Photoaffinity labeling of peroxisome proliferator binding proteins in rat hepatocytes; dehydroepiandrosterone sulfate- and bezafibrate-binding proteins

Photoaffinity labeling of peroxisome proliferator binding proteins in rat hepatocytes; dehydroepiandrosterone sulfate- and bezafibrate-binding proteins

Biochimica et Biophysica Acta 1339 Ž1997. 321–330 Photoaffinity labeling of peroxisome proliferator binding proteins in rat hepatocytes; dehydroepian...

600KB Sizes 1 Downloads 80 Views

Biochimica et Biophysica Acta 1339 Ž1997. 321–330

Photoaffinity labeling of peroxisome proliferator binding proteins in rat hepatocytes; dehydroepiandrosterone sulfate- and bezafibrate-binding proteins Hiroyuki Sugiyama a , Junji Yamada a,) , Hirotaka Takama a , Yoshitoshi Kodama b, Takafumi Watanabe a , Takeo Taguchi b, Tetsuya Suga a a

b

Department of Clinical Biochemistry, Tokyo UniÕersity of Pharmacy and Life Science, 1432-1 Horinouchi, Hachioji, Tokyo 192-03, Japan Department of Organic Synthesis, Tokyo UniÕersity of Pharmacy and Life Science, 1432-1 Horinouchi, Hachioji, Tokyo 192-03, Japan Received 28 May 1996; accepted 3 January 1997

Abstract To detect the cellular sites which directly interact with peroxisome proliferators ŽPPs. and mediate their inducing effect on peroxisomal enzymes in rat hepatocytes, two kinds of radiolabeled ligands, AD12 Ž7a-N-Ž4-azido-2-hydroxy-5iodow 125 Ixbenzyl.-aminomethyl-5-androstene-3 b-ol-17-one-O-3-sulfate. and BZ5 Ž2-w p-w2-Ž4X-azido-3X ,5X-diiodow 125 Ixbenzamido-2X-hydroxy.ethylxphenoxyx-2-methylpropionic acid., were developed for photoaffinity labeling. These compounds were derivatives of dehydroepiandrosterone sulfate ŽDHEAS. and bezafibrate, respectively, with an azido group as the photoreactive functional group. Upon UV-irradiation following incubation with rat liver cytosol and nuclei, both the ligands effectively radiolabeled several proteins analyzed by SDS-polyacrylamide gel electrophoresisrradioluminography. When w 125 IxAD12 was used at a concentration of 0.2 m M, two cytosolic proteins with molecular masses of 55 and 28 kDa and a nuclear protein of 40 kDa were specifically labeled, as coincubation with a 1000-fold excess of DHEAS inhibited labeling. Photoaffinity labeling of the cytosolic 28-kDa protein was also affected by Wy-14,643, but not by unsulfated dehydroepiandrosterone or androsterone sulfate, consistent with our previous findings obtained in competitive binding studies of w 3 HxDHEAS-binding detected in rat liver cytosol ŽYamada et al. Ž1994. Biochim. Biophys. Acta 1224, 139–146.. On the other hand, w 125 IxBZ5 specifically labeled a cytosolic protein of 31 kDa, which was inhibited by coincubation with bezafibrate, clofibric acid and Wy-14,643, but not with DHEAS. Thus, w 125 IxAD12 and w 125 IxBZ5 labeled several proteins which recognized DHEAS and bezafibrate, respectively, in rat liver cytosol and nuclei, providing a useful means to investigate PP-binding proteins. Keywords: Dehydroepiandrosterone sulfate; Bezafibrate; Binding protein; Photoaffinity labeling; Peroxisome proliferator; ŽRat liver.

Abbreviations: DHEA, dehydroepiandrosterone; PP, peroxisome proliferator; DHEAS, dehydroepiandrosterone sulfate; PPAR, peroxisome proliferator-activated receptor; AD12, 7a-N-Ž4-azido-2-hydroxy-5-iodobenzoyl.aminomethyl-5-androstene-3 b-ol-17-one-O-3sulfate; BZ5, 2-w p-w2-Ž4X-azido-3X ,5X-diiodobenzamido-2X-hydroxy-.ethylxphenoxyx-2-methylpropionic acid; PAGE, polyacrylamide gel electrophoresis; FABP, fatty acid binding protein ) Corresponding author. Fax: q81 426 765679. 0167-4838r97r$17.00 Copyright q 1997 Elsevier Science B.V. All rights reserved. PII S 0 1 6 7 - 4 8 3 8 Ž 9 7 . 0 0 0 1 6 - 2

322

H. Sugiyama et al.r Biochimica et Biophysica Acta 1339 (1997) 321–330

1. Introduction Dehydroepiandrosterone Ž DHEA. is a steroid secreted by the adrenal cortex in mammals. Although the physiological role of this steroid has not been fully elucidated, DHEA administered to rats and mice brings various pharmacological effects, e.g., as an antiobesity and antidiabetic agent w1,2x. We have previously characterized DHEA as a peroxisome proliferator ŽPP.. DHEA induces a pleiotropic response characterized by hepatomegaly and hepatic peroxisome proliferation with the induction of peroxisomal b-oxidation enzymes and several other lipid-metabolizing enzymes w3–8x. DHEA, like other PPs such as fibrate hypolipidemic drugs and phthalate ester plasticizers, also induces hepatocarcinogenesis in rats after long-term administration w9–11x. It was also revealed that administered DHEA was metabolized in hepatocytes to its sulfate conjugate, DHEA-3-sulfate ŽDHEAS. , which is responsible for the inducing effect of DHEA on peroxisomal enzymes w12–14x. Comparisons between the characteristics of the effects of DHEA and other PPs have indicated that DHEA is one of typical PPs, and suggested a common mechanism of action of these agents w3– 12,15,16x. To date, it has been demonstrated that induction of hepatic peroxisomal enzymes by PPs is mediated by the PP-activated receptor Ž PPAR a ., a member of the nuclear receptor superfamily. Therefore, this receptor plays a central role in the PP signaling pathway leading to the transcriptional activation of target genes w17–20x. However, the direct binding of PPs to

PPAR a has not yet been demonstrated, and the mechanism of PP activation of PPAR a is not clear. In addition, DHEAS does not activate the PPAR a , whereas all the other PPs examined to date can do so w17,19,20x, raising the possibility that a mechanismŽ s. other than that mediated by PPAR a is involved in the DHEA induction of peroxisomal enzymes. It is also possible that an additional stepŽ s. preceding the activation of PPAR a is involved in DHEAS signaling, which might have been impaired in the assay systems used in previous studies w17,19,20x. To address these issues, we analyzed the direct interactions of PPs with cellular components, the initial steps in the PP signaling pathway leading to enzyme induction. Previously, we found a binding activity highly specific to DHEAS in rat liver cytosol, which appeared to mediate the induction of peroxisomal enzymes by DHEA w21x. In this study, to further characterize PP binding in rat hepatocytes, we developed photoaffinity ligands, AD12 and BZ5 ŽFig. 1., derivatives of DHEAS and bezafibrate, respectively. 2. Materials and methods 2.1. Materials Na125 I Žcarrier-free, 13.5 MBqrm l. was obtained from Du Pont-New England Nuclear. Wy-14,643 was custom-synthesized by Tokyo Kasei ŽTokyo, Japan. . Bezafibrate and 2-w p-Ž 2X-aminoethyl. phenoxyx-2methylpropionic acid were gifts from Kissei ŽMatsumoto, Japan. . Other chemicals were of the highest grade commercially available.

Fig. 1. Chemical structures of photoaffinity ligands and their parent compounds.

H. Sugiyama et al.r Biochimica et Biophysica Acta 1339 (1997) 321–330

2.2. Chemical synthesis of photoaffinity ligands The strategies for the chemical synthesis of AD12 and BZ5 are illustrated in Scheme 1.

323

2.2.1. AD12 4-Androstene 3,17-dione as the starting material was dehydrogenated by reaction with tetrachloro1,2-benzoquinone Ž o-chloranil. w22x in acetic acid to

Scheme 1. Chemical syntheses of AD12 ŽA. and BZ5 ŽB..

324

H. Sugiyama et al.r Biochimica et Biophysica Acta 1339 (1997) 321–330

form a double bond at position C6, and the 17carbonyl group was selectively ketalized with ethylene glycol in the presence of p-toluenesulfonic acid w23x, giving compound AD3 ŽScheme 1A.. AD3 was subjected to hydrocyanation with diethylaluminium cyanide w24x, and the product AD4 was reduced by lithium aluminium hydride w25x which simultaneously produced an aminomethyl group at position C7 and a hydroxyl group at C3 to give AD5. After protection of the amino group with di-tert-butyl dicarbonate ŽŽBoc. 2 O. w26x, the 3-hydroxyl group of AD5 was oxidized by chromic trioxide to a carbonyl group w27x. The resulting compound, AD7, was treated with potassium tert-butoxide, then with acetic acid. The product was reduced by K-Selectride ŽAldrich, Inc.. and treated sequentially with 2 M sodium hydroxide and 30% hydrogen peroxide w28x to produce AD8. To remove the 3 a-hydroxy epimer as a byproduct, AD8 was purified by HPLC using a TSKgel SILIKA-60 column ŽTosoh, Tokyo, Japan. with hexanerisopropyl alcohol Ž 35:1. as the elution solvent, following which the protecting group Ž Boc. of the 7aminomethyl group was cleaved with 6 M hydrochloric acid w29x and N-hydroxysuccinimidyl-4-azidosalicylic acid ŽNHS-ASA, Pierce. was coupled to the exposed amino group w30x to produce AD10. AD10 was purified by silica gel column chromatography, desiccated and stored in the dark at 48C. For preparation of AD12, AD10 was iodinated with sodium iodide in the presence of Chloramine T w31x, sulfated with sulfur trioxide-pyridine complex w32x, and purified by TLC using Kieselgel 60 F254 Ž Merck. with methylene chloridermethanol Ž 5:1. as the developing solvent. The prepared compound AD12 was immediately used for photoaffinity labeling. w 125 IxAD12 was synthesized as described above, except using a mixture of NaI and Na125 I Ž 12.6: 1 by mol. for iodination of AD10. The final preparation of w 125 IxAD12 had a specific radioactivity of about 25 000 cpmrpmol and its yield was 30% based on the incorporation of 125 I. The analytical data of the products at key steps of the synthesis were as follows: AD3, m.p.: 132–1348C; IR ŽKBr. y cmy1 : 2953, 2234, 1679, 1649, 1164; 1 H-NMR ŽCDCl 3 . d : 0.94 Ž3H, s, CH 3 at C18., 1.11 Ž3H, s, CH 3 at C19., 3.88 Ž4H, m, OCH 2 CH 2 O. , 5.67 Ž1H, s, olefinic at C4., 6.12 Ž 2H, s, olefinic at C6 and 7.; EI-MS m r z: 328.3 ŽMq..

AD4, m.p.: 218–2198C; IR ŽKBr. y cmy1 : 2953, 2234, 1679, 1649, 1164; 1 H-NMR ŽCDCl 3 . d : 0.89 Ž3H, s, CH 3 at C18., 1.20 Ž3H, s, CH 3 at C19., 3.01 Ž1H, m, CH at C7., 3.88 Ž4H, m, OCH 2 CH 2 O., 5.87 Ž1H, s, olefinic at C4. ; EI-MS m r z: 355.4 ŽMq.. AD5, m.p.: 82–848C; IR ŽKBr. y cmy1 : 3396, 2941, 1561, 1169; 1 H-NMR Ž CDCl 3 . d : 0.87 Ž3H, s, CH 3 at C18., 1.08 Ž3H, s, CH 3 at C19., 2.64 Ž2H, m, NCH 2 ., 3.86 Ž4H, m, OCH 2 CH 2 O., 4.16 Ž1H, m, CH at C3., 5.35 Ž1H, s, olefinic at C4.; EI-MS m r z: 361.4 ŽMq.. AD7, m.p.: 181–1838C; IR ŽKBr. y cmy1 : 3355, 2952, 1715, 1676, 1247, 1172; 1 H-NMR ŽCDCl 3 . d : 0.87 Ž3H, s, CH 3 at C18., 1.20 Ž3H, s, CH 3 at C19. , 1.42 Ž9H, s, CŽCH 3 . 3 ., 2.70 Ž1H, m, NCH., 3.34 Ž1H, br, NH. , 3.85 Ž4H, m, OCH 2 CH 2 O.; EI-MS m r z: 459.2 ŽMq.. AD9, IR ŽKBr. y cmy1 : 3589, 2941, 1740, 1637; 1 H-NMR ŽCDCl 3 . d : 0.82 Ž 3H, s, CH 3 at C18. , 1.00 Ž3H, s, CH 3 at C19., 3.32 Ž2H, m, NCH 2 ., 4.04 Ž1H, m, OCH at C3., 5.38 Ž1H, br, olefinic at C6. ; EI-MS m r z: 317.4 ŽMq.. AD10, IR ŽKBr. y cmy1 : 3569, 2933, 2116, 1775, 1641, 1594, 1278; 1 H-NMR Ž CDCl 3 . d : 0.88 Ž3H, s, CH 3 at C18., 1.06 Ž3H, s, CH 3 at C19., 3.73 Ž2H, m, NCH 2 ., 4.18 Ž1H, m, OCH at C3., 5.46 Ž 1H, d, olefinic at C6. , 6.45 Ž 1H, dd, aromatic., 6.61 Ž1H, d, aromatic., 7.56 Ž1H, d, aromatic.; ESI-MS m r z: 479.4 ŽMqq 1.. AD12, m.p.: 169–1728C; lmax Žethanol.: 224 nm Ž e 20984., 271 nm Ž e 9519. ; IR ŽKBr. y cmy1 : 3569, 2954, 2119, 1736, 1702, 1686, 1277, 1243; 1 H-NMR ŽCDCl 3 . d : 0.93 Ž3H, s, CH 3 at C18. , 1.13 Ž 3H, s, CH 3 at C19., 3.73 Ž 2H, m, NCH 2 ., 4.71 Ž 1H, m, OCH at C3., 5.55 Ž1H, d, olefinic at C6. , 6.80 Ž1H, s, aromatic., 8.29 Ž1H, s, aromatic.; ESI-MS m r z: 729.0 ŽMqq 1. as disodium salt. 2.2.2. BZ5 The starting material, 2-w p-Ž2X-aminoethyl.phenoxyx-2-methylpropionic acid, was esterified using thionyl chloride and methanol w33x to protect the carboxyl group, and then coupled with NHS-ASA at the 2X-amino group w32x, following which the ester group was saponified by treatment with potassium hydroxide to produce BZ4 ŽScheme 1B.. BZ4 was purified by silica gel column chromatography, and stored. For preparation of BZ5, BZ4 was iodinated as

H. Sugiyama et al.r Biochimica et Biophysica Acta 1339 (1997) 321–330

described above, and purified by TLC using Kieselgel 60 F254 with methylene chloridermethanol Ž5: 1. as the developing solvent. w 125 IxBZ5 was synthesized as described above, except a mixture of NaI and Na125 I Ž 12.6: 1 by mol. was used for iodination. The final preparation of w 125 IxBZ5 had a specific radioactivity of about 45 000 cpmrpmol and its yield was 50% based on the incorporation of 125 I. The analytical data of BZ5 were as follows: m.p.: 152–1548C; lmax Ž ethanol.: 289 nm Ž e 19187. ; IR ŽKBr. y cmy1 : 3589, 3100-2900, 2120, 1636, 1572, 1149; 1 H-NMR Ž CD 3 OD. d : 1.57 Ž6H, s, CŽCH 3 . 2 . , 2.87 Ž2H, t, phenyl-CH 2 ., 3.58 Ž 2H, dd, NCH 2 ., 6.88 Ž2H, d, aromatic., 7.15 Ž2H, d, aromatic., 8.22 Ž1H, s, aromatic.; ESI-MS m r z: 385.3 ŽMqq 1. as disodium salt. 2.3. Preparation of liÕer cytosol and nuclei Male Wistar rats Ž about 250 g. were anesthetized with diethyl ether, and the livers were perfused in situ with saline through the portal vein. The livers were excised, rinsed in ice-cold 0.25 M sucrose, and homogenized in 20 volumes of 10 mM Tris-HCl Ž pH 7.5. containing 5 mM EDTA Ž TE buffer. , using a Potter-Elvehjem glass homogenizer with a Teflon pestle. After centrifugation of the homogenates at 25 000 = g for 10 min, the supernatants were further centrifuged at 105 000 = g for 60 min. The resultant supernatants were diluted with TE buffer, and used as the cytosol. For preparation of nuclei, the liver homogenates were filtered through nylon mesh Ž150 m m. and centrifuged at 600 = g for 10 min. The precipitates were washed with TE buffer, then with that containing 0.5%Žwrw. Triton X-100, three times each, and finally resuspended in TE buffer. Protein was determined by the method of Lowry et al. w34x. 2.4. Photoaffinity labeling Unless otherwise noted, the 125 I-labeled photoaffinity ligands dissolved in dimethylformamide at a concentration of 4 m M were added in a volume of 1 m l to 20 m l aliquots of the liver cytosol or nuclei, and incubated at 08C for 8 h in the dark. Competitors of the ligands were dissolved in dimethylformamide and added at a 1000-fold excess so that the total amount of the solvent added to the incubation mix-

325

tures was 1 m l. After incubation, the mixtures placed on ice were irradiated by UV for 2 min using a Black-light Ž15 watt, FL-15 BL-B Stanley. at a distance of 4 cm, and diluted with the same volume of SDS-PAGE sample buffer containing 3% SDS, 15% 2-mercaptoethanol, 180 mM Tris-HCl ŽpH 6.8., 0.003% bromophenol blue and 30% glycerol, then boiled for 2 min. These samples were subjected to SDS-PAGE as described by Laemmli w35x. Whether the same level of radioactivities were added to each incubation mixture was routinely examined by g-scintillation counting of the boiled samples. After electrophoresis, the gels were stained with Coomassie Brilliant Blue R-250 and dried. The radioactivity level of the photoaffinity ligands bound to the proteins on the gels was analyzed by radioluminography using a BAS 2000 bioimaging analyzer ŽFuji Photo Film, Tokyo, Japan.. Although the data presented here are typical of several independent experiments, these results were reproducible.

3. Results and discussion 3.1. Structural properties of AD12 and BZ5 We previously studied the structure-activity relationships of DHEAS with regard to its ability to induce peroxisomal b-oxidation enzymes using cultured rat hepatocytes w12–14x. To induce the enzymes, the 3-sulfuric and 17-carbonyl groups of DHEAS were shown to be important. Although the presence of a double bond at position C5 was not a primary determinant, the double bond served to confer the relatively planar conformation on the steroidal hydrophobic backbone, a critical requirement for enzyme induction. On the other hand, introduction of bulky substituents to position C17 or aromatization of ring A led to a loss of inducing activity. In this study, based on these findings, AD12 was designed and synthesized as a photoaffinity ligand to probe the DHEAS binding protein in rat hepatocytes Ž Fig. 1.. Since AD12 supports the whole chemical structure of DHEAS and has N-Ž 4-azido-2-hydroxy-5-iodobenzoyl.aminomethyl group which contains a photoreactive azido group at position C7 in the a-configuration, its chemical structure is compatible with the structural requirements for DHEAS induction of per-

326

H. Sugiyama et al.r Biochimica et Biophysica Acta 1339 (1997) 321–330

oxisomal enzymes. Another photoaffinity ligand used in this study, BZ5, is a derivative of bezafibrate, a typical PP and a PPAR a activator w36–38x. In BZ5, the 4-chlorobenzoyl group of bezafibrate is substituted by a 4-azido-2-hydroxy-3,5-diiodobenzoyl group ŽFig. 1.. This derivatization of the parent compound is expected to minimize modification of the interaction between bezafibrate and its binding proteins. We also synthesized another DHEAS derivative, ND9, which partially supports the DHEAS structure ŽFig. 1.. This compound was designed to limit the steric hindrance caused by the photoreactive azidocontaining substituent, and so it supports only the part of rings A and B of DHEAS. However, we detected no proteins which bound to this photoreactive compound Ž data not shown.. Therefore, no further experiments were performed using ND9. 3.2. Characterization of photoaffinity labeling with AD12 and BZ5 In preliminary experiments, AD12 was irradiated by UV at 354 nm using a Black-light Ž 15 watt. and the time course of its photolysis was examined by spectrophotometry. When the absorbance of AD12 at a lmax 271 nm was monitored, it decreased by 50% only 10 s after the UV irradiation, and after 3 min the absorbance became almost equivalent to the basal level. TLC analysis on silica with methylene chloridermethanol Ž5: 1. as a developing solvent confirmed the comparable degradation of AD12 at each time point. Similar results were obtained when the time course of photolysis of BZ5 was examined by TLC Žsilica; methylene chloridermethanol, 5:1. Ž data not shown.. Therefore, we expected that the UV-induced labeling reaction by AD12 and BZ5 would be accomplished within 3 min. In fact, UV irradiation for 1–3 min was sufficient to covalently bind these photoaffinity ligands to rat liver proteins ŽFig. 2.. When w 125 IxAD12 was incubated with rat liver cytosol at 08C for 8 h and irradiated by UV for 10 s, four faint bands were detected on the radioluminograms of SDS-PAGE gels. Prolonged UV irradiation intensified the signals, but irradiation for over 60 s showed no further effect. No bands were detected without UV irradiation. Similar results were obtained with w 125 IxBZ5, although the

Fig. 2. Time courses of photoaffinity labeling. w 125 IxAD12 and w 125 IxBZ5 Ž50 nM. were incubated with rat liver cytosol Ž5.0 mg proteinrml. at 08C for 8 h, and irradiated by UV for the time periods indicated. The photoaffinity-labeled proteins were separated by SDS-PAGE Ž10% gel. and their radioactivity was detected by radioluminography. The major proteins labeled are indicated by arrowheads with their molecular masses ŽkDa..

intensities of the detected bands appeared to be saturated even after 10 s of UV irradiation ŽFig. 2.. In either case, the band patterns visualized on radioluminograms did not differ upon UV irradiation for 60 and 180 s. In addition, there were no appreciable changes after UV irradiation for 3 min in the protein band pattern visualized on SDS-PAGE gels stained with Coomassie brilliant blue Ždata not shown., suggesting the absence of UV-induced damage in the proteins. It was also confirmed that the band intensity on the radioluminograms was elevated with increasing concentrations of cytosolic protein in the range of 1–10 mgrml Ždata not shown.. These results indicate that a period of 1–3 min in the UV irradiation is an optimum for the photoaffinity labeling with AD12 and BZ5. 3.3. Photoaffinity labeling with [ 125I]AD12 w 125 IxAD12 was incubated with rat liver cytosol or nuclei at 08C for 8 h, then irradiated with UV for 2 min ŽFig. 3. . In our previous study of a binding activity highly specific to DHEAS in rat liver cytosol, the DHEAS-binding was most stable at 08C and required 6 h to reach equilibrium at that temperature

H. Sugiyama et al.r Biochimica et Biophysica Acta 1339 (1997) 321–330

Fig. 3. Photoaffinity labeling of rat liver cytosol and nuclei with w 125 IxAD12. w 125 IxAD12 Ž0.2 m M. was incubated with the cytosol Ž7.5 mg proteinrml. or nuclei Ž5.6 mg proteinrml. in the absence Žlanes 1 and 6. or presence of a 1000-fold excess of DHEAS Žlanes 2 and 7., DHEA Žlanes 3 and 8., androsterone sulfate Žlanes 4 and 9. and Wy-14,643 Žlanes 5 and 10. at 08C for 8 h, and irradiated by UV for 2 min. The samples were subjected to SDS-PAGE Ž10% gel. and the radioactivity of the labeled proteins were analyzed by radioluminography. A, radioluminograms of the SDS-PAGE gels. The major proteins labeled are indicated by arrowheads with their molecular masses. B, radioactivity level of the photoaffinity ligand bound to selected proteins. Results are expressed as relative values compared with the data obtained in the incubation without any competitors Žlanes 2–5 vs. lane 1, lane 7–10 vs. lane 6..

w21x. Therefore, in this study, incubation was carried out as described above, the conditions of which were also applicable to the nuclei. Moreover, based on the dissociation constant Ž72 nM. of DHEAS-binding w21x, w 125 IxAD12 was used at a concentration of 0.2 m M.

327

In the cytosol, four kinds of proteins with molecular masses of 66, 55, 44 and 28 kDa were photoaffinity-labeled with w 125 IxAD12 Ž Fig. 3A, lane 1.. However, labeling of the 55- and 28-kDa proteins was markedly inhibited by coincubation with a 1000-fold excess of DHEAS ŽFig. 3A, lane 2. , indicating the specificity of w 125 IxAD12-binding to the DHEAS-recognition sites of these proteins. Labeling of the 66- and 44-kDa proteins was not affected by DHEAS, and therefore was nonspecific. On the other hand, DHEA used as a competitor did not affect labeling of the 55- or 28-kDa proteins ŽFig. 3A, lane 3., confirming the requirement of the sulfuric group for the binding. Moreover, when androsterone sulfate and Wy-14,643 were used, labeling of the 55-kDa protein was inhibited by both the competitors, whereas that of 28-kDa protein was inhibited only by Wy14,643 ŽFig. 3A, lanes 4, 5.. Since the DHEAS-binding activity detected previously in rat liver cytosol was also inhibited by Wy-14,643 but not by androsterone sulfate w21x, the 28-kDa protein is suggested to be the most important for DHEAS-binding. As described previously, the 3-sulfuric group of DHEAS is a structural requirement for induction of peroxisomal enzymes by DHEAS w13,14x, but its configuration Ž b . is not a primary determinant w13x. In fact, androsterone sulfate which has a 3-sulfuric group in the a-configuration induces peroxisomal b-oxidation to an extent comparable to that seen by DHEAS in cultured rat hepatocytes w13x. However, the DHEAS-binding activity in the rat liver cytosol is specific to the 3 b-sulfated structure because it was not competed by androsterone sulfate w21x. Therefore, before our previous hypothesis that the cytosolic DHEAS-binding mediates the inducing effect of DHEAS on peroxisomal b-oxidation w21x could be confirmed, there was a discrepancy to be explained between the configurational requirements of a 3sulfuric group of DHEAS for binding Žthe b-configuration is exclusive. and enzyme induction Žeither configuration is permissible.. To explain this, we speculated that the configurational conversion of 3 asulfated steroids to their 3 b-counterparts may occur in hepatocytes, or that there may be another recognition site mediating the action of 3 a-sulfated steroids as an inducer of peroxisomal enzymes w21x. If it is assumed that the enzyme induction by androsterone sulfate is mediated by the 55-kDa protein which

328

H. Sugiyama et al.r Biochimica et Biophysica Acta 1339 (1997) 321–330

shows binding specificity to both 3 a- and 3 b-sulfated structures ŽFig. 3A, lanes 2, 4., this would be consistent with the latter speculation. Therefore, in addition to the 28-kDa protein, the 55-kDa protein is also an intriguing candidate as a cytosolic factor mediating the PP action of DHEAS. On incubation of w 125 IxAD12 with the nucleus, five proteins with molecular masses of 48, 44, 40, 29 and 27 kDa were photoaffinity-labeled ŽFig. 3A, lane 6.. Among these, only the 40-kDa protein was specific, as a 1000-fold excess of DHEAS inhibited the labeling of this protein, but not that of the other species ŽFig. 3A, lane 7.. Competitive binding experiments revealed that the ligand specificity of the 40-kDa protein was similar to that of the cytosolic 55-kDa protein, as the labeling of this protein was significantly inhibited by both androsterone sulfate and Wy-14,643, although it was to some extent affected by DHEA unlike the labeling of the latter protein ŽFig. 3A, lanes 8–10.. Thus, using w 125 IxAD12, three kinds of DHEAS-binding proteins, two in the cytosol and one in the nucleus, were detected in the rat liver. These results were also confirmed by quantitative analysis using BAS-2000 Ž Fig. 3B. , whereas the extent of the competition by 1000-fold excess ligand in Fig. 3B was smaller than that in Fig. 3A. This difference might arise from a technical problem in getting a background level for quantitation. When the plasma membrane was examined, several proteins were labeled by w 125 IxAD12, but none of these were binding proteins specific to DHEAS as DHEAS did not compete with the labeling Ždata not shown.. 3.4. Photoaffinity labeling with [ 125I]BZ5 When the rat liver cytosol was examined using w 125 IxBZ5 as a photoaffinity ligand, several proteins Ž110, 63, 54, 31 and 28 kDa. were labeled, but labeling of only the 63- and 31-kDa proteins was inhibited by coincubation with an excess of bezafibrate ŽFig. 4A, lanes 1, 2.. Bezafibrate is a fibrate-type hypolipidemic PP, phenoxyisobutyrate-derivatives including clofibric acid, ciprofibrate and nafenopin which are thought to exert their PP effects through a common mechanism, albeit with different potencies. Therefore, we examined the competitive inhibition by clofibric acid of photoaffinity labeling of the 63- and

Fig. 4. Photoaffinity labeling of rat liver cytosol and nuclei with w 125 IxBZ5. w 125 IxBZ5 Ž0.2 m M. was incubated with the cytosol Ž5.6 mg proteinrml. or nuclei Ž8.5 mg proteinrml. in the absence Žlanes 1 and 6. or presence of a 1,000-fold excess of bezafibrate Žlanes 2 and 7., clofibric acid Žlanes 3 and 8., Wy-14,643 Žlanes 4 and 9. and DHEAS Žlanes 5 and 10. at 08C for 8 h, and irradiated by UV for 2 min. The samples were subjected to SDS-PAGE Ž10% gel.. A, radioluminograms. B, radioactivity level of the photoaffinity ligand bound to selected proteins. Results are expressed as relative values Žlanes 2–5 vs. lane 1..

31-kDa proteins. As shown in Fig. 4A Žlane 3., clofibric acid inhibited the labeling of the 31-kDa protein as significantly as bezafibrate, but not labeling of the 63-kDa protein. Thus, the 31-kDa protein seems to be a more possible candidate than the 63-kDa protein as a binding site mediating the PP effect of fibrates. Wy-14,643 also inhibited labeling of the 31-kDa protein, whereas DHEAS did not ŽFig. 4A, lanes 4, 5.. When the nucleus Ž Fig. 4A, lanes 6–10. and plasma membrane Ždata not shown. were examined, several proteins were labeled by w 125 IxBZ5, but none of these were binding proteins specific to bezafibrate as bezafibrate did not compete with the labeling.

H. Sugiyama et al.r Biochimica et Biophysica Acta 1339 (1997) 321–330

Previously, Alvares et al. w39x reported a member of the heat shock protein HSP70 family, P72, as a PP-binding protein in the rat liver cytosol which can bind clofibric acid, ciprofibrate and nafenopin, and the binding was not competed by Wy-14,643. Although P72 is likely to bind bezafibrate, w 125 IxBZ5 failed to detect any protein corresponding to P72 under the experimental conditions used in this study. On the other hand, Cannon and Eacho w40x demonstrated that structurally diverse PPs including LY171883, bezafibrate, clofibric acid and Wy-14,643 can bind to fatty acid binding protein Ž FABP. from rat liver cytosol and displace an endogenous fatty acid from its binding site, and suggested that the interaction of PPs with FABP may be involved in perturbations of fatty acid metabolism caused by these agents. Therefore, we also focused on the low molecular mass proteins photoaffinity-labeled with w 125 IxBZ5. As shown in Fig. 5 Žlanes 1, 2, 4–7. , a protein with a molecular mass of 14 kDa equal to that of FABP was labeled with w 125 IxBZ5, and this labeling was significantly competed by an excess of bezafibrate, Wy-14,643, DHEAS and palmitic acid. Although the competition by clofibric acid was only slight ŽFig. 5A, lane 3., this is compatible with the finding that the affinity of clofibric acid to FABP is very low w35x. Thus, w 125 IxBZ5 also detected FABP as a PP-binding protein in rat liver cytosol as previously reported w35x, although further identification of the 14-kDa protein is required. Interestingly, DHEAS could also compete with the fatty acid for binding site of the 14-kDa FABP Ž Fig. 5A, lane 5.. If the fatty acids replaced and released from FABP by PPs influence the activities of fatty acid-dependent transcription factors including PPAR, the PP signals should converge after FABP to commonly alter gene expression. This is the simplest explanation for the mechanism by which structurally diverse PPs elicit common cellular responses. However, this raises the question of why only DHEAS, unlike other PPs, does not activate the PPAR. 3.5. Summary In this study, to detect the cellular sites which directly interact with PPs and mediate their inducing effect on peroxisomal enzymes in rat hepatocytes, we developed w 125 IxAD12 and w 125 IxBZ5 as radiolabeled

329

Fig. 5. Photoaffinity labeling of rat liver cytosol with w 125 IxBZ5. w 125 IxBZ5 Ž0.2 m M. was incubated with the cytosol Ž5.2 mg proteinrml for lanes 1–5, 5.0 mg proteinrml for lanes 6 and 7. in the absence Žlanes 1 and 6. or presence of a 1000-fold excess of bezafibrate Žlanes 2., clofibric acid Žlanes 3., Wy-14,643 Žlanes 4., DHEAS Žlanes 5. and palmitic acid Žlane 7. at 08C for 8 h, and irradiated with UV for 2 min. The samples were subjected to SDS-PAGE Ž15% gel.. A, radioluminograms. B, radioactivity level of the photoaffinity ligand bound to a 14-kDa protein. Results are expressed as relative values Žlanes 2–5 vs. lane 1, lane 7 vs. lane 6..

ligands for photoaffinity labeling. Using these ligands, several proteins were photoaffinity-labeled in the cytosol and nucleus, many of which had similar or identical molecular masses. For example, both the ligands labeled proteins of 55-54 and 28 kDa in the cytosol and those of 48, 44, 29 and 27 kDa in the nucleus. Among these, the cytosolic 55- and 28-kDa proteins and nuclear 40-kDa protein Žfor DHEAS. and the cytosolic 31-kDa protein Ž for fibrates. were shown to be PP-binding proteins by competitive binding studies. However, it remains to be determined whether these binding proteins are identical to pro-

330

H. Sugiyama et al.r Biochimica et Biophysica Acta 1339 (1997) 321–330

teins already known and functional in the PP signaling pathway leading to peroxisomal enzyme induction, although a role of FABP which was also detected as a PP-binding protein was suggested in the mechanism of the PP-induced hepatocellular response w40x. Therefore, to identify and characterize the PPbinding proteins, further studies are required. Due to its simplicity, the photoaffinity labeling technique enables easy and quick detection of relevant binding proteins. Moreover, the proteins to which the radiolabeled ligands covalently bind can be dealt with under denaturing conditions and easily monitored, facilitating their isolation followed by structural analysis for identification. Until now, no similar experiments looking for PP-binding proteins have been performed in this field. Therefore, w 125 IxAD12 and w 125 IxBZ5 provided a useful means to analyze PP-binding proteins. It will be of interest to examine the possibility of direct interactions between the PPAR and PPs using these photoaffinity ligands. References w1x Schwartz, A. Ž1985. Basic Life Sci. 35, 181–191. w2x Regelson, W., Loria, R. and Kalimi, M. Ž1988. Ann. N. Y. Acad. Sci. 521, 260–273. w3x Frankel, R.A., Slaughter, C.A., Orth, K., Moomaw, C.R., Hicks, S.H., Snyder, J.M., Prough, R.A., Putnam, R.S. and Milewich, L. Ž1990. J. Steroid Biochem. 35, 333–342. w4x Yamada, J., Sakuma, M., Ikeda, T., Fukuda, K. and Suga, T. Ž1991. Biochim. Biophys. Acta 1092, 233–243. w5x Yamada, J., Sakuma, M. and Suga, T. Ž1991. Anal. Biochem. 199, 132–136. w6x Sakuma, M., Yamada, J. and Suga, T. Ž1992. Biochem. Pharmacol. 43, 1269–1273. w7x Rao, M.S., Musuunuri, S. and Reddy, J.K. Ž1992. Pathobiology 60, 82–86. w8x Prough, R.A., Webb, S.J., Wu, H.-Q., Lapenson, D.P. and Waxman, D.J. Ž1994. Cancer Res. 54, 2878–2886. w9x Hayashi, F., Tamura, H., Yamada, J., Kasai, H. and Suga, T. Ž1994. Carcinogenesis 15, 2215–2219. w10x Rao, M.S., Subbarao, V., Yeldanti, A.V. and Reddy, J.K. Ž1992. Cancer Res. 52, 2977–2979. w11x Metzger, C., Mayer, D., Hoffmann, H., Bocker, T., Hobe, G., Benner, A. and Bannasch, P. Ž1995. Toxicol. Pathol. 23, 591–605. w12x Yamada, J., Sakuma, M. and Suga, T. Ž1992. Biochim. Biophys. Acta 1137, 231–236. w13x Sakuma, M., Yamada, J. and Suga, T. Ž1993. Biochim. Biophys. Acta 1169, 66–72.

w14x Yamada, J., Sakuma, M., Ikeda, T. and Suga, T. Ž1994. Arch. Biochem. Biophys. 313, 379–381. w15x Yamada, J., Sugiyama, H., Tamura, H. and Suga, T. Ž1994. Arch. Biochem. Biophys. 315, 555–557. w16x Yamada, J., Sugiyama, H., Watanabe, T. and Suga, T. Ž1996. Biol. Pharm. Bull. 19, 630–631. w17x Issemann, I. and Green, S. Ž1990. Nature 347, 645–650. w18x Dreyer, C., Krey, G., Keller, H., Givel, F., Helftenbein, G. and Wahli, W. Ž1992. Cell 68, 879–887. w19x Gottlicher, M., Widmark, E., Li, Q. and Gustafsson, J.A. Ž1992. Proc. Natl. Acad. Sci. USA 89, 4653–4657. w20x Issemann, I., Prince, R.A., Tugwood, J.D. and Green, S. Ž1993. J. Mol. Endocrinol. 11, 37–47. w21x Yamada, J., Sugiyama, H., Sakuma, M. and Suga, T. Ž1994. Biochim. Biophys. Acta 1224, 139–146. w22x Brueggemeier, R.W., Floyed, E.E. and Councell, R.E. Ž1978. J. Med. Chem. 21, 1007– 1011. w23x Heathcock, C.H. and Ratcliffe, R. Ž1971. J. Am. Chem. Soc. 93, 1746–1757. w24x Nagata, W., Yoshida, M. and Murakami, M. Ž1972. J. Am. Chem. Soc. 94, 4654–4783. w25x Malek, J. Ž1987. Org. React. 36, 249–263. w26x Tarbell, D.S., Yamamoto, Y. and Pope, B.M. Ž1972. Proc. Natl. Acad. Sci. USA 69, 730–736. w27x Collins, J.C., Hess, W.W. and Frank, F.J. Ž1968. Tetrahedron Lett. 30, 3363–3366. w28x Moss, R.A. and Chen, E.Y. Ž1981. J. Org. Chem. 46, 1466–1469. w29x Stahl, G.L., Walter, R. and Smith, C.W. Ž1978. J. Org. Chem. 43, 2285–2288. w30x Poland, A., Glover, E., Ebetino, F.H. and Kende, A.S. Ž1986. J. Biol. Chem. 261, 6352–6365. w31x Ji, T.H. and Ji, I. Ž1982. Anal. Biochem. 121, 286–289. w32x McCall, J.M., Aiken, J.W., Chidester, C.G., Ducharme, D.W. and Wendling, M.G. Ž1983. J. Med. Chem. 26, 1791– 1793. w33x Brenner, M. and Huber, W. Ž1953. Helv. Chem. Acta 36, 1109–1111. w34x Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. Ž1951. J. Biol. Chem. 193, 265–275. w35x Laemmli, U.K. Ž1970. Nature 227, 680–686. w36x Reddy, J.K. and Lalwani, N.D. Ž1983. CRC Crit. Rev. Toxicol. 12, 1–58. w37x Watanabe, T., Horie, S., Yamada, J., Isaji, M., Nishigaki, T., Naito, J. and Suga, T. Ž1989. Biochem. Pharmacol. 38, 367–371. w38x Hertz, R., Berman, I. and Ber-tana, J. Ž1994. Eur. J. Biochem. 221, 611–615. w39x Alvares, K., Carrillo, A., Yuan, P.M., Kawano, H., Morimoto, R.I. and Reddy, J.K. Ž1990. Proc. Natl. Acad. Sci. USA 87, 5293–5297. w40x Cannon, J.R. and Eacho, P.I. Ž1991. Biochem. J. 280, 387–391.