A prototype of the mammalian sulfotransferase 1 (SULT1) family in Xenopus laevis : molecular and enzymatic properties of Xl SULT1B.S

The cytosolic sulfotransferase 1 (SULT1) proteins are a family of highly divergent proteins that show variable expansion in different species during vertebrate evolution. To clarify the evolutionary origin of the mammalian lineage of the SULT1 family, we compiled Xenopus laevis and X. tropicalis SULT1 ( X SULT1) sequences from public databases. The X SULT1 family was found to comprise at least six subfamilies, which corresponded in part to five mammalian SULT1 subfamilies but only poorly to zebrafish SULT1 subfamilies. SULT1C was most highly expanded, and could be divided into at least five subfamilies. A cDNA for X. laevis SULT1B ( Xl SULT1B.S), a homolog of mammalian SULT1B1, was cloned and its recombinant protein was expressed in a bacterial system. Xl SULT1B.S, unlike mammalian SULT1B1, was found to be a monomeric protein of ~34 kDa, and displayed sulfonating activity toward 2-naphthol and p -nitrophenol ( p NP). However, we could not detect such sulfonating activity toward any endogenous compounds including thyroid hormones, steroid hormones and dopamine, despite the fact that X. laevis and Rana catesbeiana liver cytosols contained sulfonating activity toward most of these endogenous compounds. At optimum pH (6.4), the Michaelis–Menten con-stant ( K m ) for p NP was two orders of magnitude greater in Xl SULT1B.S (1.04 mM) than in the cytosol preparations (8–15 μ M). Our results indicate that Xenopus possesses a prototype of the mammalian SULT1 family, and that Xl SULT1B.S showed overall similarities in primary sequence to, and significant differences in molecular and enzymatic properties from, mammalian SULT1B1.


INTRODUCTION
Sulfonation is a conjugation reaction in metabolic pathways that is responsible for the detoxification of a wide variety of xenobiotics with hydroxyl or amino groups, and is catalyzed by the cytosolic sulfotransferase (SULT) superfamily of enzymes (Blanchard et al., 2004;Coughtrie, 2016). SULTs can also sulfonate endogenous compounds such as thyroid hormones (THs), steroid hormones and neurotransmitters (Strott, 2002). In this reaction, 3'-phosphoadenosine 5'-phosphosulfate (PAPS) is utilized as a sulfonate donor and 3'-phosphoadenosine 5'-phosphate (PAP) is generated as a reaction product (Klaassen and Boles, 1997). Since sulfonation reactions generally lead to an increase in the polarity of acceptor substrates, they can accelerate urinary and biliary excretion of their substrates (Zamek-Gliszczynski et al., 2006;Lim et al., 2016).
The mammalian SULT superfamily contains at least seven families, SULT1 to SULT7 (Suiko et al., 2017), among which SULT1, a phenol SULT family, has been extensively investigated. SULT1 consists of five subfamilies, namely SULT1A (with substrate preference for phenolic compounds and dopamine), SULT1B (for THs), SULT1C (for hydroxyarylamine or acetylaminofluorene), SULT1D (for tyrosine ester) and SULT1E (for estrogens), with somewhat overlapping substrate profiles. SULT1A and SULT1B are the major enzymes that act in detoxification of many xenobiotics in human. This nomenclature system is based on amino acid sequence identity, being greater than 45% within a family and at least 60% within a subfamily (Weinshilboum et al., 1997;Blanchard et al., 2004).
Apart from mammalian SULT1, zebrafish SULT1 has been intensively investigated during the past two decades. Nine different cDNAs belonging to the SULT1 family have been cloned and expressed in Escherichia coli to date (Mohammed et al., 2012), and their products have been characterized enzymatically (Sugahara et al., 2003a(Sugahara et al., , 2003bLiu et al., 2005Liu et al., , 2008Yasuda et al., 2005aYasuda et al., , 2005b. The members of zebrafish SULT1 showed approximately 45-56% amino acid identity to some members of the mammalian SULT1 family (Liu et al., 2005;Yasuda et al., 2005b), and more than 43% amino acid identity was found among the members of zebrafish SULT1 (Yasuda et al., 2005a;Mohammed et al., 2012). Therefore, the members of zebrafish SULT1 are not classified into any of the present mammalian SULT1 subfamilies. Phylogenetic analysis assigned the nine zebrafish SULT1s to two novel groups: (1) SULT1ST5 and SULT1ST6, and (2) the other seven SULT1s. The former group is closer to the present mammalian SULT1 subfamilies than the latter group (Liu et al., 2008).
As a first step toward addressing the evolutionary and functional origin of the mammalian SULT1 family and the relationship between the fish and mammalian SULT1 subfamilies, we focused on amphibian SULT1 members. Initial BLAST searches, using public databases, detected 17 Xenopus laevis and 15 X. tropicalis SULT1 subtypes. As human SULT1B1 is an important SULT1 with a substrate preference for THs, in this study we have cloned and characterized a X. laevis homolog of mammalian SULT1B1.
Substrates of SULTs were dissolved in dimethyl sulfoxide to a final concentration of 5 mM, except for pNP (500 mM). These dissolved substrates were then diluted with enzyme reaction buffers, to give a final concentration of less than 1% (v/v) solvent.
Animal care and ethics American bullfrogs (Rana catesbeiana), 456 g males and 254-478 g females, were collected from ponds in Saitama Prefecture, Japan, in September 2018. African clawed frogs (X. laevis), weighing 40-60 g for both males and females, and tadpoles were obtained from a commercial supplier. The animals were maintained under laboratory conditions at 20-25 °C for more than one week, after which they were anesthetized by immersion in 0.1% 3-aminobenzoic acid ethyl ester.
All animal housing and experiments were conducted in accordance with the guidelines for the care and use of laboratory animals of Shizuoka University (permit #29F-8) according to international guidelines, covered by the Act on Welfare and Management of Animals (Ministry of the Environment of Japan).
Preparation of liver cytosol Livers were dissected and then perfused with ice-cold frog Ringer solution (111 mM NaCl, 3.4 mM KCl, 2 mM CaCl 2 , 2.3 mM NaHCO 3 ) containing 0.2 mg/ml heparin, as described previously (Yamauchi and Tata, 1994). The liver was minced with scissors in ice-cold frog Ringer's without heparin and then rinsed in the same medium several times. Minced tissues were homogenized in 4.5 vol. of 0.25 M sucrose, 10 mM Tris-HCl, pH 7.5, 1 mM ethylenediaminetetraacetic acid (EDTA), 1 mM MgCl 2 , 1 mM dithiothreitol, 1 mM phenylmethanesulfonyl fluoride and 1 mM benzamidine hydrochloride, with a Potter-Elvehjem homogenizer. The homogenate was then centrifuged at 1,200 × g for 15 min at 4 °C to remove the nuclear pellet. The post-nuclear supernatant was further centrifuged at 12,000 × g for 20 min at 4 °C to separate the crude mitochondrial/lysosomal pellet from the supernatant. The resulting postmitochondrial/lysosomal supernatant was centrifuged at 105,000 × g for 2 h at 4 °C to obtain the clear supernatant (cytosol), which was then stored in 10% glycerol at -80 °C until required.
Phylogenetic analysis Protein sequences of SULT1 family members were collected from public databases (NCBI, http://www.ncbi.nlm.nih.gov/; Ensembl, http://www. ensembl.org/index.html; Xenbase, http://www.xenbase.org/ entry/) (Supplementary Table S1). Since functional protein annotations in Xenopus databases were incomplete, we selected possible functional sequences, based on the sequence alignment. Unusual sequences that had insertions or deletions in central or terminal regions were excluded. As an out-group, we also collected SULT2 sequences.
Protein sequences of the SULT1 family were aligned using ClustalW ver. 2.1 (Thompson et al., 1994) in the DNA Data Bank of Japan (https://www.ddbj.nig.ac.jp/services. html). Phylogenetic trees of amino acid sequences of the SULT family were constructed using the maximum likelihood method (WAG model), the neighbor joining method (JTT model) and the minimum evolution method (JTT model) in the MEGAX program (Kumar et al., 2018). cDNA cloning Total RNA was extracted from the trunk region of X. laevis tadpoles using the LiCl-urea method (Auffray and Rougeon, 1980), as previously described (Kudo et al., 2006). RNA integrity was evaluated by electrophoresis in an agarose gel containing 2.6 M formaldehyde, and 28S and 18S ribosomal RNAs were stained with ethidium bromide and then visualized using an image analyzer (LAS-4000 miniEPUV, GE Healthcare, Little Chalfont, UK). Total RNA was reverse-transcribed into cDNAs using reverse transcriptase (TaqMan Reverse Transcription Reagents, Applied Biosystems, Foster City, CA, USA) and oligo (dT) 16 primers (2.5 μM) according to the manufacturer's instructions. Polymerase chain reaction (PCR) of the cDNAs was then carried out using Taq DNA polymerase (TaKaRa Ex Taq, Takara, Otsu, Shiga, Japan) with a specific primer set (each at 1.0 μM; see Supplementary Table S2), which was designed using the cDNA sequence (GenBank accession no. BC053792) for X. laevis hypothetical protein MGC64389 (renamed XlSULT1B.S in the present study). An approximately 1.1-kbp PCR-amplified product was then ligated into a T-vector (pT7Blue, Takara). Lastly, plasmid DNAs were purified and the insert DNAs were sequenced.
Real-time PCR Total RNA was extracted from the liver, kidney, stomach, intestine and skeletal muscle of 1-year-old young adult X. laevis (males n = 2 and females n = 2) using the LiCl-urea method (Auffray and Rougeon, 1980). The quantity of specific RNA species in each sample was estimated by real-time PCR using SYBR Green Master Mix and an ABI Prism 7000 (Applied Biosystems) after the RNA samples had been treated with reverse transcriptase (TaqMan Reverse Transcription Reagents, Applied Biosystems), as previously described (Kudo et al., 2006). Each PCR was run in duplicate to control for PCR variation. Detailed information of primer sets is shown in Supplementary Table S2. Primer specificity was confirmed by BLAST searches, the appearance of a single band on gel electrophoresis, and melting curve analysis. The thermocycler program included a step of 50 °C (2 min) and 95 °C (10 min), followed by 40 cycles of amplification at 95 °C (15 sec) and 60 °C (1 min), and then a final step at 50 °C (2 min). The endpoint used for realtime PCR quantification, threshold cycle (C T ), was defined as the PCR cycle number that crosses an arbitrarily placed signal threshold, and is a function of the amount of target DNA present in the starting material. Control reactions lacking reverse transcriptase were tested for residual genomic DNA, and contamination was evaluated by non-template controls where no RNA was present in the cDNA synthesis step. All negative controls had undetermined C T or C T > 36. Because there was less variation in the C T values of the ribosomal protein L8 gene (rpl8) transcript among the tissues examined (C T = 22.38 ± 0.69, n = 5) than in those of the glyceraldehyde-3-phosphate dehydrogenase gene (C T = 22.42 ± 4.26, n = 5) and eukaryotic elongation factor 1α gene (C T = 35.06 ± 1.74, n = 5) transcripts, rpl8 was used as a reference transcript to standardize each experiment. The transcript amount for XlSULT1B.S in each sample was divided by the amount of rpl8 transcript in the same sample by the 2 − ΔΔC T method (Livak and Schmittgen, 2001).
Purification of recombinant XlSULT1B.S The cDNA encoding XlSULT1B.S was amplified by PCR using specific primers (Supplementary Table S2). After digestion with BamHI and SalI, the cDNA was ligated into the BamHI/ SalI site of pGEX-6P-3 (GE Healthcare). XlSULT1B.S was expressed as a glutathione S-transferase (GST) fusion protein in E. coli Rosetta 2(DE3)pLysS (Merck Millipore, Billerica, MA, USA) cells in 25 ml LB medium in the presence of 0.1 mM isopropyl-1-thio-β-D-galactopyranoside, incubated overnight at 16 °C, and purified from the bacterial extracts by affinity chromatography on a glutathione-Sepharose matrix (GE Healthcare) according to the manufacturer's instructions. After washing the matrix (25 μl bed volume) with phosphate-buffered saline (10 mM Na 2 HPO 4 , 1.8 mM KH 2 PO 4 , pH 7.3, 2.7 mM KCl, 140 mM NaCl) three times, and with PreScission buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM dithiothreitol), the fusion protein was digested with PreScission protease (4 U) for 4 h at 5 °C. The XlSULT1B.S portion was recovered in the supernatant after centrifugation at 500 × g for 5 min at 4 °C.
Protein analyses Proteins were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) in a 10% acrylamide gel according to Laemmli (1970) with molecular weight markers (LMW Calibration Kit for SDS Electrophoresis, GE Healthcare), and visualized by staining with Coomassie Brilliant Blue R-250. Protein concentration was determined by the micro-Lowry method (Jain et al., 2002), using bovine serum albumin (BSA) as the standard, and read with a microplate reader (Multiskan FC, Thermo Fisher Scientific, Waltham, MA, USA).
Sulfotransferase assay Sulfonation of pNP by recombinant XlSULT1B.S (1 μg) was performed in 14.9 μl of 100 mM KH 2 PO 4 -Na 2 HPO 4 buffer, pH 6.4, 4.8 mM p-NP, 7.6 mM 2-mercaptoethanol, in the presence or absence of 50 μM PAPS, for 10 min at 25 °C (Ramaswamy and Jakoby, 1987), unless otherwise noted. The reaction was stopped by the addition of an equal volume of 1% acetic acid in methanol. After centrifugation at 17,800 × g for 10 min, the supernatant containing pNPS was analyzed by reverse phase-HPLC on a C18 analytical column (Mightysil RP-18 GP, 250 mm × 4.6 mm, 5 μm particle diameter, Kanto Chemical, Tokyo, Japan) at a flow rate of 1.0 ml/min at 40 °C under isocratic conditions (methanol: distilled water: acetic acid = 44:55:1). Chromatographic elution was monitored at 279 nm. Retention times for pNP and pNPS were 7.98 ± 0.09 (n = 5) and 9.30 ± 0.01 min (n = 5), respectively. The amount of pNPS was quantified by comparison with the pNPS standards (20-400 pmol). The detection limit was 1.3 pmol, which was evaluated by measuring the baseline noise and pNPS amounts that gave S/N = 3.
For sulfonation of 50 μM substrates (T3, rT3, T4, estrone, β-estradiol, dehydroxyepiandrosterone, pNP and 2-naphthol) and an amine (dopamine) by liver cytosol (25 μg), the enzyme reaction was carried out in 100 mM KH 2 PO 4 -Na 2 HPO 4 buffer, pH 7.2, 7.6 mM 2-mercaptoethanol, in the presence or absence of 50 μM PAPS, for 10 min at 25 °C. In this experiment, the amount of PAP that was converted from PAPS in the sulfonation reactions of phenol or amine compounds was analyzed by reverse phase-HPLC (Sheng et al., 2001) on a C18 analytical column at a flow rate of 1.5 ml/min at 40 °C under isocratic conditions (methanol: buffer [75 mM KH 2 PO 4 , 75 mM NH 4 Cl, pH 5.45, 845 μM n-octylamine] = 15.8:84.2) at 259 nm. Retention times for PAPS and PAP were 19.13 ± 0.13 (n = 5) and 11.26 ± 0.07 (n = 5) min. The amount of PAP was quantified by comparison with PAP standards (20-200 pmol). The detection limits of PAP varied depending on the enzyme solutions used. These were determined to be ~1 and 3.8 pmol in the assays of the recombinant XlSULT1B.S and the liver cytosols, respectively.
Statistics The data are presented as the mean ± standard deviation (n ≥ 3), unless otherwise noted. Differences between groups were analyzed by a one-way analysis of variance, with the Scheffe multiple comparisons test. P values < 0.05 were considered statistically significant. Statistical analyses were conducted using Microsoft Excel 2003 Data Analysis Software (SSRI, Tokyo, Japan).

RESULTS
Xenopus and other vertebrate SULT1 family members A sequence-based BLAST search and a textbased keyword search in public databases detected 17 X. laevis and 15 X. tropicalis SULT1 (XSULT1) members, of which four X. laevis and five X. tropicalis proteins had an additional or deleted region, or were products of pseudogenes. Ultimately, we selected 13 X. laevis and 10 X. tropicalis SULT1 members as possible functional proteins. Phylogenetic analysis was conducted using 23 XSULT1 and other vertebrate SULT1 (19 mammalian and chicken, 15 fish SULT1) sequences with 17 SULT2 sequences as an out-group (Supplementary Table S1). Figure 1 indicates that the SULT1 family is more highly diverged than has hitherto been considered. The SULT1 family consists of three major groups: 1C, tetrapod 1A/1B/1D/1E (1A/1B/1D/1E t ) and fish 1A/1B/1D/1E (1A/1B/1D/1E f ). Amino acid identities within SULT1 sequences (Supplementary Table S3) revealed that group 1C can be divided into mammalian 1C (1C m ), three Xenopus 1C (1Ca x , 1Cb x and 1Cc x ) and two fish 1C (1C f ; 1ST5 f and 1ST6 f ) subfamilies. The evolutionary relationships among the subfamilies, except for that between the two fish subfamilies, were not clear because the bootstrap values among the nodes of the subfamilies 1C m , 1Ca x and 1Cb x and between the nodes of the 1Cc x and fish 1C f subfamilies were less than 50%. Nevertheless, SULT1Ca x shared the highest amino acid identities (49-61%) with SULT1C m .
The XSULT1 family consists of at least six subfamilies (Fig. 1). The members of the XSULT1 family shared at least 44% and 58% amino acid identities within a family and a subfamily, respectively (Supplementary Table S3), and shared 29-38% with the members of the XSULT2 family (data not shown). XSULT1C consists of three subfamiles, accompanied by 3-5 members possessing an additional or deleted region in both Xenopus species, whereas the other XSULT1 subfamilies have one or two members. Figure 2 shows the alignment of the deduced amino acid sequences of the XSULT1B subfamily with the higher vertebrate SULT1B subfamily. Amino acid sequence identities were   Supplementary Table S1. As the protein annotations were somewhat controversial and incomplete in the Xenopus species, tentative names of proteins and subfamilies were used in this study. The X. laevis sequences with post-fixed letters S and L indicate homeologous proteins. Similar tree topologies were obtained using the neighbor joining and minimum evolution methods. SULT1 is composed of three major groups: 1C, and tetrapod and fish 1A/1B/1D/1E (1A/1B/1D/1E t and A/1B/1D/1E f , respectively). 1C consists of one higher vertebrate (1C m ), three Xenopus (1Ca x , 1Cb x and 1Cc x ), and two fish 1C (1C f ; 1ST5 f and 1ST6 f ) subfamilies, whereas 1A/1B/1D/1E t consists of tetrapod 1B (1B t , to which 1B m and 1B x belong), higher vertebrate 1A, 1D and 1E (1A m , 1D m and 1E m , respectively), and Xenopus 1A/1D x and 1A/1D/1E x subfamilies. Nodes with bootstrap values higher than 50% are represented by bold lines. 86-87% among the members of the XSULT1B subfamily (XtSULT1B, XlSULT1B.S and XlSULT1B.L), and 59-72% among the members of the higher vertebrate SULT1B subfamily. These proteins, like other mammalian SULTs (Weinshilboum et al., 1997), had sequences that resemble the signature sequences (Kakuta et al., 1998), YPKSGTxW (residues 46-53 in Fig. 2, for the 5'-phosphosulfate binding site) and RNAKDxVVSYY (residues 131-141, for the 3'-phosphate binding site), and also the conserved amino acid sequence RKGxxGDWKNxFT (residues 258-270) (Weinshilboum et al., 1997). The KTVE motif that has been proposed as a dimerization motif in mammalian SULT1 sequences (Petrochenko et al., 2001) was conserved in XtSULT1B and XlSULT1B.L as well as human SULT1B1 (HsSULT1B1), partially conserved in mouse and chicken SULT1Bs (MmSULT1B1 and GgSULT1B), but not conserved in XlSULT1B.S. cDNA cloning and tissue distribution of xlsult1b.s transcript A cDNA of 1,030 bp was obtained by conventional reverse transcription-PCR. It contained one open reading frame consisting of 294 amino acids. A BLAST search using a public database (NCBI, http://www.ncbi. nlm.nih.gov/) identified it as a cDNA for XlSULT1B.S. Its transcript was found to be abundant in the kidney and liver, and was also detected in the stomach and intestine (Fig. 3). The amounts of xlsult1b.s transcript in muscle were at background levels. xlsult1b.s transcript levels in kidney, stomach and intestine were approximately twofold higher in females than in males; however, there was no sexual difference in transcript levels observed in the male and female liver.  Residues that are perfectly conserved among these sequences are marked by asterisks. Two signature sequences (Kakuta et al., 1998), in the N-terminal (for the 5'-phosphosulfate binding site) and middle regions (for the 3'-phosphate binding site), and a conserved sequence in the C-terminal region (Weinshilboum et al., 1997) are represented by horizontal lines above the human sequence. The KTVE dimerization motif overlapping the C-terminal region (Petrotchenko et al., 2001) is boxed in the human sequence.  Total RNA was extracted from tissues of adult females (n = 2) and males (n = 2), reverse-transcribed using random primers and then amplified by real-time PCR. Relative expression levels for xlsult1b.s were normalized against the rpl8 expression level, and set at 1.0 in the female liver. Data for each tissue represent the mean ± SD in four assays (four RNA preparations from two animals).

Characterization of XSULT1B sequences
ble GST-fusion protein (~58 kDa) in E. coli (Fig. 4A). The XlSULT1B.S moiety (34 kDa) was purified from GST-Sepharose after digestion with PreScission protease. The observed molecular mass was in agreement with that calculated from the predicted amino acid sequences (34,626 Da). The yield of the recombinant XlSULT1B.S was approximately 100 μg/25 ml bacterial culture. SULT1s are known to be generally homodimers (~68 kDa) in solution (Petrotchenko et al., 2001). Although molecular size prediction by gel filtration is not accurate, the recombinant XlSULT1B.S is clearly smaller than BSA (68 kDa) and OA (45 kDa), but larger than Mb (17 kDa) (Fig. 4B); i.e., the recombinant XlSULT1B.S does not form homodimers.
Enzymatic properties of recombinant XlSULT1B.S pNP sulfonation, i.e., the amount of PAPS-dependent production of pNPS, was found to be XlSULT1B.Sdependent. The product increased steadily as the protein amount was increased up to 1 μg at 25 °C (data not shown). Examining the pH dependency of the pNPsulfonating activity revealed that XlSULT1B.S had a pH optimum of 6.4 in 0.1 M phosphate buffer (Fig. 5A). pNP sulfonation increased linearly for the initial 15 min, and then gradually reached a plateau, at 25 °C (Fig. 5B). Therefore, we performed the subsequent enzyme reactions using 1 μg of XlSULT1B.S in 0.1 M phosphate buffer, at pH 6.4, for 10 min at 25 °C. pNP-sulfonating activity was measured with pNP concentrations ranging from 0.4 to 4.8 mM in the presence of 50 μM PAPS (Fig. 5C). The double-reciprocal plot of the rate of pNP sulfonation versus the pNP concentration (Fig. 5D) indicated that the Michaelis-Menten constant (K m ) value for pNP was 1.04 mM and the maximum velocity (V max ) value was 21.3 nmol/min/mg protein (Table  1). pNP-sulfonating activity was also measured using PAPS concentrations from 2.5 to 50 μM in the presence of 4.8 mM pNP (Fig. 5E). The double-reciprocal plot of the rate of pNP sulfonation versus the PAPS concentration (Fig. 5F) gave a K m value for PAPS of 11.6 μM and a V max value of 17.3 nmol/min/mg protein (Table 1).
Substrate preference was measured with 50 μM of either an endogenous or xenobiotic compound in the presence of 50 μM PAPS. XlSULT1B.S appeared to be more active toward 2-naphthol (Table 2). pNP sulfonation by XlSULT1B.S was one order of magnitude less than 2-naphthol sulfonation under our experimental conditions. Sulfonating activities of XlSULT1B.S toward all of the endogenous compounds tested were found to be at background levels (≤ 100 pmol/min/mg protein).
Enzymatic properties of SULTs in liver cytosols pNP-sulfonating activity of SULT1 in X. laevis and R. catesbeiana liver cytosols was measured over pNP concentrations ranging from 0.4 to 4.8 mM in the presence of 50 μM PAPS or over PAPS concentrations ranging from 2.5 μM to 50 μM in the presence of 4.8 mM pNP, in 0.1 M phosphate buffer, pH 7.2 ( Table 1). The apparent K m values of both cytosol enzymes for pNP (8-15 μM) were two orders of magnitude less than that of the recombinant XlSULT1B.S, whereas the apparent K m values of   Each value is the mean ± SD (n = 3). V max , maximum velocity; K m , Michaelis-Menten constant. Different letters (a, b and c) denote significantly different means among the three groups.
both cytosol enzymes for PAPS (8-13 μM) were comparable to that of the recombinant XlSULT1B.S. The X. laevis and R. catesbeiana liver cytosols exhibited sulfonating activity against a variety of endogenous and exogenous compounds (Table 2). 2-Naphthol and pNP were the most preferred substrates among the compounds tested. In the case of endogenous compounds, the X. laevis cytosol exhibited a high sulfonating activity toward rT3, and the R. catesbeiana cytosol toward dopamine and rT3. Sulfonating activity toward dehydroxyepiandrosterone was not detected in either cytosol.

DISCUSSION
Xenopus SULT1 family and the evolution of the SULT1 family Our study indicates that the Xenopus SULT1 family comprises at least six subfamilies. The amino acid identity of the Xenopus SULT1 sequences within a subfamily was at least 58% and within a family was at least 44%. Our data also suggested that the Xenopus SULT1 family is more highly diverged than the mammalian SULT1 family.
Of the six XSULT1 subfamilies, the XSULT1B subfamily was a homolog of the mammalian SULT1B subfamily. It is also possible that the XSULT1Ca (Ca x ) subfamily is a homolog of the mammalian SULT1C subfamily. The XSULT1C subfamilies (1Ca x , 1Cb x and 1Cc x ) may have expanded to a greater extent than the higher vertebrate XSULT1C (1C m ) subfamily. Although the XSULT1Cb (1Cb x ) subfamily is annotated as "SULT1B member 1 S homeolog" (NCBI, http://www.ncbi.nlm.nih.gov/), amino acid identities and phylogenetic analyses revealed that the XSULT1Cb (1Cb x ) subfamily may not be a homolog of the mammalian SULT1B subfamily. There were also moderate sequence similarities between the Xenopus and higher vertebrate SULT1 subfamilies (1A/1D x vs. 1A m , and 1A/1D/1E x vs. 1A m , 1D m and 1E m ); however, we were not able to clearly show significant evolutionary relationships between them.
Considering the teleost SULT1 subfamilies, we cannot exclude the possibility that Xenopus has one or more novel SULT1 subfamilies that are distinct from the SULT1 subfamilies found in mammals. A previous phylogenetic analysis (Yasuda et al., 2005b) indicated that, although two zebrafish SULT1C f s (1ST5 f and 1ST6 f ) formed a monophyletic clade with all members of the mammalian SULT1 family, these were out-groups of the mammalian SULT1 family. The other zebrafish SULT1 members (1A/1B/1D/1E f ) cluster as a sister clade to the clade consisting of the mammalian SULT1s, 1ST5 f and 1ST6 f (Liu et al., 2008;Mohammed et al., 2012). Although this topology of the phylogenetic tree was not the same as ours, both analyses indicate that the zebrafish SULT1 can be divided into two groups, 1C f and 1A/1B/1D/1E f , neither of which belong to any mammalian or Xenopus SULT1 subfamilies. Therefore, there may be no zebrafish homologs of the higher vertebrate and Xenopus SULT1 subfamilies. Similar phylogenetic relationships were obtained in a catfish (Pterygoplichthys anisitsi) SULT1 family (Parente et al., 2017), where 10 SULT1 members clustered into two clades. One is more closely related to the human SULT1A subfamily, and the other is distant from any mammalian SULT1 subfamilies. In view of these studies and our analysis, a likely scenario in the evolution of the vertebrate SULT1 family is as follows: the SULT1 family greatly expanded during the early evolution of vertebrates, and some of the SULT1 subfamilies have been retained in the bony fish lineages. A prototype of the mammalian SULT1 family was gradually established during the evolution of the tetrapod vertebrates. These phylogenetic relationships also indicate that the current nomenclature of SULT1, which is largely based on the mammalian SULT1 members (Blanchard et al., 2004), cannot be simply applied to the classification of SULT1 subfamilies of other vertebrates.
Molecular, transcriptional and enzymatic properties of recombinant XlSULT1B.S XlSULT1B.S has three conserved regions containing two signature sequences of the SULT superfamily for the 5'-phosphosulfate and 3'-phosphate binding sites (Kakuta et al., 1998). As is the case in mouse SULT1E1, XlSULT1B.S was found to be monomeric in solution, although SULT1s are known to be generally homodimers in solution. The dimerization motif (KTVE motif, KxxxTVxxxE) is conserved in many mammalian SULT1s, except for mouse SULT1E1 with KxxxPExxxE (Petrotchenko et al., 2001). Each value is the mean ± SD (n = 3). Specific activity refers to pmol substrate/min/mg protein. ND, Specific activity observed is lower than the detection limit (estimated to be < 100 pmol/min/mg for purified XlSULT1B.S, and < 15 pmol/ min/mg for liver cytosols).
As XlSULT1B.S has RxxxTExxxE, a single substitution of Val to Glu in this motif may be critical for a monomeric form, as has been shown in the mutation study of human SULT1E1 (Petrotchenko et al., 2001). We detected overall similarities and significant differences in tissue-specific expression patterns and enzymatic properties between XlSULT1B.S and human SULT1B1. The xlsult1b.s transcript was abundant in kidney > liver > stomach/intestine, whereas human sult1b1 expression was shown in liver, colon and intestine, while almost no expression was seen in kidney (Wang et al., 1998). XlSULT1B.S may thus have a specific role in kidney. There were similarities in substrate preference between XlSULT1B.S and human SULT1B1, although we could not detect sulfonating activity of XlSULT1B.S toward any endogenous compounds. XlSULT1B.S exhibited a high sulfonating activity for 2-naphthol and a moderate activity for pNP, whereas human SULT1B1 had at least one or two orders of magnitude higher sulfonating activity for 2-naphthol and pNP than XlSULT1B.S, with a preference for 2-naphthol > pNP > rT3 > T4 > T3 (Fujita et al., 1997). As both the K m and V max values of XlSULT1B.S for pNP (1.04 mM and 21.3 nmol/min/mg) were greater than that of human SULT1B1 (24.1 μM and 5.1 nmol/min/mg), XlSULT1B may function at higher concentrations of pNP than human SULT1B1.
Although we detected significant TH-sulfonating activity in X. laevis cytosol as well as R. catesbeiana cytosol, the recombinant XlSULT1B.S showed little or no TH-sulfonating activity. As the detection limit of THsulfonating activity was relatively high (~100 pmol/min/ mg) for purified recombinant XlSULT1B.S in our SULT assay using the HPLC system, we cannot exclude the possibility that SULT activities with endogenous substrates were below the detection limit. Alternatively, it is likely that the recombinant enzyme could not reproduce its natural structure, resulting in lower activity than the cytosolic native enzyme. We also detected a hundred-fold greater K d value for pNP in the recombinant XlSULT1B.S compared to SULTs of the cytosol. This result may also support the above possibilities.
In conclusion, Xenopus was found to have at least six SULT1 subfamilies, whose amino acid sequences were moderately related to those of the mammalian SULT1 subfamilies. Amino acid identities and phylogenetic analysis demonstrated that the vertebrate SULT1 family consists of more highly diverged subfamilies than have hitherto been recognized, and that the XSULT1B subfamily is a homolog of the mammalian SULT1B subfamily. Transcriptional and functional analysis revealed overall similarities and significant functional differences between XlSULT1B.S and human SULT1B1. XlSULT1B.S may have sulfonating activity at least for 2-naphthol and pNP.

None.
Supplementary Subfamilies with subscripts represent those of mammals and chicken (m), Xenopus (x) and fish (f).