Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

Sulfation plays key roles in detoxification and activation of various endogenous and exogenous compounds such as hormones, neurotransmitters, drugs, and carcinogens (De Meio, 1975; Jakoby et al., 1980; Yamazoe and Kato, 1995). These reactions are catalyzed by cytosolic sulfotransferases (STs¹) in mammals, which transfer SO₃ (sulfate moiety) from 3'-phosphoadenosine-5'-phosphosulfate (PAPS) to substrates.

Species differences are often observed on sulfations of chemicals among human and experimental animal species. For example, a model substrate, 7-hydroxycoumarin (umbelliferone), was preferentially conjugated with sulfate in precision-cut liver slices from rats and mice. The rates of glucuronidation and sulfation were, however, similar in guinea pig, monkey, and human (Steensma et al., 1994). Mechanisms yielding species differences remained unclear and thus hamper the prediction of the metabolic property of chemicals in humans from experimental animal data.

STs are known to constitute a gene superfamily. This superfamily contains at least five different classes, ST1, ST2, ST3, ST4, and ST5 families in mammals, which are based on their similarities of deduced amino acid sequences (Yamazoe et al., 1994; Weinshilboum et al., 1997; Nagata and Yamazoe, 2000). ST1 family is further subdivided into five subfamilies: ST1A, ST1B, ST1C, ST1D, and ST1E.

ST1A forms mainly mediate sulfations of phenols. From human-derived cDNA libraries, three distinct cDNAs of ST1A forms, ST1A2 (also called STP2 or SULT1A2), ST1A3 (STP1, TS-PST, or SULT1A1), and ST1A5 (STM, TL-PST, or SULT1A3), have been isolated and characterized (Zhu et al., 1993; Ozawa et al., 1994; Dooley and Huang, 1996). ST1A2 and ST1A3 catalyze the sulfations of simple phenolic chemicals such as p-nitrophenol (p-NP). ST1A3 shows higher catalytic activity and affinity for p-NP than does ST1A2. In contrast, ST1A5 shows a trivial activity for p-NP, but has high affinity for dopamine (Veronese et al., 1994; Lewis et al., 1996; Fujita et al., 1999b). These data indicate the distinct substrate specificities of ST1A form in spite of their high extents of sequence similarity.

Several cDNAs corresponding to ST1A forms have been isolated from experimental animal species (Fig. 1). Enzymatic properties of these forms have not been characterized except rat ST1A1 (Ozawa et al., 1993), bovine ST1A6 (Henry et al., 1996), and dog ST1A7 (Oddy et al., 1997). Rat ST1A1 was isolated and characterized as the first ST1A form by our laboratory (Ozawa et al., 1990). Mouse St1a4 cDNA was also isolated (Kong et al., 1993), but the enzymatic property remains yet to be characterized. We have recently isolated a new ST1A cDNA from a male rabbit liver library and termed it ST1A8.

View larger version (16K): [in this window] [in a new window]   Fig. 1.   Dendrogram of ST1A forms. The tree was produced by the unweighted pair group method with arithmetic mean using the GeneWorks software (IntelliGenetics, Mountain View, CA). ST, used as prefix, represents sulfotransferase. The names of individual forms are based on the proposed nomenclature system as described (Nagata and Yamazoe, 2000).

Enzymatic properties of human and experimental animal ST1A forms have been characterized using model substrates such as p-NP and dopamine, but not simultaneously compared using recombinant ST1A forms. Different from human forms, only a single form of ST1A has been reported so far from each experimental animal species. These phenomena may imply the functional difference of ST1A forms among experimental animal species and humans. In addition, absolute tissue contents of each ST1A form have not been determined. Therefore, to address the underlying mechanism of apparent species difference in sulfations of chemicals, the present study has been done using the recombinant enzymes.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Materials. Restriction endonucleases, DNA modifying enzymes, and TaKaRa EX Taq were purchased from Takara Shuzo (Kyoto, Japan). A gt11 cDNA library of a male rabbit liver was obtained from CLONTECH (Palo Alto, CA). Enterokinase was obtained from Biozyme Laboratories, Ltd. (Gwent, UK). 2-Amino-4'-hydroxy-1-methyl-6-phenylimidazo[4,5-b]pyridine (4'-OH-PhIP) was kindly donated by Dr. K. Wakabayashi (National Cancer Center Research Institute, Tokyo, Japan) and 3-hydroxybenzo[a]pyrene (3-OH-B[a]P) was a generous gift from Dr. H. V. Gelboin (National Cancer Institute, Bethesda, MD). Sakuranetin and 5,7-dihydroxyflavanone were kindly donated from Prof. T. Kurokawa (School of Medicine, Tohoku University, Sendai, Japan). Isopropyl -D-thiogalactopyranoside, dithiothreitol, alkalinephosphatase-conjugated goat anti-rabbit IgG, 5-bromo-4-chloro-3-indolylphosphate, nitro blue tetrazonium, mefenamic acid, and phenolic chemicals used as substrates were purchased from Sigma Chemical Co. (St. Louis, MO). [³⁵S]PAPS (2000 mCi/mmol) was from New England Nuclear (Boston, MA). QIAexpress and Ni-nitrilotriacetic acid agarose were the product of Qiagen (Chatsworth, CA). Bio-Rad protein assay kit and SDS-polyacrylamide gel electrophoresis (PAGE) molecular weight standards (low range) were from Bio-Rad (Richmond, CA). All other chemicals used were of the highest grade available.

Methods.

Isolation of rabbit ST1A form (ST1A8) cDNA A gt11 cDNA library of a male rabbit liver was immunoscreened with anti-His-rat ST1A1 polyclonal antibody by the modified method as reported previously (Yoshinari et al., 1998). After the third screening, positive clones were isolated and purified. Then, DNA sequences of each clone were determined separately using dye primers and Thermo Sequenase with ABI373A DNA sequencer (Perkin Elmer Japan, Tokyo, Japan) according to the dideoxy method in conjugation with M13 phage cloning as described (Sambrook et al., 1989). The DNA sequences of the positive clones coincided with each other. Two oligonucleotides (rab.ST1A-5': GCGGATCCGATGACGATGACAAAATGGAGCTCATCCAGGACACGTCCCGC, and rab.ST1A-3': GCGCATGCCCCCTCACAGCTCTGAACGGAAGG) were designated to construct the expression vector. Oligonucleotides have BamHI and SphI restriction sites, respectively. The designated (rabbit ST1A8²) cDNA fragment was obtained by PCR. The PCR reaction mixture (50 µl) contained 5 ng of the template cDNA; 10 pmol of each 5' and 3' primers; 0.2 mM each of dATP, dCTP, dTTP, and dGTP; 0.5 units of TAKARA Ex Taq; and the Ex Taq buffer. After an initial denaturation at 94°C for 3 min, the amplification was performed for 25 cycles, with 1 min at 94°C for denaturation, 30 s at 55°C for annealing, 1 min at 72°C for extension, and a final extension period of 2 min at 72°C.

Construction of expression vectors, expression and purification of recombinant ST1A proteins. Designated ST1A cDNA fragments contained nucleotides encoding seven additional amino acid residues (GlySerAspAspAspAspLys), in which was included a sequence of recognition sites of enterokinase next to the N-terminal methionine of the native form. Designated human ST1A3, ST1A5, and rat ST1A1 cDNA fragments were obtained by PCR as described (Fujita et al., 1999a,b). The mouse St1a4 fragment was also obtained by PCR from a male mouse liver cDNA library using oligonucleotides as the primers (mST1A-5': GCGGATCCGATGACAAAATGGCTCAGAACCCCAGC, and mST1A-3': GCGTCGACCAGTGTTAGGACTGATGGC). They have BamHI and SalI restriction sites, respectively.

Antibody preparation and immunoblot analysis. Japanese White rabbits (2.5 kg, female) were immunized intradermally with 20 to 50 µg of each purified His-ST1A protein in complete Freund's adjuvant, and immunity was boosted intravenously with 20 to 50 µg of the protein 3 weeks later. One week after the boost, antisera were obtained and kept at 80°C until use.

Assay of sulfation. Sulfating activities were determined by the radioactivities of the metabolites obtained with [³⁵S]PAPS as a sulfate donor after thin layer chromatography (Yoshinari et al., 1998). A typical incubation mixture consisted of 50 mM Tris-HCl buffer (pH 7.4), 1 mM dithiothreitol, 20 mM MgCl₂, 10 µM substrate, 125 µM [³⁵S]PAPS (0.1-0.2 Ci/mmol), and 50 ng of His-ST1A protein in a final volume of 10 µl. The reaction was initiated by addition of [³⁵S]PAPS and terminated by addition of 5 µl of chilled acetonitrile after incubation at 37°C for 20 min. A portion (10 µl) of the reaction mixtures was applied to a thin layer plate (chromatogram sheet 13255; Kodak, Rochester, NY; or thin layer chromatography aluminum plate silica gel 60; Merck, Darmstadt, Germany). Metabolites on the chromatogram were developed with a solvent system of n-propanol/ammonia/water (6:3:1). The radioactive spots were analyzed by a BAS1000 image analyzer (FujiFilm, Tokyo, Japan). All the substrates dissolved in dimethyl sulfoxide (DMSO) were added to make the final DMSO concentration 0.01 to 0.04%. In the case of inhibitory assays, the final 0.08% DMSO concentration was used. Sulfating activities showed less than 3% difference between experiments with 0.04 and 0.08% DMSO toward salicylic acid, 6-hydroxymelatonin, harmol, naringenin, and dopamine. Molecular weights of 36,160 for His-ST1A5; 36,174 for His-ST1A3; 35,870 for His-ST1A1; 36,680 for His-St1a4; and 35,805 for His-ST1A8, which were derived from their deduced amino acid sequences, were used for the determination of each sulfating activity. ST1 families' K_m values for PAPS are about 0.5 to 1.0 µM (Fujita et al., 1999b), and there are species differences in PAPS tissue concentrations (Klaassen and Boles, 1997). In the present study, we used a high PAPS concentration (125 µM) and relatively short-periods of incubation to minimize the deviation of the reaction. Deviations start to occur over 20 µM p-NP and troglitazone under our assay conditions.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Comparison of Deduced Amino Acid Sequences. The deduced amino acid sequences of ST1A forms examined in the present study (human ST1A3, ST1A5, rat ST1A1, mouse St1a4, and rabbit ST1A8) and the percentage of identities are shown in Fig. 2 and Table 1, respectively. These ST1A forms share more than 69% identity with each other. The highest homology is observed between ST1A3 and ST1A5 (93%). All the experimental animal ST1A forms show slightly higher identities with human ST1A3 than ST1A5. Especially rabbit ST1A8, newly isolated from a male rabbit liver, is more closely related (more than 81% identities) to the human ST1A forms than equivalents in rat and mouse (Table 1). As shown in Fig. 2, arrows indicating 121Gln, 185Thr, and 267Thr of ST1A3 and corresponding positions of other ST1A forms are selective residues to ST1A forms. Site A (L^A/_SLLP^Q/_E^T/_SLLDQKV^V/_IKV^V/_IY), site B (L^S/_R^R/_HTHPV), and site C (SLPEET) are highly conserved within ST1A forms, but differ from all the other families or subfamilies of cytosolic sulfotransferases reported (ST1 to ST5 families). The newly isolated ST1A8 also conserves the corresponding positions and regions. Among these sites, site A has been recognized as a variable region among cytosolic sulfotransferases (Yamazoe et al., 1994; Varin et al., 1995; Weinshilboum et al., 1997).

View larger version (98K): [in this window] [in a new window]   Fig. 2.   Alignment of deduced amino acid sequences among the ST1A forms. Deduced amino acid sequences of ST1A5, ST1A3, ST1A1, St1a4, and ST1A8 were aligned. Boxes indicate conserved regions among these ST1A forms. Arrows represent position 121, 185, and 267 of human ST1A3 and corresponding positions of the other forms, which are the ST1A-specific residues. Sites (site A: LA/SLLPQ/ET/SLLDQKV/IKVV/IY; site B: LS/RR/HTHPV; and site C: SLPEET) are highly conserved within ST1A forms.

Contents of ST1A Proteins in Liver Cytosols. To clarify the enzymatic properties of ST1A forms, recombinant ST1A proteins, termed His-ST1As, were expressed in E. coli and purified by nickel-affinity chromatography. Contents of ST1A forms in liver cytosols were determined by immunoblot analyses using each anti-His-ST1A antibody and corresponding His-ST1A protein as the standard (Table 2). The antibodies did not cross-react with other ST1 subfamilies (ST1B, ST1C, ST1D, and ST1E forms) and ST2A forms. ST1A3 and ST1A5 are immunostained with anti-His-ST1A5 antibodies, but distinguished by their different mobilities on SDS-PAGE.

Sulfating Activities (Comparison of Substrate Specificities). The His-ST1A proteins have 17 additional amino acid residues (MetArgGlySerHisHisHisHisHisHisGlySerAspAspAspAspLys), including a histidine tag and a sequence for the recognition site of enterokinase next to the N-terminal methionine of the native form. The additional peptide fused to the N terminus of STs has been shown to have minimal influence on kinetic parameters (Marsolais and Varin et al., 1995; Fujita et al., 1999a), although a slight difference was observed on isoproterenol sulfation (Lewis et al., 1996). In our experiments using His-ST1A3 and His-ST1A3, both proteins showed consistent results on p-NP sulfation (K_m = 3.00 and 2.62 µM, V_max = 3.98 and 3.47 nmol/nmol/min, V_max/K_m = 1.32 and 1.33, respectively). Thus, His-ST1A was used for our present experiments as shown below.

Influence of Mefenamic Acid on ST1A-Mediated Sulfations. Recently, mefenamic acid has been shown to inhibit p-NP sulfation in human livers (Vietri et al., 2000). Therefore, the influence of this chemical was assessed using recombinant ST1A forms from human and experimental animal species (Table 8). Sulfating activities toward all the chemicals (salicylamide, 6-hydroxymelatonin, naringenin, harmol, and dopamine) of human ST1A3 and experimental animal ST1A forms were decreased in the presence of 1 µM mefenamic acid, although activities catalyzed by human ST1A5 were not altered in the presence of 10 or 100 µM inhibitor. Extents of the inhibition at 10 µM mefenamic acid were largely consistent throughout the substrates examined for each ST1A form (percentage of control activities: ST1A3, 2.3-10%; ST1A1, 0.9-9.8%; St1a4, 5.8-16.3%; and ST1A8, 49.7-74.1%, respectively). Activities of ST1A3, ST1A1, and St1a4 were diminished to about 5 to 10% of controls in the presence of 10 µM mefenamic acid, whereas the ST1A8 activities remained 50 to 70% of the control.

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

The present study using recombinant ST1A sulfotransferases has shown that all the experimental animal ST1A forms (rat ST1A1, mouse St1a4, and rabbit ST1A8) have substrate specificities similar to those of human ST1A3 compared with ST1A5. These forms exhibited substrate preferences for simple phenolic chemicals rather than for catecholamines (Tables 3 and 5). As shown in Table 4, experimental animal ST1A forms also showed substrate specificities comparable with human ST1A3 toward endogenous indoles. The activities catalyzed by human ST1A3 and experimental animal ST1A forms were decreased drastically in the presence of 10 µM mefenamic acid, but sulfating activities catalyzed by human ST1A5 were not inhibited by 100 µM mefenamic acid (Table 8). Hepatic contents were similar among human ST1A3 and experimental animal ST1A forms, whereas the content of human ST1A5 was about 19 times less than that of ST1A3 (Table 2).

To examine in detail the substrate specificities, apparent kinetic parameters for p-NP, dopamine, and 6-HM were determined and compared among these ST1A forms (Table 6). The K_m value of human ST1A3 for p-NP was lower than that of human ST1A5 and the K_m values for dopamine were opposite between the two forms. Their values were largely consistent with those previously reported (Veronese et al., 1994; Lewis et al., 1996; Fujita et al., 1999b). The ST1A forms from experimental animals also showed low K_m values for p-NP, and they showed high K_m values toward dopamine. Thus, these data also support the functional similarity of experimental animal ST1A forms to human ST1A3.

Although the relationship between the structure and function remains obscure, the residue 146 is proposed to be crucial in determining the substrate specificity of both human ST1A3 and ST1A5 (Dajani et al., 1998). Corresponding positions of all experimental animal ST1A forms are alanine as well as human ST1A3, whereas that of human ST1A5 is glutamic acid (Fig. 2). A new ST1A form, ST1A8, has been isolated from a male rabbit liver library and characterized in the present study. At amino acid sequence level, ST1A8 is more related to the human ST1A forms than equivalents in the other experimental animal species. Its enzymatic properties were similar, but not identical with human ST1A3, rat ST1A1, and mouse St1a4. For instance, K_m value of ST1A8 for dopamine was much higher than that of human ST1A3, although K_m values for p-NP were nearly the same between ST1A8 and ST1A3. In addition, the sulfating activities of ST1A3, ST1A1, and St1a4 were inhibited about 90 to 95% in the presence of 10 µM mefenamic acid, whereas those of ST1A8 were 30 to 50% inhibited (Table 8). The present study also shows first the enzymatic properties of mouse St1a4. This form shared enzymatic properties similar with rat ST1A1, but the gender difference pattern of hepatic contents was opposite from ST1A1.

As shown in Table 7, experimental animal ST1A forms showed clear differences on enzymatic properties from human ST1A3 toward chemicals such as troglitazone and 4'-OH-PhIP. These results suggest that human ST1A3 has broader substrate specificities than experimental animal ST1A forms. Human ST1A3 showed higher activities toward estradiol and troglitazone than experimental animal ST1A forms. These data suggest that ST1A3 is able to catalyze the sulfation of the phenolic chemicals with large molecular sizes.

On the other hand, human ST1A5 exhibited substrate preferences for biogenic amines such as dopamine, norepinephrine, and normetanephrine. But toward 4-hydroxy-3-methoxyphenylglycol, a metabolite of normetanephrine by monoamine oxidase A, ST1A3, ST1A1, and St1a4 also showed similar activities. Moreover, the sulfating activity of ST1A5 toward 3,4-dihydroxyphenylacetic acid was lower than for the other ST1A forms. The amine residue of catecholamine is thus likely to affect the substrate specificities of ST1A5 strongly. In addition to catecholamines, ST1A5 catalyzed sulfations of harmol, 6-hydroxymelatonin, non-nitrogen-containing naringenin, and its derivative 5,7-dihydroxyflavanone effectively.

6-Sulfoxymelatonin is known to be a major urinary metabolite of melatonin (Arendt et al., 1985). All the ST1A forms showed low K_m and high V_max values for 6-hydroxymelatonin, which suggests that these ST1A forms may contribute to an excretion of melatonin.

Toward HIAA, sulfating activities catalyzed by all ST1A forms were not detected. We also examined for 5-hydroxy-L-tryptophan, L-dopa, and salicylic acid, but sulfating activities toward these chemicals were not detected or very limited (data not shown). Thus, these data suggest phenolic chemicals having carboxyl group are not preferred substrates for ST1A forms.

Human ST1A3 and ST1A5 have been studied in the present study, although another form, ST1A2, is known to express in human liver (Ozawa et al., 1995). ST1A2 shows similar enzymatic properties with ST1A3 and is a minor form compared with ST1A3. Human ST1A3 mRNA was found to be the major transcript form in the liver, representing 43 to 89% of the three related ST1A mRNAs (Ozawa et al., 1998).

The dog ST1A form (arbitrarily termed ST1A7) is reported to sulfate both simple phenols such as p-NP and catecholamines such as dopamine (Oddy et al., 1997). It shows, however, low specificities toward tyramine and serotonin. Thus, the dog ST1A7 seems to be an ortholog of human ST1A3 rather than ST1A5. In our preliminary examination, genomic Southern blot analysis using full-length St1a4 cDNA probe suggests a single gene copy of ST1A form in mice (data not shown). Southern blot analysis of rat genomic DNA also indicates the presence of a single gene (ST1A1 gene) copy per haploid genome (Khan et al., 1993). The present simultaneous comparison on substrate specificities and hepatic contents indicates little functional similarity between human ST1A5 and experimental animal ST1A forms. Distribution of human ST1A5 is mainly in extrahepatic tissues, such as brain, platelet, and small intestine (Young et al., 1984; Van Loon and Weinshilboum, 1984; Aksoy and Weinshilboum, 1995). Human plasma contains the highest concentration of catecholamines compared with those observed in most experimental animal species (Dousa and Tyce, 1988). These phenomena may imply the recent evolution of ST1A5 in primates for response to high demands of the metabolism.

In conclusion, experimental animal (rat, mouse, and rabbit) ST1A forms showed substrate specificities similar to human ST1A3 rather than ST1A5 for several phenolic chemicals and thus they are likely to be ST1A3 orthologs functionally. These forms, however, showed enzymatic properties distinct from human ST1A3 on the sulfations of some chemicals such as troglitazone and 4'-OH-PhIP.