Screening of novel nuclear receptor agonists by a convenient reporter gene assay system using green fluorescent protein derivatives.
Nuclear receptors represent a very good family of protein targets for the prevention and treatment of diverse diseases. In this study, we screened natural compounds and their derivatives, and discovered ligands for the retinoic acid receptors (RARs) and the farnesoid X receptor (FXR). In the reporter assay systems of nuclear receptors presented here, two fluorescent proteins, enhanced yellow fluorescent protein (EYFP) and enhanced cyan fluorescent protein (ECFP), were used for detection of a ligand-based induction and as an internal control, respectively. By optimizing the conditions (e.g., of hormone response elements and promoter genes for reporter plasmids), we established a battery of assay systems for ligands of RARs, retinoid X receptor (RXR) and FXR. The screening using the reporter assay system can be carried out without the addition of co-factors or substrates. As a result of screening of more than 140 compounds, several compounds were detected which activate RARs and/or FXR. Caffeic acid phenylethyl ester (CAPE), known as a component of propolis from honeybee hives, and other derivatives of caffeic acid up-regulated the expression of reporter gene for RARs. Grifolin and ginkgolic acids, which are non-steroidal skeleton compounds purified from mushroom or ginkgo leaves, up-regulated the expression of the reporter gene for FXR.
[c] 2005 Elsevier GmbH. All rights reserved.
Keywords: FXR; RAR; Reporter assay; Fluorescence; GFP; Caffeic acid; Ginkgolic acid; Grifolin
Nuclear hormone receptors are ligand-activated transcription factors that are involved in a variety of physiological, developmental, and toxicological processes. The nuclear hormone receptor superfamily includes receptors for thyroid and steroid hormones, retinoids and vitamin D, as well as receptors for unknown ligands. These receptors share a highly conserved DNA-binding domain and a discrete ligand-binding domain, and bind to hormone response elements (HREs) on the DNA during the formation of homodimers, heterodimers, or monomers. This ligand binding to nuclear receptors leads to conformational change of these receptors and the recruitment of coactivator complexes, resulting in transcriptional activation (Khorasanizadeh and Rastinejad, 2001). Their ligand-dependent activity makes nuclear receptors good pharmacological targets.
Nuclear receptors form a superfamily of phylogenetically related proteins encoded by 48 genes in the human genome. Three isotypes of retinoic acid receptors (RARs: RAR[alpha], RAR[beta] and RAR[gamma]) are receptors for retinoids such as all-trans-retinoic acid (ATRA) (Petkovich et al., 1987; Brand et al., 1988; Krust et al., 1989). RAR[alpha] is associated with differentiation therapy for human acute promyelocytic leukemia (Hansen et al., 2000). RAR[beta] plays a central role in limiting the growth of different cell types (reviewed in Hansen et al., 2000), and is thus a possible target for the treatment of breast and other cancers. RAR[gamma] is also primarily expressed in the skin and is involved in skin photoaging and carcinogenesis, and in skin diseases such as psoriasis and acne (Fisher et al., 1996).
The farnesoid X receptor (FXR) is a receptor for bile acids such as chenodeoxycholic acid (CDCA), deoxycholic acid, cholic acid, and their conjugates. Bile acids are synthesized in the liver and secreted into the intestine, where their physical properties facilitate the absorption of fats and vitamins through micelle formation. Cholesterol disposal from the liver is also dependent on the bile acid composition of the secreted bile. Bile acids bind to FXR to activate and regulate the transcription of FXR target genes. FXR controls the expression of critical genes in bile acid and cholesterol homeostasis (Makishima et al., 1999; Parks et al., 1999; Wang et al., 1999). FXR-null mice show elevated serum cholesterol and triglyceride levels (Sinal et al., 2000), and an FXR agonist has been shown to reduce serum triglyceride levels (Maloney et al., 2000). FXR is thus an attractive pharmacological target for the treatment of hyperlipidemia. Moreover, an FXR agonist has been reported to confer hepatoprotection in a rat model of cholestasis (Liu et al., 2003).
The retinoid X receptor (RXR) is a common heterodimeric partner for many receptors, including thyroid hormone receptor (TR), RAR, vitamin [D.sub.3] receptor (VDR), peroxisome proliferator-activated receptor (PPAR), liver X receptor (LXR), and FXR, in addition to functioning as a receptor for 9-cis-retinoic acid (9CRA) during formation of a homodimer.
To determine ligands for these nuclear receptors, we developed a reporter assay system using GFP derivatives. To study the promoter and enhancer control of gene expression, firefly luciferase is widely used as a reporter protein because it has high sensitivity and a broad linear range. In the commonly used reporter assay, [beta]-galactosidase, a well-characterized bacterial enzyme, or renilla luciferase is usually used in conjunction with firefly luciferase to normalize the transfection efficiency of the reporter gene (Sherf et al., 1996; Martin et al., 1996). In such cases, the activity of the two reporter proteins must be measured in different ways (e.g., absorptiometry and luminescence photometry) or by using two substrates. In the reporter assay presented here, we used two species derived from green fluorescent protein (GFP), one (enhanced yellow fluorescent protein, (EYFP)) to measure the promotion and enhancement of gene expression, and the other (enhanced cyan fluorescent protein, (ECFP)) to normalize the transfection, and were thus able to measure the fluorescent protein signals simultaneously without any co-factor or substrates. As a result of screening of more than 140 compounds, it was found that several compounds activate RARs and/or FXR.
Materials and methods
Chenodeoxycholic acid was purchased from Sigma-Aldrich (St. Louis, MI, USA), and ATRA and 9CRA were from Wako (Osaka, Japan). Ginkgolic acid 17:1, 15:0, and 13:0 were purchased from Nagara Science (Gifu, Japan).
Purification and synthesis of test compounds
Ginkgolic acid 15:1 was purified from Ginkgo biloba L. var. diptera according to Morimoto et al. (1968). 2-Methyl ginkgolic acid methyl ester was prepared by methylation of the ginkgolic acid with methyl iodide and [K.sub.2]C[O.sub.3] (Paul and Yeddanapalli, 1956; Begum et al., 2002). Grifolin was purified from Albatrellus confluens and Albatrellus ovinus (Ishii et al., 1988; Nukata et al., 2002). We isolated bazzanenyl caffeate from the liverwort Bazzania fauriana (Toyota and Asakawa, 1988). We synthesized caffeic acid phenethyl ester (CAPE), farnesyl caffeate and geranyl caffeate for acquirement in quantity. The synthesis of CAPE by coupling reactions of caffeic acid and [beta]-phenylethyl bromide was reported by Hashimoto et al. (1988), and the details of the synthesis of farnesyl and geranyl caffeates are described below. The purity of the compounds for the bioactivation test was shown to be over 95% by [.sup.1.H] and [.sup.13.C] NMR spectra.
Synthesis of farnesyl caffeate
Twenty-five percent NaOH (2.5 ml) was added to a solution of caffeic acid (3,4-dihydroxycinnamic acid) (2.10 g) in HMPA (hexamethylphosphoric triamide) (150 ml), and the mixture was stirred for 1 h under [N.sub.2] at room temperature. A solution of farnesyl bromide (4.98 g) in HMPA (20 ml) was added dropwise for 10 min to the reaction mixture. The reaction mixture was stirred for 24 h at room temperature, and poured in ice cold [H.sub.2]O (300 ml). The organic layer, which was extracted with [Et.sub.2]O (200 ml x 2), was washed with brine (300 ml), dried (MgS[O.sub.4]) and evaporated under reduced pressure to an oil (6.75 g). The oil was chromatographed on silica gel (200 g) with a gradient solvent system of CH[Cl.sub.3]-EtOAc, increasing the amount of 2% portions EtOAc stepwise to give 32 fractions. Farnesyl caffeate (1.435 g; Y. 43.2%) was obtained from 10% EtOAc-n-hexane eluate (Fr. 12-18) as a pure white powder. Caffeic acid (1.025 g; Y. 48.8%), the starting material, was recovered from 20% EtOAc-n-hexane eluate (Fr. 25-31).
Farnesyl caffeate: EI-MS: m/z 384 ([M.sup.+],5%), 315, 204, 180, 163 (100%), 135, 93, 69; HR-MS: m/z 384.2307, [C.sub.24][H.sub.32][O.sub.4] requires 384.2300; anal. calcd. for [C.sub.24][H.sub.32][O.sub.4]: C, 74.97; H, 8.39. Found: C, 74.85; H, 8.30; FT-IR (KBr) [cm.sup.-1]: 3480 (OH), 3301 (OH), 1678 (C = O), 1600, 1278, 1183; UV (EtOH) [[lambda].sub.max] nm (log [epsilon]): 333 (4.15), 303 (4.00), 248 (3.90), 220 (4.03); [.sup.1.H] NMR (acetone-[d.sub.6]): [delta]1.56 (3H, s, C[H.sub.3]), 1.62 (3H, s, C[H.sub.3]), 1.65 (3H, s, C[H.sub.3]), 1.76 (3H, s, C[H.sub.3]), 4.68 (1H, d, J = 7.0 Hz, H-1'), 5.12 (2H, m, H-6' and H-10'), 5.41 (1H, t, J = 7.0 Hz, H-2'), 6.26 (1H, d, J = 15.9 Hz, H-[beta]), 6.87 (1H, d, J = 8.2 Hz, H-5), 7.03 (1H, dd, J = 1.8, 8.2 Hz, H-6), 7.15 (1H, d, J = 1.8 Hz, H-2), 7.53 (1H, d, J = 15.9 Hz, H-[alpha]), 8.26 (1H, br.s, -OH), 8.49 (1H, br.s, -OH); [.sup.13.C] NMR ((acetone-[d.sub.6]): [delta]16.1 (q, C[H.sub.3]), 16.4 (q, C[H.sub.3]), 17.7 (q, C[H.sub.3]), 25.8 (q, C[H.sub.3]), 26.8 (t, C[H.sub.2]), 27.4 (t, C[H.sub.2]), 40.1 (t, C[H.sub.2]), 40.4 (t, C[H.sub.2]), 61.3 (t, C[H.sub.2]), 115.1 (d, CH), 115.7 (d, CH), 116.3 (d, CH), 120.1 (d, CH), 122.4 (d, CH), 124.6 (d, CH), 125.1 (d, CH), 127.6 (s, C), 131.6 (s, C), 135.9 (s, C), 142.1 (s, C), 145.6 (d, CH), 146.3 (s, C), 148.7 (s, C), 167.3 (s, -COO)).
Synthesis of geranyl caffeate
Twenty-five percent NaOH (2.1 ml) was added to a solution of caffeic acid (2.00 g) in HMPA (150 ml), and the mixture was stirred for 1 h under [N.sub.2] at room temperature. A solution of geranyl bromide (3.10 g) in HMPA (20 ml) was added dropwise for 10 min to the reaction mixture. The reaction mixture was treated further as described above to afford geranyl caffeate (1.48 g; Y. 61.4%) as a white powder, and caffeic acid (0.56 g; Y. 28.0%).
Geranyl caffeate: EI-MS: m/z 316 ([M.sup.+], 10%), 247, 180, 163 (100%), 136, 69; HR-MS: m/z 316.1682, [C.sub.19][H.sub.24][O.sub.4] requires 316.1674; anal. calcd. for [C.sub.19][H.sub.24][O.sub.4]: C., 72.12; H, 7.65. Found: C, 72.01; H, 7.68; FT-IR (KBr) [cm.sup.-1]: 3483 (OH), 3295 (OH), 1678 (C=O), 1599, 1278, 1183; UV (EtOH) [[lambda].sub.max] nm (log [epsilon]): 334 (4.16), 302 (4.05), 249 (3.93), 222 (4.01); [.sup.1.H] NMR (acetone-[d.sub.6]): [delta]1.60 (3H, s, C[H.sub.3]), 1.66 (3H, s, C[H.sub.3]), 1.75 (3H, s, C[H.sub.3]), 4.68 (1H, d, J = 7.0 Hz, H-1'), 5.12 (1H, t, J = 7.0 Hz, H-6'), 5.40 (1H, t, J = 7.0 Hz, H-2'), 6.27 (1H, d, J = 15.9 Hz, H-[beta]), 6.87 (1H, d, J = 8.2 Hz, H-5), 7.03 (1H, dd, J = 2.0, 8.2 Hz, H-6), 7.16 (1H, d, J = 2.0 Hz, H-2), 7.55 (1H, d, J = 15.9 Hz, H-[alpha]), 8.28 (1H, br.s, -OH), 8.50 (1H, br.s, -OH); [.sup.13.C] NMR ((acetone-[d.sub.6]): [delta]16.4 (q, C[H.sub.3]), 17.7(q, C[H.sub.3]), 25.8 (q, C[H.sub.3]), 27.0 (t, C[H.sub.2]), 40.1 (t, C[H.sub.2]), 61.3 (t, C[H.sub.2]), 115.1 (d, CH), 115.6 (d, CH), 116.3 (d, CH), 120.0 (d, CH), 122.4 (d, CH), 124.6 (d, CH), 127.6 (s, C), 132.0 (s, C), 142.1 (s, C), 145.6 (d, CH), 146.3 (s, C), 148.7 (s, C), 167.3 (s, -COO)).
Plasmids were constructed for the expression of RXR[alpha], FXR and RARs. The ORF regions of human RXR[alpha], human FXR, mouse RAR[alpha]1, mouse RAR[beta]2, and mouse RAR[gamma]1 (accession numbers X52773, U68233, X57528, S56660, X15848) were amplified by PCR and inserted into pcDNA3.1 (Invitrogen, Carlsbad, CA, USA), respectively. For reporter plasmids, the luciferase region of the pGL3-Control Vector (Promega, Madison, WI, USA) was replaced with the EYFP fragment of pEYFP-N1 or the ECFP fragment of pECFP-N1 (Clontech, Palo Alto, CA, USA) using NcoI and Xbal sites. Subsequently, the simian virus 40 (SV40) early promoter was cut out with BglII and HindIII, and replaced with the thymidine kinase (TK) promoter of the pRL-TK vector (Promega) or one of several other promoters (the 3' region of the TK promoter, the cytomegalovirus (CMV) promoter, or the minimal CMV promoter and the 3' region of the CMV promoter (201 and 265 bp)) amplified using the following PCR primers:
5'-ggagatctggccccgcccagcgtcttgtc-3' and 5'-ggaagcttgcggcacgctgttgacgctgttaagcgggtcgctgcaggg-3' (3' region of the TK promoter);
5'-ccagatcttagttattaatagtaatcaattacggggtc-3' and 5'-ccaagcttgatctgacggttcactaaaccagc-3' (CMV promoter); 5'-ccagatcttaggcgtgtacggtgggagg-3' and 5'-ccaagcttaggctggatcggtcccggtg-3' (minimal CMV promoter);
5'-ccagatcttgggagtttgttttggcacc-3' and reverse primer of CMV promoter (CMV 201); and
5'-ccagatcttcaatgggcgtggatagcgg-3' and reverse primer of CMV promoter (CMV265).
Double-stranded oligonucleotides containing HREs (RXRE, RARE and FXRE; shown in Fig. 1B) were ligated into the upstream region of these promoters using MluI and BglII sites. The sequences of the constructed plasmids were confirmed by sequencing using an ABI PRISM 310 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA).
[FIGURE 1 OMITTED]
Cotransfection and reporter assay
A monkey kidney cell line, COS-7, was kept in DMEM with 10% FBS. Transfections were performed using an Effectene transfection reagent (Qiagen, Valencia, CA) according to the manufacturer's instructions. The ratio of the reporter plasmid, receptor expression plasmids (for example, the RAR[alpha] and RXR[alpha] expression plasmids for assay of RAR[alpha] ligands) and the internal control plasmid was 4:1:1:1. The culture medium was replaced with DMEM without phenol red (Gibco BRL, Gaithersburg, MD) supplemented with 10% charcoal-treated FBS (Hyclone, Logan, UT) when the transfections were performed. At 15h after transfections, the cells were treated with trypsin/penicillin reagent and divided among wells of a black, 96-well plate with 100 [micro]l of the culture medium. At 6h after division among wells, the cells were treated with chemicals. After a 40-h incubation, the medium was eliminated by decantation, the cells were washed twice with PBS, and the wells were filled with 200 [micro]l PBS. Fluorescence was detected using a microplate reader (ARVO; Perkin Elmer, Fremont, CA, USA). The fluorescence of EYFP was detected with an excitation filter of 485 nm and an emission filter of 545 nm, and that of ECFP was detected with filters of 420 and 486 nm (Perkin Elmer), respectively. The auto-fluorescence in COS-7 cells was subtracted from each of the detected fluorescences, and the EYFP/ECFP ratio was calculated using the resulting values.
Reporter assay system
In the present reporter assay, EYFP and ECFP were selected as a reporter protein and an internal control for normalization of transfection, respectively. These two fluorescent proteins were chosen, because the peaks of their excitation and emission wavelengths are sufficiently different (a difference of 80 and 50 nm, respectively) so that they can be detected simultaneously without cross-detection. The considerable cross-detection between EYFP and ECFP could be prevented using a set of optical filters (see Materials and methods). The EYFP/ECFP ratio was calculated after the autofluorescence of COS-7 cells was subtracted from the fluorescence intensities of EYFP and ECFP, because the autofluorescence was not negligible.
The reporter plasmids were constructed as shown in Fig. 1A. As HREs for FXR (FXR-RXR heterodimer), RAR (RAR-RXR heterodimer) and RXR (RXR homodimer), the fragments shown in Fig. 1B were used. In order to amplify signals, we employed three copies of DR5 (direct repeat with 5 bp of spacing) and four copies of DR1 as RAR and RXR response elements (RARE and RXRE). For the FXR response element (FXRE), four copies of the response element (inverted repeat) existing in the upstream region of the phospholipid transfer protein (PLTP) gene were employed. The tandem repeats in HREs elevated the response to a sufficient degree to detect the chemicals that activated the receptor. Then, an appropriate promoter for enhancing the fluorescent signal while retaining the response to the chemicals was selected from among seven promoters (Fig. 1C). Since the SV40 or CMV promoter caused a high fluorescence intensity with or without ligands, the responses to the ligands were not strong. The response of the RAR reporter plasmid with the SV40 promoter was about ten-fold. However, the apparent rate of the response was enhanced by interference of the expression of ECFP by the expression of EYFP, because the same promoter was employed for the reporter plasmid and the internal control plasmid. Therefore, the rate did not reflect a real response, and had a large SD. The TK promoter, the 3' region of the TK promoter and the minimal CMV promoter caused strong responses, but the expression in the control plasmid was too low for quantitative measurement. The expression of reporter proteins with the 3' region of the CMV promoter was higher than that with TK or the minimal CMV promoter, maintaining the induction rate by the ligands. Based on a comparison between the 3' regions of the CMV promoters, we selected the CMV201 (201 bp of the CMV promoter) promoter for use in the experiments below, since the response of CMV201 was stronger than that of CMV265.
In addition to the promoter for reporter plasmids, the promoter for the internal control plasmid and the expression plasmids of nuclear receptors were examined in order to establish an appropriate assay system of the nuclear receptor ligands. When the SV40 promoter was employed for the expression of ECFP in the internal control plasmid, the SV40 promoter for nuclear receptor expression interfered with the expression of ECFP (data not shown). Therefore, the CMV promoter was employed for nuclear receptor expression plasmids. Finally, we established the following plasmid set as the reporter assay system: a reporter plasmid containing the EYFP gene, whose expression was regulated by the HRE and CMV201 promoter; an internal control plasmid containing the ECFP gene expressed by the SV40 promoter; and the expression plasmid of the nuclear receptor containing each nuclear receptor gene expressed by the CMV promoter.
Fig. 2A shows the response to typical agonists for FXR, RARs and RXR[alpha] in the screening system. For screening of RAR ligands, three subtypes of RARs (RAR[alpha]1, RAR[beta]2, RAR[gamma]1) were expressed in the cells independently. Although endogenous RARs co-exists in the cell, the preference for the subtype of compounds could be detected. Fig. 2B and C show the dose-dependence of the assay system of FXR and RAR ligands, respectively. RARs were activated by 100 pM of ATRA. E[D.sub.50] values were estimated to be about 1-10 nM for RAR[alpha] and 0.1-1 nM for RAR[beta] and RAR[gamma] (only the result of RAR[alpha] is shown in Fig. 2B). On the other hand, activation of FXR was seen in 3-10 [micro]M CDCA and greater activation was observed at 100 [micro]M CDCA (Fig. 2C). These dose-dependent response patterns were comparable to those reported previously (Brand et al., 1988; Parks et al., 1999), indicating that these assays could be used for quantitative measurement of the activation by ligands. The established method of the reporter assay was described in Materials and methods.
Screening of a novel ligand for nuclear receptors
Using the established screening system, we found some natural compounds and their derivatives which acted as agonists for RARs and FXR. In the screening, there was a possibility that unexpected factors may have changed the signal responses (in the present assay system, the transcriptional efficiency may be changed irrespective of the nuclear receptor, the tested chemicals may have their own fluorescence, and so forth). Therefore, another reporter plasmid without HRE was also constructed to eliminate these unexpected factors. As this plasmid was used in place of the reporter plasmid, the compounds that regulated the expression of EYFP without HRE were eliminated. Some results of the response for each nuclear receptor are shown in Fig. 3 (RAR, upper panel; FXR, middle panel; control, lower panel). The results for RAR[beta] are presented as representative of those for RARs. Ten millimolar of each compound referring to the stock solution in DMSO was added to the culture medium of the transfected COS-7 cells at a final concentration of 30 [micro]M (Fig. 3, Nos. 1-26). Compound Nos. 27, 28, and 29 were 3 [micro]M ATRA, 30 [micro]M CDCA, and vehicle, respectively. ATRA also slightly activated the FXR-RXR heterodimer, due to the activation of RXR. Although, for example, Nos. 16, 18, 19, and 25 enhanced the relative EYFP/ECFP ratio, these compounds also enhanced the control that was used with the reporter plasmid without HRE. Thus it was concluded that these compounds were not ligands for the nuclear receptors.
As a result of screening more than 140 compounds (a part of the results is shown in Fig. 3), five compounds were found as ligands for the nuclear receptors. CAPE (compound No. 20 in Fig. 3), geranyl caffeate (No. 21), and farnesyl caffeate (not shown in Fig. 3) were found to be RAR agonists. Ginkgolic acid 15:1 (No. 12), geranyl caffeate (No. 21), and grifolin (No. 26) were found to be FXR agonists.
[FIGURE 2 OMITTED]
The structures of the caffeic acid derivatives tested in the screening are shown in Fig. 4A. CAPE, known as an active compound of propolis from honeybee hives, was synthesized from caffeic acid and [beta]-phenylethyl bromide and other caffeic acids were purified and synthesized as described in Materials and methods. Three of these compounds (i.e., all of those tested except for bazzanenyl caffeate) activated RARs (Fig. 4B). The cells treated with over 30 [micro]M of these compounds were removed from wells by washing of the reporter assay, because these compounds were toxic to the cell. Therefore, the results shown are for a reporter assay conducted using lower concentrations. Although the activation of RARs could be hardly detected by a low concentration of caffeic acid-derivatives, the activation by the compounds 10-30 [micro]M was comparable to maximum activation by ATRA. As shown in Fig. 4B, CAPE activated RAR[beta] to a greater degree than RAR[alpha] or RAR[gamma].
As FXR agonists, geranyl caffeate, ginkgolic acid 15:1 and grifolin were found. Geranyl caffeate, the RAR agonist, highly activated FXR (Fig. 3, No. 21), but the activation of the RXR homodimer was not detected (data not shown). It could not be determined whether or not farnesyl caffeate, a compound similar to geranyl caffeate, activated FXR, because 30 [micro]M of these compounds showed toxicity for cells. The structures of ginkgolic acids and grifolin are shown in Fig. 5A. It has been reported that ginkgolic acid 15:1 was present in ginkgolic leaves (Ahlemeyer et al., 2001), and grifolin in mushrooms (Hirata and Nakanishi, 1949; Sugiyama et al., 1992). The activations of FXR by ginkgolic acid 15:1 and geranyl caffeate were comparable to that by CDCA, the most potent endogenous bile acid. Ginkgolic acids 17:1, 15:0 and 13:0 (described in Fig. 5A) were also investigated as the other ginkgolic acids of ginkgo leaves (Fig. 5B). Ginkgolic acid 17:1 activated FXR more strongly than did 15:1, and ginkgolic acids with an alkyl chain (13:0, 15:0) activated FXR at concentrations of more than 20 [micro]M. It seemed that the double bond and length of the carbon chain had an influence on FXR activation. Moreover, the structures except for the carbon chain were also important for FXR activation, because the methylated compound of ginkgolic acid 15:1 (2-methyl ginkgolic acid methyl ester, Fig. 5A) had no potency for FXR activation (Fig. 3, No. 13).
[FIGURE 3 OMITTED]
[FIGURE 4 OMITTED]
To discover ligands for the nuclear receptors, we developed a battery of reporter assay systems incorporating the advantages of fluorescent proteins. The disadvantage of GFP (low sensitivity) could be overcome by modifications. The present screening system using fluorescent proteins has clear merits of a high efficiency, convenience and low cost, because the two fluorescent signals can be measured simultaneously without addition of any co-factors. Moreover, the fluorescent signal was stable for more than 2 h after the wash. Considering these merits, this reporter assay system with fluorescent proteins might be advantageous for automatic high-throughput screening. If the expression of the fluorescent protein can be increased, the measurement of fluorescence can be carried out in culture medium, and the signal can be measured by time-course without any treatment. Moreover, the use of three fluorescent proteins (for example, DsRed with EYFP and ECFP) would enable us to carry out more efficient measurement.
[FIGURE 5 OMITTED]
Using this assay system, several compounds that induce expression of the reporter gene for RARs and/or FXR were identified. These compounds were described as ligands in this report, although there is a possibility that these compounds are metabolized and their metabolites bind to the receptors as ligands.
Three new ligands for RARs were identified: CAPE, geranyl caffeate, and farnesyl caffeate. The whole structure of these compounds may be needed for RAR-activation, because caffeic acid, a constituent compound of the compounds, did not activate RARs (data not shown). CAPE has been reported to have antioxidant, antiviral, anti-inflammatory and immunomodulatory activities (Grunberger et al., 1988), and has also been shown to inhibit the growth of different types of oncogene-transformed cells and to induce apoptosis (Grunberger et al., 1988; Burke et al., 1995; Su et al., 1994; Watabe et al., 2004). Since RARs have been reported to mediate many biological processes, it is possible that some of the diverse activities are due to their binding to RARs. Since geranyl and farnesyl caffeate have also been reported to exert antioxidant effects and to inhibit the growth of cancer cells (Inoue et al., 2004), the three compounds may suppress the growth of cancer by at least two pathways: induction of RAR and antioxidant effects. Considering its preferential activation of RAR[beta] (Fig. 4B), CAPE may inhibit cancer (e.g., lung cancer) growth more selectively without substantial toxicity, such as the triglyceride elevation associated with RAR[alpha], and the skin, bone and teratogenic toxicity associated with RAR[gamma] Thus, especially CAPE could be assumed to be a seed for the development of an anti-cancer drug.
We also found that two natural compounds, ginkgolic acids and grifolin, activated FXR. Grifolin was first isolated as an antibiotic constituent of a mushroom, Grifola confluens (Hirata and Nakanishi, 1949). In 1992, it was reported that grifolin decreased liver cholesterol content, plasma total cholesterol levels, and plasma (very low-density lipoprotein (VLDL) + low-density lipoprotein (LDL)) cholesterol levels, and increased plasma high-density lipoprotein (HDL) cholesterol and plasma triglyceride levels (Sugiyama et al., 1992). It has been suggested that the effect of grifolin might be elicited, at least in part, by the augmented excretion of cholesterol into the faces (Sugiyama et al., 1994). On the other hand, FXR controls the expression of critical genes in bile acid and cholesterol homeostasis. In fact, FXR-null mice show elevated serum cholesterol and triglyceride levels (Sinal et al., 2000), and an FXR agonist has been shown to reduce serum triglyceride levels (Maloney et al., 2000). Moreover, FXR induces the expression of the gene of PLTP, which plays a role in HDL metabolism (Urizar et al., 2000). It seems that the cholesterol-lowering and HDL-cholesterol-increasing effects of grifolin are related to FXR activation, although grifolin's enhancement of triglyceride production was not consistent with its down-regulation of FXR agonists.
The FXR agonists found in this study are all non-steroidal compounds, whereas the well-known ligand of FXR, bile acid, is a steroidal one. The common characteristic of the structure of the ligands is their long carbon chains (i.e., geranyl, farnesyl and pentadecenyl), and farnesol has been shown to be a FXR ligand (Forman et al., 1995). However, aspects of the structures other than the carbon chains also appear to be important for FXR activation, because geraniol, a constituent compound of geranyl caffeate, has been reported not to activate FXR (Forman et al., 1995), and the methylated compound of ginkgolic acid 15:1 had no potency for FXR activation in the present study.
Several compounds, such as TTNPB, GW4064, Farnesoid, Forskolin, Fexaramine, AGN29 and AGN31, have been reported as non-steroidal agonists (Maloney et al., 2000; Howard et al., 2000; Downes et al., 2003; Dussault et al., 2003). The non-steroidal ligands may be important tools for studying the pharmacology of the receptor, because they may not have the property of bile acids and are not metabolized to form harmful lithocholic acid (Fischer et al., 1996; Javitt, 1966). In the present study, ginkgolic acids and geranyl caffeate strongly activated FXR, and both had structures quite different from bile acids, so that they could be good tools in this sense. Moreover, the importance of identifying gene-selective modulators that regulate a subset of FXR-specific genes as therapeutic agents has been recognized (Cui et al., 2003; Dussault et al., 2003). The gene-selective modulators of estrogen receptor, selective estrogen receptor modulators (SERMs), have been well studied (reviewed in McDonnell et al., 2002), and some compounds with a structure divergent from that of estrogen have been identified and applied to therapies of breast cancer and osteoporosis. The non-steroidal compounds could also be good tools for studying the selective response of FXR target genes.
In this report, we developed a new method for screening novel nuclear receptor agonists, and used it to identify new candidate ligands for FXR and RARs. We expect that these new ligands will be good pharmacological tools. Since the compound whose structure is much different from bile acids is expected to possess a specific effect as a ligand, we continue to screen various ligands from natural compounds with a wide variety of structures.
This work was supported by a grant-in-aid (MF-16) from the Pharmaceuticals and Medical Device Agency, a grant-in-aid for Research on Health Sciences Focusing on Drug Innovation from the Japan Health Science Foundation, and a grant-in-aid for Research on Advanced Medical Technology from the Ministry of Health, Labour and Welfare of Japan.
Ahlemeyer, B., Selke, D., Schaper, C., Klumpp, S., Krieglstein, J., 2001. Ginkgolic acids induce neuronal death and activate protein phosphatase type-2C. Eur. J. Pharmacol. 430, 1-7.
Begum, P., Hashidoko, Y., Islam, M.T., Ogawa, Y., Tahara, S., 2002. Zoosporicidal activities of anacardic acids against Aphanomyces cochlioides. Z. Naturforsch. [C] 57, 874-882.
Brand, N., Petkovich, M., Krust, A., Chambon, P., de The, H., Marchio, A., Tiollais, P., Dejean, A., 1988. Identification of a second human retinoic acid receptor. Nature 332, 850-853.
Burke Jr., T.R., Fesen, M.R., Mazumder, A., Wang, J., Carothers, A.M., Grunberger, D., Driscoll, J., Kohn, K., Pommier, Y., 1995. Hydroxylated aromatic inhibitors of HIV-1 intergrase. J. Med. Chem. 38, 4171-4178.
Cui, J., Huang, L., Zhao, A., Lew, J.-L., Yu, J., Sahoo, S., Meinke, P.T., Royo, I., Pelaez, F., Wright, S.D., 2003. Guggulsterone is a farnesoid X receptor antagonist in coactivator association assays but acts to enhance transcription of bile salt export pump. J. Biol. Chem. 278, 10214-10220.
Downes, M., Verdecia, M.A., Roecker, A.J., Hughes, R., Hogenesch, J.B., Kast-Woelbern, H.R., Bowman, M.E., Ferrer, J.-L., Anisfeld, A.M., Edwards, P.A., Rosenfeld, J.M., Alvarez, J.G.A., Noel, J.P., Nicolaou, K.C., Evans, R.M., 2003. A chemical, genetic, and structural analysis of the nuclear bile acid receptor FXR. Mol. Cell 11, 1079-1092.
Dussault, I., Beard, R., Lin, M., Hollister, K., Chen, J., Xiao, J.H., Chandraratna, R., Forman, B.M., 2003. Identification of gene-selective modulators of the bile acid receptor FXR. J. Biol. Chem. 278, 7027-7033.
Fischer, S., Beuers, U., Spengler, U., Zwiebel, F.M., Koebe, H.G., 1996. Hepatic levels of bile acids in end-stage chronic cholestatic liver disease. Clin. Chim. Acta 251, 173-186.
Fisher, G.J., Voorhees, J.J., 1996. Molecular mechanisms of retinoid actions in skin. FASEB J. 10, 1002-1013.
Forman, B.M., Goode, E., Chen, J., Oro, A.E., Bradley, D.J., Perlmann, T., Noonan, D.J., Burka, L.T., McMorris, T., Lamph, W.W., Evans, R.M., Weinberger, C., 1995. Identification of a nuclear receptor that is activated by farnesol metabolites. Cell 81, 687-693.
Grunberger, D., Banerjee, R., Eisinger, K., Oltz, E.M., Efros, L., Caldwell, M., Estevez, V., Nakanishi, K., 1988. Preferential cytotoxity on tumor cells by caffeic acid phenethyl ester isolated from propolis. Experientia 44, 230-232.
Hansen, L.A., Sigman, C.C., Andreola, F., Ross, S.A., Kelloff, G.J., De Luca, L.M., 2000. Retinoids in chemoprevention and differentiation therapy. Carcinogenesis 21, 1271-1279.
Hashimoto, T., Tori, M., Asakawa, Y., Wollenweber, E., 1988. Synthesis of two allergenic constituents of propolis and poplar bud excretion. Z. Naturforsch. [C] 43, 470-472.
Hirata, Y., Nakanishi, K., 1949. Grifolin, an antibiotic from a basidiomycete. J. Biol. Chem. 184, 135-143.
Howard, W.R., Pospisil, J.A., Njolito, E., Noonan, D.J., 2000. Catabolites of cholesterol synthesis pathways and forskolin as activators of the farnesoid X-activated nuclear receptor. Toxicol. Appl. Pharmacol. 163, 195-202.
Inoue, K., Nishitani, N., Tanabe, A., Yamanaka, H., Hashimoto, T., Asakawa, Y., Fujiki, H., 2004. Significance of structural specificity in the antioxidant activity of hydrangenol and caffeic acid-derivatives. Abstract for 124th Congress on Nihon yakugakukai, Osaka, Japan, vol. 2, p. 173.
Ishii, N., Takahashi, A., Kusano, G., Nozoe, S., 1988. Studies on the constituents of Polyporus dispansus and P. confluens. Chem. Pharm. Bull. 36, 2918-2924.
Javitt, N.B., 1966. Cholestasis in rats induced by taurolithocholate. Nature 210, 1262-1263.
Khorasanizadeh, S., Rastinejad, F., 2001. Nuclear-receptor interactions on DNA-response elements. Trends Biochem. Sci. 26, 384-390.
Krust, A., Kastner. P.H., Petkovich, M., Zelent, A., Chambon, P., 1989. A third human retinoic acid receptor, hRAR[gamma]. Proc. Natl. Acad. Sci. USA 86, 5310-5314.
Liu, Y., Binz, J., Numerick, M.J., Dennis, S., Luo, G., Desai, B., MacKenzie, K.I., Mansfield, T.A., Kliewer, S.A., Goodwin, B., Jones, S.A., 2003. Hepatoprotection by the farnesoid X receptor agonist GW4064 in rat models of intra- and extrahepatic cholestasis. J. Clin, Invest. 112, 1678-1687.
Makishima, M., Okamoto, A.Y., Repa, J.J., Tu, H., Learned, R.M., Luk, A., Hull, M.V., Lustig, K.D., Mangelsdorf, D.J., Shan, B., 1999. Identification of a nuclear receptor for bile acids. Science 284, 1362-1365.
Maloney, P.R., Parks, D.J., Haffner, C.D., Fivush, A.M., Chandra, G., Plunket, K.D., Creech, K.L., Moore, L.B., Wilson, J.G., Lewis, M.C., Jones, S.A., Willson, T.M., 2000. Identification of a chemical tool for the orphan nuclear receptor FXR. J. Med. Chem. 43, 2971-2974.
Martin, C.S., Wight, P.A., Dobretsova, A., Bronstein, I., 1996. Dual luminescence-based reporter gene assay for luciferase and [beta]-galactosidase. Biotechniques 21, 520-524.
McDonnell, D.P., Connor, C.E., Wijayaratne, A., Chang, C.Y., Norris, J.D., 2002. Definition of the molecular and cellular mechanisms underlying the tissue-selective agonist/antagonist activities of selective estrogen receptor modulators. Recent Prog. Horm. Res. 57, 295-316.
Morimoto, H., Kawamatsu, Y., Sugihara, H., 1968. Stereostructure of toxin from the fruit of Ginkgo biloba L. Chem. Pharm. Bull. 16, 2282-2286.
Nukata, M., Hashimoto, T., Yamamoto, I., Iwasaki, N., Tanaka, M., Asakawa, Y., 2002. Neogrifolin derivatives possessing anti-oxidative activity from the mushroom Albatrellus ovinus. Phytochemistry 59, 731-737.
Parks, D.J., Blanchard, S.G., Bledsoe, R.K., Chandra, G., Consler, T.G., Kliewer, S.A., Stimmel, J.B., Willson, T.M., Zavacki, A.M., Moore, D.D., Lehmann, J.M., 1999. Bile acids: natural ligands for an orphan nuclear receptor. Science 284, 1365-1368.
Paul, V.J., Yeddanapalli, L.M., 1956. On the olefinic nature of anacardic acid from Indian cashew nut shell liquid. J. Am. Chem. Soc. 78, 5675-5678.
Petkovich, M., Brand, N.J., Krust, A., Chambon, P., 1987. A human retinoic acid receptor which belongs to the family of nuclear receptors. Nature 330, 444-450.
Sherf, B.A., Navarro, S.L., Hannah, R.R., Wood, K.V., 1996. Dual-luciferase[TM] reporter assay: an advanced co-reporter technology integrating firefly and Renilla luciferase assays. Promega Notes 57, 2.
Sinal, C.J., Tohkin, M., Miyata, M., Ward, J.M., Lambert, G., Gonzalez, F.J., 2000. Targeted disruption of the nuclear receptor FXR/BAR impairs bile acid and lipid homeostasis. Cell 102, 731-744.
Su, Z.-Z., Lin, J., Grunberger, D., Fisher, P.B., 1994. Growth suppression and toxicity induced by caffeic acid phenethyl ester (CAPE) in type 5 adenovirus-transformed rat embryo cells correlate directly with transformation progression. Cancer Res. 54, 1865-1870.
Sugiyama, K., Kawagishi, H., Tanaka, A., Saeki, S., Yoshida, S., Sakamoto, H., Ishiguro, Y., 1992. Isolation of plasma cholesterol-lowering components from Ningyotake (Polyporus confluens) mushroom. J. Nutr. Sci. Vitaminol. 38, 335-342.
Sugiyama, K., Tanaka, A., Kawagishi, H., Ojima, F., Sakamoto, H., Ishiguro, Y., 1994. Hypocholesterolemic action of dietary grifolin on rats fed with a high-cholesterol diet. Biosci. Biotechnol. Biochem. 58, 211-212.
Toyota, M., Asakawa, Y., 1988. Sesquiterpenoids from the liverwort Bazzania fauriana. Phytochemistry 27, 2155-2159.
Urizar, N.L., Dowhan, D.H., Moore, D.D., 2000. The farnesoid X-activated receptor mediates bile acid activation of phospholipid transfer protein gene expression. J. Biol. Chem. 275, 39313-39317.
Wang, H., Chen, J., Hollister, K., Sowers, L.C., Forman, B.M., 1999. Endogenous bile acids are ligands for the nuclear receptor FXR/BAR. Mol. Cell 3, 543-553.
Watabe, M., Hishikawa, K., Takayanagi, A., Shimizu, N., Nakaki, T., 2004. Caffeic acid phenethyl ester induces apoptosis by inhibition of NF[kappa]B and activation of Fas in human breast cancer MCF-7 cells. J. Biol. Chem. 279, 6017-6026.
T. Suzuki (a,b), T. Nishimaki-Mogami (a), H. Kawai (a), T. Kobayashi (a), Y. Shinozaki (a), Y. Sato (a), T. Hashimoto (c), Y. Asakawa (c), K. Inoue (a), Y. Ohno (a), T. Hayakawa (a), T. Kawanishi (a,*)
(a) National Institute of Health Sciences, Tokyo, Japan
(b) Pharmaceuticals and Medical Device Agency, Tokyo, Japan
(c) Faculty of Pharmaceutical Sciences, Tokushima Bunri University, Tokushima, Japan
Received 27 September 2004; accepted 6 April 2005
*Corresponding author. Tel.: + 81 3 3700 9064; fax: +81 3 3700 9084.
E-mail address: firstname.lastname@example.org (T. Kawanishi).