Bioluminescent Probes That Detect KID−KIX Interaction in Living Cells Tetsuya Ishimoto,† Hiroki Mano,† Takeaki Ozawa,‡ and Hisashi Mori*,† †

Department of Molecular Neuroscience, Graduate School of Medicine and Pharmaceutical Sciences, University of Toyama, Toyama, Japan ‡ Department of Chemistry, Graduate School of Science, The University of Tokyo, Bunkyo-Ku, Tokyo, Japan S Supporting Information *

ABSTRACT: The cyclic adenosine monophosphate response element-binding protein (CREB) is a transcription factor that contributes to memory formation. The transcriptional activity of CREB is induced by its phosphorylation at Ser-133 and subsequent interaction with the CREB-binding protein (CBP)/p300. We designed and optimized firefly split luciferase probe proteins that detect the interaction of the kinase-inducible domain (KID) of CREB and the KIX domain of CBP/p300. The increase in the light intensity of the probe proteins results from the phosphorylation of the responsible serine corresponding to Ser-133 of CREB. Because these proteins have a high signal-to-noise ratio and are nontoxic, it has become possible for the first time to carry out long-term measurement of KID−KIX interaction in living cells. Furthermore, we examined the usefulness of the probe proteins for future high-throughput cell-based drug screening and found several herbal extracts that activated CREB.

■

INTRODUCTION The cyclic adenosine monophosphate (cAMP) response element-binding protein (CREB) was first identified as a protein that binds to the upstream sequence of somatostatin (TGACGTCA), which was referred to as the cAMP response element (CRE).1,2 Later, it was found that the phosphorylation at a serine residue in the kinase-inducible domain (KID) of CREB increases the transcriptional activity of CREB following the recruitment of the CREB-binding protein (CBP) or its paralogue p300, a transcription coactivator that has histone acetyltransferase activity.3 We refer to this phosphorylation site (133rd serine in CREB) as S133 in this study. Many studies have suggested the relationship of CREB with synaptic plasticity and memory formation.4−6 Thus, monitoring the phosphorylation of CREB and subsequent KID-KIX interaction during memory formation is important to elucidate the mechanism underlying memory formation. The screening for CREB activators may lead to the development of therapies for memory disorders. Although CREB was originally reported to be phosphorylated by protein kinase A (PKA), it has been revealed that S133 of CREB is phosphorylated not only by PKA but also by many other kinases such as protein kinase C (PKC), calcium calmodulin-dependent protein kinase (CaMK), and mitogenactivated protein kinase kinase (MEK).7 This means that measuring PKA activity is not sufficient for measuring the © 2012 American Chemical Society

CREB activation level. There are some PKA probes in which catalytic, regulatory subunit of PKA and fluorescent or luminescent proteins are used;8−11 however, they detect cAMP level but not CREB phosphorylation and subsequent CREB-CBP/p300 binding. To detect the phosphorylation of CREB specifically, Western blotting and immunohistochemistry using the antibody specific to phosphorylated S133 are widely used in in vitro studies. However, these methods do not enable the detection of the phosphorylation of CREB in living cells. CRE-dependent transcription can be detected by CREluciferase reporter assay in living cells. However, this assay not only detects CREB-dependent transcription, but is also influenced by other transcription-regulating proteins that bind to CRE, such as AP-1.12 A detection method for CREB phosphorylation using fluorescence resonance energy transfer (FRET) has been reported.13 This method was designed to detect changes in emission properties induced by the interaction of two distinct fluorescent proteins induced by the KID−KIX interaction. Under the conditions in which fluorescent proteins can be excited by light and autofluorescence noise is very low, this method enables the measurement of phosphorylation level Received: September 7, 2011 Revised: December 28, 2011 Published: April 16, 2012 923

dx.doi.org/10.1021/bc200491j | Bioconjugate Chem. 2012, 23, 923−932

Bioconjugate Chemistry

Article

Figure 1. Schematic diagram of split luciferase for detecting interaction between KID and KIX. (A) Phosphorylation-dependent interaction between KID and KIX domains. When the S133 residue of CREB is phosphorylated (represented as ‘P’ in the figure), the KID interacts with the KIX domain of CBP. (B) Schema of measurement of the KID−KIX interaction with light emission. Probe proteins consist of the N-terminal and C-terminal domains of luciferase (LucN and LucC) and KID and KIX domains attached through short peptide linkers. Both fusion proteins contain nuclear localization signals (NLS) that allow nuclear translocation. Once S133 of the KID domain is phosphorylated, two probe proteins interact with each other and luciferase activity is restored to emit light.

clones. The SV40 enhancer promoter sequence amplified from pGL4.13 (Promega) was inserted in 5′ entry clones, and then the start codon and nuclear localization signal (DPKKKRKVDPKKKRKV) were inserted using a KOD plus mutagenesis kit (Toyobo). The amino-terminal part (nucleotide numbers 1−1245) of firefly luciferase ORF (LucN), carboxyl-terminal part of firefly luciferase (LucC) sequence corresponding to nucleotide numbers 1246−1650 of the luciferase ORF, nucleotide numbers 417−596 of CREB cDNA (Genbank: BC021649) (KID), and nucleotide numbers 1923-2165 of CBP cDNA (Genbank: BC072594) (KIX) were amplified and inserted into the center and 3′ entry clones. The obtained clones were reacted with a destination vector pDEST R4-R3 using LR clonase plus II (Invitrogen) to generate the expression clones A1, A2, B1, B2, C1, C2, D1, and D2. The linker sequences of these clones were modified using a KOD plus mutagenesis kit (Toyobo). All the expression clones are illustrated in Figure 2A. All the constructs were confirmed by DNA sequencing using ABI3100 (Applied Biosystems). For the CRE-luc assay, pGL4.29 (Promega), and a plasmid which expressed enhanced green fluorescent protein (EGFP) driven by SV40 promoter were used. Cell Culture and DNA Transfection. HEK293T (0.27 × 105 cells/cm2) cells were cultured in DMEM supplemented with 10% FCS at 37 °C in 5% CO2 on 24-well or 96-well dishes (Falcon). The cells were transiently transfected using Lipofectamine 2000 (Invitrogen) in accordance with the manufacturer’s instructions. Cell Viability Assay. HEK293T cells 0, 1, and 2 days after transfection were used for the assay. Cell viability was determined by 3-(4,5-dimethyl-2-thizolyl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) assay in accordance with the manual

using living cells. However, changes in emission ratio are usually small and are principally less quantitative in the FRET method. Furthermore, excitation light is toxic to cells. These disadvantages make long-term measurement of CREB phosphorylation level difficult in living cells. Therefore, a simple and quantitative method of measuring CREB activation in living cells has been desired. Complementation and reconstitution assay of firefly split luciferase has been established in this decade.14−16 Because this method has the advantages of viability of cells and linearity of signal intensity, we attempted to develop a new probe protein that can be used for the detection of CREB activation in real time and for a long period by employing firefly split luciferase technique. On the other hand, the optimization of domain combination and a linker sequence are required to establish practical split luciferase probes that have a sufficient light intensity from living cells.17 In this study, we designed and optimized the probe proteins with a high light intensity depending on phosphorylation at S133 and subsequent KID− KIX interaction. Then, we determined whether long-term observation of KID−KIX interaction is possible in living cells. Furthermore, we examined the usability of our probe proteins in cell-based drug screening using a herbal extract library and found some extracts that activated CREB.

■

EXPERIMENTAL PROCEDURES Plasmid Construction. Plasmids that coded probe proteins were constructed using a MultiSite Gateway Three-Fragment Vector Construction kit (Invitrogen) as described in previous reports.17−19 In accordance with the manufacturer’s instructions, PCR-amplified DNA fragments of interest were inserted by BP reaction in a defined order of 5′, center, and 3′ entry 924

dx.doi.org/10.1021/bc200491j | Bioconjugate Chem. 2012, 23, 923−932

Bioconjugate Chemistry

Article

Figure 2. Development of probe proteins to measure KID−KIX interaction. (A) Schematic structures of probe proteins A1−D6, in which the linker sequences, domain combinations, and domain location were modified. Combinations of probe proteins from groups A and B or groups C and D were coexpressed in HEK293T cells, and the emitted photons were counted for 10 min using an image intensifier/CCD camera. (B) Numbers of photons emitted from all the combinations of probe proteins for 10 min. Data are presented as mean ± SEM, n = 4.

Western Blotting. Cells were harvested and lysed using MPER solution (Thermo Scientific). Proteins separated by SDSPAGE were electrically transferred onto a PVDF membrane (GE Healthcare). The membranes were incubated with goat anti-luciferase (1:1000, Promega), rabbit anti-phospho CREB (1:500, Upstate), rabbit anti-CREB (1:500, Cell Signaling Technology), or rabbit anti-actin (1:1000, Santa Cruz) primary antibody in the Can Get Signal solution (Toyobo) at 4 °C

attached to the MTT cell viability assay kit (Biotium Inc.). After incubation with the tetrazolium salt MTT (1 mg/mL) for 2 h at room temperature, the harvested cells were solubilized with 1 mL of dimethyl sulfoxide. We measured the absorbance at wavelengths of 570 and 630 nm using a spectrophotometer (Gene Quant 1300, GE Healthcare). Sample signal intensity was obtained by subtraction of OD at 630 nm from OD at 570 nm and indicated as a percentage of the control value. 925

dx.doi.org/10.1021/bc200491j | Bioconjugate Chem. 2012, 23, 923−932

Bioconjugate Chemistry

Article

Figure 3. Characterization of KID−KIX interaction probe proteins expressed in cells. (A) Effect of probe protein expression on cell proliferation. HEK293T cells were transfected with the plasmids encoding probe proteins (C6 + D5) or wild-type luciferase (Luc2) using lipofectamine 2000 (lipo). Cell viability was measured by the MTT method 1−2 days after transfection. Data are presented as mean ± SEM, n = 4. (B) Effect of probe protein expression on CRE-dependent gene expression. HEK293T cells were transfected with CRE-luc reporter plasmid with the expression plasmid of C6, D5, or EGFP. The relative increase in light intensity in response to forskolin treatment was measured for 12 h. Dotted lines represent control experiments without forskolin treatment. (C) Effect of point mutation at phosphorylation site of probe proteins on light emission. The probe proteins (C6 + D5) and point-mutated C6 probe C6 (S133A) with D5 were expressed in HEK293T cells. Light intensity was measured 1 h after forskolin treatment. (D) Western blot analysis of probe proteins. The expression levels of probe proteins in experiment (C) were analyzed using the anti-luc antibody. The bands corresponding to the expected sizes of D5 and C6 are indicated by arrows. The locations of protein size markers (kDa) are indicated on the right side. Data are presented as mean ± SEM, n = 4 in (A, B, and C).

overnight. The membranes were reacted with secondary rabbit anti-goat IgG HRP-conjugated and goat anti-rabbit IgG HRPconjugated antibodies (1:2000, Bio Rad) and visualized using an ECL plus Western blotting reagent (GE Healthcare). Chemiluminescence signals were detected using the LAS-4000 system (Fujifilm). Measuring Light Intensity from Cultured Cells. The Aequoria-2D/C8600 system (Hamamatsu Photonics) and Wasabi software (U9304−02) were used for data acquisition and analysis, as described in a previous report.17 Briefly, cultured cells were taken out from a CO2 incubator 48 h after transfection. The culture medium was replaced with L15 medium containing 0.5 mM luciferin EF (Promega). The culture plate was placed in a dark box for 30 min. Then, the number of photons emitted from one culture dish well was determined using a CCD camera with an image intensifier attached to the top of the dark box. Light intensity was measured every 5 min continuously for 12 h after the stimulation. The numbers of emitted photons from wells were normalized to those from nontreated wells (control) and were plotted as relative intensity. All the measurement of light intensity was done at room temperature. Herbal Extracts. The library of herbal extract was a kind gift from the Institute of Natural Medicine, University of Toyama. Each herb (45 g in 900 mL water) was boiled for 70 min and filtered. Freeze−dried extracts were resuspended in

water at 10 mg/mL. The cells in the 96-well dish were treated with the extracts (500 μg/mL). Statistical Analyses. Data are presented as mean ± standard error of the mean (SEM). In Figures 3A,C and 5A, a two-tailed Student’s t test was used to determine statistical significance. Two-factor repeated measure ANOVA was performed for Figure 3B.

■

RESULTS Design of Split Luciferase Probes for Detecting KID− KIX Interaction. KID of CREB interacts with KIX domain of CBP/p3003. Phosphorylation of KID at S133 of the CREB residue triggers interaction with KIX (Figure 1A), and the KID−KIX interaction activates the subsequent transcription machinery.20 On the basis of these features, we designed split luciferase probe proteins for detecting KID−KIX interaction (Figure 1B). The LucN (1−415 aa) and LucC (416−550 aa) domains of firefly (Photinus pyralis) luciferase were attached to the KID and KIX domains with short linker peptides. Each probe protein contains a nuclear localization signal. The interaction between KID and KIX leads to complementation of LucN and LucC, restoration of luciferase activity, and subsequent photon emission. Development of Probe Proteins to Measure KID−KIX Interaction Level. Because the linker sequence and domain combination markedly affect the ability of photon emission from split luciferase probes, as described in a previous report,17 926

dx.doi.org/10.1021/bc200491j | Bioconjugate Chem. 2012, 23, 923−932

Bioconjugate Chemistry

Article

Figure 4. Long-term measurement of KID−KIX interaction in living cells. (A) Change in light intensity of cells expressing probe proteins after forskolin treatment. HEK293T cells were transfected with C6+D5 probe proteins. Forskolin was added to the culture medium at time 0, and the relative change in light intensity was measured for 12 h. (B, C, D) Same measurements as in (A) were performed using the cells transfected with Luc2, C6 (S133A)+D5, and LucN+LucC. Data in A, B, C, and D are presented as mean ± SEM, n = 4. (E) Western blotting of endogenous CREB, phosphorylated CREB (pCREB), and actin in HEK293T cells after forskolin treatment. HEK293T cells treated with forskolin were analyzed by Western blotting to detect the phosphorylation of endogenous CREB.

we constructed several plasmids that expressed fusion proteins with modifications in the linker sequence and domain combination, as shown in Figure 2A. A MultiSite Gateway Three-Fragment Vector Construction kit was employed to obtain A1, A2, B1, B2, C1, C2, D1, and D2 probe expression plasmids. The linkers of these plasmids were modified by the inverse PCR method to construct the other plasmids. Each combination of probes from groups A and B or groups C and D was expressed in HEK293T cells, and the emitted photons were counted for 10 min using an image intensifier/charge-coupled

device (CCD) camera. Among all combinations of probe proteins, we found that the cells expressed the combination C6+D5 emitted more photons than any other combination of probes (Figure 2B) including control probes LucN and LucC (Supporting Information Figure S1A). The expression levels of probe proteins (C1, C6, D1, and D5) were not significantly different (Figure S1B). The probe protein combination C6+D5 was used in the following experiments. Characterization of KID−KIX Interaction Probe Proteins Expressed in Cells. Since split luciferase probes are 927

dx.doi.org/10.1021/bc200491j | Bioconjugate Chem. 2012, 23, 923−932

Bioconjugate Chemistry

Article

intensity from each culture dish was measured continuously. We found that the temporal increase in light intensity from the cells expressing C6+D5 (Figure 4A) peaked at 2 h, whereas other cells expressing Luc2 (Figure 4B), C6 (S133A)+D5 (Figure 4C), and LucN+LucC (Figure 4D) did not show any increase in light intensity in response to forskolin. We examined whether the phosphorylation of endogenous CREB level increases in response to forskolin. HEK293T cells were treated with 10 μM forskolin and analyzed by Western blotting. Phosphorylated CREB level increased within 2 min and peaked at 5 min; this peaking remained for 6 h after the treatment. The total amount of CREB was unchanged during the treatment with forskolin. These findings indicate that our probes enable long-term evaluation of KID−KIX interaction in living cells. We also confirmed that a sufficient amount of luciferin for the activity of luciferase remained in the culture medium after 12 h measurement (Figure S3A). This indicates that light intensity was not decreased because of luciferin depletion. We also observed that the expression level of Luc2 driven by SV40 promoter was unchanged during measurement (Figure S3B), which indicated changes in light intensity in Figure 4, was not due to the expression level of proteins. Effect of Kinase Inhibitors on Light Intensity. The photon emission shown in Figure 4A is considered to be a result of PKA activation, because forskolin increases cAMP concentration. We examined whether our probe proteins are capable of evaluating the level of CREB phosphorylation when two or more kinases induce the phosphorylation in living cells. HEK293T cells expressing the C6+D5 probe proteins were treated with H89 (a PKA inhibitor), KN93 (a CaMK II inhibitor), calphostin C (a PKC inhibitor), and a mixture of three inhibitors for 8 h. After the treatment, we measured light intensity and found H89 and KN93 but not calphostin C significantly decreased light intensity. An additional decrease in light intensity was observed when the cells were treated with a mixture of inhibitors (Figure 5A). To determine whether the change in light intensity correlated with endogenous CREB phosphorylation, cells treated with inhibitors were analyzed by Western blotting. We found that H89, KN93, and the mixture of inhibitors significantly attenuated pCREB signals (Figure 5B). On the other hand, cells treated with calphostin C showed only slight attenuation of pCREB signals. In the control experiments, we found that kinase inhibitor treatment resulted in slight decreases in MTT signals of HEK293T cells when KN93 and inhibitor mixture were added (Figure S4A). Since these changes in MTT signal are smaller than the changes in light intensity shown in Figure 5A, a major cause of reduced light intensity in Figure 5A was not the toxic effects of inhibitors. We also found that forskolin treatment together with kinase inhibitors canceled the forskolin-induced upregulation of light intensity (Figure S4B). These findings show that CREB is phosphorylated by mainly PKA and CaMKII and not by PKC in nontreated HEK293T cells, and our probe proteins are useful for evaluating the level of CREB phosphorylation when other kinases contribute to phosphorylation simultaneously. Screening for CREB Activators Using Herbal Extract Library. Some herbal extracts or molecules purified from herbal extracts show the activity to phosphorylate CREB.21−24 To determine the capability of our probe proteins to screen for molecules that induce CREB phosphorylation, we used a library consisting of 120 herbal extracts. Each extract was added to a culture dish of cells expressing probe proteins, and the light intensity from each dish was measured at times 0 and 4 h. Most

exogenous proteins, they may interfere with some biological processes in cells. In order to test these interferences of our probe proteins, we determined whether the probes affect on HEK293T proliferation rate. HEK293T cells were transfected with probe proteins or wild-type luciferase (Luc) expression plasmids, and cell viability was measured after transfection for 2 days. We did not find any difference in the intensity of the 3(4,5-dimethyl-2-thizolyl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) signal between the probe-protein- or Luc-expressing cells and the cells treated with transfection reagent alone (Figure 3A). This finding demonstrates that the probe proteins do not affect cell proliferation. Next, we examined whether the expression of probe proteins interferes with CRE-mediated transcription in HEK293T cells. Expression plasmid of C6, D5, or EGFP was co-transfected with CRE-luc plasmid that could indicate CRE-mediated transcription by photon emission. Two days after transfection, we found that at least 90% of HEK293T cells expressed EGFP (Figure S2). Under this condition, cells were treated with 10 μM forskolin, a compound that upregulates cAMP production followed by PKA activation and CRE-dependent transcription, and light intensity was measured for 12 h. All the CRE-lucexpressing cells showed an increased light intensity following forskolin treatment. We did not find a significant inhibition of upregulation of light emission when each probe protein (C6 or D5) was coexpressed with CRE-luc, as compared with EGFP coexpression with CRE-luc (Figure 3B). Furthermore, no increase in light intensity was observed without forskolin treatment. These findings indicate that probe proteins do not have any effect on CRE-mediated transcription. We determined whether forskolin enhanced the photon emission from C6+D5 probe proteins in HEK293T cells. We found that 1 h incubation with 10 μM forskolin markedly enhanced the photon emission from the cells expressing C6+D5 probe proteins (Figure 3C). Next, we examined whether the point mutation at the phosphorylation site diminished the response to forskolin. S133 of C6 probe was replaced with alanine, which was termed C6 (S133A), and was expressed in HEK293T cells with the D5 probe. We found that HEK293T cells expressing C6(S133A)+D5 probes exhibited no increase in light intensity in response to the forskolin treatment. This result strongly suggests that our probes upregulate the light intensity by binding of phosphorylated KID and KIX domain. Furthermore, the light intensity of the cells expressing C6 (S133A)+D5 proteins was lower than that of C6+D5expressing cells without forskolin treatment. We then compared the amount of expressed C6 and C6 (S133A) probe proteins. The cells expressing the probes were analyzed by Western blotting, and we found no significant difference in the intensity of band signals between C6 and C6 (S133A) (Figure 3D). These findings indicate that the difference in light intensity seen in Figure 3C is not due to the difference in expression level between C6 and C6 (S133A). From the results shown in Figure 3C and D, the forskolin-induced upregulation of light emission from the probes was considered to be induced by phosphorylation at S133 and subsequent KID−KIX interaction. Long-Term Measurement of Light Intensity from KID−KIX Interaction Probes in Living Cells. We then determined whether the long-term measurement of KID−KIX interaction was possible using living cells. HEK293T cells were transfected with C6+D5, Luc2, C6 (S133A)+D5, or LucN +LucC. Then, the cells were treated with 10 μM forskolin for 12 h at room temperature. During the treatment, the light 928

dx.doi.org/10.1021/bc200491j | Bioconjugate Chem. 2012, 23, 923−932

Bioconjugate Chemistry

Article

sequence modulation (C1+D1). Because the expression levels of C1, D1 probe proteins and C6,D5 probe proteins in HEK293T cells were not significantly different (Figure S1B), this increase in light intensity is probably caused by the optimal interaction between LucN and LucC. As the (GGGGS)3 linker26 worked well in other split luciferase probes,17 using flexible peptides may be important for the development of split luciferase. We did not find any interference with normal cellular activities such as proliferation and CRE-dependent transcription caused by probe protein expression (Figure 3). These findings confirm that our probe proteins do not have any harmful effects on normal cellular activities. Characterization of Developed Probe Proteins. HEK293T cells expressing probe proteins showed an increased light intensity when treated with forskolin (Figure 3C). Because forskolin activates protein kinase A by upregulating intracellular cAMP production, this increase in light intensity is considered to be the result of KID−KIX interaction initiated by phosphorylation of probe proteins. Furthermore, we found that the single amino acid mutation (S133A) in the C6 probe protein abolished the increase in light intensity in response to forskolin treatment. Thus, the photon emission of our probe proteins originates from single amino acid phosphorylation at S133 in the C6 probe protein and subsequent interaction with D5. We found that the time course of phosphorylation of endogenous CREB after forskolin treatment (Figure 4E) was correlated with that of light intensity from probes (Figure 4A). Furthermore, levels of phosphorylated CREB match the light intensity from probes (Figure 5). These results indicate that the time course of CREB phosphorylation overlaps well with that of the KID−KIX interaction; thus, the photon emission from probes is considered to reflect the phosphorylation state of CREB. The level of phosphorylated CREB is considered to be low in nonstimulated neurons; however, we found photon emission in HEK293T cells expressing probe proteins under nonstimulated condition. It is assumed that the photon emission is caused by constitutive phosphorylation of some probe proteins in HEK293T cells. Consistent with this idea, endogenous CREB is phosphorylated in nonstimulated HEK293T cells (Figure 4E). A significant increase in light intensity from HEK293T cells expressing probe proteins was observed from 15 min and peaked 1−2 h after forskolin treatment (Figure 4A). This time course is similar to that of chromatin binding of p300 reported previously, peaked 1 h after serum treatment of T98G cells, as determined by chromatin immunoprecipitation (ChIP) assay.27 Because endogenous CREB phosphorylation occurred 2 min after forskolin treatment (Figure 4E), the photon emission from probe proteins was delayed 10−15 min compared with endogenous CREB phosphorylation. On the other hand, CREdependent transcription is upregulated 4 h after forskolin treatment and it occurs much later than photon emission from probe proteins (Figure 3B). This might be because transcriptional activation after CREB-CBP/p300 interaction requires 4 h. We conclude that the photon emission of our probe proteins originates from the phosphorylation of S133 and the timing of photon emission is similar to that of CREB-CBP/p300 interaction. Specificity and Advantages of Our Probe Proteins. There are some probes that detect PKA activity.8−11 These probes are based on the interaction between catalytic and regulatory subunits of PKA. Thus, these probes detect the enhancement of cAMP production and subsequent activation

Figure 5. Effect of kinase inhibitors on light intensity. (A) Changes in light intensity of cells expressing probe proteins after the treatment with protein kinase inhibitors. HEK293T cells transfected with C6+D5 probes were treated with kinase inhibitors (10 μM H89, 10 μM KN93, 100 nM calphostin C, and mixture of all) for 8 h. Control cells were treated with DMSO at a concentration equivalent to that of inhibitor mixture. Data are presented as relative light intensity (mean ± SEM, n = 4). (B) Western blotting of endogenous CREB and phosphorylated CREB (pCREB) after the treatment with protein kinase inhibitors. Inhibitor-treated cells are analyzed by Western blotting using antipCREB, anti-CREB, and anti-actin antibodies.

of the cells treated with an extract and control cells did not show an increased light intensity. We chose 8 extracts that increased light intensity from among the 120 extracts, whose ID numbers are indicated on the bars, and further analysis was carried out (Figure 6A). The cells treated with these 8 extracts were analyzed by Western blotting using anti-pCREB and -CREB antibodies. Seven extracts (Chrysanthemi flos, Schisandrae f ructus, Cimicif ugae rhizoma, Caryophylli f los, Moutan cortex, Viticis f ructus, and Myrrha) showed increased pCREB signal, whereas one extract (Aurantii nobilis pericarpium) did not show an increased signal intensity. Three extracts (Chrysanthemi flos, Cimicifugae rhizome, and Caryophylli flos) increased the expression level of the CREB protein. Because the results of photon measurement were consistent with those of Western blotting with high probability, our probe proteins are therefore useful for screening for molecules that affect CREB phosphorylation.

■

DISCUSSION Development of Split Luciferase Probes for Detection of KID−KIX Interaction. We developed probe proteins for detection of KID−KIX interaction in living cells (Figure 2). On the basis of the reported optimized cleavage site of firefly luciferase,16,17,25 we constructed various fusion probe proteins. Light intensity from probe proteins (C6+D5) was more than 100-fold higher than that of probe proteins without linker 929

dx.doi.org/10.1021/bc200491j | Bioconjugate Chem. 2012, 23, 923−932

Bioconjugate Chemistry

Article

Figure 6. Screening for CREB activators using herbal extract library. (A) Changes in light intensity of cells expressing probe proteins after treatment with herbal extracts. HEK293T cells expressing C6+D5 probe proteins were treated with oriental herbal extracts for 4 h. Each bar represents changes in light intensity induced by each extract. Black bars are control experiments without extract treatment. (B) Western blotting of endogenous CREB, pCREB, and actin after treatment with herbal extracts. Cells with increased light intensity numbered in (A) were analyzed 4 h after the treatment by Western blotting using anti-pCREB, anti-CREB, and anti-actin antibodies.

priate for measuring CREB activity for a long time in living cells. On the other hand, our split luciferase probe proteins have little noise and no phototoxicity because no excitation light is required. These advantages enable us to carry out longterm measurement of KID−KIX interaction in living cells (Figure 4A). Screening Herbal Extracts That Induce CREB Activation. In this report, we screened and analyzed the extracts that increased the light intensity of our probe proteins. Because CREB phosphorylation in neurons is an important event for plastic changes in synaptic transmission, the extracts that activate CREB are expected to have the potential as drugs that improve memory and cognition.28,29 On the other hand, CREB phosphorylation is considered to be important in oncogenesis and some tumor cells show overexpressed or constitutively activated CREB.30,31 Hence, screening for molecules that

of PKA. Although CREB phosphorylation was considered to be induced mainly by PKA, it is widely accepted that many kinases phosphorylate S133 of CREB and increase CREB transcriptional activity depending on the cell type and stimulus used.7 This means that PKA-detecting probes are not sufficient for measuring CREB activity. On the other hand, our probes are based on KID phosphorylation and subsequent KID−KIX interaction. Even when KID is phosphorylated by kinases other than PKA, our probes can emit light. Hence, we conclude that our probes are useful for the specific detection of CREB activation. As shown in Figure 5, our probe proteins can detect KID−KIX interaction induced by CaMKII. A FRET probe that detects CREB phosphorylation has been reported.13 The FRET probe is capable of detecting CREB activation with a high spatiotemporal resolution. However, methods based on FRET have the limitations of autofluorescence noise and phototoxicity, which makes them inappro930

dx.doi.org/10.1021/bc200491j | Bioconjugate Chem. 2012, 23, 923−932

Bioconjugate Chemistry

■

decrease light intensity from the probe proteins may contribute to the search for the antitumor compounds. We analyzed 8 extracts that increased the light intensity from the probe proteins. Among them, some extracts (Schisandrae fructus, Moutan cortex, Viticis f ructus, and Myrrha) induced CREB phosphorylation, which is probably caused by the activation of kinases induced by some constituents in the extracts. Other extracts (Chrysanthemi flos, Cimicifugae rhizoma, and Caryophylli flos) induced not only the phosphorylation but also the expression of CREB. In the case of these extracts, the increase in kinase activity and CREB expression level occurred simultaneously. This might be due to the upregulation of transcriptional activation induced by some extract constituents. As an increased expression level of CREB in leukemia cells has been reported,32 the CREB upregulation seen in leukemia cells and HEK293T cells treated with these herbal extracts may share the same signal transduction pathway. Because each herbal extract consists of many kinds of molecules, we presently cannot specify which constituent is responsible for CREB activation. Some studies suggest that candidates responsible molecules. Schizandrin33 and Gomisin A,34 compounds purified from Schisandrae fructus, are reported to reverse the scopolamine-induced memory impairment. Paeonol from Moutan cortex is reported to improve the learning ability of mice with Dgalactose-induced aging in the Morris water maze test.35 Because memory formation is closely related to CREB activation, it is of interest to investigate whether these molecules induce improvements of memory via CREB phosphorylation. Interestingly, the extract of Aurantii nobilis pericarpium that increased the light intensity of probe proteins did not show an increased CREB phosphorylation level in Western blotting. Although we do not have any data that can explain why this pseudopositive reaction occurred, some constituents in Aurantii nobilis pericarpium extract may affect LucN and LucC directly and increase light intensity. Taken together, we conclude that our probe protein is useful for screening for molecules that affect CREB phosphorylation with high probability, simplicity in method, and good signal-to-noise ratio compared with the FRET technique. In addition, this cellbased approach can exclude compounds that are toxic to cells and identify the molecules that permeate across the cell membrane and activate CREB. Furthermore, high-throughput screening will be possible if sample application and photon detection are automated. For the future application of our probe proteins, we expect to apply them in the monitoring of CREB activation in the brain of living animals. Imaging the activity of a particular protein in the brain of a living animal has been impossible using fluorescence-based methods such as FRET because animal tissues have a marked autofluorescence and a change in FRET signal intensity is undetectable. Because our probes do not require excitation light, we do not have to consider the problem of autofluorescence. Furthermore, the luciferase substrate luciferin can reach the brain even when administered by intraperitoneal injection.36 Using transgenic mice that express the probe proteins, behavioral analysis in parallel with the observation of CREB activation in the brain will be possible. This experiment will elucidate when and where CREB activation is required for memory formation.

Article

ASSOCIATED CONTENT

* Supporting Information S

Additional figures as described. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*Tel: +81-76-434-7230, Fax: +81-76-434-5015, E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by PRESTO of the Japan Science and Technology Corporation (JST, No. 2602), Japan Society for the Promotion of Science (JSPS, No. 2070331), and Grant-inAid for the Cooperative Research Project from Joint Usage/ Research Center (Joint Usage/Research Center for ScienceBased Natural Medicine), Institute of Natural Medicine, University of Toyama in 2010.

■

REFERENCES

(1) Montminy, M. R., Sevarino, K. A., Wagner, J. A., Mandel, G., and Goodman, R. H. (1986) Identification of a cyclic-AMP-responsive element within the rat somatostatin gene. Proc. Natl. Acad. Sci. U. S. A. 83, 6682−6. (2) Montminy, M. R., and Bilezikjian, L. M. (1987) Binding of a nuclear protein to the cyclic-AMP response element of the somatostatin gene. Nature 328, 175−8. (3) Parker, D., Ferreri, K., Nakajima, T., LaMorte, V. J., Evans, R., Koerber, S. C., Hoeger, C., and Montminy, M. R. (1996) Phosphorylation of CREB at Ser-133 induces complex formation with CREB-binding protein via a direct mechanism. Mol. Cell. Biol. 16, 694−703. (4) Benito, E., and Barco, A. (2010) CREB’s control of intrinsic and synaptic plasticity: implications for CREB-dependent memory models. Trend. Neurosci. 33, 230−40. (5) Bourtchuladze, R., Frenguelli, B., Blendy, J., Cioffi, D., Schutz, G., and Silva, A. J. (1994) Deficient long-term memory in mice with a targeted mutation of the cAMP-responsive element-binding protein. Cell 79, 59−68. (6) Lonze, B. E., and Ginty, D. D. (2002) Function and regulation of CREB family transcription factors in the nervous system. Neuron 35, 605−23. (7) Johannessen, M., Delghandi, M. P., and Moens, U. (2004) What turns CREB on? Cell. Signal. 16, 1211−27. (8) Adams, S. R., Harootunian, A. T., Buechler, Y. J., Taylor, S. S., and Tsien, R. Y. (1991) Fluorescence ratio imaging of cyclic AMP in single cells. Nature 349, 694−7. (9) Bagorda, A., Das, S., Rericha, E. C., Chen, D., Davidson, J., and Parent, C. A. (2009) Real-time measurements of cAMP production in live Dictyostelium cells. J. Cell Sci. 122, 3907−14. (10) Zaccolo, M., De Giorgi, F., Cho, C. Y., Feng, L., Knapp, T., Negulescu, P. A., Taylor, S. S., Tsien, R. Y., and Pozzan, T. (2000) A genetically encoded, fluorescent indicator for cyclic AMP in living cells. Nat. Cell Biol. 2, 25−9. (11) Stefan, E., Aquin, S., Berger, N., Landry, C. R., Nyfeler, B., Bouvier, M., and Michnick, S. W. (2007) Quantification of dynamic protein complexes using Renilla luciferase fragment complementation applied to protein kinase A activities in vivo. Proc. Natl. Acad. Sci. U. S. A. 104, 16916−21. (12) Hai, T., and Curran, T. (1991) Cross-family dimerization of transcription factors Fos/Jun and ATF/CREB alters DNA binding specificity. Proc. Natl. Acad. Sci. U. S. A. 88, 3720−4. 931

dx.doi.org/10.1021/bc200491j | Bioconjugate Chem. 2012, 23, 923−932

Bioconjugate Chemistry

Article

(13) Friedrich, M. W., Aramuni, G., Mank, M., Mackinnon, J. A. G., and Griesbeck, O. (2010) Imaging CREB activation in living cells. J. Biol. Chem. 285, 23285−95. (14) 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, 15608−13. (15) Ozawa, T., Kaihara, A., Sato, M., Tachihara, K., and Umezawa, Y. (2001) Split luciferase as an optical probe for detecting proteinprotein interactions in mammalian cells based on protein splicing. Anal. Chem. 73, 2516−21. (16) Luker, K. E., Smith, M. C., Luker, G. D., Gammon, S. T., Piwnica-Worms, H., and Piwnica-Worms, D. (2004) Kinetics of regulated protein-protein interactions revealed with firefly luciferase complementation imaging in cells and living animals. Proc. Natl. Acad. Sci. U. S. A. 101, 12288−93. (17) Ishimoto, T., Ozawa, T., and Mori, H. (2011) Real-time monitoring of actin polymerization in living cells using split luciferase. Bioconjugate Chem. 22, 1136−44. (18) Hartley, J. L., Temple, G. F., and Brasch, M. A. (2000) DNA cloning using in vitro site-specific recombination. Genome Res. 10, 1788−95. (19) Sasaki, Y., Sone, T., Yoshida, S., Yahata, K., Hotta, J., Chesnut, J. D., Honda, T., and Imamoto, F. (2004) Evidence for high specificity and efficiency of multiple recombination signals in mixed DNA cloning by the Multisite Gateway system. J. Biotechnol. 107, 233−43. (20) Vo, N., and Goodman, R. H. (2001) CREB-binding protein and p300 in transcriptional regulation. J. Biol. Chem. 276, 13505−8. (21) Cheung, W. M., Hui, W. S., Chu, P. W., Chiu, S. W., and Ip, N. Y. (2000) Ganoderma extract activates MAP kinases and induces the neuronal differentiation of rat pheochromocytoma PC12 cells. FEBS Lett. 486, 291−6. (22) Lin, Y. L., Lee, Y. C., Huang, C. L., Lai, W. L., Lin, Y. R., and Huang, N. K. (2007) Ligusticum chuanxiong prevents rat pheochromocytoma cells from serum deprivation-induced apoptosis through a protein kinase A-dependent pathway. J. Ethnopharmacol. 109, 428−34. (23) Kim, D. H., Kim, S., Jeon, S. J., Son, K. H., Lee, S., Yoon, B. H., Cheong, J. H., Ko, K. H., and Ryu, J. H. (2009) Tanshinone I enhances learning and memory, and ameliorates memory impairment in mice via the extracellular signal-regulated kinase signalling pathway. Br. J. Pharmacol. 158, 1131−42. (24) Hou, Y., Aboukhatwa, M. A., Lei, D. L., Manaye, K., Khan, I., and Luo, Y. (2010) Anti-depressant natural flavonols modulate BDNF and beta amyloid in neurons and hippocampus of double TgAD mice. Neuropharmacology 58, 911−20. (25) 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, 484−92. (26) Huston, J. S., Levinson, D., Mudgett-Hunter, M., Tai, M. S., Novotny, J., Margolies, M. N., Ridge, R. J., Bruccoleri, R. E., Haber, E., Crea, R., et al. (1988) Protein engineering of antibody binding sites: recovery of specific activity in an anti-digoxin single-chain Fv analogue produced in Escherichia coli. Proc. Natl. Acad. Sci. U. S. A. 85, 5879− 83. (27) Ramos, Y. F., Hestand, M. S., Verlaan, M., Krabbendam, E., Ariyurek, Y., van Galen, M., van Dam, H., van Ommen, G. J., den Dunnen, J. T., Zantema, A., and Hoen, P. A. (2010) Genome-wide assessment of differential roles for p300 and CBP in transcription regulation. Nucleic Acids Res. 38, 5396−408. (28) Glannon, W. (2006) Psychopharmacology and memory. J. Med. Ethics 32, 74−8. (29) Scott, R., Bourtchuladze, R., Gossweiler, S., Dubnau, J., and Tully, T. (2002) CREB and the discovery of cognitive enhancers. J. Mol. Neurosci. 19, 171−7. (30) Siu, Y. T., and Jin, D. Y. (2007) CREB--a real culprit in oncogenesis. FEBS J. 274, 3224−32.

(31) Shankar, D. B., Cheng, J. C., and Sakamoto, K. M. (2005) Role of cyclic AMP response element binding protein in human leukemias. Cancer 104, 1819−24. (32) Crans-Vargas, H. N., Landaw, E. M., Bhatia, S., Sandusky, G., Moore, T. B., and Sakamoto, K. M. (2002) Expression of cyclic adenosine monophosphate response-element binding protein in acute leukemia. Blood 99, 2617−9. (33) Egashira, N., Kurauchi, K., Iwasaki, K., Mishima, K., Orito, K., Oishi, R., and Fujiwara, M. (2008) Schizandrin reverses memory impairment in rats. Phytother. Res. 22, 49−52. (34) Kim, D. H., Hung, T. M., Bae, K. H., Jung, J. W., Lee, S., Yoon, B. H., Cheong, J. H., Ko, K. H., and Ryu, J. H. (2006) Gomisin A improves scopolamine-induced memory impairment in mice. Eur. J. Pharmacol. 542, 129−35. (35) Zhong, S. Z., Ge, Q. H., Qu, R., Li, Q., and Ma, S. P. (2009) Paeonol attenuates neurotoxicity and ameliorates cognitive impairment induced by d-galactose in ICR mice. J. Neurol. Sci. 277, 58−64. (36) Izumi, H., Ishimoto, T., Yamamoto, H., Nishijo, H., and Mori, H. (2011) Bioluminescence imaging of Arc expression enables detection of activity-dependent and plastic changes in the visual cortex of adult mice. Brain Struct. Funct. 216, 91−104.

932

dx.doi.org/10.1021/bc200491j | Bioconjugate Chem. 2012, 23, 923−932