Cloning and Characterization of Cytosolic Sulfotranferases from Channel Catfish Liver

Material Information

Cloning and Characterization of Cytosolic Sulfotranferases from Channel Catfish Liver
Merritt,Kristen Kay
Place of Publication:
[Gainesville, Fla.]
University of Florida
Publication Date:
Physical Description:
1 online resource (137 p.)

Thesis/Dissertation Information

Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Pharmaceutical Sciences
Medicinal Chemistry
Committee Chair:
James, Margaret O
Committee Members:
Sloan, Kenneth B
Luesch, Hendrik
Barber, David S
Graduation Date:


Subjects / Keywords:
Amino acids ( jstor )
Catfish ( jstor )
CDNA libraries ( jstor )
Enzyme substrates ( jstor )
Enzymes ( jstor )
Estrogens ( jstor )
Gels ( jstor )
Liver ( jstor )
Polymerase chain reaction ( jstor )
Protein isoforms ( jstor )
Medicinal Chemistry -- Dissertations, Academic -- UF
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Pharmaceutical Sciences thesis, Ph.D.


Sulfotransferase (SULT) enzymes are responsible for Phase II detoxification of xenobiotics, activation or deactivation of drugs, and regulation of many endogenous compounds, including estrogen and androgen. Studies suggest that the sulfotransferase superfamily may play an important role in early development. The enzymes may also be involved in endocrine disruption of aquatic species, as evidenced by disruption of reproductive systems and sexual differentiation, following exposure to organic pollutants that inhibit sulfotransferases. Despite the potential importance of these enzymes to aquatic pollution, sulfotransferases in fish have been little studied. Here, sulfotransferases from liver of channel catfish (Ictalurus punctatus) were cloned and expressed. Total RNA was extracted from catfish liver tissue and used to construct cDNA libraries. Specific primers were designed based on published expressed sequence tag (EST) sequences, and full sequences amplified via Rapid Amplification of cDNA Ends Polymerase Chain Reaction (RACE PCR). Three complete sulfotransferase sequences have been fully cloned from cDNA libraries, two SULT1 family members and one SULT2. Additionally, a partial sequence of SULT6B1 was cloned. Each SULT1 sequence encodes 301 amino acid proteins with predicted molecular masses of ~ 35 kDa; the SULT2 sequence encodes 287 amino acids with a predicted molecular mass of 34 kDa. Both SULT1 sequences have high identity (47-52%) with multiple 1-family isoforms, the highest being mouse SULT1D1 (52%) and human SULT1C4 (51%). The SULT2 sequence has highest identity with human SULT2B1 (44%). The SULT2 isoform was expressed with the pMAL system and purified with affinity and DEAE chromatography. Activity assays were carried out with the ?free? form of the enzyme from which the fusion-binding protein had been removed. The recombinant sulfotransferase exhibited activity with DHEA, with Km of 43.7 ?M and Vmax of 516 pmol/min/mg. No detectable activity was observed with endogenous compounds testosterone, 17?-estradiol or xenobiotics 6-hydroxymethly-benzo[a]pyrene, 7-hydroxymethyl-12-methylbenzanthracene, 3-hydroxy-benzo[a]pyrene, or with any of four hydroxy-polychlorinated biphenyls. These studies provide evidence for at least four distinct hepatic SULT enzymes belonging to three different families (SULT1, SULT2, and SULT6) in the catfish liver. Supported in part by the US Public Health Service, ES 07375. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis (Ph.D.)--University of Florida, 2011.
Adviser: James, Margaret O.
Electronic Access:
Statement of Responsibility:
by Kristen Kay Merritt.

Record Information

Source Institution:
Rights Management:
Copyright Merritt,Kristen Kay. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
LD1780 2011 ( lcc )


This item has the following downloads:

Full Text




2 2011 Kristen Belcher Merritt


3 Dedicated to my son, Joshua, who graciously shared his mom with her schooling during his formative years, And to my father, from whom I got this wonderful, crazy scientific brain; Dedicated to his memory (Lee Belcher, 1940-2007)


4 ACKNOWLEDGMENTS I would like to thank m y committee member s, Drs. David Barber, Hendrik Luesch, Kenneth Sloan, and especially my sponsor, Dr. Margaret James, for giving me this opportunity and for her patience. Thanks also to Dr. Pete r Anderson, Director of Whitney Laboratory, for giving me a place there to do my research. Ev eryone at Whitney contributed to my success, answering my many questions, givi ng me guidance, and I would like to express my appreciation to them all. In particular, thank you to Dr. Mi ke Matz for teaching me how to make a really great cDNA library, and to Andr ea Kuhn for showing me how to approach a PCR challenge in ingenious ways until the full sequence is obtained. Great appreciation goes to artist and friend Larry Behnke, for his tremendous support and encouragement and for his graphics assistance, part icularly the drawing for the title page (that I wasnt able to use).


5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ............................................................................................................... 4 LIST OF TABLES ...........................................................................................................................8 LIST OF FIGURES .........................................................................................................................9 LIST OF ABBRE VIATIONS....11 ABSTRACT ...................................................................................................................... .............13 CHAP TER 1 INTRODUCTION .................................................................................................................. 15 SULT Details ..........................................................................................................................16 Structure ..................................................................................................................... .....16 Nomenclature .................................................................................................................. 23 Known SULTs .................................................................................................................24 Substrates .................................................................................................................... .....27 Bioactivation ................................................................................................................. ...27 Benzylic alcohols .....................................................................................................28 Aromatic amines ...................................................................................................... 30 Allylic alcohols ........................................................................................................31 Secondary nitroalkanes ............................................................................................ 32 Regulatory Roles ............................................................................................................. 32 Neurosteroids ...........................................................................................................32 Catecholamines ........................................................................................................ 33 Estrogen ....................................................................................................................34 Thyroid hormones ....................................................................................................37 Endocrine Disruption .......................................................................................................39 Research Aims .................................................................................................................43 2 GENE IDENTIFICATION ..................................................................................................... 46 Materials and Methods ...........................................................................................................46 Materials ..................................................................................................................... .....46 Methods ...........................................................................................................................46 cDNA library ............................................................................................................46 Total RNA extraction ...............................................................................................47 RNA LiCl cleanup .................................................................................................... 47 cDNA first strand synthesis ...................................................................................... 49 cDNA amplification ................................................................................................. 49


6 Primer design ............................................................................................................ 50 Experiments ................................................................................................................... ..51 Standard protocol for cloning ...................................................................................51 Gel extraction, ligation, cu lture of PCR product ......................................................53 Colony screen ...........................................................................................................54 Sequencing ............................................................................................................... 54 Catfish Sulfotransferase Isoforms ................................................................................... 56 Results and Discussion ........................................................................................................ ...56 Results .............................................................................................................................56 cDNA libraries .........................................................................................................56 PCR experiments ......................................................................................................57 Discussion .................................................................................................................... ....69 SULT1v1 and SULT1v2 ..........................................................................................70 SULT2 ......................................................................................................................78 Previously determined catfish SULT fragments ...................................................... 87 3 RECOMBINANT PROTEIN EXPRESSION ........................................................................ 89 Materials and Methods ...........................................................................................................89 Materials ..................................................................................................................... .....89 Methods ...........................................................................................................................90 Generation of coding sequence insert ...................................................................... 91 Preparing insert and vector for ligation ....................................................................91 Pilot experiments ...................................................................................................... 93 Scaled-up expression, separation, and purification ..................................................98 Experiments and Results ....................................................................................................... 101 Generating Full-length Insert Sequence ........................................................................ 101 Insert and Vector Preparation, Ligation ........................................................................ 103 S1v1 ligation ..........................................................................................................104 SULT2 ligation .......................................................................................................106 4 ENZYME ASSAYS ............................................................................................................. 111 Materials and Methods .........................................................................................................111 Materials ..................................................................................................................... ...111 Methods .........................................................................................................................112 Phenolic substrate 3OH-BaP ..................................................................................112 Steroid substrates estrog en and testosterone ..........................................................114 Alcoholic and OH-PCB substrates ......................................................................... 114 Results ...................................................................................................................................116 3OH-BaP ....................................................................................................................... 116 Endogenous Substrates ..................................................................................................116 Xenobiotic Substrates .................................................................................................... 117 Discussion .................................................................................................................... .........119 5 SUMMARY ....................................................................................................................... ...122


7 LIST OF REFERENCES .............................................................................................................129 BIOGRAPHICAL SKETCH .......................................................................................................137


8 LIST OF TABLES Table page 1-1 Known mammalian sulfotransferases, their m ost common substrates, and tissues they are found in. .......................................................................................................................45 2-1 Primers used to make catfish cDNA libraries .................................................................... 50 2-2 Primers designed at the start and end of each corresponding EST sequence .. ................. 55 2-3 Summary of screening P CR experiments based on seve n EST sequences mined from the NCBI database .............................................................................................................58 2-4 Primers designed to obtain the entire sequence for SULT1v1. ...................................... 60 2-5 Primers designed to obtain the entire SULT1v2 sequence from channel catfish liver cDNA libraries. ......................................................................................................... .63 2-6 Primers designed to obtain the entire SULT1v2 sequence from channel catfish liver cDNA libraries. ......................................................................................................... .66 3-1 Primers used to prepare recombinant channel catfish sulfotransferase enzymes for expression. ................................................................................................................... ......92 3-2 Conditions of digests and cl eanup for S1v1and SULT2 inserts ......................................109 3-3 Conditions of digest an d cleanup for pMAL vector ........................................................110 4-1 Summary of varying experimental conditi ons for activity assays with recombinant catfish SULT2 and 3OH-BaP. .........................................................................................116 4-2 Activity of xenobiotic substrates with purified recombinant channel catfish SULT2 and positive control (human)SULT2A1. ......................................................................... 118


9 LIST OF FIGURES Figure page 1-1 Reaction catalyzed by the sulfotransferase superfamily. ................................................ 15 1-2 Representative structure of (h uman)SULT1A1 in complex with 3phosphoadenosine-5-phosphate (PAP) and two p-nitrophenol ( pNP) molecules ............ 17 1-3 Proposed reaction mechanism for sulfotransferases.. ........................................................ 20 1-4 Comparison of crystal structures of SULT1 isoforms complexed with 3phosphoadenosine-5-phosphate (PAP) and substrates pnitrophenol ( pNP) and 17 estradiol (E2) ..................................................................................................................... 21 1-5 Dissection diagrams of a representative cytosolic sulfotransferase (SULT) name, shown as splice variant or allele. ..................................................................................... 24 1-6 Structures of representational chemicals bioactivated by cytosolic sulfotransferases. ......29 1-7 Sulfonation of the primary catecholamines to the form in which they predominately exist in circulation. ......................................................................................................... ....34 1-8 Structure of mouse SULT1E1 in th e presence of PAP and substrate 17 -estradiol. ......... 35 1-9 Regulation of thyroid hormones T4, T3, and diiodothyronine (DIT) ................................ 38 1-10 Zebrafish offspring affected by parent al intake of 2,3,7,8-te trachlorodibenzo-pdioxin. ....................................................................................................................... .........41 2-1 Flowchart of steps involved in obtaining sequence data for a gene, beginning with the creation of a cDNA library. ..........................................................................................48 2-2 Nondenaturing agar gel of total RNA ex tracted from channel catfish liver. ...................49 2-3 PCR products for two cDNA libraries, CAP T-30 and TRSA. ......................................51 2-4 Amino Acid sequence of the coding region of a putative family 1 sulfotransferase, SULT1v1, from channel catfish liver tissue. ................................................................62 2-5 Amino acid sequence of a sulfotransfe rase from channel catfish liver tissue putatively identified as Sult1v2. ........................................................................................65 2-6 Translated amino acid sequence of a sulf otransferase from channel catfish liver tissue. ..................................................................................................................... ..........67 2-7 Alignment (ClustalW2) of the translat ed partial gene sequence obtained from channel catfish liver with othe r bony fish SULT6B1 sequences. ......................................71


10 2-8 Alignment of translated channel ca tfish sequences SULT1v1 and SULT1v2 with human sulfotransferases 1A1a and 1E1.. ...........................................................................74 2-9 Alignment (Clustal W2) of channel catfish SULT1 sequences with bonyfish SULT1 sequences sharing highest percent identity ........................................................................77 2-10 Alignment of the channel catfish puta tive SULT2 translated amino acid sequence with the human SULT2 isoforms sh aring highest % identity. .........................................80 2-11 Conserved motif 3PB determined from the alignments of channel catfish SULT2 amino acid sequence and the mammalian and bony fish sequences sharing highest % identity. ...........................................................................................................................81 2-12 Alignment of the channel catfish puta tive SULT2 amino acid sequence with the bony fish SULT2 isoforms shar ing highest identity.. .................................................................83 2-13 Alignment of sequence fragments from a native sulfotransferase purified from channel catfish liver.. .........................................................................................................87 3-1 Flow chart of steps taken to prepare pM AL expression vector and insert sequence. ........90 3-2 Pilot expression experiment with S2 samples #12 and #14. .............................................. 93 3-3 SDS-PAGE gel of pilot pur ification of channel catfish SULT2 recombinant protein with the pMal vector expression system. ........................................................................ 96 3-4 SDS-PAGE gels of pilot cleavage using SULT2 recombinant protein, expressed with the pMAL vector system. ................................................................................................. 97 3-5 Expression of recombinant SULT2 under various growth conditions, run on a denaturing SDS-PAGE gel.. ..............................................................................................99 3-6 Concentration estimate of S1v1 and S2 PCR products ....................................................103 3-7 Example of successful ligation for S1v1 and pMAL vector. ......................................... 104 3-8 A280 measurements of fract ions collected during affin ity column purification and DEAE column purification of recomb inant catfish SULT2 enzyme. ............................ 108 4-1 Structures of xenobiotics tested for enzymatic activity with recombinant channel catfish SULT2 sulfotransferase. ................................................................................113 4-2 Michaelis-Menten plot of catfish SULT2 activity with substrate DHEA. .......................118


11 LIST OF ABBREVIATIONS 1SS First-strand synthesis product 3OH-BaP 3-Hydroxy-benzo[a]pyrene 6OHM-BaP 6-Hydroxymethyl-benzo[a]pyrene 7OH-12MBA 7-Hydroxymethyl, 12methyl-benzo[a]anthracene AHR Aryl hydrocarbon receptor -Me -Mercaptoethanol BSA Bovine serum albumin DEAE Diethylaminoethyl anion exchanger DHEA Dehydroepiandrosterone dNTPs deoxyribonucleotide phosphates DTT Dithiothreitol E2 17estradiol EC Estrogenic compounds EDTA Ethylenediaminetetraacetic acid ER Estrogen receptor EST Expressed sequence tags IPTG Isopropyl -D-thiogalactopyranoside Kan Kanamycin LB Luria broth LiCl Lithium chloride LA Taq Long Acting DNA polymerase (thermostable) MBP Maltose-binding protein MeOH Methanol MgCl2 Magnesium chloride


12 MnCl2 Manganese chloride MWCO Molecular weight cutoff NaCl Sodium chloride NaOH Sodium hydroxide NCBI National Center for Biotechnology Information OD600 Optical density at 600 nm wavelength OH-MPACs Hydroxy-methyl polyc yclic aromatic hydrocarbons OHPCBs Hydroxy-polychl orinated biphenyls PAGE Polyacrylamide gel electrophoresis PAPS 3-Phosphoadenosine-5-phosphosulfate PCR Polymerase chain reaction P(H)AH Polycyclic (halogena ted) aromatic hydrocarbons PMSF Phenylmethylsulfonyl fluoride pNP p-nitrophenol RACE-PCR Rapid Amp lification of cDNA Ends Polymerase Chain Reaction SDS Sodium dodecyl sulfate SOC Super Optimal broth with glucose TBE Tris-Borate-EDTA Xgal 5-Bromo-4-chloro-3-indolyl -D-galactopyranoside


13 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy CLONING AND CHARACTERIZATION OF CY TOSOLIC SULFOTRANSFERASES FROM CHANNEL CATFISH LIVER By Kristen Belcher Merritt August 2011 Chair: Margaret O. James Major: Pharmaceutical Sciences Sulfotransferase (SULT) enzymes are responsible for Phase II detoxification of xenobiotics, activation or deactivation of drugs, and re gulation of many endogenous compounds, including estrogen and androgen. Studies s uggest that the sulfotransfera se superfamily may play an important role in early development. The enzymes may also be involved in endocrine disruption of aquatic species, as evidenced by disruption of reproductive systems a nd sexual differentiation, following exposure to organic pollutants that inhi bit sulfotransferases. Despite the potential importance of these enzymes to aquatic pollution, sulfotransferases in fish have been little studied. Here, sulfotransferases from liver of channel catfish ( Ictalurus punctatus ) were cloned and expressed. Total RNA was extracted from cat fish liver tissue and used to construct cDNA libraries. Specific primers were designed ba sed on published expressed sequence tag (EST) sequences, and full sequences amplified via Rapi d Amplification of cDNA Ends Polymerase Chain Reaction (RACE PCR). Three complete su lfotransferase sequences have been fully cloned from cDNA libraries, two SULT1 family members and one SULT2. Additionally, a partial sequence of SULT6B1 was cloned. E ach SULT1 sequence encodes 301 amino acid proteins with predicted molecular masses of ~ 35 kDa; the SULT2 sequence encodes 287 amino


14 acids with a predicted molecular mass of 34 kDa. Both SULT1 sequences have high identity (47-52%) with multiple 1-family isoforms, the highest being mouse SULT1D1 (52%) and human SULT1C4 (51%). The SULT2 sequence has highes t identity with human SULT2B1 (44%). The SULT2 isoform was expressed with the pMAL sy stem and purified with affinity and DEAE chromatography. Activity assays were carried ou t with the free form of the enzyme from which the fusion-binding protein had been remove d. The recombinant sulfotransferase exhibited activity with DHEA, with Km of 43.7 M and Vmax of 516 pmol/min/mg. No detectable activity was observed with endogenous compounds testosterone, 17 -estradiol or xenobiotics 6hydroxymethly-benzo[a]pyrene, 7-hydroxymethyl-12-methylbenzanthracene, 3-hydroxybenzo[a]pyrene, or with any of four hydroxy-polychlorinated biphe nyls. These studies provide evidence for at least four distinct hepatic SULT enzymes belonging to three different families (SULT1, SULT2, and SULT6) in the catfish liver. Supported in part by the US Public Health Service, ES 07375.


15 CHAPTER 1 INTRODUCTION Sulfotransferases (SULTs) are a superfa mily of Phase II me tabolic enzymes involved in detoxification of both xenobiotic and endogenous chemicals as we ll as regulation of endogenous compounds. The enzymes catalyze the transfer of a sulfuryl group from a cofactor 3phosphoadenosine 5-phosphosulfate (PAPS) to nucle ophilic groups of substrates (Figure 1-1). Substrates are both endogenous and non-endogenous and are generally phenols, primary and secondary alcohols, and amines (Glatt, et al., 2001). Two mammalian classes of sulfotransferases are known: one is comprise d of membrane-bound forms located in the Golgi apparatus, which metabolizes endogenous macromol ecular structures; the ot her is composed of cytosolic forms that metabolize both xenobi otic and small endogenous compounds, such as hormones and neurotransmitters (Strott 2002). Th is research project addressed only cytosolic sulfotransferases. Figure 1-1. Reaction catalyzed by the sulfotrans ferase superfamily. A sulfuryl group is transferred from cofactor 3-phosphosdenosine 5-phosphosulfate (PAPS) to substrates with a phenol, alcohol, or amine group.


16 Addition of a sulfuryl group to a substrate can have varying results: it can increase a substrates solubility and enhance excretion, or can inactivate an endogenous compound, such as sex or thyroid hormones. In some cases, a non-endogenous compound is bioactivated with beneficial results, such as for the drug minoxid il, or detrimentally to a compound that can form DNA adducts (Gamage, et al., 2006). In recent years, studies have revealed that sulfotransferases play significant regulatory roles in thyroid and endocrine systems (Strott 2 002). Studies also suggest the enzymes may play important roles in embryonic de velopment, at least in mammal s (Wood, et al., 2003; Duanmu, et al., 2005). Chemical interference with, or inhibition of, the re gulatory enzymes involved could, potentially, result in developmenta l or reproductive abnormalities or deficiencies (Colburn, et al., 1996; Krimsky 2001). It has been shown that fish exposed to polychlorin ated biphenyls (PCBs) and hydroxylated PCBs suffer reproductive or developmental abnormalities (Walker and Peterson, 1991; Carlson and Williams, 2001). It has been suggested that inhibition of sulfotransferases by environmental pollutants is an indirect mechanism of endocrine disruption (Kester, et al ., 2002), as exhibited by aquatic and am phibious organisms (Jurgella, et al ., 2006; Wang and James 2007; Thibaut and Porte 2004). Thus, a better understanding of fish sulfotransferases is needed to investigate mechanis ms of toxicity that may be associated with the sulfonation pathway. SULT Details Structure Cytosolic su lfotransferases exist as homodi mers, with individual units comprised of approximately 280 to 300 amino acids, depending on the SULT family. Sulfotransferase crystal


17 structures reveal a characteristic / motif with a central sheet composed of five parallel strands surrounded by -helices on both sides. This region forms the PAPS binding site and catalytic center, including a conserved 5-phosphosulfate-b inding (PSB) loop involved in binding the cofactor (Lee, et al ., 2003). A representative stru cture of (human)SULT1A1 in complex with PAP and substrate is shown in Figure 1-2. The PSB-loop is comprised of a loop sequence connecting the firs t strand of the central -sheet with the first helix, composed of the conserved amino acid sequence TYPKSGT, which forms specific bond interactions with the 5phosphate of the PAP molecule (Pedersen, et al., 2000). A crystal structure of DHEAsulfotransferase in complex with subs trate is reported to have only four -strands making up the central sheet (Rehse, et al., 2002). Figure 1-2. Representative structure of (human)SULT1 A1 in complex with 3phosphoadenosine-5-phosphate (PAP) and two p-nitrophenol ( pNP) molecules, reprinted with permission from Gamage et al., 2006. The original caption states secondary structural elements are depict ed as coils for helices and arrows for strands. The bound ligands are shown as spherical atomic models; PAP, green; pNP1, orange; pNP2, yellow.


18 Crystal structures of SULTs bound with PAPS/PA P cofactor, with or without substrate, have led to a fairly good understanding of the amino acid residues responsible for PAPS binding/interaction, which are highly conser ved among all SULTs. Less is known, however, about the substrate binding regions and factors leading to substrat e specificity. Insight has also been gained from crystal structures regarding the phenomenon of substrat e inhibition seen with some isoenzymes, where sulfotransferase enzymes are active against a substrate at low concentrations but are inhi bited at high concentrations of the same molecule. A crystal structure of human SULT2A1 (ori ginally reported as SULT2A3) with PAP (3phosphoadenosine 5-phosphate) revealed the PSB-loop interaction and a 3-phosphate binding site involving residues Arg121 a nd Ser129, as well as backbone interactions with Lys248 and Gly249 near the enzyme carboxy terminal (Pedersen, et al ., 2000). Amino acid residues His99 and Lys44 were reportedly in a pos ition to assist in the dissociat ion of the sulfuryl group from PAPS. The substrate binding pocket was determined to be primarily constituted by residues Pro14-Ser20, Glu79-Ile82, and Asn136-Lys144. Th e four residues Glu79-Ile82 form a loop forming the opening of SULT2A1 (the authors term), which is compared to a similar loop of SULT1E1 that contains nine residues from Glu83 to Asn91. Similar comparisons between human dehydroepiandrosterone (DHEA) sulfotransferase and estrogen SULT were made by Rehse, et al who crystallized DHEA SULT/estradiol comp lex, absent cofactor (Rehse, et al., 2002). They observed that residue His99 form ed a hydrogen bond with the steroid O-3, which correlates with His108 for estrogen sulfotransfera se. They also observed two orientations of DHEA in the substrate binding poc ket, only one of which was catalytic, the other potentially leading to enzyme inhibition.


19 Crystal structures of human SULT2B1a and 2B1b isoforms with PAP and pregnenolone were examined to determine the basis for differen ces in substrate specificity (Lee, et al., 2003). SULT2B1a readily sulfonates pre gnenolone, but its activity with ch olesterol is minimal, whereas SULT2B1b, which can sulfonate both steroids, pr eferentially sulfonates cholesterol. Both isoforms are capable of sulfonating DHEA, but do so with relatively low efficiencies. The specificities for cholesterol and pregnenolone were determined to be due to an amino-terminal helix, comprising residues Asp19 to Lys26; in particular, residues 19DISEI23 are responsible for the ability of SULT2B1b to sulfonate chol esterol. In a crystal structure of mouse estrogen sulf otransferase complexed with PAP and estradiol substrate, conserved residues Lys48 and Ser138 were observed to directly interact with 5and 3-phosphate of PAP, respectively; the Ser138 inte raction most likely corr elates with the Ser129PAP interaction of SULT2A1 (Pedersen, et al., 2002). In addition, the aforementioned His108 directly coordinated to the acceptor group of the 17 -estradiol molecule. Ser138 is conserved in all known SULTS with no exception (Pedersen, et al., 2002), and its importance has been confirmed with site-directed mutagenesis studies, as was the importance of Lys48. In all known crystal structures the conserve d serine forms a hydrogen bond with the 3-phosphate of the PAP molecule, but no evidence has been seen for its i nvolvement in catalysis. A crystal structure of human estrogen SULT-PAPS complex revealed similar interactions of (mouse)SULT1E1 residues Ser138, Lys48, and His108 for human enzyme residues Ser137, Lys47, and His107 (Pedersen, et al., 2002). The inte raction with the 3-phosphate gr oup may be involved in specific binding and stability of PAP, or may play a role in regulati ng the action of Lys47 (Lys48 in mouse SULT) in controlling the dissociation of the 5sulfate group from PAPS in the absence of substrate. In the presen ce of substrate and upon His107 initiation of catalysis, Lys47


20 undergoes a conformational cha nge and interacts with the bridging oxygen of the phosphatesulfate bond of PAPS to promot e the dissociation of sulfuryl leaving group. The authors proposed a transfer reaction that proceeds through an SN2-like in-line displacement mechanism in which the conserved lysine and his tidine play essential roles (Figure 1-3 ). Figure 1-3. Proposed reaction mechanism (Pedersen et al., 2002, reprinted with permission) for sulfotransferases. In human SULT1E1 conserved residues Ser137, Lys47, and His107 coordinate to advan ce the sulfonyl tran sfer reaction. U pon binding of PAPS, Lys47 interacts with Ser137, preventing PAPS hydrolysis. The cata lytic base His107 deprotonates the acceptor 3-hydroxyl of the estradiol molecule and increases the nucleophilic character of this hydroxyl gr oup. Subsequently, the nucleophilic group attacks the sulfur atom, which builds up a partial negative ch arge on the bridging oxygen of the PAPS molecule. The side chai n of Lys47 then switches interactions from Ser137 to the bridging oxygen, aiding in sulfate dissociati on and transfer to substrate.


21 Figure 1-4. Comparison of crys tal structures of SULT1 isoforms complexed with 3phosphoadenosine-5-phosphate (PAP) (a,c) and substrates pnitrophenol ( pNP) and 17 -estradiol (E2), as noted. Reprinted with permission from Gamage, et al., 2003; the original caption is as follows: C omparison of SULT1 crystal structures. Structures of human SULT1A1 complexed with PAP ( green ), pNP1 ( orange ), and pNP2 ( blue ) ( a); human SULT1A3 complexed with sulfate ( pink ) (b ); and mouse SULT1E1 complexed with PAP ( green) and E2 ( pink ) ( c ). The blue regions in the SULT1A1 structure indicate parts of the prot ein that are disordered in SULT1A3. The dotted lines indicate disordered region s of SULT1A3 and SULT1E1. The sulfonation of estradiol (E2) can be cat alyzed by both (human)SULT1A1 and 1E1, but at much different efficienciesA1 catalyzes E2 at M levels while 1E1 is active with only nM


22 levels of substrate (Falany, et al., 1994 and 1995). In addition, E2 is sulfonated by a human DHEA sulfotransferase, SULT2A1, at M levels, without exhibiti ng substrate inhibition at concentrations as high as 6 M (Falany, et al., 1989; Wang and James 2005). Gamage, et al., (2003) examined crystal structures of (human)SULT1A1 in complex with 17 -estradiol (E2) and PAP, and separately with p-nitrophenol ( pNP) to understand the mechanisms of substrate inhibition and the ability of this enzyme to accomm odate such different substrates (Figure 1-4). In the latter complex, the conserved PSB-loop was noted at residues 45 to 51 (45TYPKSGT51). The 3-phosphate of PAPS was observed to interact with two conserved regions: 257RKG259 from the G XX G XXK SULT motif and residues Arg130 and Se r138. The adenine ring of PAP formed stacking and T-shaped interactions wi th conserved residues Trp53 and Phe229 and was stabilized by hydrogen bond in teractions between N-6 and Thr227, and N-3 and Tyr193. Substrate specificity was determined in part to be due to Ala146; a single-site mutation to glutamate (A146E) reduced affinity of SULT1A1 for pNP 400-fold (Gamage, et al., 2003). Substrate inhibition was explai ned by the presence of two pNP molecules in a single large L-shaped binding site, one of which was bound in a catalytically competent manner ( pNP#1). The phenol-hydroxy group of pNP#1 formed hydrogen bonds with side chains of catalytic residues His108 and Lys106 and with a well-ordere d water molecule. The substrate nitro group interacted with a water molecu le and formed van der Waals in teractions with Val148, Phe247, and Met248. The second, inactive pNP molecule was more weakly bound in its site and formed no interactions with catalytic residues. Kinetic studies showed s ubstrate inhibition with this substrate at c oncentrations above 2 M. Plasticity of the substrate binding pocket wa s seen with the SULT1A1-E2S-PAP complex (Gamage, et al., 2005). Two loop regions that close tightly over the SULT1A1pNP complex


23 open up for E2 binding, increasing the space availabl e for the larger fused-ring substrate. The two regions are residues 146-154, between two -helices, and residues 8490, just preceding an -helix. A substrate access gate formed by Phe142 and Phe81 permits binding of planar substrates, such as E2, only at the catalytic site; this formation is also present in (mouse)SULT1E1 (Gamage, et al., 2003). The su lfonated estradiol in this SULT1A1 crystal structure complex was bound in a catalytically incomp etent manner, which forms a dead-end complex; this formation, the authors theorize, explains substrate inhibition by estradiol. At high concentrations of substrate, an estradiol mol ecule could bind to SULT before the desulfonated PAP cofactor is released, rendering the enzyme inactive until both molecules are released. The SULT1A1 L-shaped substrate binding po cket is well-ordered and very hydrophobic (Gamage, et al., 2003). Its plas ticity, generated by conformationa l change, allows a substrate range greatly varied in size and shape, as exhibited by the complexes with pNP and E2. An additional example of large substrate binding is that of thyroid hormone T2 (Gamage, et al., 2003). A catalytically competent binding conf ormation model orients the substrate 4-OH toward catalytic residues and the two phenyl ring s (of T2) adopt an L-shaped conformation with respect to each other, which enable s each ring to occupy one of the two pNP binding sites. To accommodate this molecule, Phe247 would adopt a different (though still favorable) rotamer conformation to prevent a steric clash with one of the iodine atoms of the substrate. Nomenclature In 2004, guidelines were presented for nam ing cytosolic sulfotransferase enzymes, which have since been adopted (Blanchard, et al., 2004). The accepted superfamily name is SULT, followed by alternating Arabic num erals and capital letters desi gnating family and subfamily. All enzymes sharing at least 45% amino acid sequen ce identity are grouped into the same family;


24 within a family, enzymes with shared amino acid sequence identity of 65% share the same subfamily. Unique isoforms within a subfamily are identified with an Arabic numeral following the subfamily letter. Figure 1-5 shows dissection diagrams of the new naming system, reprinted from the original article. Species name code s are in parenthesis im mediately before SULT; codes used are three-to-five le tter code designations found at According to this lis t, the species code for channel catfish, Ictalurus punctatus is ICTPU. Figure 1-5. Dissection diagrams of a representa tive cytosolic sulfotransferase (SULT) name, shown as splice variant or allele. Reprin ted with permission from Blanchard, et al., 2004. Splice variants and alleles are specified with extensions at the end of the enzyme name. Protein sequences that differ due to alternative splicing are di fferentiated by _vx where x designates the sequential variant number (for example, (HUMAN)SULT2B1_v1 and _v2). Alleles should be named with an asterisk and Arabic numeral following the gene isoform number; alleles with synonym ous SNPs (encoding the same amino acid sequence) would be assigned a final capital le tter (e.g. *1A). Pseudogenes are identified by a P following the gene name. Known SULTs As of 2004, at least 56 distin ct eukaryotic SULT isoform s have been identified and functionally characterized (Blanc hard, et al., 2004), the major ity of them from mammalian


25 systems. To date, 11 human sulfotransferases have been identified, belonging to three families: SULT1, SULT2, and SULT4. Of these, three en zymes have known allelic variants (SULT1A1, 1A2, and 2A1) and one isoform is variant (SU LT2B1a and 2B1b) due to alternative splicing. The most extensive group of the human cytoso lic SULTs is the SULT1 family, which includes SULTs 1A1, 1A2, 1A3, 1B1, 1C1, 1C2, and 1E1. SULT4A1 is expressed specifically in the brain and has been identified in human, mouse, and rat tissue (Falany, et al., 2000; Sakakibara, et al., 2002), but no substrate has yet been found against which it is active. Some mammalian SULTs not found in humans include families 3 and 5 (e.g. 3A1 found in mouse and rabbit, 5A1 in mouse) and SULT1D1 found in mouse, ra t, and dog (Blanchard, et al., 2004). The majority of the fish sulfotransferases iden tified thus far have been from zebrafish, as a model for studying the enzyme. A review of th e National Center for Biotechnology Information (NCBI) protein database, accessed November 2009, lis ts 309 fish sulfotransferase sequences, all but 95 of which are Danio rerio. Of those 95 entries, 57 are id entified as unnamed protein product from the Tetraodon nigroviridis genome, putatively identified as SULT because they contain the domain sequence common to all members of the enzyme superfamily. The remaining 38 protein sequences, with few excep tions, are translated from mRNA sequences identified as putative sulfotransferases of va rying family designations, none of which have been confirmed or characterized. The seven ex ceptions consist of a sequence fragment from I. punctatus liver, submitted by th is lab; a 2006 entry of Leuciscus cephalus (European chub) characterized and identified as SULT1 isof orm 3; a non-cytosolic sulfotransferase for Oreochromis mossambicus (tilapia); a 191-base sequence fo r carbohydrate sulfotransferase from I. punctatus, unpublished data; two cytosolic sulfotransferases for Oryzias latipes (Japanese medaka) identified as SULT1 isoforms 2 and 3, unpublished data; a 191-amino acid sequence


26 from I. punctatus identified as a carbohydr ate sulfotransferase, unpublished data; and a 2002 submission from I punctatus for a 356-amino acid sequence identified as tyrosylprotein sulfotransferase. Therefore, the only charac terized sulfotransferase for a food-fish is a single isoform for the chub (Assem, et al., 2006). None of the characterized zebrafish and chub sulfotransferases have been similar enough to any named mammalian SULTs to be named to a subfamily; they have instead been given a family name (e.g. SULT1) followed by a numerical designation based on the order of discovery (e.g. Sulfotransferase family, cytosolic sulfotransferase 6). A current review of the NCBI protein database, accessed July 2011, revealed two major entries for bonyfish sequences. Leong, et al. (2010) submitted sequences for Atlantic salmon, Salmo salar and northern pike, Esox Lucius, obtained to study evolutionary changes. Chen, et al. (2010) published sequences for channel catfish, Ictalurus punctatus and blue catfish, Ictalurus furcatus, for genetic comparisons of the two closely related catfish species. All sequences were generated from cDNA libraries, from multiple organs and identified by BLASTX analyses. The corresponding proteins were not expressed or characterized. The entries include fifteen cytosolic sulfotransferases: Atlantic salmon sequen ces are identified as SULT2B1, SULT4A1, SULT6B1, sulfotransferase 1, sulfotransferase 2, and sulfotransferase 3; northern pike sequences are identified as SULT2 B1, SULT6B1 and sulfotransferase 3; blue catfish sequences are identified as SULT4A1, SULT6B1 and sulfotransferase 3, and channel catfish sequences are identified as SULT4A1, SULT2B1 and sulfotransferase 3. Three amino acid sequence fragments were obtai ned for a sulfotransferase isolated from channel catfish liver and intestine (Tong and James 2000), and the protein characterized by enzymatic assay. The purified enzyme, with an estimated molecular mass of 41 kDa, was


27 preferentially active with pheno lic substrates. It efficiently catalyzed the sulfonation of 9hydroxybenzo[a]pyrene, 2-naphthol, 4-nitropheno l, and 4-methylumbelliferone, and showed activity with male and female sex steroids dehydroepiandrosterone (DHEA), estrone, and 17 estradiol. The research proposed here is built on these findings. Substrates Although SULTs have a wide and varied range of substrates, there are a few generalizations that can be m ade according to family. The SULT1 family predominately catalyzes phenolic substrates, particularly estrogens; SULT1A3 exhibits selectivity toward catecholamines (Strott 2002). The preferred substrate for SULT1E1 is 17 -estradiol, with a Km in the low nM range (Zhang, et al., 1998), and for SULT1A3 is dopamine, with a Km of 1 M (Falany 1997). The SULT2 family preferentially catalyzes substrates w ith an alcohol moiety, particularly androgens. The SU LT2B1 variants, _v1 and _v2, differ in their activities toward the steroids pregnenolone and cholesterol; _v1 pref erentially sulfonates pregnenolone and only minimally sulfonates cholesterol, whereas _v2 is able to sulfonate both steroi ds but preferentially sulfonates cholesterol and with grea ter efficiency. Details of subs trate selectivity and expression in tissues for the enzymes that have been characterized are given in Table 1-1. Bioactivation While sulfotransferase-cataly zed reactions result in a m ore hydrophilic product that can be excreted via urine or bile, in some cases the su lfate ester produced acts as an excellent leaving group. In these cases, highly reactive carbenium or nitrenium ion species result, which can form protein or DNA, adducts (Banoglu 2000; Turesk y 2004). The most common chemical classes found to be bioactivated to car cinogenic and/or mutagenic form s are benzylic alcohols from polycyclic aromatic hydrocarbons (PAHs), N -hydroxy-aromatic amines (arylamines), allylic


28 alcohols, and secondary nitroalkan es. A few drugs, such as minoxidil, become active only after sulfotransferase catalysis. Benzylic alcohols These com pounds are commonly formed by metabolic hydroxylation of PAHs; carcinogenic or mutagenic PAHs may bear a prim ary or secondary benzylic hydroxyl functional group. Many lab results have shown the re quired metabolic processes for ultimate carcinogenicity of many alkyl-substituted PAHs to be hydroxylation at the benzylic position followed by sulfuric acid ester formation, ma king SULTS the main en zymes catalyzing the metabolic activation of hydroxylated PAHs. A se ries of studies in which female SpragueDawley rats were subjected to repeated subcutaneous injec tion of PAHs bearing primary benzylic alcohol groups (mainly forms of anthr acene and pyrene) indicate that the electrophilic sulfooxy metabolites of hydroxymethyl PAHs account for the complete carcinogenicity of these intermediary metabolites. Examples of primar y and secondary benzylic alcohols (BAs) subject to bioactivation are shown in Figure 1-6. Prim ary BAs, such as hydroxymethylbenzo[a]pyrenes, are primarily sulfonated by alcohol sulfotrans ferases (DHEA SULTs; SULT2 family), while competitive inhibition studies point to aryl SULTs (SULT1 family) being primarily responsible for sulfonating secondary BAs. Most carcinogenic/mutagenic PAHs bearing a secondary benzylic alcohol group are in the form of cyclic derivatives. The most studi ed PAHs in this class are the mono-and dihydroxy derivatives of cyclopenta[cd ]pyrene (CPP), a ubiquitous environmental and occupational pollutant. Other secondary BAs exhibiting muta genicity after activation by SULTs include 1OH-3-methylcholanthrene and oxidated metabolites of benzo[a]pyrene, particularly dihydrodiol and tetrol derivatives. Me thylene-bridged PAHs, which are universal pollutants,


29 Figure 1-6. Structures of repres entational chemicals bioactivated by cytosolic sulfotransferases. B(a)P: benzo[a]pyrene; C PP: cyclopenta[cd]pyrene produced mutagenicity in Ames tests utilizing S. typhimurium after incubations with rat and/or human liver cytosol (fortified with PAPS), or with recombinant rat SULT. In addition, noncyclic secondary BAs derived from PAHs such as 1-hydroxyethylpyrene and 6hydroxyethylbenzo[a]pyrene were also reporte d to induce SULT-catalyzed mutagenicity. Glatt, et al. (1994 and 1995) obs erved differences between species and gender for activation of BAs. Several benzylic alcohols were activ ated more efficiently by cytosol from rat liver, whereas others were more efficiently activated by human liver cytosol. In addition, strong sex differences in bioactivation were observed in th e rat but not in humans. For example, 1-(1-


30 pyrenyl)ethanol (1-HEP) was more efficiently bioactivated by hepatic cytosol from females than by preparations from males, and 1-(6-benzo[a]pyrenyl)ethanol, which showed activity in cytosol from females, showed no activity in hepatic cytosol from males. In addition, 1-HEP was bioactivated 67 times greater by human than by rat hydroxysteroid SULT (SULT2 family), while 7-hydroxymethyl-12-methylbenz[a ]anthracene was bioactivat ed 27 times greater by rat hydroxysteroid SULT. Aromatic amines It is well establish ed that N-hydroxy metabolite s of carcinogenic aromatic amines (AAs) and their heterocyclic derivatives (HAAs) require pha se II enzyme activation, in which human liver SULTs have been shown to have a substant ial role. Both AAs and HAAs are initially metabolized to their N-hydroxylated forms, prim arily by cytochrome P450 (CYP) (specifically, CYP1A2 is implicated in HAA bi oactivation); subsequent cataly sis by N-acetyltransferase or sulfotransferase forms highly reactive esters that bind readily to DNA bases. In some cases, the initial CYP-catalyzed metabolites of HAA can bind to DNA directly without further bioactivation. The most extensively studied compounds, and the first compounds for which the mechanism of bioactivation was discovered, are 2-ami nofluorene (2-AF) and its derivative, 2acetylaminofluorine (2-AAF). Another known AA bioactivated by SULT is N-hydroxy-4aminobiphenyl (Chou, et al., 1995), which is activated in vitro by TS-PST (SULT1 family). HAAs are formed during cooking of meats and poultry (see reviews Sugimura 1997 and Turesky 2004) and exist also in tobacco/cigarette smoke condensate, and belong to two classes: those formed from pyrolysis of amino acids and protei ns through radical reactio ns and those created by heating mixtures of creatine, sugars and ami no acids. The amounts of HAAs generated depend on the method of cooking, and increase greatly with well-done grilled meats or barbecued


31 poultry. Frequent consumption of well-done me ats that contain HAAs has been reported to elevate the risk of developing several common hum an cancers, including co lorectal, prostate, and breast. The most prevalent and most studied mu tagenic metabolite is that of 2-amino-1-methyl6-phenylimidazo(4,5-b)pyridine (PhIP), which is formed during cooking of meat products. In vivo, PhIP has been indicated to induce both co lon and mammary tumors in rodents and may play a role in development of breast cancer in humans. Huma n sulfotransferases 1A2, 1A3, and 1E1 all show activity with this compound, as do rat SULTs 1A1, 1B1, and 1C1, but with differences: human SULT isoforms were 5-6xs more active than rat SULTs against PhIP. A polymorphism of hSULT1A1 (G A at codon 213) that replaces arginine with histidine exhibited a substantial decrease in Ph IP catalysis resu lting in DNA binding. Allylic alcohols The m ost extensively studied allylic alcohol s are 1-hydroxysafrole,1-hydroxyestragole, 5hydroxymethylfurfural (5-HF), and -hydroxytamoxifen ( -OH-TAM). Safrole (3,4methylenedioxy-allylbenzene) and estragol e (4-methoxy-allylbenzen e) are the natural ingredients of many essential oils and flavors. DNA adducts formed from the hydroxy metabolites of these two compounds have been characterized, and the hepatic aryl SULT has been implicated in their bioactivation. Tamoxifen is a non-steroidal antie strogen widely used for the treatment of breast cancer, as well as a chemopreventative agent for women at high risk for the disease (Kim, et al., 2004 and 2005). Long-term use, however, has been linked to an increased risk of developing endometrial cancer, and was listed in 1996 as a human carcinogen by the Intern ational Agency of Research on Cancer. It has also been shown to cause cancer in rat liver at significantly higher rates than in human liver. The major allylic alcohol metabolite of tamoxifen, -OH-TAM, is believed to be


32 the metabolite responsible for the ultimate toxi city. Tamoxifen is hydroxylated by cytochrome P450 to various metabolites; the -hydroxylated form is further metabolized by alcohol (hydroxysteroid) sulfotransferases (SULT2 family) to reactive este rs that form DNA adducts. The species difference in liver tumor formation c ould be due to reduced s ubstrate specificity or protein levels of human hydroxysteroid SULT compared to rat hydroxysteroid SULT. Secondary nitroalkanes 2-Nitropropane (2-NP) is a constituent of ciga rette sm oke and a widely used industrial solvent that is reported to be a potent hepatocarcinog en in rats and to be genotoxic. Hepatic SULT activity is required for the metabolic activati on of 2-NP and its anionic form, propane 2nitronate, as well as other secondary nitroalkanes such as nitrocyclohexane and nitrocyclopentane, to reactive species capable of genotoxicity. Specifi cally, aryl SULT (SULT1 family) has been shown to have a role in bioa ctivated secondary nitr oalkanes leading to DNA adducts. Rat SULT1A1 and SULT1C1 were able to catalyze the activation of 2-NP and its anionic form; it has not yet been determined whether human SULTs can activate secondary nitroalkanes. Regulatory Roles Neurosteroids Regulation of neurochem icals is mediated vi a sulfonation of the endogenous molecules, transforming activity and may affect transport across membranes. Wood, et al., 2003, demonstrated that sulfonated estrone injected into fetal plasma was concentrated in fetal brain tissue, evidence of the compounds ability to cross the blood-brain ba rrier. The authors state that the mechanism of crossing is unknown, and sugge st that it may occur via organic acid transporters known to trans port sulfoconjugated steroids. High levels of sulfonated neurosteroids found throughout the brain might also be due to de novo neurosteroid synthesis


33 within the cell (Mellon and Griffin 2001). Cytoso lic sulfotransferases expressed in the human brain include 1A1, 1A3, 4A1, and 1E1 (Glatt, et al., 2001; Gamage, et al., 2006). While these previously identified enzymes ar e known to exist in the brain, little information is available about the distinct localization a nd level of expression of each. SULT 2A1 has been identified in cytosolic fractions of rat brain. The proposed function of this isomer is to sulfonate neurosteroids, endogenous molecules associated with learning and memory. Neurosteroids are also proposed to have antistre ss and anxiolytic functions. A new SULT isomer, 4A, has been identified and characterized in human, mouse and rat brain tissues (Alnou ti and Klaassen 2006; Falany, et al., 2000). Interestingly, this sulfotransferase is found exclusively in brai n tissues, and shows less than 35% amino acid sequence identity with the othe r human cytosolic SULTs. The re gion-specific expression pattern in the brain suggests a function in the central nervous system, but the specific role and substrate selectivity are, as yet, unknown. Catecholamines Metabolism of catecholamines is probably th e area of greatest species differences in sulfotransferase activity on endogenous molecules. Conjugation represents a major metabolizing mechanism for catecholamines to the extent that approximately 84% of total epinephrine, 73% of total norepinephrine, and 97% of total dopamine circulate in conjugated forms (Figure 1-7). In humans, the conjugation is almost entirely sulfoconjugation, in cont rast to rats, in which it is predominately glucuronidation. To date, human is the only species in which a catecholaminespecific SULT isoform, 1A3, has b een identified (Strott, 2002). Sulfonation of catecholamines can have a stabi lizing effect or lead to elimination. Free catecholamines exhibit a short plasma half-life of 1-3 minutes, in contrast to their sulfate


34 Figure 1-7. Sulfonation of the pr imary catecholamines to the form in which they predominately exist in circulation. (Reprinted with permission from Strott 2002) conjugates, which have a plasma half-life of 3-4 hours. The extr emely high ratios of conjugatedto-unconjugated catecholamines in the circulator y system suggest they provide a ready store of easily activated forms as well as easily-excreted. The much more stable conjugate can travel through the bloodstream to a target organ, to be immediately ac tivated by sulfatase when needed. Catecholamines are important mediators of the bodys response to physio logical stresses and may play a role in immediate adaptation to extrauterine life (just after bi rth). Animal studies indicate that stresses such as hypoxia and birth produce dramatic increases in fetal catecholamine production (Richard, et al., 2001). In human fetuses, SULT1A3 is expressed at high levels in the liver, but adult hepatic expression is essentially absentinstead, th e gastrointestinal tract is the major site, along with high dopamine sulfate production. Estrogen SULT1E1, t he estrogen sulfotransfe rase, displays a particularly high affinity (in the lower nM


35 Figure 1-8. Structure of mouse SULT1E1 in the presence of PAP and substrate 17 -estradiol, obtained from the Protein Data Bank (, entry 1AQU). A representation of the structure of a S ULT1E1 monomer is shown, with -helical sections displayed as red cylinders and -sheets as light blue arrows. (Reprinted with permission from Coughtrie 2002) range) for its natural substrate, 17 -estradiol, and is believed to play an important role in estrogen regulation (Coughtrie, 2002). Crystal structure for mouse SULT1E1 with PAP and estradiol is shown in Figure 1-8. The enzyme is expressed in th e endometrium and is regulated during the menstrual cycle, suggesting involvement in successful im plantation of fertilized ova. Estrogen exists in the human circulation system primarily as the sulfonated form, which may serve as storage or transport between organs, since the conjug ate is not active with estrogen receptors. Environmental pollutants, hydroxylated PCBs (polychlorinated biphenyls), have been identified as potent i nhibitors of SULT1E1 (K i in the pM range), which suggests a mechanism


36 by which the xenobiotics exert th eir well-documented endocrine disrupting effects (Wang and James 2006). Estradiol-3-sulfate levels in plasma of pregnant women increase as a function of fetal gestation, but the fetal plasma le vels are unknown. The most abundant forms of estrogen in fetal plasma, however, are known to be sulfoconjugate d. Concentrations of unconjugated estrogens increase only after the beginning of the increase in fetal plasma cortisol and corticotropin. Research by Wood, et al. (2003) measured the leve ls and effects of estrogen sulfotransferases in ovine fetal plasma. Significant levels were m easured in hypothalamus and brainstem throughout the last trimester of gestation, as well as postn atally. They hypothesized that high levels of circulating 17 -estradiol-3-sulfate provide a ready sour ce of estradiol for the portions of fetal brain that control the hypothalamus-pituitary-ad renal (HPA) axis. Inte raction between the HPA axis and placental estrogen production produces a positive feedback loop that leads to an increased estrogen-to-progesterone ratio, uterine contraction, a nd initiation of labor and delivery. Exogenous infusions (0.25 mg/day or 1 mg/day, fi ve days) of estradiol-3-sulfate resulted in increased levels of fetal plasma cortisol and cort icotropin. An increase in estrogen action in the cerebellum was also measured. Sulfoconjugated es trogen, in the form of estrone sulfate, was taken up by fetal brain, through an undetermined mechanism. The authors concluded that estradiol-3-sulfate circulates in hi gh concentrations in fetal plasma and is directly available to the fetal brain. There it is deconj ugated by estrogen sulfatase and i nvolved in HPA axis stimulation in late gestation and induction of uterine contractility and labor. Qian, et al., (2001) observed the effects of es trogen sulfotransferase inhibition in male knockout mice, in which the estrogen-specific sulfot ransferase gene was di srupted. Results were primarily characterized based on testis abnormalities (Leydig cells in particular) and


37 reproductive success (litter size a nd frequency). While the reproduc tive systems of younger (3-6 months) knockout male mice appeared normal, older mice developed Leydig cell hypertrophy/hyperplasia and seminiferous tubu le damage. Younger knockout mice developed similar structural lesions when given exogenous E2 (estradiol), whereas their wild type agematched counterparts did not. The lesions were attributed to chronic and progressive damage from long-term estrogen exposure. In wild type mice, estrogen sulfotra nsferase is expressed abundantly in the Leydig cells of testis. Estroge n biosynthesis is high in the testis, and several tissues within the male reproductive system are established estrogen site s. Therefore, since sulfonated estrogen is unavailable for receptor binding, the enzyme action probably serves a protective effect, its absence marked by reproductive tissue damage seen. Although caudal sperm number was not significa ntly decreased in 18to 22-month-old knockout mice, in comparison with 3-month-old knockout mice, sperm motility was reduced. Both forward and total motility decreased in older mice by 80% and 60%, respectively. Litter frequency over a 2-month period was similar to that of age-matched wild type (WT) mice, but the litter size was significan tly smaller (avg. 8.5 WT, 5 KO). The effect of estrogen sulfotransferase di sruption in female mi ce is currently under investigation; preliminary studi es show that female fertility is reduced, as evidenced by significantly reduced litter size. Thyroid hormones Sulfonation is im portant for both synthesis and metabolism of thyroid hormones, and probably plays a key role in regulating the amount of active thyroid hormone (T3) available to cellular receptors (Strott 2002; Richard, et al., 2001). Thyroglobulin is the major protein produced by the thyroid gland and is the macromol ecular precursor of thyroid hormones. During


38 Figure 1-9. Regulation of thyroid hormones T4, T3, and diiodot hyronine (DIT); T4 is the main secretory product of thyroid follicular cells and is converted to T3, the biologically active form of thyroid hormone. Potentia l sulfonation sites ar e shown by SO3-. Reprinted with permission from Strott, 2002. IRD: inner-ring deiodination; ORD: outer-ring deiodination synthesis and processing, t hyroglobulin undergoes extensive posttranslational modification, including sulfonation (Figure 19). One modification, tyrosine sulfonation, is commonly found throughout the vertebrate phylum, suggesting it wa s acquired at an early stage in thyroid evolution. The exact purpose of sulfate modifi cations of thyroglobulin is not well understood and under continued investigation. Sulfoconjugation of iodothyronines T4 and T3 has a significant effect on their further metabolism. T4, the main secretory product of thyroid follicular cells, is converted in extrathyroidal tissues to T3, the biologically active form of thyroid hormone. If T4 is sulfonated before conversion, it can only be transformed into the sulfonated,


39 biologically inactive form of T3. Once sulfonated, T3 does not bind to nuclear receptors and is, therefore, biologically unav ailable. Disposal of T3 is primarily via sulfate formation--type 1 deiodination of T3 sulfate in human liver homogenates occurs at a rate that is 30-fold higher as compared with unconjugated T3. It has been proposed that one function of sulfonation is to inactivate thyroid hormone so that iodine can be reused for thyroid hormone synthesis. Sulfonated T3 is present at very high levels in the fetal circulation, its co ncentration increasing with fetal age, well into postn atal life (Richard, et al., 2001). This supports a role for sulfotransferase as a way of protecting the fetus from excessive active thyroid hormone. Endocrine Disruption Certain classes of xenobiotic chem icals, termed endocrine disruptors, ar e believed to interfere with an organisms sex steroid homeostasis by ac ting directly as hormone mimics or indirectly by disrupting the biosynthesis/regulatory pathways. In general, the use and distribution of these chemicals is widespread and ubiquitous, an thropomorphic/synthetic, and are from pharmaceutical, industrial, and agricultural sources (Mills and Chichester 2005). Some examples are synthetic estrogens, biodegrad ation alkyl polyethoxylate deterg ents (particularly nonylphenol and octylphenol), and polychlorinated biphenyls. Adding to the xenobiotic effect is that many of the chemicals have long half-lives so are very slow to decompos e in the environment. Although some are present at low levels in the soil, sedi ment, or water, they can be bioaccumulated to toxic levels and can have a combined effect w ith other endocrine disrupting chemicals in an additive manner (Brian, et al., 2005). Evidence of endocrine disruption in aquatic sp ecies has been observed both in the wild and the laboratory. The observe d effects include, but are not li mited to, reduced egg production and/or fertilization, altered se x organs, intersex (having both male and female reproductive structures), skewed sex ratios, and altered be havior (Arukwe 2001; Mill s and Chichester 2005).


40 Wild populations of riverine fish (roach; Rutilus rutilus ) near discharge from sewage treatment plants exhibited high levels of intersex and vite llogenin (a female protein for egg production) in males (Jobling, et al., 1998). Populations of Pallid sturgeon in the Mississippi River ( Scaphirhynchus albus ) and Chinook salmon ( Oncorhynchus tshawytcha ) from the Columbia River (northwestern US) have decr eased significantly, resulting in their listings as endangered (Nagler, et al., 2001; Harshbarger, et al., 2000). Both populations ha ve exhibited decreased reproduction capacity and intersex, and reside in waterways contaminated with compounds associated with endocrine disruption. Of th e wild salmon tested, 84% of the phenotypically female fish were positive for a male-specific genetic marker, while 29% of the males had both mature sperm and ovigerous lamellae, fema le reproductive tissue. The dogwhelk ( Nucella lapillus) and common whelk (Buccinum undatum ) have locally disappeared from the Scheldt estuary due to tributyltin-induced imposex (Versl ycke, et al., 2005). Tributyltin (TBT) is an antifouling agent used on ships and has been show n to induce reproductive toxicity at very low levels. Florida alligators ( Alligator mississippiensis ) in lakes polluted by old-use organochlorinated pesticides (OCPs) have incr eased mortality and shown reproductive toxicity linked to OCPs (Seplveda, et al., 2006). Repor ted endocrine-disrupting effects have included altered secondary sex characteristics, endocrine status, and se x differentiation in hatchlings and juveniles. Zebrafish ( Danio rerio) in the lab exposed to 2,3,7,8-tetrachlorodibenzop-dioxin (TCDD) produced offspring with morphological anom alies (Figure 1-10) as a result of maternal transfer (Heiden, et al., 2005). The parent zebrafish TCDD concen tration levels increased in a dose-dependent manner due to bioaccumulation; there was a correlati ve decrease in the ovosomatic index (ovary weight/body weight) and offspring survival 24 hours post-hatch.


41 Figure 1-10. Zebrafish offspring affected by parental intake of 2,3,7,8tetrachlorodibenzo-pdioxin. Reprinted with permission from He iden, et al., 2005; orig inal caption states Representative zebrafish larvae (6 dpf) from females fed diets containing TCDD. Larvae exhibited one or more of the follo wing morphological anoma lies: Cranial (ce), pericardial (pe), or yolk sac edema (yse), uninflated swim bladder (*), subcutaneous hemorrhage (not shown), shortened jaw (j), and tail necrosis (not shown). A = control, B = 10 ppb diet, C = 40 ppb diet, and D = 100 ppb diet. Scale bar represents 0.01 mm. The exact mode of action(s) for endocrin e modulators is not fully understood. The original hypothesis was of dir ect-acting estrogen mimics, but there was often poor correlation between effects and chemical leve ls or estrogen receptor affinity. More recent research has led to a second theory of indirect-a cting endocrine disruptors, which have effect by interfering with enzymes involved in the sex ster oids biosynthesis and regulator y pathways (Kester, et al., 2002). Specifically for sulfotransferases, since its catalytic activity toward s a hormone renders it inactive, inhibition of the enzyme would increas e levels of active estrogens and androgens and upset the delicate balance of sex steroids.


42 The effect of this was seen in male SU LT1E1-knockout mice (Qian, et al., 2001). The knockout males were fertile and phenotypically nor mal initially, but developed age-dependent Leydig cell hypertrophy/hyperplasia and seminifer ous tubule damage, reduced sperm motility, and smaller litters, when compared to age-matched wild-type males. Sulfotransferases are inhibited by many envi ronmental chemicals considered endocrine disruptors (see review Wang and James 2006), including hydroxylated PCBs (OH-PCBs), dioxins, dibenzofurans, and pentachlorophenol. Kester, et al. tested various OH-PCB compounds with recombinant human estrogen sulfotransferase and observed noncompetitive inhibition for a number of congeners. The most potent structures, which inhibited SULT at subnanomolar IC50 levels, were the para-hydroxyla ted compounds, particularly 4-OH-3,5dichloro substitution pattern. The same pattern of inhibition was seen in channel catfish liver cytosol treated with hydroxy-PCBs (van den Hur k, et al., 2002). Structures with 4, 5, and 6 chlorine substitutions and the phenolic group in the ortho, meta, and para positions were examined; the most potent sulfotransferase inhibitors were those with the 4-OH-3,5-dichloro substitution pattern. Formation of E2-3-sulfate a nd, to a lesser extent, E2-17-sulfate in catfish liver cytosol was inhibited by the therapeutic drug celecoxib (Wang and James, 2007). Inhibition of human estrogen SULT was also seen with monoand dihydroxylated metabolites of dibrominated biphenyl (van Lipzig, et al., 2005). Phenol sulfotransfe rase (SULT1 family) was inhibited by 2,6-dichloro-4-nit rophenol (DCNP), which in hu man SULTs is a preferential inhibitor of SULT1A1 but not SULT1A3 (Seah and Wong 1994). Rat SULT1s, while inhibited by DCNP, did not show the same selectivityboth 1A1 and 1A3 isoforms (P form and M form) were inhibited. Rat aryl SULTs (SULT1 family) were efficien tly inhibited by transdihydrodiol derivatives of benzo[a]pyre ne (B[a]P) in a competitive manner (Ki ~ 4 M), while


43 phenolic derivatives 7-OH-B[a]P and 8-OH-B[a] P were good substrates (Rao and Duffel 1992). High body burdens/levels of many of these chemicals have been measured in fish, wildlife, and humans (Harshbarger, et al., 2000; Rauschenbe rger, et al., 2009; Facemore and Gross 1995; Jones and de Voogt 1999; Bonefeld-Jorgensen 2010). Research Aims The potential role of sulfotransferases in e ndocrine disruption is great, but knowledge of nonm ammalian SULTs is limited, at best. Of the cyto solic fish SULTs characterized, only one is for a food-fishthe others are from zebrafish, for use as a laboratory model. Considering the commercial importance of fisheries and sports fi shing, this leaves a large informational void about fish health and well-being. The channel catfish is the most important spec ies of aquatic animal commercially cultured in the United States (Wellborn, 1988). The species has been successfully introduced throughout the U.S. and the world, popular both for fish ing (the world record channel catfish was 58 pounds) and raised in commercial hatcheries. Channel catfish are omnivorous, with a diet ranging from aquatic insects to clams to algae to smaller fish. They use their barbels to locate food on the bottom, particularly at night, although they are not strictly bottom feeders. Because they are bottom feeders, however, they have a greater risk of exposur e to pollutants in the sediment. A greater understanding of their metabolism of xenobiotics and, in particular, the sulfotransferase enzymes, is needed. Thus, rese arch described in this dissertation was directed toward the following aims and hypothesis: Overall Objective: Gain a greater understanding of Phase II metabolism in channel catfish, specifically of sulfotransferase enzymes. Specific Aim #1: Obtain full-length cytosolic sulfotransferase cDNA sequences from a cDNA library made from channel catfish liver.


44 Specific Aim #2: Express and purify r ecombinant proteins from the full-length channel catfish sulfotransfera se cDNA sequences obtained. Specific Aim #3: Characterize expresse d, purified recombinant sulfotransferases from channel catfish liver tissue with en zyme assays using known SULT substrates. Research Hypothesis: Cytoso lic sulfotransferases in chan nel catfish are active toward substrate types with the same selectivity as mammalian; specifically, that Family 1 isoforms will preferentially sulfonate phenolic substrates, and Family 2 will be selective toward alcoholic substrates.


45 Table 1-1. Known mammalian sulfotrans ferases, their most common substrates, and tissues they are found in. Isoform Substrates Tissues 1A1 Phenols Highly expressed in liver; also in brain, breast, intestine, endometrium, kidney, lung, platelets (human) Acetaminophen, minoxidil, 17-ethinylestradiol, 17estradiol, iodothyronines High affinity for p-nitrophenol 1A2 Phenols cDNA, not expressed protein, in liver and colon (human) Lower affinity for p-nitrophenol 1A3 Catecholamines Highly expressed in intestinal ti ssue, brain; also in platelets; neglible in adult liver (human only) Neurosteroids: norepinephrine, epiniphrine, dopamine High affinity for dopamine Only found in humans so far 1E1 3-hydroxy group of estrogens Several steroid hormone-responsive tissues: endometrium, testis, breast, adrenal gland, placenta; also in liver, small intestine, depending on species Endogenous and xenobiotic estrogens; estrone, estradiol, 17-ethinylestradiol 1B Thyroid hormones (human)1B1 mRNA found in liver, colon, small intestine, blood leukocytes (human) 2-naphthol, dopamine Not steroid hormones 1C Phenols (rat)1C1liver specific (rat and human) Procarcinogen AAP (N-OH-2-acetylaminofluoren e) (human)1C2-thyroid, stomach, kidney p -Nitrophenol (human)1C4-predominatey in fetal kidney and lung, also in adult ovary and brain 1D Small phenolic or amine-containing molecules (mouse)1D1-kidney and uterus; not liver, brain, lung, intestine, testicular or glandular tissue (rat, mouse, dog) Not steroid or thyroid hormones 2A and 2B 3-hydroxy steroid groups (human)2A1: adrenal cortex, liver, brain, intestine DHEA, androsterone, allopregnanolone (human)2B1: prostate, placenta, trachea 2A only: phenolic OH group on 3-position of estrogens; bioactivation of polycyclic aromatic hydrocarbons 2B1 only in mouse, human so far 2B1 more selective for 3--OH steroids 2B1_v1 and _v2 _v1 preferentially catalyzes pregnenolone (mouse)2B1: intestine, epididymis, uterus _v2 catalyzes both pregnenolone and cholesterol


46 CHAPTER 2 GENE IDENTIFICATION Materials and Methods Materials Prim ers were constructed by MWG-Biotech (High Point, NC). LA Taq was from TaKaRa Bio Inc. (Madison, WI). DNA ladders (100 bp, 1 kb), 6X loading dye, and Taq DNA Polymerase were purchased from Promega (Madison, WI). TOPO cloning vector, TOP10 competent cells, and SOC medium were obtained from Invitrogen (Carlsbad, CA). RNAqueous kit and RNA later were purchased from Ambion (Austin, TX). QIAprep Spin Miniprep Kit, QIAGEN Plasmid Midi Kit, and QIAquick Gel Extraction Kit were from Qiagen, Inc. (Valencia, CA). Advantage 2 PCR Enzyme System and Powe rScript First-Strand reverse transcriptase kit were purchased from Clontech (P alo Alto, CA). Sequencing reag ent Big Dye version 1.1 and 5X buffer were from Applied Biosystems (F oster City, CA). Edge gel filtration cartridge s were obtained from Edge Biosys tems (Gaithersburg, MD). Agar, LB Broth, agarose, and 10X TBE (Tris-Borate-EDTA) buffer were purchased from Fisher Scientific. Ethidium Bromide, ampicillin sodium salt, and carbenicillin disodium salt were obtained from Sigma-Aldrich. Data we re analyzed with Lasergene SeqMan, from DNAStar (Madison, WI), DNAMAN by Lynnon Corporation (Quebec, Canada), ClustalW ( ools/clustalw2/index.htm l ), and GeneRunner ( ). Methods cDNA library Two different cDNA libraries, with 5 adapte rs Lu4TRSA a nd CAPT30 (Table 2-1), were created from channel catfish liver tissue, with tw o different 3 adapters; an overview of steps is


47 shown in Figure 2-1. Total RNA was extracted with the RNAqueous k it, per manufacturers protocol, and cDNA libraries constructed per the methods of Matz (2003, and personal communication), template-switch method B. Details of RNA extraction and library construction follow; during all procedures, glove s and counter surfaces were kept Rnase-free with liberal use of RNaseZap. Total RNA extraction Catfish liver tissue was weighed out to 70 m g and homogenized manually with an RNasefree pestle in a 1.5 mL microcentrifuge tube in 900 L lysis solution (provided in RNAqueous kit), added in ~200 L increments. After homogenizing, sm all pieces of tissue were broken up by triturating with a cut pipette tip until comple te. Sample was centrifuged for 10 minutes at 13,000 rpm at room temperature and the supernatant transferred to a new 1.5 mL microcentrifuge tube. An equal volume of 64% ethanol, 500 L, was added to the supernatant; sample was mixed by inverting, and the entire volume transf erred equally to two RNAqueous filter units. The filters were centrifuged for 1.5 minutes at 12,000 rpm, the filtrate disposed of, and the filters washed with kit buffers as directed (700 L wash solution #1, centrifuge 1 minute max, 500 L wash solution #2/3, centrifuge, repeat). All filtrates were discarded. Tota l RNA was eluted from the filter into a single tube by treating with 40, 30, and 10 L warmed (80 C) Elution Solution (from the RNAqueous kit), with centrif ugation for 30 seconds, 12,000 rpm for each. The RNA quality was checked by loading a 2 L sample on a nondenaturing 1% agarose gel; clear bands could be seen for 28S and 18S ribosomal RNA as shown in Figure 2-2. RNA LiCl cleanup Eluted to tal RNA was further cleaned up with lithium ch loride precipitation which removes carbohydrates and gross DNA contamination; this was necessary per the protocol


48 Figure 2-1. Flowchart of steps involved in obtaining sequence data for a gene, beginning with the creation of a cDNA library. followed (Matz 2003 and personal communication) because the RNA concentration was high (minimum 0.2 g/ L). One-half volume of LiCl Precip itation Solution (fro m RNAqueous kit), 40 L, was added to the eluted RNA, mixed well and incubated at C for 30 minutes, then centrifuged for 15 minutes at 13,000 rpm. Supernatan t was discarded and the pellet washed with cold 70% ethanol, 500 L, the ethanol poured off and the pellet allowed to air-dry. The pellet was resuspended in 40 L Milli-Q water and 2 L was loaded onto a nondenaturing agarose gel, as before, to check quality. RNA concentration (~300 g/mL) and quality (ratio ~1.6) were calculated by spectrophotometer. Homogenized catfish liver tissue 1. Extract total RNA 2. LiCl cleanup 3. Incubate with RT, primer 1st-strand synthesis product (1SS) Amplified cDNA library with end adapters cDNA library working stock PCR with adapter primers (5)TS-PCR and (3) TRsa or CAP-T30 Dilute 1:50 (aq) 1. PCR 2. Check on gel; add cycles if needed 3. Nest PCR; repeat step 2 4. Gel purify oligomer PCR Product 1. Ligate into TOPO vector 2. Transfect cells, culture 3. Colony screen 4. Harvest plasmid by mini-prep Selected plasmid with PCR insert S e q u e n c e r e a c t i o n s Multiple Sequences (~300-400 bp each) 1. BLAST data 2. Align into contig 3. Translate contig Amino acid sequence BLAST Complete sequence? 1. Design new primers 2. Go to PCR or sequence step NO YES Repeat 2X from PCR step for confirmation


49 28S 18SMRNA M Figure 2-2. Nondenaturing agar ge l of total RNA extracted from channel catfish liver. Clear bands for 28S and 18S ribosomal RNA are s een, an indication of intact RNA as per Matz, 2003 (journal article a nd personal conversation). DNA standards of 1kb and 100 bp are to the left and right of the RNA lane, respectively. cDNA first strand synthesis To 4 L RNA solution in water (~1 g total RN A) was added 1 L of 10 M primer TRsa (primer sequences listed in Tabl e 2-1); this was incubated at 65 C for 3 minutes, and then put on ice. To this was added 2 L 5X First-Strand reverse transcriptase buffer, 1 L 0.1M DTT (provided with reverse transcriptase), 1 L 5 M TS-oligo, 1 L PowerScript reverse transcriptase, and 0.5 L dNTPs, 10 mM. This was incubated at 42 C for 1 hour, at which time 1 L of 20 mM MnCl2 was added. The sample was furt her incubated for 15 minutes at 42 C, and the reaction stopped by heating to 65 C for 3 minutes. cDNA amplification PCR was set up to am plify the first-strand synt hesis product (1SS) and attach the 5 and 3 adapters (see Table 2-1). P CR mixture was composed of 3 L 10X Advantage 2 PCR buffer, 1 L of 10mM dNTPs, 1.5 L of 2 M TS-PCR 5 adapter primer, 1.5 L of 2 M 3 adapter primer (TRsa or CAP-T30), 1.5 L of 1:5 dilution 1SS, 21 L Milli-Q water, and 0.6 L


50 Advantage 2 Taq polymerase. This was cycled as follows: 94 C for 30 seconds, 65 C for 1 minute, and 72 C for 2 minutes. After 17 cycles (the minimum recommended), 2 L of each PCR product was checked on a 1% agarose gel, and cycles added as needed. A chase step was performed by holding the samples at 72 C, adding 2.5 L each of additional primer, and going through two PCR cycles (same PCR program). All st eps taken were accordi ng to the protocol of Matz, 2003 (and personal communication). Worki ng stocks of 1:50 dilutions with Millipore water were made and all samples stored at C. Table 2-1. Primers used to make catfish cDNA libraries Primer Sequence (5 3) Type TS-PCR AAGCAGTGGTATCAACGCAGAGT 5 adapter TS-oligo AAGCAGTGGTATCAACGCAGAGTACGCrGrGrG 5 adapter CAP-T30* AAGCAGTGGTATCAACGCAGAGTACT(30)VN 3 adapter TRsa CGCAGTCGGTACT(13) 3 adapter Lu4TRsa CGACGTGGACTATCCATGAACGCA CGCAGTCGGTACT(13) 3 adapter *V = A,G,or C Quality of the final libraries was tested by PCR with a primer pair that had previously proven successful (SULF-1 and SULF-2, Table 2-2). The PCR mixture was composed of 1 L template (1:50 TRsa or CAP-T30 library); 0.5 L each primer, 10 M starting concentration; 8 L LA Taq dNTPs; 5 L LA Taq 10X buffer; and 35 L Milli-Q water. After a hot start cycle (95 C 5 minutes, 58 C 2 minutes, 72 C 3 minutes), 0.5 L LA Taq polymerase was added and 30 cycles performed of 95 C 30 seconds, 58 C 30 seconds, 72 C 1 minute. A negative control was included with water as template. Products were run on an agarose gel and both libraries showed excellent, strong bands at th e known product length (Figure 2-3a). Primer design Gene-specific prim ers were designed based on pu tative catfish sulfotransferase Expressed Sequence Tags (ESTs) obtained from the NCBI database. The primers were designed


51 Figure 2-3. PCR products for two cDNA libraries CAP T-30 and TRSA. The libraries were tested with a primer pair, SULT-1/SULT-2, that produced a band at ~500 bp. Each library was tested in duplicate with a negative control (water) and 1kb marker (M). according to the published guidelines (Matz, et al ., 2003 and personal communication) of at least 40% G/C content, no A/T at the 3 end, minimu m 20 bp in length, and annealing temperature of at least 55 C. Primer sequences were checked for dimerization and other potential problems using Gene Runner and DNAma n software programs and BLASTed on NCBI to check specificity. Nesting primers we re designed following the same guidelines, with the additional requirement of being ~50 bp downstream of the original primers in the respective DNA sequence. Experiments Standard protocol for cloning Much of the research project involved P CR experim ents, followed by selection and preparation of PCR products l eading to sequence analysis. The same basic protocol was followed for each step, with minor alterations in temperature, primers used, etc. Standard protocols are described below, and deviations or specific details (such as primers) will be noted for particular experiments. The PCR mix prepared consisted of 34 L Milli-Q water, 1 L template, 8 L LA Taq dNTPs mix, 5 L LA Taq 5X polymerase buffer, and 0.5 L each primer (sense and antisense). After a hot start at 95 C, 0.5 L LA Taq polymerase (TaKaRa) was added. Primer concentrations depended on type: sequenc e-specific primers were used at 10 M concentrations, TRSA M CAP T-30


52 while adapter-specific primers were 2 M starting concentrations. The template was usually amplified cDNA library, 1:50 diluti on (working stock). For nes ting PCR, template was a 1:50 dilution of the first PCR product. For a large nu mber of samples, a master mix was prepared of water, dNTPs, Taq buffer, and, sometimes, te mplate or primer (depending on the experiment); this was vortexed and pipetted at appropriate volume into the 0.5 mL tubes, followed by individual addition of primer a nd/or template, as needed. Initial PCR cycles included a hot start and fi nal extension, but subse quent cycles (when a few more were needed to obtai n the product) included neither. PCR cycling was performed by either of two thermal cyclers: Mastercycler gradient (eppendorf) or PTC-100 (MJ Research, Inc.), both featuring block-heating. A program for a complete PCR cycle was as follows: Hot start (1 cycle): 95 C, 5 minutes TA, 2 minutes 72 C, 3 minutes 25-30 cycles: 95 C, 30 seconds TA, 30 seconds 72 C, 1 minute Final extension: 72 C, 10 minutes TA = primer annealing temperature, calcul ated by the formula 4(G+C) + 2(A+T) + 3 ( Matz, et al., 2003) After cycles were completed, pr ogress was checked by loading 3 L PCR product (with 12 L loading dye) on a 1% agarose gel. If additional cycles were needed, the sample was returned to the thermal cycler for additional cycles without hot start or 10-minute extension period. Usually a second nesti ng PCR would be carried out on a 1:50 dilution of the first PCR


53 product, using a second set of specific primers. These primers would be designed to be ~50 bp from the initial primers in the sequence, the purpose of the nesting PCR being to improve specificity of product and eliminate some of th e more general gene sequences pulled up in the first PCR experiment. Gel extraction, ligation, culture of PCR product Once PCR product had been obtained, 15 L was run on a 1% agaros e g el, and selected bands either cut out with a razor or picked out with a glass Past eur pipet. Razor-extracted gel slices were treated with the Qi agen Gel Extraction kit, accordi ng to manufacturers instructions, and cDNA eluted with 30 L Elution Buffer supplied with the k it. Gel picked out with a pipet was transferred to 30 L Milli-Q water and frozen; when th awed, the water contained sufficient cDNA diffused from the gel to use for ligation. The isolated cDNA sequence was inserted into a TOPO commercial cloning vector, either pCR 4-TOPO or pCRII-TOPO, according to even tual use (sequencing only or recombinant protein expression). Regardless of the vector used, the method was the same: 4 L cDNA solution, 1 L vector, and 1 L salt solution (supplied with v ector) were combined, mixed by tapping, and kept at room temp erature for 5-30 minutes. Comp etent TOPO TOP-10 cells were transformed with 2.5 L of ligate mixture, which was added with a long pipette tip; cells were stirred gently with the tip and kept on ice. After 30 minutes the cells were heat-shocked at 42 C for 45 seconds, returned to the ice, and 250 L SOC medium was added. The cell culture was incubated, with shaking, at 37 C for 1 hour, and then distributed between two agar plates (Luria Broth medium + carbenicillin 50 g/mL). Plates were incubated overnight at 37 C and the colonies screened for insert.


54 Colony screen Selected colonies were transferred to a master plate with a sterilized toothpick which was subsequently dropped into 5 mL Luria Broth (LB) media with either carbenicillin (50 g/mL) or ampicillin (100 g/mL) to culture overnight at 37 C. The master plate was incubated overnight at 37 C and a scraping of each colony transferred with a sterile toothpick to 50 L Milli-Q water and heated at 95 C for 5 minutes. Lysed cells were spun for 2-3 minutes on a tabletop centrifuge (mini-centrifuge) and 5 L added to a PCR mixture consisting of 8.25 L water, 10 L 5X loading dye, 5 L Promega Taq polymerase 10X buffer, 0.5 L dNTPs (20 mM), 0.5 L each primer (25 M), and 0.25 L Promega Taq polymerase. PCR was performed for 30 cycles of 94 C, 30 seconds; 57 C, 30 seconds; 72 C 1 minute. Samples were checked by loading 12 L on an agarose gel; because loading dye was part of the PCR mix, no additional dye was necessary. For a large number of colonies be ing screened, a master mix was prepared of water, loading dye, polymerase buffer, dNTPs, buffer, and Taq polymerase, at the above concentrations, which was vortexed and pipetted in appropriate volume into 0.5 mL tubes, followed by addition of cell lysate. Sequencing Cultures f rom positive screens were pelleted and the plasmids extracted using the QIAprep Spin Miniprep Kit, according to the microcentrif uge protocol included with the kit. Sequencing reactions were carried out using Big Dye version 1.1 as follows: mixtures of 4 L Big Dye; 2 L Big Dye 5X buffer; 7 L Milli-Q water; 4 L primer (1 M); and 3 L plasmid underwent 25 cycles in a thermal cycler of 96 C for 30 sec.; 50 C for 15 sec.; 60 C for 4 min.; then kept at 4C until processed further. The samples were purified with Edge Gel Filtration Cartridges (Biosystems) according to the manufacturers protocol, dried under vacuum, and kept at


55 Table 2-2. Primers designed at the start a nd end of each corresponding EST sequence reported to NCBI to be channel catfish sulfotransfe rase. The primers were used to screen a channel catfish liver cDNA library for th e presence of each EST sequence. n denotes a nest primer. Primer Sequence 5'-3' EST name sense/anti SULF-1 GTCGGATTGGTGAGACTGGAAGAATCACTT 1C2 sense SULF-2 TTCTGCTTGAGCATGCTTGTAACATTCCA 1C2 antisense SULF-3 CGGTTTGATGAACACTACTGGCAGAA 1C2 nest sense SULF-4 CAGTCATGTTCTAG GATTGTATACGTTGCT 1C2 nest antisense SULF-5 GTCTGCAGCGATAAGTGAAAACGTA Estrogen-like, isoform 2 (1C1) sense SULF-6 CCATATACTGTACCTTGCAGTTCTGTT Estr ogen-like, isoform 2 (1C1) antisense SULF-7 TGGAGGGAGTTAATACTTCTTCGAT Es trogen-like, isoform 2 (1C1) n sense SULF-8 GCACCAACTGAACAGGTAGATGA Estrogen-like, isoform 2 (1C1) n antisense SULF-9 CACGTCTTCGCCATCACAA Sultx3 sense SULF-10 GTAGTAGGAGACCACCAGGTCTTTA Sultx3 antisense SULF-11 GACTAAGCTACGTGATCTCAAATGCAT Sultx3 nest sense SULF-12 AAGCGGTACGGCAGGTGACTTT Sultx3 nest antisense SULF-13 ACCAGACGATGTCCTTATTGTCACCTAT Spleen sense SULF-14 AGGATTCGTTCATCATGCCAT AGTGA Spleen antisense SULF-15 GCACAACATGGATGCAGGAGATT Spleen nest sense SULF-16 GCCTTTGCTCCAGATTGAGTATCTTA Spleen nest antisense SULF-17 AACAGGAACAAGATGTCTGCTCCAA Intestine sense SULF-18 ACTAAGAGACTCACTCGGCTGTCATA Intestine antisense SULF-19 GCATGGAAATGAAGGATGAGGAGAA Intestine nest sense SULF-20 GCTGAGAAGAGAGAGAGAAAAGCAA Intestine nest antisense SULF-21 AGGTTCTTCATTGTACGAGAGCCATT Pituitary HNK-1 sense SULF-22 CGCAGAGTTCGGCGTATGTTAA Pituitary HNK-1 antisense SULF-23 GTTCAAGGACAAGTTTGTGGAGAA Pituitary HNK-1 nest sense SULF-24 TGGATGACGTGCTCTCCGAA Pituitary HNK-1 nest antisense SULF-25 GCTGTGATAAGTAGCATGACGGAA Tyrosine-ester sense SULF-26 CACATGATCATACCAAGGTCCAAA Tyrosine-ester antisense SULF-27 CATCATGGAAGGATCTGATGTT Tyrosine-ester nest sense SULF-28 CTGGATCAGGTTGTAGGTGATT Tyrosine-ester nest antisense SULF-30 AAGTGATTCTTCCAGTCTCACCAATCCGAC 1C2 RC of sulf-1 for 5' nest antisense SULF-32 TTCTGCCAGTAGTGTTCATCAAACCG 1C2 RC of sulf-3 for 5' antisense SULF-31 ATGTGACTGGCTACTGGGAGAAGAA Tyrosine-ester 3' sense SULF-33 ACTGATCGTGAGGTGGAGCGTCTTT Tyrosine-ester 3' nest sense


56 4C until sequenced. Sequence results were compared using Blast programs on the NCBI website ( ) and contigs created with SeqMan software. Other software program s used included DNAman, GeneRunner, and ClustalW. Catfish Sulfotransferase Isoforms A search of the NCBI database turned up a num ber of Expressed Sequence Tags (ESTs) for channel catfish, between 400-600 bp in le ngth, seven of which were identified as sulfotransferases. None of the sequences had been translated or publis hed; the group generating the ESTs is interested in them for use in mi croarrays. The sequences were used to design specific primers as previously described; primer sequences are listed in Table 2-2. These primers were used to screen the catfish cDNA library fo r presence of the particular EST sequences. Positive sequences were pursued further, with additional primers, to pull up the complete sequence. Results and Discussion Results cDNA libraries Successful libraries were constructed on tw o separate occasions, with som e method alterations the second time (such as adding a LiCl precipitation step) to impr ove overall quality. Both libraries were constructed with tissue from the same catfish liver. The first library was made with only the CAP-T30 3 adapter; the sec ond (successful) set of libraries was made with two different 3 adapters, CAP-T30 and TRsa. Th e libraries were fully amplified in 20 or fewer cycles. Working stock dilutions (1:50) were made and checked for quality by a PCR experiment using SULF-1 and -2 primer pairs. Products from all libraries showed strong bands on a 1% agar


57 gel at the correct molecular weight (~500 bp), indicating that the cDNA libraries were of good quality. PCR experiments cDNA library #1, with the CAP-T30 adapter, an d the second library with the TRsa adapter were screened for sequences corresponding to the seven ESTs found in the NCBI database. Prim er pairs specific to the 5 and 3 of each EST sequence were designed for the screen, a positive endpoint being a PCR product of the si ze of the corresponding EST. In some cases, PCR products were sequenced to confirm a matc h with its EST, while in other cases it was decided to proceed directly to RACE PCR to ge t additional sequence information. Details of the library screens are given in Table 2-3. Of the se ven EST fragments, five were determined to be present in the catfish liver cDNA library; of these, three were select ed for continued PCR experiments to obtain complete sequences. These were the ESTs identified as Spleen, C1, and Tyrosine ester, which were eventually renamed (according to the BLAST hits of the extended sequences) to SULT2, SULT1v1, and SULT1v2, abbreviated as S2, S1v1, and S1v2, respectively. SULT6B1 partial sequence Of the multiple bands obtained from RACE PCR to extend the Intestine EST sequence, three bands (~1400 bp, ~1300 bp, and ~ 500 bp) were selected for ligation into PCR 4-TOPO vector Transfection of competent ce lls, plating, and colony screens were carried out as described previously. Based on the screen, one (for the ~1400 bp sample) and two (~1300 and ~500 bp samples) colonies were cultured and the plasmids harvested for sequencing with forward and reverse vector primer s T3 and T7. Additionally, the ligated vector itself that had not been transfected into competen t cells for replication wa s sequenced with genespecific primer SULF-19.


58 Table 2-3. Summary of scr eening PCR experiments based on seven EST sequences mined from the NCBI database *Adapter primer Lu4-TRsa used as pair for all 2nd Library PCRs 1st Library PCR 2nd Library PCR (RACE) EST name, source (NCBI ID#) Primer pair (nesting pair) # Cycles PCR product(s) Specific primer* (nest primer) # Cycles PCR product(s) 1C1; liver (BM438392) SULF-5; -6 (SULF-7; -8) 30 (12) Band of EST size; confirmed with seq. SULF-5 (SULF-7) 30 (12) 3 bands 1000-1400bp 1C2; liver (BM438489) SULF-1; -2 30 Bands of EST size; confirmed with seq. Not screened 30 Bands of EST size Intestine (CF261660) SULF-17; -18 (SULF-19; -20) 30 (21) Band with initial PCR, none with nesting PCR SULF-17 (SULF-19) 37 (12) Multiple bands for both PCRs (seq. ID SULT6B1) HNK-1; pituitary (CF262806) SULF-21; -22 30 Band of EST size SULF-21 (SULF-23) 37 (20) Multiple faint bands with initial PCR, none with nest Spleen (BM425065) SULF-13; -14 40 Band of EST size SULF-13 (SULF-15) 30 (16) Multiple bands with initial PCR, one with nest ~1200bp Tyrosine-ester; liver (BM438574) SULF-25; -26 (SULF-27; -28) 30 (12) Band of EST size; confirmed with seq. Not screened ----Sultx3; brain (BE212978) SULF-9; -10 (SULF-11; -12) 45 (23) Bands near EST size; none SULT per seq. SULF-9 (SULF-11) 37 (20) Faint band with initial PCR, none with nesting PCR


59 The sequences obtained were aligned into a sing le contig of 531 nucleotides, translating to 177 amino acids. The highest percent identity matc hes, per the NCBI blastx program, were with Danio rerio SULT6B1 (67%) and Salmo salar SULT6B1 (61%). At 177 amino acids, the catfish fragment is approximately two-th irds of a complete gene sequence; the D. rerio and S. salar proteins are 296 and 277 am ino acids, respectively. The catfish SULT6B1 partial cDNA sequen ce and its amino acid translation are: AACAAGATGT CTGCTCCAAC TTTTGCAGCA AACATAAAGT CCAACATGGA GCGGGGCATG GAAATGAAGG 70 ATGAGGAGAA GCTTTACAAG CGAGACGGGA TCCTCTACTC CACGATCATG AGCCCACCGG AGAACCTAGA 140 TGCCCTTAAG GACCTGGAGG CCAGGGAGGA TGATGTCATG CTGGTAGCAT ACCCCAAATG TGGCTGTAAC 210 TGGATGGTAG GAGTGTTGAG GAAAATTATG ACTACATGTG GATACACTCT CTCTGAAAGG CCCCCTCTGA 280 TCGAGTTCCA CTCTCCAGAC GCACAGAAGA ATGCAGCCCA GATGCCATCC CGGCGTTTTT TTGCAACTCA 350 TTTGCATCCT GATTACATCC CTGTTTCATT TAAGACCAAT AAAACAAAGA TGTTGGTGGT GTTTCGAAAC 420 CCCAAAGACA CCGTCGTCTC TTATTACCAT TTCATGAACA AAAACCCAGT GTTGCCCAAA GCCGAGTCCT 490 GGGATAAATT CTTCTCTGAC TTCATGTCCG GTGAAGTGGG C 531 NKMSAPTFAA NIKSNMERGM EMKDEEKLYK RDGILYSTIM SPPENLDALK DLEAREDDVM LVAYPKCGCN 70 WMVGVLRKIM TTCGYTLSER PPLIEFHSPD AQKNAAQMPS RRFFATHLHP DYIPVSFKTN KTKMLVVFRN 140 PKDTVVSYYH FMNKNPVLPK AESWDKFFSD FMSGEVG 177 SULT1v1 A band of the proper size for the EST (named 1C1 by the EST author) was obtained from the first cDNA library (CAPT-30 adapter) by PCR with specific primers SULF5/SULF-6, 30 cycles, and nest PCR with pr imers SULF-7/SULT-8, 12 cycles. A sequence reaction with primers SULF-5 and SULF-6 confirmed the PCR product as the 1C1 EST sequence. RACE PCR with the same cDNA library as template, adapter primer pairs SULT5/TRsa (3 end) and SULF-6/TS-PCR (5 end), for 27 cycles at TA 60 C resulted in smears for the 3 end and two bands for the 5 end, at ~600 and 1900bp. Nest PCR was carried out on a 1:50 dilution of the 5 prod ucts with primers SULF-8/TS-PCR, 58 C. No product was evident after 10 cycles or 15 cycles; after 25 cycl es, only a smear was seen on the gel. This was attributed, possibly, to the fact that a 10 M concentration of adapter primer had been used; a much lower concentration, 2 M, would be better; it was also an indication that the cDNA library was not of high-enough quality. A seco nd RACE PCR was carried out with the new


60 TRsa library as template, specific primer SULF-5 and adapter primer Lu4-TRsa, for 30 cycles at 58 C. Table 2-4. Primers designed to obtai n the entire sequence for SULT1v1. Primer Sequence (5 3) Type SULF-71 GCTCATC AAAACTCATCTACCTGT TCAGTT sense SULF-72 TCATGCTCGG TGTTCGCTAC CATATCTT antisense SULF-73 TCACAATTCAGCA CCAGAGCGTCAGA sense SULF-74 TGATATCCTG GAGTGGAA antisense SULF-76 GTCTTTGGCA TTGCGAGCTA CATA antisense SULF-78 GAGGATATCA TTTGGTCTTG CTT antisense STEXP-71 CACCATGGAGGGA GTTAATACTTCT sense STEXP-73 CTCTGTGTGGAACTGTAGGGTTGTGTT antisense Three bands were seen, between 1100-1400 bp. Nest PCR with specific primer SULF-7 and adapter primer Lu4-TRsa gave three bands also, between 1000-1400 bp in length. Sequencing with vector primers T3 and T7 resu lted in two separate contigs, 798 and 519 bp in length that had highest BLAST identity w ith various mammalian SULT1 isoforms and uncharacterized Danio rerio Family 1 sulfotransferase sequences Since it was believed that the two contigs were the unconnected 5 and 3 re gions of one gene, new primers were designed (SULF-71 to ) to sequence the missing data; prim er details are given in Table 2-4. Sequence reactions were carried out w ith new primers SULF-73, -71, and on the plasmid DNA from the three bands (PCR products) obta ined previously. The sequences derived connected the two separate contigs obtained earlier, for one comp lete cDNA sequence that included the entire coding region and some untransla ted region (UTR) as well. The complete nucleotide seque nce composing both coding and untranslated regions, with the coding region in blue and start/stop codons underlined, is as follows; the translated amino acid sequence is given in Figure 2-4:




62 AA sequence (coding/ORF only):MEGVNTSSMK STCRPELFDF EGISMVHYFT DNWEKVQKFQ ARPNDILIAT YPKAGTTWVS YILDLLYFHN 70 SAPERQTSLP IFVRVPFLEA VFPEMPTGVD LADKLPNTPR LIKTHLPVQL VPKSFWEQNC KVVYVARNAK 140 DNAVSYFHFA RMINLLPEPG NWNTFLQSFM DGKLVSGPWY DHVTGYWEKK QTYSNLHYMF FEDMVANTEH 210 ELEQLCSFLG LSTPAEERER ITKCVHFDVM KQNNMTNHSS FSHMDFKISP FMRKGKVGDW KNHFTVAQNE 280 RFDEHYWQKM KNTTLQFHTE I 301Species Family, subfamily Identity Accession # Channel catfish 1, isoform 3 99% NP_001187436.1 Blue catfish 1, isoform 3 96% ADO28255.1 Zebrafish 1 (multiple isoforms) 65-77% a European chub 1, isoform 3 75% gb_ABJ98761.1 Salmon 1, isoform 3 73% NP_001134825.1 Northern pike 1, isoform 3 73% gb_ACO14398.1 Sablefish 1, isoform 3 68% gb_ACQ58293.1 Mouse 1D1 1C2 49% 48% NP_058051.3 NP_081211.3 Rat 1D1 49% NP_068537.1 Dog 1C2 50% XP_532396.1 Human 1A1a 1E1 47% 46% NP_001046.2 NP_005411.1 a: NP_899191.2, NP_899190.2, NP_891986.1, NP_001132953.1, NP_001132954.1, NP_991183.1 Figure 2-4. Amino Acid sequence of the coding re gion of a putative family 1 sulfotransferase, SULT1v1, from channel catfish liver tissue. Sulfotransferase amino acid sequences sharing the highest % identity are listed, as determined by the NCBI Blastp program (nr and refseq databases). was that of the channel catfis h tyrosine ester EST, which was at the genes 5 end. PCR experiments were carried out to get additional 3 sequence data a nd to test two different vector primer concentrations: 10 M and 1 M. Primer pair SULF-25/CAP-T30 was used for both concentrations of CAPT-30, with annealing temperature of 58 C with 2-minute extension time. After 30 cycles no bands were seen for the lower concentration, and only a very faint band (~900 bp) at the higher concentration. Ten cycles were added for each concentration, resulting in multiple bands for the 10 M sample and only a single very faint band for the 1 M


63 sample (~900 bp). Because of the number of ba nds obtained, the PCR experiment was repeated with 10 M CAP-T30, 32 cycles, 58 C TA, 2 minute extension time. Although a number of bands were again obtained, two were very strong (~850 and 1750 bp) and were selected for eventual sequencing. After liga tion and culturing of selected se quences, seven colonies were screened as described previously using commercial vector primers T3 and T7. Only the lower band (~850 bp) samples gave positive results; nothi ng was seen in the colony screen gel for the higher (~1750 bp) band. Sequence results (using vector primers M13-Forward and Reverse) matched a 3 D. rerio sulfotransferase gene (but not to th e stop codon) and the catfish tyrosine ester EST, respectively. Thus far a contig was formed of 255 amino acids (including some 5 UTR), which was used to design new primers, SULF-31 and to obtain the remaining 3 sequence of the gene; primer details ar e given in Table 25 and Table 2-1. Using the second CAP-T30 library as template PCR was carried out with primer pair SULF-31/CAPT-30, TA 58 C, 1.5 minute extension time, and 2 M vector primer. A faint band was seen after 31 cycles, and a nest PCR was done with 1:25 dilution of that product, same conditions as before, with primer pair SULF -33/CAPT-30. Three bands were seen after 20 cycles, at ~200, 750, and 1000 bp. The smallest band was selected for even tual sequencing (as it was the correct size for the length of missing 3 sequence) with T3 and T7 primers, and completed the S1v2 contig at the 3 end. Table 2-5. Primers designed to obtain the entire SULT1v2 sequence from channel catfish liver cDNA libraries. Primer Sequence (5 3) Type SULF-25 GCTGTGATAAGTA GCATGACGGAA sense SULF-26 CACATGATCATACCAAGGTCCAAA antisense SULF-27 CATCATGGAA GGATCTGATGTT sense SULF-28 CTGGATCAGGTTGTAGGTGATT antisense SULF-31 ATGTGACTGGCTACTGGGAGAAGAA sense SULF-33 ACTGATCGTGA GGTGGAGCGTCTTT sense


64 The complete nucleotide seque nce composing both coding and untranslated regions, with the coding region in blue type a nd the start/stop codons underlined, is as follows; the translated amino acid sequence is given in Figure 2-5: SULT1v2 GAAGTCAGCT GTGATAAGTA GCATGACGGA AGTGTGTCCA GGTCAAAAGT TTTGAGGGAA AGAAACTGAG 70 TGACAAAGAC AAAATTGCAA TTTCTTATTT CATC ATG GAA GGGTCTGATG TTTCTTTGAT GACGTCGACC 140 AGCAGACCAG AACTGTTTGA CTTTGAAGGC ATCTCTATGG TGCATTACTT TACTGATGAC TGGGAAAATG 210 TTCAAAACTT CCAGGCAAGA CCAGATGACA TCCTCATTGC TACTTACCCC AAAGCAGGCA CCACCTGGGT 280 TTCCTACATC TTAGACCTTA TTTATTTTGG CAATACAGCA CCAGAGCGTC AGACCTCAAT ACCCATCTAC 350 TTGCGAGTGC CTTTCCTTGA GGCGGTTATT CCTAAGATAG CTACAGGGGT TGAACTAGTA AATAATTTAT 420 CCACCACACC ACGACTCATC AAAACTCATT TACCTGTTCA GCTGGTGCCC AAGTCCTTCT GGGAACAGAA 490 CTGCAAGGTT GTCTATGTAG CTCGTAATGC CAAAGACAAT GTTGTGTCGT ATTTCCACTT TGAGCGCATG 560 AATCACCTAC AACCTGATCC AGGAGACTGG AACAATTATC TACAGATGTT CATGGATGGA AAGAAGGTGT 630 TTGGACCTTG GTATGATCAT GTGACTGGCT ACTGGGAGAA GAAGCAGACA TACTCTAATC TTCACTACAT 700 GTTCTTTGAA GATATGGTGG AGAACACTGA TCGTGAGGTG GAGCGTCTTT CTTCTTTCCT CGGTTTATCT 770 ACACCTGCGG AAGAGAGGGA GAGAATTACA AAATGCGTTC ACTTTGATGT TATGAAGCAG AACAACATGA 840 CCAACCATTC CTCATTCTCA CACATGGACT TCAAGATCTC ACCATTCATG CGTAAAGGTA AAGTTGGAGA 910 CTGGAAGAAT CACTTCACTG TGGCTCAAAA TGAGCGGTTT GATGAACACT ACTGGCAGAA GATGAAGAAC 980 ACAACCCTAC AGTTCCACAC AGAGATT TAA GAATATAATT TACAATATAT TACACAGCAG TTTAGTCAAG 1050 ATTTTTTGTC CACATGTGCT TTACATAATC ATTAAAAGAC ATAATGAAAG GAATGGGGTT TTTTTTAGAT 1120 GCGTGCGTTC ATGGATAGTC CACGTCGAAG GGCGAG 1156 SULT2. A single band of the proper size for the corresponding EST (Spleen) was obtained from PCR using the first cDNA library (CAPT-30 adapter) as template, with specific primers SULF-13/SULF-14, 40 cycl es (30 initially, 10 added), TA 60 C, followed immediately with RACE PCR. RACE PCR wa s carried out with the same c DNA library, with primer pairs SULF -13/LT and SULF-14/TS-PCR, 30 cycles with hot start, TA 58 C, resulted in multiple bands. Nest PCR was carried out on a 1:50 diluti on of the first PCR product, with primer pairs SULF-15/Lu4TRsa and SULF-16/TS-PCR, keeping the same conditions. A band of ~400bp was obtained for the SULF-16/TS-PCR sample afte r 12 cycles and a band of ~ 1200 bp for the SULF-15/Lu4TRsa sample after 16 cycles. These two bands were picked from the gel with a glass Pasteur pipet and ligated into TOPO vector. The ligates were transfected into TOPO TO P10 competent cells and plated as described previously. Five colonies pe r plate were screened for inse rt using primers M13-forward


65 Figure 2-5. Amino acid sequence of a sulfotransfe rase from channel catfish liver tissue. The enzyme, putatively identified as Sult1v2, has highest percent identity with bony fishes, and significantly lower match w ith mammalian sulfotransferases, as determined by the NCBI protein Blast program (blastp; nr and refseq databases). and reverse. No colonies from the SULF-16/TS-PCR sample contained insert, while a single colony from the SULF-15/Lu4TRsa sample contained an insert. Th e corresponding colony was cultured overnight, mini-prepped, and the plasmi d sequenced using primers T3 and T7. The sequence obtained (526 nt), with the original EST sequence (379 nt), formed a contig that included the start codon. Due to the sequenci ng instruments limitations, new primers were needed to determine the rest of the oligomer sequence. Primers SULF-35 and SULF-37 were designed (Table 2-6) to obtain the remaining seque nce data. The resulting sequences fit with the initial contig and ex tended to the 3 end (the stop codon). AA sequence (coding/ORF only):MEGSDVSLMT STSRPELFDF EGISMVHYFT DDWENVQNFQ ARPDDILIAT YPKAGTTWVS YILDLIYFGN 70 TAPERQTSIP IYLRVPFLEA VIPKIATGVE LVNNLSTTPR LIKTHLPVQL VPKSFWEQNC KVVYVARNAK 140 DNVVSYFHFE RMNHLQPDPG DWNNYLQMFM DGKKVFGPWY DHVTGYWEKK QTYSNLHYMF FEDMVENTDR 210 EVERLSSFLG LSTPAEERER ITKCVHFDVM KQNNMTNHSS FSHMDFKISP FMRKGKVGDW KNHFTVAQNE 280 RFDEHYWQKM KNTTLQFHTE I 301Species Family, subfamil y Identity A ccession # Channel catfish 1, isoform 3 84% NP_001187436.1 Blue catfish 1, isoform 3 84% ADO28255.1 Zebrafish 1 (multiple isoforms) 66-78% a European chub 1, isoform 3 76% gb_ABJ98761.1 Salmon 1, isoform 3 74% NP_001134825.1 Northern pike 1, isoform 3 72% gb_ACO14398.1 Rainbow smelt 1, isoform 2 70% gb_ACO08931.1 Mouse 1D1 52% NP_058051.3 Rat 1D1 51% NP_068537.1 Dog 1A1 49% NP_001003223.1 Cow 1C2 49% NP_001074388.1 Human 1A1a, 1A2, 1A3 1E1 48% 45% b NP_005411.1 a: NP_899191.2, NP_899190.2, NP_891986.1, NP_001132953.1, NP_991183.1, NP_001132954.1 b: NP_001046.2, NP_001045.1, NP_003157.1




67 AA sequence (coding/ORF only):MTEADLYTEY KGVYVPKNLH PSNSLKYYED FTFRPDDVLI VTYPKSGTTW MQEIVPLIHS EGDLTPVHTI 70 PNWDRVPWLE EHRAKILNLE QRPSPRIFAT HFHYGMMNES YFQVKPKVIY TMRNPKDVFT SSFHYYGMAS 140 YLVNPGTVDE FLEKFLNGKI MFGSWFDHVK GWLKAKDKDH IFYISYEEMI EDLNGCVTKL AQFLEKPLSP 210 EVIEKISENC LFKNMKKNKM SNYSLVPEEF MNQKKSEFLR KGVAGDWKNF FTEAQTEKFN TVYKDKMKDV 280 TFRFVWD. 288Species Family, subfamily Identity Accession # Zebrafish 2, isoform 2 2, isoform 1 74% 72% NP_001071637.1 NP_944596.2 Salmon 2 74% NP_00113490.1 Northern pike 2 69% gb_AC014375.1 Zebrafish 2, isoform 3 52% NP_001071636.2 Channel catfish 2B1 43% NP_001187828.1 Cow 2B1 44% NP_001075195.1 Mouse 2B1 43% NP_059493.2 African frog 5 44% NP_001090347.1 Human 2B1a 2B1b 43% NP_004596.2 NP_814444.1 Rabbit 2B1 42% NP_001075645.1 Rat 2B1 41% NP_001034754.1 Figure 2-6. Translated amino aci d sequence of a sulfotransferase from channel catfish liver tissue. The enzyme, putatively identified as SULT2, has highest pe rcent identity with zebrafish and other bonyfish, and signif icantly lower match with mammalian sulfotransferases, as determined by th e NCBI protein BLAST program (blastp, nr and refseq databases). New PCR experiments were undertaken to obtai n the complete oligomer and confirm the putativeSULT2 sequence. Using as template the TRsa cDNA library, experiments with three primer pairs were set up: gene -specific primers SULF-13, -35, and -14, each paired with vectorspecific primer Lu4-TRsa (seque nce in Table 2-1). PCR conditi ons were hot start, then 30 cycles at TA 60 C, 1.5-minute extension, and a 10-minut e final extension. A single band, ~500 bp, was seen for the SULF-14 sample, but nothing for the other two. After 10 additional cycles (increments of 5 cycles) a band of ~1050 bp for the SULF-35 sample and of ~1200 bp for the


68 SULF-13 sample were obtained. Nest PCR was th en performed with 1:50 d ilutions of the initial PCR products as templates. Primer pairs fo r the SULF-14, -35, and 13 samples were SULF16/Lu4, SULF-37/Lu4, and SULF-15/Lu4, respectively. PCR conditions were the same, except the nest PCR was run for 12 cycles. A single band was obtained for each sample: ~1100 bp for SULF-15, ~1050 bp for SULF-37, and ~500 bp for SULF -16. Because 40 cycles were needed to obtain bands for the SULF-13 and -35 samples, th e PCR was repeated with 35 initial cycles, all other conditions unchanged. A single band was seen for each sample: ~1400 bp for the SULF13 sample and ~1000 for SULF-35 sample. Nest PCR was done on 1:50 dilutions of these PCR products; all conditions the same as the previous nest PCR. After 12 cycles a band of ~1000 bp was seen for the SULF-37 sample, and after 15 total cycles a band of~ 1400 bp for the SULF-15 sample. At this point there we re five nest PCR products one for the SULF-16/Lu4 nest (~500 bp), two for the SULF-15/Lu4 samples (~1100 a nd ~1400 bp), and two for the SULF-37/Lu4 nest (~1050 and ~1000 bp). All samples were ligated into TOP10 cells, plated, and five colonies selected from each sample for screening. Nearly all colonies containe d the desired insert; a representative selection was chosen and the corresponding cultures mini-prepped. The plas mids were sequenced using vector primers T3 and T7, which were included in the TOPO vector kit. The sequences obtai ned were aligned into the existing full-length contig. The new seque nces matched, confirming the cDNA sequence of SULT2. Additional sequences were obtained, giving a tota l of seven discrete contigs from separate PCR experiments. Alignment of the amino acid tr anslations of these sequences reveals a singlenucleotide polymorphism (SNP) of c/a at nucleotide 573 of SULT2. This translates to an amino acid change of D191E.


69 Discussion In summ ary, three complete and one partial (approximately two-thirds) sulfotransferase gene sequences were obtained from channel catfish liver cDNA. Two of the complete genes, based on NCBI BLAST results, have the highest percent identity with family one sulfotransferases; the third has gr eatest percent identity with fa mily two sulfotransferases, and the partial sequence has highest percent identity with SULT6B1. Based on these results, the sequences have been putatively termed SU LT1v1 (S1v1), SULT1v2 (S1v2), SULT2 (S2), and SULT6B1. The informal use of v1 and v2 is for version, not to be confused with the formal use of _v for variant. SULT6B1. Of the four fish SULT6B1 isoforms with which the catfish sequence is aligned in Figure 2-7, only that of Danio rerio has been characterized (Sugahara, et al., 2003). Among the endogenous compounds tested as substrates, the D. rerio enzyme showed greater specific activity toward dopamine (3.9 nmol/min/mg), weak er activities toward thyroid hormones T3 and T4 0.7 and 1.2 nmol/min/mg, respectively), and no measureable activity toward DHEA, estrone, or dopa. Of the xenobiotic compounds tested, the authors report partic ularly strong activity (specific activity 169 nmol/min/mg) toward n-pro pyl gallate, an antioxi dant food additive. Rather strong activities (1.2 to 19.5 nmol/min/mg) were seen toward most of the other xenobiotic compounds tested, including flavonoids, isofla vonoids, and other phenolic compounds, which the authors state is in line with the potential role of this enzyme in the detoxification and excret ion of xenobiotics. A zebrafish SULT6B1 oligomer was included in a biomarker array (Lam et al., 2008) for expression-based chemogenomics using whole-adult organisms. The purpose of the array is for the identification of biomarkers for potent aryl hydrocarbon receptor (AHR) and estrogen receptor (ER) agonists, focusing on polycyclic (h alogenated) aromatic hydrocarbons (P(H)AHs)


70 and estrogenic compounds (ECs). When P(H) AHs and ECs were tested that were known AHR and ER agonists, respectively, the zebrafish SULT6B1 was significantly upregulated in the P(H)AH group, but showed no significant response in the EC group. Therefore, it can be concluded that the enzyme is associated with, and perhaps regulat ed by, aryl hydrocarbon receptors but not estrogen receptors. The NCBI database tool, HomoloGene, identifi es eight putative SULT6B1 homologs, from 265 to 304 amino acids in length. The bonyfish is oforms range from 277 to 297 amino acids in length. Because of the differences in lengths of known SULT6B1 isoforms it can only be estimated how much of the catfish SULT6B1 is needed to complete the coding sequence, which is approximately 100 amino acids (300 nucleotides ). Based on an alignment with other fish SULT6B1s, the translated catfish sequence obtained thus far appears to be just upstream of the start codon (Figure 2-7). Sense and antisense primers from the mid-region of the gene (corresponding to nucleotide region 450-550) should be designed, with nest primer pairs, to obtain the complete gene sequence with RACE PCR. SULT1v1 and SULT1v2 Two cDNA sequences encoding di stinct sulfotransferase isof orm s having greater than 45% identity with known Family 1 enzymes have been obtained from cha nnel catfish liver tissue. The open-reading frame (ORF) of each isoform encompasses 903 nucleotides and codes for a 301-amino acid polypeptide, and share 85% identity and 91% similarity. A BLAST search of the NCBI databases revealed that, while the tw o sequences share high id entity (68-78%) with other non-catfish Family 1 fish sulfotransferases, they display much lower identity with known mammalian Family 1 SULTs (maximum 52%). Two Family 1 catfish sequences, from channel catfish and blue catfish, were nearly identi cal to the SULT1v1 and SULT1v2 sequences. The


71 Catfish NKMSAPTFAANIKSNMERGMEMKDEEKLYKRDGILYSTIMSPPENLDALKDLEAREDDVM 60 Zebrafish --------MSQMKSRMETAAKMKDEDKLYRRDGILYSTVLSPPETLDKLKDLQAREDDLI 52 Rainbow_smelt ---MDQQLKSHGMSKMEEAKNVKDEDKLYKYEGVLYSSIMSPQENLKGLENFEARPDDLL 57 Northern_pike -------MTSLFSAKIQLAKEMKEEDKSYRYNGVLYSVIMSPEENLKAIESLEARADDVV 53 : :.:: ::*:*:* *: :*:*** ::** *.*. ::.::** **:: Catfish LVAYPKCGCNWMVGVLRKIMTTCG---YTLSERPPLIEFHSPDAQKNAAQMPSRRFFATH 117 Zebrafish LVAYPKCGFNWMVAVLRKIINASTGKDEKPPERPPLVEFLPPTVQEEMAQMPPPRLLGTH 112 Rainbow_smelt LVAYPKCGFNWMVSVIRKIRTASSG-QNEMPSGPPLIEFFPPEVQQLAKERPSPRFLGTH 116 Northern_pike LVAYPKCGFNWMVAVLRKIIVAATG-EKTDSKYNALIEFCGPQMQEVIHQAPAPRLLGTH 112 ******** ****.*:*** :. .. .*:** *: : *. *::.** Catfish LHPDYIPVSFKTNKTKMLVVFRNPKDTVVSYYHFMNKNPVLPKAESWDKFFSDFMSGEVG 177 Zebrafish LHPDNMPATFFTKKPKILVVFRNPKDTLVSYYHFMNKNPVLPNAESWDKFFSDFMTGDVS 172 Rainbow_smelt LHPDNIPSSFIAKKTKMLVVFRNPKDTVVSYYHFMNNNPVLPNTKSWDAFFTDFMKGEVA 176 Northern_pike LHPDNLPLSINTKKTKLLVMFRNPKDTVVSYYHFSNNNPVLPTPESWDSFFSDFLSGQVP 172 **** :* :: ::*.*:**:*******:****** *:*****..:*** **:**:.*:* Catfish -----------------------------------------------------------Zebrafish WGSYFDHALAWEKRIDDPNVMIVMYEDLKQNLPEGVKKISEFFSLPLTDEQVSSIAGQST 232 Rainbow_smelt WGSYFDHALAWEKRMDGPNVMIVTYEQLKENLVDGVRNISRFFGFTLTDEQVQTIANEST 236 Northern_pike WGSYFDHALGWEKRMEDSNVMIVTFEELKQDLSEGVRRVSEFFGFSLSDAQVQAIAGESN 232 Catfish -----------------------------------------------------------Zebrafish FSAMVENSQKSHGNFGSIFFRKGEVGDWKNHFSEAQSKQMDELYHSKLAGTKLAARMNYD 292 Rainbow_smelt FNAMKESSKNSHGQHGNVFFRKGEVGDWKNNFSEAQSKQMDEEFQKLLAVLV-------288 Northern_pike FKAMKESSKESYGHMSNVIFRKGEVGDWKNHFTPAQSQEMDAAFEKHLAGTKLGAKLKYD 292 Catfish ----Zebrafish LYCQ296 Rainbow_smelt ----Northern_pike LYCRD 297 Sequence Accession Ids: Zebrafish ( D. rerio ) NP_999851.1; Rainbow smelt ( O. mordax ) ACO09698.1; Northern pike ( E. lucius ) ACO13268.1 Figure 2-7. Alignment (ClustalW2) of the translated partial gene sequence obtained from channel catfish liver with othe r bony fish SULT6B1 sequences. matched sequences, identified as cytosolic sulf otransferase 3, shared 96-99% identity with SULT1v1 and 84% identity with SULT1v2. Theref ore, according to nomenclature guidelines, the catfish sequences can be considered member s of Family 1 but not yet assigned a subfamily designation. Conserved motifs. Both sequences contain re gions highly conserved among all sulfotransferases, considered signature seque nces. These include the 5-phosphate binding loop (PSB-loop), two 3-PAPS binding sites (one designated 3PB), and the dimerization sequence, as well as conserved individual ami no acids involved in cata lysis and/or binding.


72 The first of these signature se quences, the PSB-loop, is locate d at the N-terminal region and is composed of the conserved motif TYPKSGT x W (Kakuta, et al., 1997). This region, which is present in human sulfotransferase s 1E1 and 1A1a at residues 44-52 and 45-53, respectively (Figure 2-7), is present in both catfish SULT1 isoforms as 50TYPKAGTTW58. The identical sequence is present in Danio rerio SULT1 sequences isoforms 1, 2, 3, and 4, which have the highest identity with the catfish se quences (78-68%). A reaction mechanism involving the lysine residue (Lys47 for human SULT1E1) within this region has been proposed by Pedersen, et al (2002), in which the lysine, along with highly conserved residues Ser137 and His107 (for human 1E1), promotes the dissociati on of the sulfuryl leaving group. These residues are conserved in both catfis h SULT1 isoforms at Lys53, Ser145, and His115. Mouse1E1 residues His108 and Lys106 (corresponding to human1E1 His107 and Lys105) form hydrogen bonds with the acceptor hydroxyl (O-3) of estradiol substrate (Kakuta, et al ., 1997), residues that are conserved in the catfish SULT1 isoforms at His115 and Lys113. This histidine has been implicated as a general base that abstracts a proton from the 3-OH of estradiol, thereby activating it for attack. A second signature sequence is present at the C-terminal end of the enzyme, at 263RKGKVGDWKNHFT275 for both catfish SULT1 isoforms. This RKG xx GDWK xxFT conserved motif is critical for the binding of PAPS (Komatsu, et al ., 1994) and incorporates a conserved G xxG xxK P-loop sequence found in ATP and GT P binding proteins (Saraste, et al ., 1990). The first residue, Arg263, corresponds to Arg256 in humanSULT1E1, Arg257 in mouse1E1, and Arg276 in flavonol 3-sulfotransfe rase, and is involved in binding to the 3phosphate of PAPS (Yoshinari, et al., 2001). In addition, the adenine of the PAP molecule is in a


73 parallel ring-stacking arrange ment with conserved residues Trp58 and Phe237 (which correspond to Trp52 and Phe228 in humanSULT1E1). Another 3-PAPS bindi ng region is present at 135VARNAKDNAVSY146 for catfish SULT1v1 and 135VARNAKDNVVSY146 for catfish SULT1v2, designated as 3PB (Yoshinari, et al ., 2001). In particular, Arg137 and Ser145 (w hich correspond to mouse1E1 Arg130 and Ser138) directly interact w ith the oxygen atoms of the 3 -phosphate group (Kakuta, et al ., 1997). The authors state that, together, the two residues neutra lize the negative charge of the phosphate. A dimerization motif overlaps the C-terminal 3-PAPS binding region at motifs that are underlined. The sequences were aligned with ClustalW2. The motif sequence is 271KNHFTVAQNE280 for both catfish SULT1 is oforms (Petrotchenko, et al., 2001). Between two sulfotransferase proteins the KTVE motif (conserved structure K xxx TV xxx E) forms a hydrophobic zipper-like anti-paralle l structure that is enforced by ion pairs at each end. Residues Asn272 to Glu280 from each mono mer form four backbone hydrogen bonds, and are flanked by ionic interactions between si de-chains Glu280 and Lys271 at both ends of the loop. The presence of this motif indicates that the catfish SULT1 isofor ms behave as dimers (either homoor hetero-) in vivo This is in contrast to the mouseSULT1E1, which is a monomer (Petrotchenko, et al., 2001), and in which the conserved re sidues TV of the KTVE dimerization motif are substituted with residues PE. Substrate binding region The crystal structur e of humanSULT1A1pnitrophenol (pNP) complex (Gamage, et al ., 2003) shows two pNP molecules in the substr ate binding pocket. The L-shaped binding pocket that accommodates the two molecules is very hydrophobic; predominant aromatic residues making it up are Phe81, Phe84, Phe142, His149, Phe247, Phe255,


74 Figure 2-8. Alignment of translated channe l catfish sequences SULT1v1 and SULT1v2 with human sulfotransferases 1A1a and 1E1. Critic al and conserved residues are indicated with arrows, with the corresponding catfish residues noted in the key. Sequence identities are: Human_1A1a: NP_899191.2, Human_1E1: NP_005411.1. Tyr240, Tyr169, and Phe24. Predominate alipha tic residues are Ile89, Val148, Met248, Ile21, Met77, Val243, Pro90, and Ala146. As in the mouseSULT1E1-E2 complex, Phe142 and Phe81 form a substrate access gate that permits binding of pl anar substrates only at the catalytic site. Two loop regions that close tightly over the SULT1A1pNP complex are opened up in the SULT1A1-E2 complex, increasing the space ava ilable for E2 binding (Gamage, et al., 2005); Catfish_1v1 MEGVNTSSMKSTCRPELFDFEGISMVHYFTDNWEKVQKFQARPNDILIATYP K AGTTW VS 60 catfish_1v2 MEGSDVSLMTSTSRPELFDFEGISMVHYFTDDWENVQNFQARPDDILIATYP K AGTTW VS 60 Human_1A1a -----MELIQDTSRPPLEYVKGVPLIKYFAEALGPLQSFQARPDDLLISTYP K SGTTW VS 55 Human_1E1 ------MNSELDYYEKFEEVHGILMYKDFVKYWDNVEAFQARPDDLVIATYP K SGTTW VS 54 : ..*: : : *.. :: *****:*::*:****:****** Catfish_1v1 YILDLLYFHNSAPERQTSLPIFVRVPFLEAVFPEMPTGVDLADKLPNTPRLI K T H LPVQL 120 catfish_1v2 YILDLIYFGNTAPERQTSIPIYLRVPFLEAVIPKIATGVELVNNLSTTPRLI K T H LPVQL 120 Human_1A1a QILDMIYQG-GDLEKCHRAPIFMRVP F LEFKAPGIPSGMETLKDTP-APRLL K T H LPLAL 113 Human_1E1 EIVYMIYKE-GDVEKCKEDVIFNRIPFLECRKENLMNGVKQLDEMN-SPRIV K T H LPPEL 112 *: ::* *: *: *:**** : .*:. .. :**::***** Catfish_1v1 VPKSFWEQNCKVVYVARNAKDNAV S YFHFARMINLLPEPGNWNTFLQSFMDGKLVSGPWY 180 catfish_1v2 VPKSFWEQNCKVVYVARNAKDNVV S YFHFERMNHLQPDPGDWNNYLQMFMDGKKVFGPWY 180 Human_1A1a LPQTLLDQKVKVVYVARNAKDVAV S YYHF YHMAKVHPEPGTWDSFLEKFMVGEVSYGSWY 173 Human_1E1 LPASFWEKDCKIIYLCRNAKDVAV S FYYFFLMVAGHPNPGSFPEFVEKFMQGQVPYGSWY 172 :* :: ::. *::*:.***** .**:::* *:** : ::: ** *: *.** Catfish_1v1 DHVTGYWEKKQTYSNLHYMFFEDMVANTEHELEQLCSFLGLSTPAEERERITKCVH F DVM 240 catfish_1v2 DHVTGYWEKKQTYSNLHYMFFEDMVENTDREVERLSSFLGLSTPAEERERITKCVH F DVM 240 Human_1A1a QHVQEWWELSRTHP-VLYLFYEDMKENPKREIQKILEFVGRSLPEETVDFVVQHTS F KEM 232 Human_1E1 KHVKSWWEKGKSPR-VLFLFYEDLKEDIRKEVIKLIHFLERKPSEELVDRIIHHTS F QEM 231 .** :** :: : ::*:**: : :*: :: *: . : : : *. Catfish_1v1 KQNNMTNHSSFS--HMDFKISPFMRKGKVGDW K NHFTVAQN E RFDEHYWQKMKNTTLQFH 298 catfish_1v2 KQNNMTNHSSFS--HMDFKISPFMRKGKVGDW K NHFTVAQN E RFDEHYWQKMKNTTLQFH 298 Human_1A1a KKNPMTNYTTVPQE F MDHSISPFMRKGMAGDW K TTFTVAQN E RFDADYAEKMAGCSLSFR 292 Human_1E1 KNNPSTNYTTLPDEIMNQKLSPFMRKGITGDW K NHFTVALN E KFDKHYEQQMKESTLKFR 291 *:* **:::.. *: .:******* .****. **** **:** .* ::* :*.*: Catfish_1v1 TEI 301 catfish_1v2 TEI 301 Human_1A1a SEL 295 Human_1E1 TEI 294 :*: Key: Arg263Thr275 Ser145, Lys53 Lys113, His115 Trp58, Phe237 Lys271 Glu280


75 these regions consist of residues 146-154 between 6 and 7, and particularly residues 84-90 just preceding 4. Another flexible region is the loop that connects 12 and 13 in SULT1A1E2; this loop covers the substr ate upon binding and is involved with several direct contacts with bound E2. Phe247 is involved in binding by its ro tamer conformation; in SULT1A1-E2 complex the residue adopts a conformation diffe rent from that in the SULT1A1p NP crystal structure, thereby increasing the space available to bind the fuse d ring structure of the steroid. Plasticity of the binding region is critical for the isozymes broad substrat e range, which includes planar aromatics (phenols), large L-shaped aromatics (i odothyronines), and extended planar aromatic ring systems (estrogens, hydroxylamines, heterocyclic hydroxlamines). In mouseSULT1E1, a region at the end of a conserved sequence of amino acids forming the 6 helix harbors four of the 20 amino-acid di fferences seen between humanSULT1A1 and humanSULT1A3. This alpha-helix (no. 6) is predicted to be locat ed over the top of the substrate binding pocket of members of the phenol SULT fa mily (Kakuta, et al, 1997). HumanSULT1A1 and SULT1A3 are 93% identical but have very specific differences in substrate activity; particularly, SULT1A1 selectivel y sulfates 4-nitrophenol at low micromolar concentrations, and SULT1A3 preferentially sulfates dopamine at si milar concentrations. Mutagenesis studies by Dajani and associates (1998) determined th at a specific amino acid in this region, Glu146, governs the substrate specificity of SULT1A3 (SULT1A1 has Ala146). At physiological pH, dopamine, like other biogenic amines, carries a po sitive charge; therefore, the importance of the acidic glutamic acid suggests that an interact ion between positively charged substrates and a negatively charged amino acid in the substrate binding pocket of SULT1A3 is central to the reaction specificity associated with these compo unds. According to the authors, the amino acid at position 146 (or its equivalent) may be an important determinant within various SULT1


76 subfamilies; all SULT1A1 orthologs (mammalia n) known at that time were found to have alanine at position 146, whereas in SULT1B and SULT1C family enzymes, asparagine is found at this position, while estrogen SULTs have valine and sometimes isoleucine or methionine. A conserved methionine is presen t immediately preceding this residue. Channel catfish SULT1v1 has isoleucine at this positi on, while catfish SULT1v2 has aspa ragine, as do all other bonyfish SULT1 sequences sharing highest id entity with the catfish isoforms. Through recombinant chimeric protein experiment s, Sakakibara, et al. (1998) showed that a large central segment of humanSULT1A1 a nd humanSULT1A3, which included amino acids 143 and 146, was involved in substrate recognitio n. Two highly variable regions, 84-89 and 143-148, showed differential roles in substrate binding, catalysis, and sensitivity to inhibition by 2,6-dichloro-4-nitrophenol (DCNP) SULT1A3 is markedly more thermolabile and 3 orders of magnitude less sensitive to inhibition by DCNP, a classic (or known) sulfonation inhibitor used to distinguish SULT1A1 from SULT1A3. A crystal structure of mouseSU LT1E1 with PAP and substrate 17 -estradiol (Kakuta, et al ., 1997) showed that residues Phe142, Ile146, a nd Tyr149 contribute to binding at the E2 binding site (in humanSULT1E1, the correspondi ng residues are Phe141, Val145, and His148). Asn86 is believed to be in a position to form hydrogen bonds with the 17 -hydroxy group, while Lys106 and His108 were within H-bonding distance of the steroi ds 3-hydroxy group. According to site-directed mutagene sis studies by Petrotchenko, et al (1999), the Asn86 interaction is not essential for the sulfotransfe rase activity. Three loops referred to as binding pocket loops (BPL), were reported to form the hydrophobic substrate-binding pocket in




78 Chen and associates (1999) reprised the findings of mouseSULT1 E1 binding residues, humanSULT1A1 region 143-148, and humanSULT1A3 positi on 146 (Kakuta, et al ., 1997; Sakakibara, et al., 1998; Dajani, et al 1998) and concluded that they suggest that amino acids 140-150 comprise a region of substr ate binding common to all sulf otransferases (but actually only to Family 1 sulfotransferases). The authors also point out the detail that all SULT1s contain an acidic amino acid, either glutamate or aspa rtate, at the position between the conserved prolines in this putative binding region. They determined there was a Y xxx K xxP x P conserved motif among phenol SULTs and UGT1A6 (a phenol -catalyzing enzyme) in this region. In examining the alignment of catfish S1v1 and S 1v2 with humanSULT1E1 (which correlates to mouseSULT1E1) and humanSULT1A1 (Figure 2-8) and in alignment with other bony fish Family 1 SULTs (Figure 2-9), the motif Y xxx K xx P x P is not observed but a conserved motif of F xxM xxxx P x P is seen at catfish residues 149-159. The Phe of this motif is indicated in mouseSULT1E1 (Phe142) substrate binding interaction with E2 (Kakuta, et al ., 1997) and, with Tyr81, forms a stricture-like gate that sandwiches portions of Aand B-rings of the E2 molecule (Yoshinari, et al ., 2001). In humanSULT1A1 Phe142 forms a gate with Phe81 that may play a role in maintaining a proper structure of the substrate-binding pocket (Gamage, et al ., 2003); the motif region also encompasses the critical Glu146 resi due that defines the dopamine specificity for humanSULT1A3. The conserved prolines w ith an acidic amino acid between them are present at catfish residues 157-159. Therefore it is likely that this region is involved in substrate binding for the catfish SULT1s, as has been shown for other Family 1 sulfotransferases. SULT2 A cDNA se quence encoding a sulfotransferase enzyme having greater than 45% identity with known Family 2 enzymes has been obtained from channel catfish liver tissue. The openreading frame (ORF) encompasses 864 nucleotides and codes for a 287-amino acid polypeptide.


79 A BLAST search of the NCBI databases revealed that, while the sequence shares high identity (69-74%) with other Family 2 fish sulfotransfe rases (with the exception of a channel catfish sequence identified as SULT2b1, with which it shar es 43% identity), it displays much lower identity with known mammalian Fa mily 2 SULTs (maximum 44%). Therefore, according to nomenclature guidelines, the catfish sequence can be considered a member of Family 2 but not assigned a subfamily designation. Conserved motifs. The sequence contains signature sequences highly conserved among all sulfotransferases. These include the 5 -phosphate binding loop (PSB-loop), two 3-PAPS binding sites (one designated 3 PB), and the dimerization sequen ce. The motifs and critical amino acids are indicated in Figure 2-10. The first of these signature se quences, the PSB-loop, is locate d at the N-terminal region and consists of residues TYPKSGT x W (Kakuta, et al ., 1997). This region, which is present in human sulfotransferases 2A1 a nd 2B1b at residues 41-49 and 6775, respectively (Fuda, et al ., 2002), is present in the catfish SULT2 isoform as 42TYPKSGTTW50. Identical sequences are present in salmon, northern pike, and published channel catfish Fam ily 2 sulfotransferases, and nearly identical sequences (one am ino acid difference) are present in Danio rerio SULT2 sequences isoforms 1 and 2 (Figure 2-12), which have the highest identity with the catfish sequences (74-69%). The identical sequence is also present in zebrafish SULT2 isoform 3, which has 52% identity with the catfish SULT2 isoform. This region contains the highly conserved lysine residue involved in ca talysis (Pedersen, et al ., 2002; Lee, et al ., 2003). This residue, present at Lys70 in human SULT2B1b, along with highly cons erved residues Ser155 and His125 in human 2B1b, promotes the dissociation of the sulf uryl leaving group. These residues are


80 Figure 2-10. Alignment of the channel catfish putative SULT2 translated amino acid sequence with the human SULT2 isofor ms sharing highest % identi ty. Conserved regions are marked and critical amino acids highlight ed. Sequence identities: Human2B1b: NP_814444.1; Human2A1: AAH20755.1 conserved in the catfish SULT2 isoform at Lys45, Ser131, and Hi s101; a histidine residue is also present at position 103. A second signature sequence is present at the C-terminal end of the catfish SULT2 enzyme, at 250RKGVAGDWKNFFT262. This RKG xxGDWK xxFT conserved motif is critical for the binding of PAPS (Komatsu, et al ., 1994). The first residue, Arg250, corresponds to Arg274 in human2B1b and is involved in binding th e 3-phosphate oxygens of PAPS (Lee, et al ., 2003). Human2B1b MDGPAEPQIPGLWDTYEDDISEISQKLPGEYFRYKGVPFPVGLYSLESISLAENTQDVRD 60 Hum2A1 --------------------------MSDDFLWFEGIAFPTMGFRSETLRKVRDEFVIRD 34 catfish_S2 ------------------------MTEADLYTEYKGVYVPKNLHPSNSLKYYED-FTFRP 35 .. : ::*: .* ::: .: .* Human2B1b DDIFIITYP K SGTTWMIEIICLILKEGDPSWIRSVPIWERAPWCETIVG-AFSLPDQYSP 119 Hum2A1 EDVIILTYP K SGTNWLAEILCLMHSKGDAKWIQSVPIWERSPWVESEIG-YTALSETESP 93 catfish_S2 DDVLIVTYP K SGTTWMQEIVPLIHSEGDLTPVHTIPNWDRVPWLEEHRAKILNLEQRPSP 95 :*::*:*******.*: **: *: .:** ::::* *:* ** * : ** Human2B1b RLMSS H LPIQIFTKAFFSSKAKVIYMGRNPRDVVV S LYHYSKIAGQLKDPGTPDQFLRDF 179 Hum2A1 RLFSS H LPIQLFPKSFFSSKAKVIYLMRNPRDVLV S GYFFWKNMKFIKKPKSWEEYFEWF 153 catfish_S2 RIFAT H FHYGMMNESYFQVKPKVIYTMRNPKDVFT S SFHYYGMASYLVNPGTVDEFLEKF 155 *::::*: :: :::*. *.*.** ***:**..* :.: : .* : ::::. Human2B1b LKGEVQFGSWFDHIKGWLRMKGKDNFLFITYEELQQDLQGSVERICGFLGRPLGKEALGS 239 Hum2A1 CQGTVLYGSWFDHIHGWMPMREEKNFLLLSYEELKQDTGRTIEKICQFLGKTLEPEELNL 213 catfish_S2 LNGKIMFGSWFDHVKGWLKAKDKDHIFYISYEEMIEDLNGCVTKLAQFLEKPLSPEVIEK 215 :* : :******::**: : :.::: ::***: :* : ::. ** :.* : Human2B1b VVAHSTFSAMKANTMSNYTLLPPSLLDHRRGAFLRKGVCGDWKNHFTVAQSEAFDRAYRK 299 Hum2A1 ILKNSSFQSMKENKMSNYSLLSVDYVVDK-AQLLRKGVSGDWKNHFTVAQAEDFDKLFQE 272 catfish_S2 ISENCLFKNMKKNKMSNYSLVPEEFMNQKKSEFLRKGVAGDWKNFFTEAQTEKFNTVYKD 275 : :. *. ** *.****:*:. : .: :**:**.*****.** **:* *: ::. Human2B1b QMRGMP--TFPWDEDPEEDGSPDPEPSPEPEPKPSLEPNTSLEREPRPNSSPSPSPGQAS 357 Hum2A1 KMADLPRELFPWE----------------------------------------------285 catfish_S2 KMKDVT-FRFVWD----------------------------------------------287 :* .:. *: Human2B1b ETPHPRPS 365 Hum2A1 -------catfish_S2 -------PSB-loop 3 PB


81 In addition, the adenine of the PA P molecule is sandwiched between two aromatic residues, in a parallel/anti-para llel ring-stacking a rrangement with conserved residues Trp50 and Phe222 (which correspond to Trp75 and Phe246 in human 2B1b). Another 3-PAPS binding region, desi gnated as 3PB (Yoshinari, et al ., 2001), is present at 121TMRNPKDVFTSS132 for catfish SULT2. Although many residues are different from the SULT1 conserved region, critical residues Arg123 and Ser131 are present, which directly interact with the oxygen atoms of the 3-phosphate group (Kakuta, et al ., 1997). An examination of SULT2 sequences, mammalian and fish, that share highest identity with catfish SULT2 reveals the conserved motif shown in Figure 2-11 (which includes three residues upstream from the original motif). Interestingly, the mammalian sequences are almost completely conserved in some residues, while different residues are nearly completely conserved in the fish sequences at these positions. V I Y x M RNPKDVFT/I S S G R V V L Figure 2-11. Conserved motif 3PB determined fr om the alignments of channel catfish SULT2 amino acid sequence and the mammalian and bony fish sequences sharing highest % identity. Single letters ar e completely conserved, x denotes a nonconserved residue, and split residues denote conserved amino acids in fish (top, blue) and mammalian (bottom, red). The dimerization region, the KTVE motif, ove rlaps the C-terminal 3-PAPS binding region at 258KNFFTEAQTE267 for the catfish SULT2 isoform (Petrotchenko, et al ., 2001). This motif (conserved structure K xxx TV xxx E) is involved in the formation of sulfotransferase homoor hetero-dimers; however, in catfish SULT2 sequence the critical V is substituted with E. In the mouse SULT1E1 sequence, a known monomer in vivo, the critical V is substituted with P; therefore, without further inve stigation, it cannot be presumed that the ch annel catfish SULT2 enzyme behaves as a dimer in vivo


82 Substrate binding regions. Unlike the SULT1 family, which is comprised of numerous subfamily members, the mammalian SULT2 family consists of only two known subfamilies: 2A1 and 2B1, the latter of whic h has two varieties, 2B1a and 2B1b. HumanSULT2A1 shares only 37-39% identity to the human2B1 proteins. SULT2B1 isoforms are stereo-specific for hydroxy steroids (Lee, et al ., 2003) and show no detectable sulfonation with 3 -hydroxysteroids or estrogens as substrates (He and Falany, 2006; Meloche and Falany, 2001); while both 2B1 isoforms are capable of sulfonating DHEA, they do so with relatively low efficiencies (Lee, et al ., 2003). Characterization of human2B1a (Mel oche and Falany, 2001) showed activity with pregnenolone, epiandrosterone, DHEA, and androstenediol, all of which possess a 3 -hydroxy group, but no activity with androsterone, testoste rone, dihydrotestosterone, or cholesterol-which is, however, a preferred subs trate of 2B1b (Fuda, et al. 2002). SULT2B1b showed lower activity with pregnenolone than 2B1a with Km values of 7.2 ( 0.9) M versus 2.7 ( 1.1) M. SULT2B1a and 2B1b share 93% identity and are en coded by the same gene, but differ at the amino terminal as a result of an alternativ e exon 1 (Meloche and Falany, 2001; Lee, et al ., 2003). The unique amino terminal sequence is responsible for different preferred substrate specificities for each isoform: pregnenolone for 2B1a and cholesterol for 2B1b (Fuda, et al ., 2002). SULT2A1 has a broader substrate range th an 2B1, is able to sulfonate both and -hydroxy steroids and was initially termed the DHEA sulfotrans ferase for its efficiency with that substrate. Falany et al (1989) found purified human liver SULT2A1 to be active with the 3-hydroxy group of all hydroxysteroids tested as well as the 17-hydroxy group of testosterone and the 3hydroxy of -estradiol and estrone. The crystal structure of human 2A1 in co mplex with substrate DHEA was solved by Rehse, et al (2002), under the ID/name 2A3, in co mplex with PAP by Pedersen, et al 2000, and


83 Ssalar MTEAELYTEYKGVYLPTQLHPQGSLKYYEEFTFRHDDILIVTYPKSGTTWMQEIVPLVQS 60 Elucius MTEY-LYEEYKGVYLPTLIYPQESLKFYEDFTFRQDDILIVTYPKSGTTWMQEIVPLVQS 59 catfish MTEADLYTEYKGVYVPKN LH PSNSLKYYEDFTFRPDDVLIVTYPKSGTTWMQEIVPLIHS 60 Drerio MTESELYSVHKGVFVPTHLHPAESLKYYEDFIFRPDDILIVTYPKSGTIWMQEIVPLVVS 60 *** ** :***::*. ::* ***:**:* ** **:********** ********: Ssalar GGDLTPVLTVPNWDRVPWLEEHRACVLNLEQRASPRMFATHYHYNMMPASFFTVKPKVIY 120 Elucius GGDLSPVLTVPNWDRVPWLEESRARTLNLEQRESPRLFATHYQYDMMPASFFTVKPKVIY 119 catfish EGDLTPVHTIPN WDRVPWLE EHRAKILNLEQRPSPRIFATHFHYGMMNESYFQVKPKVIY 120 Drerio EGDLTLVLTVPNWDRVPWLEEHRAILLSLEQRASPRIFATHFHHQMMNPSYFKIEPRVLY 120 ***: *:*********** ** *.**** ***:****::: ** *:* ::*:.:* Ssalar VMRNPKDVFTSSYHYYGMASYLVKPGTQDQFLQKFINGKVMFGSWFDHVIGWLNAKDQDR 180 Elucius LMRNPKDVFISSYYYHGMASFLVNPGTQEEFLQKFINGEVIYGSWFDHVKGWLNAKDQDC 179 catfish TMRNPKDVFTSSFH Y YGMAS YLVNPGTVDEFLEKFLNGKIMFGSWFDHVKGWLKAKDKDH 180 Drerio VMRNPKDVFISSFHYYGMASFLVNPGTQDEFMEKFLNGNIMFGSWFDHVKGWLNAAEQEH 180 ******** **::*:****:**:*** ::*::**:**::::******* ***:* ::: Ssalar TMYISYEEMIFDLRDSVSKISQFMGKSLDSEVIEKIADHCVFKNMKQNKMSNYSLVPNEF 240 Elucius IMYISYEEMILDLKDSVSRISQFLGKTLDNEVIEKIADHCVFKNMKQNKMSNFSMVPTGF 239 catfish IFYISYEEMIEDLNGCVTKLAQFLEKPLSPEVIEKISENCLFKNMKKNKMSN Y SLVPEEF 240 Drerio ILYISYEEMINDLRASVEKIATFLGKSLSSEVVEKIADHCVFKNMKQNKMSNLSLVPEEF 240 :******** **. .* ::: *: *.*. **:***:::*:*****:***** *:** Ssalar MDHNKSEFLRKGIAGDWKNQFTVAQAEYFDAVYKKQMKDIKYKFVWD 287 Elucius MDQNKSGSLRKGIAGDWKNHFTVAQTEYFDAAYNDKMKDIKYPFVWD 286 catfish MNQKKSE FLRKGVAGDWKNFFT EAQTEKFNTVYKDKMKDVTFRFVWD 287 Drerio MDQKKSEFLRKGIAGDWKNHFSAAQEERFNAVYDDKMKDVKFKFPWD 287 *:::** **:*:****** *: ** *::.*..:***:.: ** Leu19, Trp73, Arg83 His20, Tyr141, Phe248 Trp78, Tyr135 Glu81, Tyr233, Val236 Tyr136 Ala139 Phe240 Figure 2-12. Alignment of the channel catfish putative SULT2 amino acid sequence with the bony fish SULT2 isoforms sharing highest iden tity. Critical residues discussed in the text are noted. A key for the residues is shown at bottom. with substrate ADT and PAP cofactor by Chang, et al ., 2004. Only the latter structure was solved with both substrate and cofactor in place. Rat, mouse, and human SULT2B1a and 2B1b ar e encoded by the same gene with an alternative exon 1 (Meloche and Falany, 2001). The cDNA open reading frames encode 350 and 365 amino acids for 2B1a and 2B1b, respectively, of which the final 342 amino acids are identical. Fuda, et al ., (2002) determined through mutagenesi s studies that the sequence of 18DISEI24 of SULT2B1b is crucial for functionality of cholesterol sulfonation, particularly the two Ile residues. Substitution of Ile20 or Ile23 with charged residues Glu or Lys, or with polar but uncharged Gln, resulted in near loss of catalytic activity with cholesterol. Conservative


84 substitution of either of thes e Ile residues with Leu resulted in 80% (residue 20) and 100% (residue 23) retention of choleste rol sulfotransferase activity, while substitution with methionine resulted in a partial retention (30% at residue 20, 60% at residue 23) of catalytic activity. Removal of 23 amino acids from the aminoterminal end that is unique to SULT2B1b resulted in an almost complete loss of cholesterol sulfotransfe rase activity, whereas removal of the eight amino acids from the amin o -terminal end unique to the 2B1a isoform did not significantly alter pregnenolone su lfotransferase activity. Mutati on of residue Ile51 to Asn in guinea pig SULT2 (Park et al, 1999) altered th e stereo-specificity of the enzyme from hydroxysteroids to -hydroxysteroids. Human SULT2B1s have Ile at this location, indicating its potential importance in s ubstrate specificity for -hydroxysteroids. The catfish SULT2, as well as the fish SULT2s sharing the highest identity, have Gln in this location, which, like Asn, is a polar, hydrophilic amino acid. This suggest s that the fish SULT2s may have an -hydroxy substrate specificity. SULT2A1 has Ala in this location and sulfates a variety of steroids, including both and -hydroxysteroids. Alanine is small, uncharged, hydrophobic, and nonpolar; isoleucine is very similar, with a longer, branched carbon ch ain, so is probably involved in steric hindrance with -hydroxysteroids. The carboxyl ends of the genes have a unique prolin eand serine-rich extension (He and Falany, 2006) of about 53 amino acids. The spec ialized carboxyl-terminal end of the SULT2B1s is not critical for catalysis; removal of 53 am ino acids from the end, which is common to both SULT2B1a and SULT2B1b, does not significantly redu ce the ability of eith er to sulfonate its preferred substrate (Fuda, et al ., 2002). The authors found evidence of phosphorylation in this region, particularly of phosphoserine. Their studies suggest that the extension is involved in protein stability and nuclear localization; while the native form is present in nuclei in human


85 tissues (particularly placental) and transfected Be Wo cells; a truncated form did not localize to the nuclei. This extended carboxy-terminal sequen ce rich in proline and serine residues appears to be unique to human SULT2B1 enzymes; it has not been found in other human SULT2 isoforms (He and Falany, 2006) or the guinea pi g (Meloche and Falany, 2001). The mouse and rat orthologues possess only shor t carboxy-extension with no putative phosphorylation sites (Kohjitani, et al., 2006). In the crystal structure of humanSULT2B 1b with DHEA or pregnenolone (Lee, et al ., 2003), three regions were described in the substr ate binding pocket near the steroids A ring, B and C rings, and D ring at the enzymes surface. The two steroids we re reported to have identical orientations in SULT2 B1b. Although the catalytic orie ntations of substrates in SULT2B1b were determined to be very differe nt from the orientati ons in humanSULT2A1, some of the stabilizing residues described are conserved. The correspon ding catfish residues are either identical (mostly) or si milar in the bony fish sequences having highest percent identity with the SULT2 catfish sulf otransferase (Figure 2-12). In SULT2B1b, residues lining the binding pocke t near the A ring of either DHEA or pregnenolone are Phe272, Tyr159, Gln165, Tyr44, and Trp103. Correlating residues in the SULT2 catfish sequence are Phe248, Tyr135, Ty r141, His20, and Trp78. The aromatic side chain of Trp103 in SULT2B1b is stacked paralle l to the A and B rings of the steroid; its correlating residue in SULT2A1, Tr p77, plays the same role with androsterone (Chang, et al, 2004). In SULT2A1, the A and B rings of andros terone (ADT) are sandwiched between side chains of Trp77 on one side and Phe133 and Tr p134 on the other, stabilizing the ADT A ring. Thus, the Trp residue is a conser ved feature involved in stabiliz ing the 3-hydroxyl steroid A ring for SULT2 sulfotransferases. Both catfish SULT2 and human2B1b have Tyr residues that


86 correlate with the SULT2A1 Phe133 (Tyr159 fo r SULT2B1b, Tyr135 for catfish), but where SULT2A1 has Trp134, SULT2B1b has Ser160 and catfish Tyr136. Residues near the B and C rings of ster oid in the SULT2B1b complex are Tyr257, Leu260, and Thr106. In 2A1 the corresponding residues are Tyr231, Leu234, and Ser80, while in catfish they are Tyr233, Val236, and Glu81. Looking at the alignment of the three sequences, it was observed that not only is the first Tyr completely cons erved, but the three preceding residues, MSN, are also. The Leu260 position is c onserved as well, with a similar amino acid (valine) in the catfish sequence; only the third residue near the B and C rings in 2B1b, Thr106, is completely different between the three sequences Interestingly, a charged amino acid, glutamic acid (E81), occupies this posit ion in the catfish sequence. In SULT2B1b, residues lining th e opening of the binding poc ket and at the proteins surface are Trp98, Val108, Leu43, Leu264, and Ile20. This last residue, Ile20, is part of a special cholesterol-recognition region unique to humanSU LT2B1b and has no correlating amino acids in either of the other two sequences. Trp98 is conserved as Trp72 in SULT2A1 and Trp73 in catfish SULT2. The second residue, Val108, is c onserved in SULT2A1 as Ile82, but in catfish the position is held by a charged amino acid, Arg83. Leu43 correlates to Gly17 in 2A1 and Leu19 in catfish, while Leu264 correlates to Tyr2 38 in SULT2A1, which has been indicated in substrate inhibition (Lu, et al., 2008) by ADT and DHEA, an d to Phe240 in catfish. Substrate inhibition of humanSULT2A1 by ADT and DHEA was reported by Chang, et al. (2004) as a mechanism of regulating homeost asis and metabolism of the compounds, and to maintain steroid levels. Two amino acids, Tyr2 38 and Met137, were identified by Lu, et al. (2008) through site-directed mutage nesis studies as bei ng critical to this process. Met137 causes steric hindrance when ADT is bound, preventing all but one orientation. DHEA is able to adopt


87 either of two orientations, one of which is c onsidered a dead-end (noncatalytic) orientation. Release of substrate from the s ubstrate-binding cavity in a ternar y dead-end complex is regulated by Tyr238, which acts as a gate residue. The authors report that corresponding residues in other cytosolic sulfotransferases were found to exhi bit similar function in modulating substrate inhibition, including Phe247 in humanSULT1A1 (Gamage, et al ., 2003). Previously determined catfish SULT fragments Tong and Jam es (2000) purified a phenol sulfotransferase enzyme from channel catfish liver and intestine. Sequences of three fragments were determined, shown in Figure 2-13. The fragments, which align poorly with the catfish SULT2 isoform (alignment not shown), are nearly identical to the channel catfish SULT1 isoforms obta ined in this research (Figure 2-13). The two different amino acids for sequence #1 are similar in chemical properties; one of the fragment residues was unidentified. One residue for se quence #2 and three residues for sequence #3 are different and dissimilar from the catfish SULT1 is oforms; all other residues are identical. This high degree of similarity/high % identity putatively identifies the native sulfotransferase enzyme as a Family 1 isoform. Figure 2-13. Alignment of sequence fragments from a native sulfotransferase purified from channel catfish liver. Asterisks indicate id entical amino acid residues, dashes indicate similar residues. In addition, another 3 sequence obtained during PCR experiments described above appears to be of a different SULT1 (data not shown), suggesting a minimum of four Family 1 cytosolic sulfotransferases in the channel catfis h liver. This is reasonable and not unexpected, considering that, to date, eight zebrafish SULT1 isoforms have been discovered. Further 124 141175 186 272 287 S1v1 SFWEQNCKVVYVARNAKD S1v2 SFWEQNCKVVYVARNAKD Seqs SFFEQNXKIVYVARNAKD **-*** *-********* (#1) VSGPWYDHVTGY VFGPWYDHVTGY VFGPWYDHVTGY ********** (#2) NHFTVAQNERFDEHYW NHFTVAQNERFDEHYW NHFTVAQWEQFDEHYK ******* ***** (#3)


88 research is needed to determine/discover the exte nt of the cytosolic sulfotransferase family, not only in liver, but also in all major organs.


89 CHAPTER 3 RECOMBINANT PROTEIN EXPRESSION Materials and Methods Materials All m aterials used for PCR, agarose gels, a nd sequencing were the same as detailed in chapter 2. The pMAL-c2X Protein Fusion and Purification System wa s purchased from New England Biolabs (NEB), Ipswich, MA. The pC RII-TOPO vector and TOP10 competent cells were obtained from Invitrogen (Carlsban, CA). Modifying enzymes Klenow Fragment polymerase, Hind III restriction enzyme, and T4 DNA Ligase, with their corresponding buffers, were from New England Biolabs. XL1-Blue co mpetent cells were supplied by Stratagene (La Jolla, CA). The following products were purchased from Sigma-Aldrich Co. (St. Louis, MO): Bovine serum albumin, isopropyl -D-thiogalactopyranoside (IPTG), phenylmethylsulfonyl fluoride (PMSF), pepstatin A, 5Bromo-4-chloro-3-indolyl -D-galactopyranosid e (Xgal), sodium acetate, Bradford reagent, and D-(+)-glucose. Methanol, ethanol, acetic acid, sodium chloride, sodium hydroxide and glycerol were purchased from Fisher Biotech. Di ethylaminoethyl anion exchanger DE52 (DEAE) was obtained from Whatman Inc. (Florham Park, NJ). Proteins were separated and vi sualized on SDS-PAGE gels, either precast or made fresh. Precast NuPAGE Novex 4-12% Bis-Tris gels and 20X MES buffer were purchased from Invitrogen. For making fresh gels, 10X Tris/glycine/SDS buffer, 30% acrylamide/bis solution, and sodium dodecyl sulfate (SDS) were obtained from Bio-Rad (Hercules, CA). TEMED and ammonium persulfate were from Fisher Biotech (Pittsburgh, PA). Prestained protein standards BenchMark and MultiMark were obtained from Invitrogen; nonstained SDS-PAGE molecular weight standards, low-range, were purchased fr om Bio-Rad. Coomassie Brilliant Blue R250 was


90 from Bio-Rad. Gels were dried using the DryE ase Mini-Gel Drying System and Gel-Dry Drying Solution from Invitrogen. Spectra/Por molecular porous membrane dialysis tubing #4, molecular weight cutoff (MWCO) 12-14000, 2.0 m l/cm was obtained from Spectrum. Buffer components Trizma base, triethanolamine, triethanolamine hydrochloride, ethylenediaminetetraacetic acid (EDTA), and D-(+)-maltose monohydrate were purchased from Sigma. Tris and dithiothreit ol (DTT) were from Bio-Rad. Methods An overview of experim ental protocols is below, with details given if they are universal. Because much of the work done for protein expression and purification involved trial-and-error improvements of experimental conditions, most de tails will be given un der the Experiments and Results section. PCR full-length sequence Ligateinto pCRIIvector Culture (TOP10 cells) Mini-prep, confirm sequence PCR from plasmid, adding 3 Hind III Klenowfragment digest Hind III digest Cleanup Vector digested with Klenowfragment and Hind III restriction enzyme Cleanup GeneratingInsert: 2 StagespMAL Vector Preparation Ligate prepared insert and plasmid Figure 3-1. Flow chart of steps taken to prepare pMAL expressi on vector and insert sequence.


91 Generation of coding sequence insert Full-length sequences of the coding regions were generated via PCR and prepared for ligation into the pMAL Protein Fusion and Purification System for recombinant protein expression; a flow chart of step s is shown in Figure 3-1. Prim ers (Table 3-1) were designed so as to create a blunt-ended fragment to insert into the Xmn1 site of the pMAL vector in the same translational reading frame as the vectors malE gene. The 5 primer starts exactly at the sequence start codon and the 3 primer includes the stop codon The coding sequence was initially ligated into vector pCRII-TOPO in order to incorporate a 3 Hind III restriction enzyme site in the sequence downstream of the stop codon. All PCR, lig ation, and culture steps were carried out as described in Chapter 2. Cells were colony-screened for both the presence of insert, using sense and antisense sequence-specific prim ers, and for orientation, using a 5 sequencespecific primer and commercial 3 vector-speci fic primer M13 reverse. Colonies with correctly oriented inserts were cultured and mini-prepped as described previously, and the sequences confirmed. Entire mini-prep volumes were gel-purified and the concentration of plasmid estimated by comparison on an agaros e gel to a 100 bp ladder with varied weight designations. Preparing insert and vector for ligation A second PCR reaction was set up u sing the mini-prepped pCRII-TOPO plasmids as template and a 5 sequence-specific primer with the start codon sequence and 3 vector-specific primer M13 reverse. This served to generate a 5 blunt-end insert sequence beginning at the start codon and a 3 Hind III site in the noncoding region beyond the stop codon. Inserts were treated with Klenow fragment and Hind III digestions, the pMAL vector with Hind III and Xmn 1 restriction enzymes. Digest products were cleaned up with 3M sodium acetate precipitation followed by alc ohol rinse; different strategies were tried for least loss of


92 product and are detailed in Tables 3-2 and 3-3. Prepared insert and vector were ligated using T4 DNA ligase and buffer according to the manufacturer s guidelines and inc ubated overnight at 16 C. Between 3-5 L ligate was used to transfect 50 L XL1-Blue competent cells and plated on a single LB/carbenicillin agar plate grown overnight at 37 C. Select colonies were transferred to two new plates marked with numbered grids, one with IPTG/X-gal for blue-white screening, and one without, per pMAL instructions. Positiv e cultures were colony-screened for presence and orientation of and the positive screens sequenced to confirm that they were unchanged. Perfect (unchanged) sequences were used for expression experiments. Table 3-1. Primers used to prepare recombinan t channel catfish sulfotransferase enzymes for expression. Primer s/a Sequence 5-3 Notes S1v1start STEXP-73s S1v2start SULF-34 S2start ST2EXP-3s SULF-71 SULF-35 SULF-15 SULF-76 SULF-74 SULF-14 M13 Reverse M13 Forward *sense (s) or antisense (a)


93 Pilot experiments The optim al conditions for recombinant protei n expression (length of time, temperature), purification, and cl eavage from fusion binding protein were determined with a series of pilot experiments. Pilot experiments were carried out with the channel catfish SULT2 isoform. All pilot experiments were carried out according to the pMAL instructions. Brief descriptions of each are given below. Figure 3-2. Pilot expression experiment with S2 samples #12 and #14. The outer left lane is a size marker, sizes given in kDa; the firs t two lanes for each sample are negative controls, at times 0 and 18 hours; the next two bands for each are induced sample, at 5 and 18 hours. Cells were lysed and centrifuged; no further purification was done. The sulfotransferase band is detected here by its increased intensity for I5 and I18 lanes (below arrow).


94 Pilot expression. For each experiment, an 80 mL cultu re (LB/carbenicillin) was grown to OD600 = 0.5 to 0.8 at 37 C, preferably nearer 0.5. An aliquot (500 L) was removed as a zero time-point (t0) and the culture induced with IPTG at a final concentration of 0.3 mM; a parallel culture of 10 mL with no added IPTG was used fo r a negative control. Cultures were expressed at both room temperature (~23 C) and at 19 C for up to 24 hours. Two aliquots were removed (500 L each) hourly for the first five hours, then at time-point intervals thereafter, immediately centrifuged and the pellets stored at -20 C. At the end of the expression period the remaining induced culture was pelleted and stored at -20 C for use in further studies. Time-point pellets were lysed using 3 freeze/thaw cycles in ethanol/N2(l) and 42 C hot water bath; one sample for each time point was centrifuged for 2-3 minutes at 13,000 rpm into soluble/insoluble fractions, the other kept intact as a whole cell sample. All preparations (whole ce ll, soluble, insoluble) were run on an SDS-PAGE gel (F igure 3-2) to determine the best time and temperature for expression levels as well as to minimize formation of inclusion bodies (insoluble recombinant protein masses; image not shown). Pilot purification. The induced cell pellet was resusp ended in 10 mL Column Buffer (20mM Tris-HCl, 200 mM NaCl, 1mM EDTA) and lysed by sonication with 15-second bursts/rest cycles. Aliquots were tested at each rest to determine when lysis was complete. Lysis was determined by adding 10 L lysate to 1.5 mL Bradford reagent until the color change reached a plateau, ~2.5 minutes (sonication wa s carried out for 3.5 minutes total). Maltose-binding-protein (MBP) complex affi nity to resin and pilot purification experiments were carried out by a batch-wi se method according to the manufacturers instructions. Specifically, the sample was centrifuged at 9000 x g for 20 minutes, 4 C, and the


95 supernatant removed to a fresh tube. The pelle t was resuspended in 5 mL Column Buffer. Aliquots of 5 L of each fraction (supernatant and resuspended pellet) were removed and combined with 5 L 2X SDS loading dye buffer to run on an SDS-PAGE gel. A 200 L volume of amylose resin (from the pMAL kit) was wa shed twice with Column Buffer, centrifuged between washes, and resuspended in 200 L Column Buffer. Am ylose resin slurry, 50 L, was combined with 50 L of supernatant, incubated on i ce for 15 minutes, and centrifuged for 1.5 minutes at 13,000 rpm. The supernatant was rem oved to a separate tube (W1), the resin washed with 1 mL Column Buffer, centrifuge d again, and the wash supernatant removed (W2). The protein-bound resin was then resuspended in 50 L 1X SDS loading dye. Aliquots of the two wash samples (W1, W2) were prepar ed for SDS-PAGE gel by combining equivolume with 1X SDS loading dye (10 L each) and heating the samples, along with the supernatant, protein-bound resin, and pellet fractions to 95 C for five minutes and ce ntrifuging briefly. All prepared samples were loaded onto a precast gel (Figure 3-3). The majority of the expressed sulfotransfera se was primarily in the insoluble fraction (P); however, there was sufficient expressed pr otein in the soluble fr action to proceed. The initial wash fraction (W1) s howed no significant protein, indi cating good binding to the resin. The washed resin shows a clean band for th e fusion-bound protein, with no noticeable extra protein bands. Elution/purification experiments were al so carried out according to the pMAL manufacturers instru ctions for both batch and column methods (per pMAL instructions, details not given here); the best purif ication was obtained using column chromatography, and was used for all protein expression samples. Eluted pr otein was detected by measuring the absorbance at


96 280 nm with a 1 cm cuvette. The equation C (mg/mL) = A280 / pathlength (cm) gives a rough estimate protein concentration. Figure 3-3. SDS-PAGE gel of pilot purification of channel catfish SULT2 recombinant protein with the pMal vector expression system. The far left lane shows standard protein markers, sizes noted in kDa. R ecombinant sulfotransferase/MBP complex is indicated with an arrow. The columns are insoluble fraction (P), soluble fraction (S), first wash (W1), second wash (W2), and washed resin, with bound protein complex. Pilot cleavage. Protein-containing fract ions from the pilot elution/purification experiments were used for the pilot cleavage ex periments. The suggested ratio of cleavage enzyme Factor Xa (provided with pMAL kit) to fusion protein is 1% (e.g. 1 mg Factor Xa per 100 mg protein), with a usable range of 0.1 to 5.0%; recommended concentration of sample to be cleaved is 1 mg/mL protein. Three experiment s with enzyme:protein ratios of 10%, 3% and 0.8% were set up at room temperature for 24 hours.


97 Figure 3-4. SDS-PAGE gels of pilot cleavage using SULT2 recomb inant protein, expressed with the pMAL vector system. Gel on left is time points taken, in hours, while sample was incubated with protease Factor Xa; P = uncut sample, control (far right) had no protease added. Three experiments were done with different Factor Xa to protein ratios; two are shown here (lef t gel: 0.8%, right gel 3%). Gel on right shows effect of -mercaptoethanol treatment; the only differenc e observed is for Factor Xa, which is a dimer, and is split into its component pa rts. Xa = protease, MBP = maltose binding protein standard, Cut = treated SULT2 comp lex, Unc = untreated protein complex. Far left lanes of each gel are commercial markers, with sizes given in kDa. Aliquots of 5 L were taken at intervals and combined with 5 L 2X SDS-PAGE loading dye; at completion, aliquots were run on non-de naturing SDS-PAGE gels, along with uncut sample and standards of Factor Xa and maltose -binding protein. In addition samples were run (Figure 3-4) with 1.5 L -mercaptoethanol added to determ ine its effect (if any). The maximum amount of protein cleaved, even after 24 hours, was ~50%. Based on the pilot experiments, the experiment al conditions for all pr otein expression and purification experiments were as follows: induction with 0.3 mM (f inal concentration) IPTG and expression at 19 C overnight (~ 18 hours), lysis by sonication for 3 minutes in 15-second bursts, purification by affinity column, cleavage at room temperature overnight with Factor Xa, and separation by DEAE chromatography column.


98 Scaled-up expression, separation, and purification Culture and expressio n. Cell cultures for expression of recombinant protein were prepared in 1-liter volume s as described in Chapter 2, with the addition of 2 g glucose per liter of media. Once a successful pMAL expression vect or/insert had been constructed and competent cells transfected, a 15% (v/v) sterile glycerin stock was prepared and stored at C. For each expression experiment a 100 mL star ter stock of LB media fortifie d with either ampicillin or carbenicillin was prepared with no added glucose. The stocks were inoculated with a tiny scrape of the glycerin stock using either a sterile toothp ick or sterile pipette tip, which was dropped into the culture tube. This was grown at 37 C, with shaking, until the growth reached OD600 > 1. The entire volume was then added to a liter vo lume of sterile LB medi a with ampicillin or carbenicillin and gl ucose (unless the OD600 was high, >1.8, in which case a lesser volume was used respective to cell density). The inoculated media was grown at either 37 C or room temperature (~25 C) with shaking until OD600 was in the range of 0.5 to 0.8, preferably closer to 0.5. This was based on a growth experiment using different ages (number of days postinoculation) of starter culture grown to different densities (Fig ure 3-5). The results indicated that, while the density of the starter culture is not critical, a fresh starte r culture and cells grown to OD600 nearer 0.5 gave much better expr ession of recombinant protein. Protein expression was induced with IPTG, 0.3 mM final con centration, and incubated at room temperature with shaking for ~19 hours. Cells were harvested by centrifugation for 20 minutes, 4 C, at 5000 x g, and the supernatant discarde d. Pellets were resuspended in 40 mL Column Buffer + (containing 1mM dithio threitol (DTT), 1mM phenyl methylsulfonyl(PMSF), 1 M pepstatin and 10% v/v glycerol and frozen overnight at C.


99 Figure 3-5. Expression of r ecombinant SULT2 under various growth conditions, run on a denaturing SDS-PAGE gel. Each expressed sa mple is to the left of its negative (noninduced) control. The numbers indicate the final OD600 the cells reached before being induced with 0.3 M IPTG and incubating fo r 21 hours at room temperature. The starter cultures (stock) used for each 125 mL culture were of different densities and number of days post-inoculation. From le ft to right, the conditions were (1) 13 mL fresh stock at OD600 = 0.719; (2) 7 mL stock at OD600 = 0.713, 47 days postinoculation; (3) 10 mL stock at OD600 = 0.691, 18 days post-inoc ulation; (4) 5 mL fresh stock at OD600 = 0.521. All starter cultures (stock) were stored at 4 C until used. The far right lane is a commercial marker, with sizes shown in kDa; an arrow indicates the expressed protein. For lysis, the suspended cells were thaw ed on ice water until a thin slush, and then sonicated with a probe sonicator for 3 minutes in 15-second bursts/rests. Cytosol was isolated by high-speed centrifugation for 1hour, 4 C, at 150,200 x g, and loaded onto an affinity column for the first purification step. If the cytosol was not purified i mmediately it was stored at C. Affinity chromatography. Amylose resin (provided in the pMAL vector kit) was packed into a glass 15-mL column and washed with 8 column volumes of Column Buffer. Cytosol was loaded onto the column at a flow rate of 1 mL /minute; three one-mL a liquots of Column Buffer containing 1mM DTT, 1mM PMSF, 1 M pepstatin and 10% glycerol were added to the sample container toward the end of the load to assure that all sample loaded onto the column. This was followed with a 180 mL column wash with Colu mn Buffer and subsequent elution of fusionprotein with 30 mL Elution Buff er (Column Buffer with 10 mM maltose) collected in ten 3 mL

PAGE 100

100 fractions. The load and wash fractions were collected in separate sing le flasks. Fractions containing protein were dete rmined by measuring absorbance at 280 nm and pooled. Factor Xa cleavage and dialysis Fusion-bound protein (the concentration estimated by A280) was combined with 50 g Factor Xa and incubated overn ight at room temperature. Incubation was initiall y carried out at 4 C, but very little cleavage was seen after 46 hours (details not shown). An SDS-PA GE gel was run to check the exte nt of cleavage of recombinant protein. A portion of the cut/unc ut fraction was reserved for later use in activity assays; the remainder was dialyzed against 250 mL 10 mM TEA containing 10% glycerol, 5 mM mercaptoethanol and 25 mM NaCl The 16 mm dialysis tubing was prepared by soaking in warm deionized water with occasional gentle st irring and two water changes over 2-3 hours; the last water bath contained 1 mM EDTA. The protein sample was added to the tubing and dialyzed at 4 C overnight. The dialyzed sample was pipetted into a 5 mL glass tube. DEAE chromatography. Ion-exchange resin DE-52 was prepared by washing it with 100 mM TEA buffer, pH 7.4, until the resin wash buffe r reached a matching pH. The resin was then washed with 10 mM TEA buffer (pH 7.4) until th e ionic strengths of resin buffer and 10 mM TEA buffer matched. Washes were done in a one-lit er beaker and the resin stirred gently with a glass rod to avoid breakage; each pH reading was taken after th e resin had settled. Prepared resin was packed into a glass 15 mL column in the cold room (4 C) and allowed to settle overnight. Just before loading the sample the column was washed with 120 mL 10 mM TEA buffer, pH 7.4, 25 mM NaCl. The sample was loaded ont o the column at a rate of approximately 0.5 mL/minute. The column was then washed with 40 mL of TEA buffer containing 25 mM NaCl at 1 mL/minute; 2.5 mL fractions were collected with an automatic fraction collector, increasing to

PAGE 101

101 3.5 mL after ~10 fractions. This was followe d by a 30 mL wash with 10 mM TEA buffer, 50 mM NaCl; 3.5 mL fractions were collected thro ughout. The free sulfotra nsferase protein was eluted with a gradient of 50-275 mM NaCl in 10 mM TEA buffer, with 2.5 mL fractions collected until ~15 fractions, then increased to 3.5 mL. The gradient was set up with a dual chamber stirrer, 50 mL each of TEA buffer with 500 mM and 50 mM NaCl. Protein-containing fractions were determ ined by measuring every other fraction by spectrophotometer at 280 nm. Because three prot eins are potentially present (Factor Xa, the cleaved maltose-binding protein, and the expressed sulfotransferase ), activity assays were also done to confirm the recombinant protein fractions (detailed in Chapter 4). In addition, 20 L of each fraction was taken and those with the highest A280 readings, along with the neighboring fractions, were analyzed by SDS-PAGE. Experiments and Results Generating Full-length Insert Sequence Full-length sequences were obtained with P CR, using both libraries (CAPT-30 and TRsa) as tem plate, with the primers listed in Tabl e 3-1. The initial cond itions were annealing temperature 55 C, hot start, 1-minute extension, 22 cy cles, with a final 10-minute final extension. Additional cycles were added without a hot start. PCR experiments were set up for all three genes (S1v1, S1v2, S2). Good bands were seen on a 1% agaros e gel for S1v1 after 22 cycles for the CAPT30 library and 30 cycles (tot al) for the TRsa library. The S2 sequence was pulled up after 30 cycles for both libraries; no band was seen for S1v2 after a total of 40 cycles. PCR products selected to continue with were S1v1, TRsa library, and S2, CAPT30 library; S1v2 was not pursued further.

PAGE 102

102 The PCR products, 16 L each, were cut from an agarose gel and extracted with Qiagen gel extraction kit. El ution volumes were 30 L for S2 and 10 L for S1v1 (a decision based on band intensities). Extracted cDNA was lig ated into PCRII-TOPO sequencing vector, 2.5 L of which was transfected into 250 L TOP10 competent cells for each gene. The entire cell volume for each gene was plated onto pre-warmed LB/Kanamycin (Kan) plates: 75 L on one and 175 L on another that had been previously coated with 40 L of 35 mg/mL Xgal. The plates were incubated overnight at 37 C. Blue-white screen results for S1v1 were pre dominately negative (blue) for insert, while the majority of the S2 colonies were positive (white) for insert; all were small, slow-growing colonies. Sufficient S1v1 positive colonies were pr oduced to enable transfer of 16 colonies of each gene to a master plate (LB/Kan only; no added Xgal) for subsequent screening. The cells were screened by PCR with a gene-specific 5 prim er and a vector-specific 3 primer, as listed in Table 3-1. The colony screen gel showed a few positive lanes for S1v1 and all for S2. Seven S1v1 colonies and six S2 colonies were selected from the master plat e and cultured in 5 mL LB/amp overnight with shaking. Samples were identified by gene and master plate colony number: S1v1 #1,5,8,13,14,15,16 and S2 #5,6,8,9,12,13. All cultures grew successfully; the plasmids were extracted by mini-prep, a nd sequence reactions prepared for each. Sequence reaction primers were selected such that each region of the sequence would have two replicates, and are listed in Table 31. Two primers were vect or-specific (sense and antisense) and two were gene-speci fic (sense and antisense). Of all the sequences generated, one was perfect (unchanged) for each gene: S1v1 #16 and S2 #5. The corresponding cells were cultured in 10 mL per usual protocol a nd the plasmids harvested by mini-prep.

PAGE 103

103 Insert and Vector Preparation, Ligation The PCRII-TOPO vectors with co rrect sequence inserts (S1v1 #16 and S2 #5) were subjected to PCR to generate th e coding insert with a 3 Hind III sequence. The primers (Table 3-1) were sequence-specific for th e 5 end (S1v1start and S2start) and vector specific for the 3 end (M13 Reverse). PCR conditions were as descri bed for the initial PCR to generate the insert, but for fewer cycles. S1v1 was generated with 12 cycles, and S2 with 20 cycles. The entire PCR product volumes were run on a deep -well agarose gel, and the bands cut out and extracted with Qiagen gel extraction kit, 30 L elution volume. The cDNA concentrations were estimated by comparison to a Promega 100 bp ladder corun on a 1% agarose gel; three different volumes (1, 2, and 3 L) of ladder were run against 3 L sample and the concentrations calculated. The S1v1 product was estimated to be 180 ng/ L and the S2 to be 3.3 ng/ L (Figure 3-6). Figure 3-6. Concentration estimate of S1v1 (left, 3 L) and S2 (right, 3 L) PCR products run against 1, 2, and 3 L of Promega 10 0 bp standard. Also shown is a 1.5 L sample of S1v1 insert prepared for ligation into the pMal expression vector. Inserts next were treated with Klenow fr agment and Hind III digestions, the pMAL vector with Hind III and Xmn 1 enzymes. At this point much sample was lost during the cleanup

PAGE 104

104 steps; experiments were repeated with adjustments in methods to improve final product amount. Specific details are given in Ta bles 3-2 and 3-3. Oligomer c oncentrations were estimated by comparison against three different volumes of Promega DNA ladder on an agarose gel. To obtain more pMAL vector, TOP10 competent ce lls were transfected with the commercial plasmid, cultured in one-liter volum es and harvested by mini-prep. S1v1 ligation Entire volum es of prepared S 1v1 sample (Table 3-2) and prep ared pMal vector #4 (Table 33) were combined with 5.6 L of 10X T4 ligas e buffer and 2 L T4 ligase enzyme. This was incubated overnight at 16 C, and then checked on an agarose gel. Size comparison with a previous gel of prepared, unligated pMal ve ctor (Figure 3-7) rev ealed a difference of approximately 1 kb, the correct size increase for a successful ligation. Figure 3-7. Example of successful ligation for S1v1 and pMAL vector. Ligated samples (2 L and 1 L) are on the left gel, and vector only (1 L) on the right gel; the same marker is used for both. Lines drawn show that th e ladders are aligned, a nd that the vector is significantly smaller than the ligated samples. The size difference matches that of the insert sequence.

PAGE 105

105 Stratagene XL1-Blue competent cells, 50 L, were transfected with 3 L of S1v1/pMal ligate as described and plated onto two separa te LB/carbenicillin pl ates: 100 L on one and 50 L on the other. Of the coun tless colonies (too many on the 100 L plate to sel ect individuals, only the 50 L plate was used) grown, 36 were sele cted for blue-white screening. These were transferred by sterile toothpick to two new plates marked with nu mbered grids, one plate treated with Xgal and IPTG. The toothpicks were th en each dropped in individual 5 L LB/ampicillin cultures and all incubated at 37 C overnight. Six of the colonies screened distinctly white, indicative of presence of insert (positive screen) and were selected, from the untreated pl ate, to be colony screened. Primer pairs S1v1start/ SULF-73s were used to screen for pres ence of insert and S1V1-start/M13forward for orientation of insert; conditions were as desc ribed previously. Four plasmids, #1, 30, 31, 32, screened successfully; the corresponding cultures were mini-prepped and the plasmids sequenced with primers S1v1-start, SULF-71, SULF-8, SULF-72, and M13forward. Plasmid #1 gene sequence was unchanged and selected for expression. Multiple expression experiments were done for the S1v1 isoform; in each, no band was seen at the expected size on SDS-PAGE gels. A dark band was seen, but at the wrong size (too low) and may be a different protein altogether. A review of the lab noteb ook revealed that when the colonies were blue-white screened the same toothpick was used to culture the cells; this means they were contaminated with Xgal from the screening plate. According to the pMal manual, this can cause a scrambling of se quence upstream from the start codon, involved in generating the recombinant expression. The plasmi d/insert of Sample #1 was re-sequenced, to include the upstream region, and the sequence compared with th e manufacturers information; the upstream sequence was, indeed, changed. This likely resulted in expression failure, and can

PAGE 106

106 be corrected by transfecting fresh cells with liga te product and proceeding forward from there. Due to time constraints this project was set aside and focus put on the SULT2 gene. SULT2 ligation Entire vo lumes of prepared SULT2 (Tab le 3-2) and pMal vector 3b (Table 3-3) were combined, dried under vacuum, and resuspended in 19 L of 1X T4 ligase buffer. To this was added 1 L of T4 DNA ligase and the samp le incubated overnight at 16 C. Stratagene XL1Blue competent cells (50 L) were transfected with 5 L of ligated S2/pMal vector and plated as described. Of the 20 colonies for blue-white screening, only three screened clearly negative for insert. The positive colonies were cultured a nd screened by PCR for presence and direction (orientation) of insert. The prim er pairs used to screen for pres ence of insert were S2-start/ST2EXPD3s, and for orientation were S2-start/M13f orward (Table 3-1). Polymerase reaction conditions were 55 C annealing temperature for 30 cycles. Good bands were seen for samples #9, 12, 14; the corresponding cultures were mini -prepped and the plasmids set up in sequence PCR reactions with sense primers SULF-35 and S2-start and antisense primers SULF-14 and M13forward (Table 3-1). Of the three, sa mples #12 and #14 sequences were unchanged but #9 was missing the first nucleotide of the start codon and was unusable. Pilot expression experiments of both #12 and #14 showed good expr ession, and #12 was selected for scaled-up recombinant protein expression. A one-liter culture was grown, induced, and expression carried out at room temperature for 21 hours. The harvested cells were combined with cells from pilot expression studies, lysed, and the cytosol separated by ultracentrifugation at 100,000 x g. Total cytosol protein was determined by Lowry protein assay to be 1.85 mg/mL. MBP-SULF2 complex was purified by

PAGE 107

107 affinity column, the fractions (2.5 mL each) containing eluted protein determined by A280 to be #6-10 (Figure 3-8).Fractions #7-10 were pooled and the protein concentration approximated at 7 mg (using the absorbances), and the sample treated with 70 g Factor Xa. After 46 hours incubation at 4 C a gel was run to check the progress, which was poor; incubation was continued for another 14 hours at room temperature, and progress check again on an SDS-PAGE gel and found to be improved. Glycerol was added to the sample to 5% v/v and stored at -20 C until dialysis. The cut and uncut protein mixture with Fact or Xa was dialyzed against 10 mM TEA with 10% glycerol, 5 mM -mercaptoethanol, and 25 mM NaCl ove rnight at 4 C and separated by DEAE chromatography. All fractions were measured for absorbance at 280 nm (Figure 3-8) to determine those containing protein (purified sulfotransferase, Factor Xa, maltose-binding protein, or uncut enzyme). Fractions #48-57 were considered to potentially contain purified sulfotransferase and #72-80 to contain uncut protein. SDS-PAGE gels revealed bands fo r uncut enzyme and either (or both) Factor Xa and MBP in the fractions #72-80, but no bands were seen for th e fractions #48-57, probabl y because the protein concentration was too low. An activity assay (described in chapter 4) with DHEA showed good activity in fractions #48-51 and re duced activity in a fraction ( #79) containing uncut enzyme. The active fractions, #48-51, were combined and concentrated with Amicon 4 mL concentrator tubes, MWCO 10K, from 6.5 mL to 3mL and stored in 1 mL cryogenic tubes (500 L in each) at -80 C. Protein concentration was determ ined to be 0.4 mg/mL by BioRad assay.

PAGE 108

108 0.0 0.2 0.4 0.6 0.8 12345678910111213A280Fraction #Affinity Column 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 35404550556065707580A280Fraction #DEAE Column Figure 3-8. A280 measurements of fractions collected during affi nity column purification (top) and DEAE column purification of recombinant catfish SULT2 enzyme. Fractions #710 from the affinity column were pooled a nd treated with Factor Xa to cleave the fusion protein. Proteins were then separated by DEAE column (bottom). Fractions #48-57 and #72-80 were selected for DHEA ac tivity screening; fraction #49 showed highest activity.

PAGE 109

109 Table 3-2. Conditions of digest s and cleanup for S1v1and SULT2 inserts, in prepar ation for ligation into pMAL expression vecto r Sample Klenow Cleanup Hind III Cleanup Concentrations S1v1 #1 10 L sample (~2 g DNA), dried 10u enzyme 18 L 1x NEB2 buffer with 33 M dNTPs RT 20 mins inactivated 75 C, 20 mins PCR mini-elute kit, 10 L elution 10 L DNA 5 L 10x NEB2 buffer 33 L water 2 L enzyme (40u) 37 C 1 hour PCR mini-elute kit, 10 L elution; then: 1.2 L 3M NaAc 1.2 L isopropanol -80 C 10 mins cent 15 mins, max, 4 C remove supernatant wash pellet 500 L 100% EtOH, then 250 uL 70% EtOH, dried; resuspended in 25 L TrisEDTA buffer Starting: 180 ng/ L (1.8 g in 10 L volume) Post mini-elute: ~6.7ng/ L (67 ng in 10 L volume) Final: ~0.5 ng/ L (6 ng total in 12 L volume) S1v1 #2 10 L sample (~2 g DNA), dried 2u enzyme 18 L 1x NEB2 buffer with 33 M dNTPs RT 20 mins inactivated 75 C, 20 mins None; dried sample under vacuum Dried DNA sample 40 L 1x NEB2 buffer 2 L enzyme (40u) 37 C 1 hour inactivated 65 C, 10 mins PCR mini-elute kit, 10 L elution Starting: 180 ng/ L (1.8 g in 10 L volume) Final: 45 ng/ L (382 ng total in 8.5 L volume) SULT2 90 ng DNA, dried 1u enzyme 19 L 1x NEB2 buffer with 33 M dNTPs RT 20 mins inactivated 75 C, 20 mins None Added to Klenow treatment 1 L enzyme (20u) 37 C, 1 hour PCR mini-elute kit, 11 L elution (for 10 L net) Starting: 3.3 ng/ L (90 ng DNA) Final: hard to estimate, similar to starting conc.

PAGE 110

110 Table 3-3. Conditions of dige st and cleanup for pMAL vector in preparation for ligation with S1v1 or S2 insert Sample Treatment Cleanup Concentrations pMAL #1 2.5 L vector (0.5 g DNA) 16.5 L 1x NEB2 buffer with 100 g/mL BSA 10u Xmn 1 enzyme 10u Hind III enzyme 37 C 1 hour inactivated 65 C 10 mins PCR mini-elute kit, 10 L elution followed by: 1.2 L NaAc and 1.2 L isopropanol -80 C 10 mins cent 15 mins, 4 C; removed supernatant washed with EtOH: 500 L 100%, then 250 L 70%; dried, resuspended pellet 25 L Tris-EDTA buffer Starting: 25 ng/ L (0.5 g in 20 L volume) Post mini-elute: ~20 ng/ L (200 ng in 10 L volume) Final: ~ 1 ng/ L (25 ng in 25 L volume) pMAL #2 2.5 L vector (0.5 g DNA) 16.5 L 1x NEB2 buffer with 100 g/mL BSA 10u Xmn 1 enzyme 10u Hind III enzyme 37 C 1 hour, then: added 10u each enzyme 37 C overnight 2 L 3M NaAc and 2 L isopropanol RT 10 mins cent 15 mins, RT; removed supernatant washed 70% EtOH, filled tube cent 15 mins; poured out EtOH, dried resuspended pellet 25 L Tris-EDTA buffer Starting: 25 ng/ L (0.5 g in 20 L volume) Final: too little to estimate (very faint band for 10 L on gel) pMAL #3a 2.5 L vector (0.5 g DNA) 16.5 L 1x NEB2 buffer with 100 g/mL BSA 10u Xmn 1 enzyme 10u Hind III enzyme 37 C 1 hour, then: added 10u each enzyme 37 C overnight Entire volume run on agarose gel, cut out cleaned up with gel extraction kit, 10 L elution (note: in error heated at 100 C, not 55 C) Starting: 25 ng/ L (0.5 g in 20 uL volume) Final: (hard to estimate) ~2 ng/ L (16 ng in 8 L volume) pMAL #3b 2 L 3M NaAc and 2 L isopropanol, 1 L seeDNA RT 10 mins cent 15 mins, RT; removed supernatant washed 70% EtOH, filled tube cent 15 mins; poured out EtOH, dried resuspended pellet 15 L Tris-EDTA buffer Starting: 25 ng/ L (0.5 g in 20 L volume) Final (hard to estimate) ~3 ng/ L (~39 ng in 13 L volume) pMAL #4 (cultured) 50 L vector (~750 ng DNA) 5.6 L 10x NEB2 bfr; 100 g/mL BSA 15u Xmn 1 enzyme 15u Hind III enzyme 37 C 1 hour, then: added 15u each enzyme 37 C overnight 60 L 3M NaAc and 60 L 100% EtOH RT 15 mins, then 4 C 40 mins cent 15 mins, 10 C; poured off supernatant, dried resuspended pellet in 15 L Tris-EDTA bfr Starting: ~15 ng/ L (~750 ng DNA in 50 L volume) Final: ~ 15 ng/ L (~210 ng in 14 L)

PAGE 111

111 CHAPTER 4 ENZYME ASSAYS The two m ain family groups of cytosolic sulfot ransferases exhibit substrate specificities: members of Family 1 preferentially act on substr ates with phenolic moieties, and Family 2 on those with alcohol moieties. In order to determine that th e catfish putative SULT2 enzyme follows this pattern, the recombinant enzyme wa s tested with various substrates, phenolic and alcoholic, endogenous and xenobiotic. Enzyme assay experiments were performed following established protocols (Tong and James, 2000; Wang and James, 2007). Substrates were selected according to previous use in sulfotransferase assays, particularly those used in fish SULT experiments and those diagnostic for human SULTs (such as p-nitrophenol for SULT1A1, dopamine for SULT1A3, 17estradiol for SULT1E1, and DHE A for SULT2A1). Xenobiotic substrates were additionally selected based on their structure (presence of alcohol ligand, likelihood of being a sulfotransferase 2 family subs trate) and their environm ental relevance. For example, mammalian SULT2A1 enzyme can sulfon ate polychlorinated biphenyls and benzylic alcohols (Liu, et al. ., 2006). Therefore, several hydroxy-polychl orinated biphenyls (OH-PCBs) were studied as well as hydroxy-methyl pol ycyclic aromatic hydrocarbons (OH-MPACs) Materials and Methods Materials Radiolabeled 3-phosphoadenosine-5-phosphosulfate (35S-PAPS), 1.846 Ci/nmol, was purchased from PerkinElmer Life Science (Boston, MA), as were radiolabeled 3H-estradiol, 43.8 mCi/mmole, and 14C-testosterone, 56.9 mCi/mmole. Substrates 6-hydroxymethylbenzo[a]pyrene, 7-hydroxymethyl-12-methylbe nzanthracene, and 3-hydroxy-benzo[a]pyrene (3OH-BaP) were obtained from the NCI Chemi cal Carcinogen Reference Standard Repository (Midwest Research Institute, Kansas City, MO). The 4-OH-PCBs were a gift from Dr. Larry

PAGE 112

112 Robertson, University of Iowa or were purchased from AccuSt andard (New Haven, Connecticut). Endogenous substrate dehydroepi androsterone (DHEA), bovine serum albumin (BSA), Tris-HCl, and Tris-base were from Sigma-Aldrich (St. Louis, MO). EcoLume Scintillation Cocktail, 7-mL disposable glass sc intillation vials, and glass disposable culture tubes (13 x 100 and 16 x 100) were purchased from Fisher Scientific (Atlanta, GA), as was magnesium chloride. PICA Low UV Solution was from Waters Corporati on (Milford, MA). Methods Phenolic substrate 3OH-BaP Activity with 3OH-BaP was m eas ured with fluorescent detec tion, as the sulfonated product fluoresces at 415nm. Due to light -sensitivity, the experiment was carried out with the lights off; when the sample tubes were tr ansported between rooms (and potentially subjected to hallway light) they were kept darkened with aluminum foil cover. The assay, with a final volume of 500 L, was carried out in 16 x 100 mm disposable glass tubes. The substrate, at final concentrations of either 0.1 M or 1.0 M, was pipetted carefully into the bottom of the tube and the solvent removed under nitrogen. BSA was added to a final concentration of 0.4% and the samples vortexed. Next were added Tris-Cl, pH 7.4, 0.05 M; PAPS 20 M and water to a final volume of 500 L (minus the enzyme volume to be added). The samples were vortex-mixed. Because some assays were carried out with cytosol fr actions, which contained some levels of PAPS naturally, the PAPS was added last in 15-second in tervals to start the r eaction. The reaction was allowed to proceed for 10 minutes in a 35 C water bath with shaking then stopped with the addition of 2 mL of ice-cold MeOH. For time 0 controls, methanol was added just prior to addition of enzyme. The resulting precipitati on was cleared by centri fugation with a floor centrifuge for 10 minutes (at its maximum speed, ~2500 rpm) at room temperature, and 2 mL

PAGE 113

113 was pipetted to a new test t ube. Just before fluorescence wa s measured (excitation/emission 294/415 nm), 500 L 1N NaOH was added to the sample and vortexed. The background was measured with a blank of 2 mL methanol and 500 L NaOH. OH 6-hydroxymethyl-benzo[a]pyrene 6OHM-BaPH3C OH 7-hydroxymethyl,12-methyl-benzo[a]anthracene 7OH-12MBA A B Figure 4-1. Structures of xenobi otics tested for enzymatic activ ity with recombinant channel catfish SULT2 sulfotransferase. Hydroxypolychlorinated biphenyls are at top (A), and hydroxy-methyl polycyclic aromatic hydrocarbons below (B).

PAGE 114

114 Steroid substrates estrogen and testosterone Assays for activity with endogenous substrates estradiol and testoster one were carried out as described in W ang and James, 2007. Both substrates were radiolabeled (3H-estradiol, 14Ctestosterone) and product measured by scintillation counting (rather than HPLC). Final substrate concentrations were 0.8 M for estradiol a nd 22 M for testosterone; SULT2 sample was affinity purified and Factor Xa treated (a mixture of cleaved and uncleaved enzyme). Alcoholic and OH-PCB substrates Seven substrates with alcoholic m oieties we re tested for activity with the recombinant SULT2 catfish enzyme, one endogenous and six xenobiotics. Radiolabeled cofactor 35S-PAPS was used, and sulfonated product detected by se paration of the ion pair with tetrabutyl ammonium into an organic phase which was count ed by scintillation counte r. The xenobiotic chemicals (Figure 4-1), 6-hydroxymethyl-benzo [a]pyrene (6OHM-BaP) and 7-hydroxymethyl, 12-methyl-benzo[a]anthracene (7OH-12MBA), 4-hydroxy-3-chlorobiphenyl (4OH-CB2), 4hydroxy-3,5-dichlorobiphenyl (4OH-CB14), 4hydroxy-3,4-dichlorobiphenyl (4OH-CB39), and 4-hydroxy-2,3,4,5-tetrachlor obiphenyl (4OH-CB68) were test ed at two to three substrate concentrations, while kinetics we re determined for the endogenous substrate DHEA. The same protocol was followed for all molecules tested. Three concentrations of xenobiotic substr ates 6OHM-BaP and 7OH-12MBA (2, 20, and 200 M) were examined for activity, while the range of endogenous substrate DHEA examined for kinetics determination was 1.0 to 200 M. For each determination, a no-substrate blank was included. Two concentrations of the four OH-PCB substrates, 100 M and 500 M, were examined for activity. Stock solutions of all compounds were such that aliquot volumes for each

PAGE 115

115 experimental concentrat ion were at least 10 L but no greater than 100 L. Substrates were pipetted into disposable glass 13 x 100 size tubes and the solven t completely evaporated under a low pressure N2 stream. A master mix composed of Tris-HCl, pH 7.0, MgCl2, enzyme, water and BSA fraction V was prep ared, vortexed and 90 L pipetted into each substrate tube. The final concentrations of each component were 50 M Tris-HCl, 5 mM MgCl2, 0.4% BSA and 20 L of either recombinant SULT2 enzyme or (hum an)SULT2A1 cytosol (for positive control). The tubes were then vortexed and centrifuge d at full speed in a floor model for two minutes to assure that all substrate was dissolv ed in the buffer solution, and not adhering to the tube wall as a result of the nitr ogen stream. Samples were placed in a 35 C hot water bath with shaking (80 rpm) and the reaction initiated with the addition of 10 L 35S-PAPS in 30-second intervals (5 M final concentration). After 10 minut es the reaction was stopped with the addition of 0.1 mL of a 1:1 mixt ure of 2.5% acetic acid and PIC A, and 0.3 mL water, and vortexed to mix. Immediately 2.0 mL of water-saturated ethyl ac etate was added and the sample vortexed again. Tubes were centrifuged at full speed (in the floor model centrifuge) for 10 minutes and most of the top solv ent layer carefully transferred to a 7 mL scintillation vial. One mL of ethyl acetate was added to each tube for a second extrac tion, the tubes were vortexed for 15 seconds and centrifuged for 10 minutes as before The organic layer was added to the first extracted volume with great ca re not to transfer any of the aqueous layer. Solvent was evaporated under a low air stre am to dryness or near-dryness (maximum volume 0.5 mL), and the vial filled with 4 mL Ec oLume scintillation fluid. Each concentration was done in duplicate, and a time0 nega tive control with the highest concentration of substrate was in cluded. For the time0 samples, stop-reaction chemicals acetic

PAGE 116

116 acid/PIC A and water were adde d immediately before 35S-PAPS, followed by the ethyl acetate. In addition, a blank with no substrate wa s incubated with all other components. Results 3OH-BaP Four assays were done with the xenobiotic substrate 3OH-BaP three with cell cytosol (from induced cells) containing the fusion protei n of SULT2 with maltose-binding protein and one with partially purified a nd cleaved enzyme (conditions su mmarized in Table 4-1). No activity was seen in any of the assays except for the positive control of channel catfish liver cytosol. In all conditions, good activity was seen for the positive control, but fluorescence for incubations with the sulfotransfera se samples was equivalent to the time0 samples. Therefore, it was concluded that 3OH-BaP was not a substrate for this enzyme. Table 4-1. Summary of varying experimental conditions for activity assays with recombinant catfish SULT2 and 3OH-BaP. SULT2 type (g) (+) Control (g) (-) Control (g) 3OH-BaP Induced cell cytosol (6-300 g) none No cytosol added 0.1 M Induced cell cytosol (9.25 g) Catfish liver cytosol (74 g) SULT2 cytosol of cells induced, no expression (60 g) 0.1 M Induced cell cytosol (20 g) Catfish liver cytosol (20 g) SULT2 cytosol of cells induced, no expression (20 g) 1.0 M Isolated, Xa-treated enzyme Catfish liver cytosol (20 g) SULT2 cytosol of cells induced, no expression (20 g) 1.0 M Endogenous Substrates DHEA activity assays were carried out with a 400 M final concentration of substrate and with recombinant (human)SULT2A1 as a positive control. The first assay was performed with recombinant catfish SULT2 isolated by affinity chromatography and cleaved with Factor Xa; because the sample had not been purified by DE AE column it contained both cut and uncut (maltose-bound) enzyme. Therefore, the first as say was a screen to determine if there was any

PAGE 117

117 activity with DHEA as substrate. Good activity was seen for a ll samples, the positive control and the catfish SULT2. The cleaved sample wa s dialyzed against 10 mM TEA buffer fortified with 25 mM NaCl, 5 mM -mercaptoethanol, 10% glycerol and the SULT2 enzyme purified with a DEAE column. Fractions showing signi ficant absorbance at 280 nm were screened for DHEA activity, using 20 L sample volume. Four fract ions, #48-51, were combined and concentrated with Amicon 4 mL concentrator tubes (MWCO 10,000) to a concentration of 0.4 mg/mL; this purified enzyme was used to de termine kinetics. The assay, using eight DHEA concentrations between 5 and 200 M, gave results that fit the Michaeli s-Menten equation with excellent duplication of replicates. The kinetic values for catfish SULT2 with DHEA were determined to be KM = 43.7 M and Vmax = 516 pmol/min/mg (Figure 4-2). Specific activity (with 50 M DHEA) was 298 pmol/min/mg. For the estradiol and testosterone activity as says, no significant act ivity was observed for either hormone, suggesting that they are not subs trates for this channel catfish SULT2 isoform. Xenobiotic Substrates Com pared to the (human)SULT2A1 positive control, very little activity was seen for catfish SULT2 with either 6OHM-BaP or 7OH-12MBA, or with any of the four OH-PCBs (Table 4-2). All of the measured dpms for the 4-hydroxy-pol ychlorinated biphenyl compounds were below the limit of detection (approximately 3xs the blank). The lowest activity for positive control (human)SULT2A1 with 6OHM-BaP or 7OH-12MBA was 0.7 pmol/min/mg, which was the highest activity for SULT2 with 200 M 7OH-12MBA. For recombinant catfish SULT2 with 6OHM-BaP, activity decreased as substrate conc entration increased, from 0.256 pmol/min/mg at

PAGE 118

118 [DHEA] vs. Velocity 0 25 50 75 100 125 150 175 200 0 100 200 300 400 500 VMAX KM 515.8 43.65[DHEA] ( M)V ( mol/min/mg protein) Figure 4-2. Michaelis-Menten plot of catfish SULT2 activ ity with substrate DHEA. Table 4-2. Activity of xenobiotic substrates wi th purified recombinant channel catfish SULT2 and positive control (human)SULT2A1. ND = below limit of detection, ~ 3xs blank. Substrate Concentration (M) SULT2 Activity (pmol/min/mg) hSULT2A1 Activity (pmol/min/mg) 4-OH-CB39 100 ND 8.04 500 ND 181 4-OH-CB68 100 ND 7.51 500 ND 217 4-OH-CB2 100 ND 33.0 500 ND 55.6 4-OH-CB14 100 ND 157 500 ND 132 6OHM-BaP 2 0.256 1.25 20 ND 0.720 200 ND 0.780 7OH-12-MBA 2 ND 0.710 20 ND 6.86 200 0.738 2.67

PAGE 119

119 2 M substrate to 0.100 pmol/min/mg enzyme at 200 M substrate. Activity was nearly unchanged at the lower substrate concentra tions for 7OH-12MBA and SULT2, 0.144 and 0.156 pmol/min/mg, increasing to 0.738 pmol/min/mg enzyme at the highest substrate concentration of 200 M. Very little activity was seen for the recomb inant SULT2 with any of the 4-OH-PCBs, at either the 100 or 500 M concentrations; the highest activity was 0.8 pmol/min/mg from 4-OHCB68. In contrast, the positive control, hSULT2A1, exhibited activity as high as 217 pmol/min/mg for the same substrate at 500 M. Da ta are given in Table 4-2; the consistently low (to non-significant) activity leve ls for the OH-PCBs indicate that this chemical type is not a substrate for the recombinant channel catfish SULT2 Discussion Of the three endogenous and six xenobiotic com pounds tested, catfish SULT2 showed significant activity with DHEA only. A ll of the measured dpms for the 4-hydroxypolychlorinated biphenyl compounds were below the limit of de tection (approximately 3xs the blank) and are therefore consider ed to be non-substrates for this enzyme. Of the two other xenobiotic compounds assayed, 6-hydroxymethyl -benzo[a]pyrene and 7-hydroxymethyl, 12methyl-benzo[a]anthracene, only the highe st concentration for 7OH-12MBA (200 M) and the lowest concentration for 6OHM-BaP (2 M) measured activity greater than the limit of detection. Interestingly, activity for 6OHM-BaP decreased as the substrate concentration increased, suggesting possible substr ate inhibition. Further investig ation is needed to determine whether this is occurring or whether, lik e 7OH-12MBA, the compound is simply not a significant substrate for catfish SULT2.

PAGE 120

120 The 4-OH-PCB compounds shown to have act ivity with human SULT2A1 (Liu, et al. ., 2006) had chlorine atoms flanking either side of the hydroxy group. The one compound tested that was inhibitory of 2A1, but was not a substrat e, had no neighboring chlorines. The structures selected for the assays with catfish SULT2 had at least one, and usually tw o, chlorines vicinal to the hydroxy group. Therefore, the lack of activity with the catfish SULT2 enzyme is not due to the structure of the OH-PCBs selected. Ch annel catfish liver cytosol sulfonation of 17 -estradiol was inhibited by 4OH-PCBs (Wa ng and James, 2007), but the sulf otransferase isoform involved could not be the SULT2 tested here, as it exhibited no activity with estradiol. The Km value of catfish SULT2 with DHEA was determined to be 43.7 M, the Vmax 0.516 nmol/min/mg, and the efficiency (Vmax/Km ) 0.012 mL/min/mg. No substrate inhibition was observed at the concentr ations tested (maximum 200 M). The Km value is somewhat higher than those previously determined for the sulfonation of DHEA by human SULT2A1 (25 M), SULT2B1a (2.27 M) or SULT2B1b (4.37 M) (Liu, et al. 2006; Geese and Raftogianis 2001). The Vmax is also notably lower th an the human SULT2 isoforms (122, 4.31, and 1.61 nmol/min/mg for 2A1, 2B1a, and 2B1b, respectivel y). This difference is not surprising, however, considering the low percent identity (4 3%) shared between the catfish SULT2 and its human homologues. In contrast, the kinetic constants for D. rerio SULT2 isoforms sharing the highest identity with the catfish isoform, st1 (72% identical) a nd st2 (74% identical), ar e more similar. The catfish Km value is 2 to 3 times lower than the zebrafish SULT2 st1 and st2 (102 and 177 M, respectively), while the Vmax is 4 to 5 times lower than those of zebrafish (2.10 and 2.64 nmol/min/mg, respectively). The catalytic efficiencies, however, are comparable: 0.012 mL/min/mg for catfish versus 0.021 mL/min/mg for zebrafish SULT2 st1 and 0.015 mL/min/mg

PAGE 121

121 for zebrafish SULT2 st2. The specific activities with DHEA as substrate are also comparable: 0.30 nmol/min/mg for catfish SULT2, versus 0.55 nmol/min/mg for zebrafish SULT2 st1 and 0.58 nmol/min/mg for zebrafish SULT2 st2. The kinetic differences between the human and bony fishes catalysis with SULT2 and DHEA are lik ely due to the different levels of that substrate in the bloodstream and its role. Du e to the high Km and low Vmax of DHEA for catfish SULT2, this molecule is probably not the preferred substrate for this isoenzyme, such as dopamine is for SULT1A3, cholesterol for SULT2B1b, pregnenolone for SULT2B1a, and 17 estradiol for SULT1E1. Further studies are warranted to thoro ughly characterize this isoform and determine its role in catfi sh physiology. The activity exhib ited with DHEA, which has an alcoholic moiety and is known to be a SULT2 substrate, and the la ck of activity exhibited with 3OH-BaP, which has a phenolic moiety and is known to be a SULT1 substrate, is in line with the identification of this catfish enzyme as a member of the family 2 sulfotransferases.

PAGE 122

122 CHAPTER 5 SUMMARY Three com plete and one partial (approximately two-thirds) sulfotransferase gene sequences have been obtained from channel catfish live r cDNA. Two of the complete genes, based on NCBI BLAST results, have highest percent identity with family one sulfotransferases; the third has greatest percent identity with family two sulfotransferases, and the partial sequence has highest percent identity with SULT6B1. Base d on these results, the sequences have been putatively termed SULT1v1 (S1v1), SULT1v2 (S1v2), SULT2 (S2), and SULT6B1. The informal use of v1 and v2 is for version, not to be confused with the formal use of _v for variant. A cDNA sequence encoding a sulfotransferase enzyme having greater than 45% identity with known Family 2 enzymes was obtained from channel catfish liver tissue, expressed, and characterized. The open-reading frame (ORF) encompasses 864 nucleotides and codes for a 287-amino acid polypeptide. A BLAST search of the NCBI databases revealed that, while the sequence shares high identity (69-74%) with other Fa mily 2 fish sulfotransferases, it displays much lower identity with known mamma lian Family 2 SULTs (maximum 44%). All enzymes sharing at least 45% amino acid se quence identity are grouped into the same family; within a family, enzymes with shared amino acid sequence identity of 65% share the same subfamily (Blanchard, et al ., 2004). Catfish SULT2 fits with the 45% rule for SULT2, but not with any known and named subfamilies, a char acteristic also seen in zebrafish SULTs. Therefore, according to nomenc lature guidelines, the catfish sequence can be considered a member of Family 2 but not assigned a subfam ily designation. The two catfish SULT1 isozymes follow the same pattern.

PAGE 123

123 Until recently, the majority of the fish sulf otransferases identified have been from zebrafish, as a model for studying the enzyme. The only characterized sulfotransferase for a food-fish is a single Family 1 is oform for the chub (Assem, et al ., 2006). None of the characterized zebrafish and chub sulfotransfera ses have been similar enough to any named mammalian SULTs to be named to a subfamily; they have instead been given a family name (e.g. SULT1) followed by a numerical designati on based on the order of discovery (e.g. Sulfotransferase family, cytosolic sulfotransfera se 6). In 2010, two articles were published that dealt with comparison of cDNA sequences be tween two fish species (each) and introduced thousands of fish cDNA sequences. Leong, et al (2010) compared 9,057 sequences for Atlantic salmon, Salmo salar with 1,365 sequences for northern pike, Esox Lucius to study evolutionary changes. Chen, et al. (2010) compared 1,064 cDNA sequences for channel catfish, Ictalurus punctatus, with 681 sequences for blue catfish, Ictalurus furcatus, for genetic comparisons of the two closely related catfish sp ecies. These sequences include six identified as cytosolic sulfotransferase: SULT4A1, SULT6B1 and sulfotransferase 3 for blue catfish, and SULT4A1, SULT2b1 and sulfotransferase 3 for channel catfis h. The sulfotransferase 3 sequences BLAST to Family 1 sulfotransferases, and are presumably family 1, isoform 3 sulfotransferases. The channel catfish sulfotransferase 3 is nearly identical (99%) to the SULT1v1 in this dissertation, and the blue catfish sulfotransferase 3 shares 96% identity. The published catfish sulfotransferase 3 sequences share 84% identity with the SULT1v2 in this dissertation. In contrast, the published channel catfish sequence identified as SULT2b1 shares only 43% identity with the SULT2 in this dissertation. BLAST resu lts for the SULT2b1 show a higher similarity to SULT5A1, which better fits th e low percent identity with the SULT2 discussed in this project.

PAGE 124

124 Since the onset of this disse rtation project the technology fo r gene hunting has evolved tremendously. The work discussed herein was do ne at the benchtop, the sequences obtained and confirmed through laborious and time consum ing methods, but the resulting sequences are absolute. Current methods begin with EST da ta and derive cDNA da ta from single pass sequence and re-sequence methods using 96-well plates instead of single tubes. As a result, the amount of data generated is impressive and a ma jor contribution, but the sequences can be prone to sequence errors introduced during the PCR am plification process that are not discovered without repeat (minimum of thr ee) independent PCR and sequencing steps. Therefore, there is a place in current molecular biology technique s for both the high-throughput generation of sequencing data and the more time-consuming benchtop confirmation steps to detect PCR errors, single-nucleotide polymorphisms, and splice variants. Substrate specificity characterized family 1 and family 2 sulfotransferases: SULT1 enzymes preferentially catalyze substrates w ith phenolic moieties, while SULT2 enzymes preferentially catalyze substrates with an alcohol moiety, particularly androgens. This pattern of activity with alcoholic su bstrates and not with phenol substrat e is exhibited by the catfish SULT2 enzyme. The channel catfish SULT2 translated amino acid sequence contains signature sequences highly conserved among all sulfotransferases. These include the 5-phosphate binding loop (PSB-loop) and two 3-PAPS binding sites (one designated 3PB). In the dimerization motif a critical valine residue in known mammalian SULT dimers is replaced by glutamic acid in the catfish SULT2. A different substitution (P/V) ex ists in the mouse SULT1E1, a known monomer. It cannot be presumed, therefore, that the catfish SULT2 behaves as a dimer in vivo The first of the signature sequences, the PSB-loop, is located at the N-terminal regi on and consists of

PAGE 125

125 residues TYPKSGT x W (Kakuta, et al ., 1997). This sequence in catfish SULT2, present as TYPKSGTTW, is identical to that in salmon, northern pike, and the published channel catfish Family 2 sulfotransferases, as well as zebrafish SULT2 isoform 3, and only one amino acid different from the sequence in zebrafish SULT2 isoforms 1 and 2. The region is identical in sequence to human SULT2B1b and differs by a single amino acid fr om human SULT2A1. A second signature sequence, conserved motif RKG xxGDWK xxFT, is present at the Cterminal end of the catfish SULT2 enzyme and is critical for the binding of PAPS (Komatsu, et al ., 1994). The first residue, Arg250, corresponds to Arg274 in human2B1b and is involved in binding the 3-phosphate oxygens of PAPS (Lee, et al ., 2003). In addition, the adenine of the PAP molecule is sandwiched be tween two aromatic residues, in a parallel/anti-parallel ringstacking arrangement with conserved residue s Trp50 and Phe222 (which correspond to Trp75 and Phe246 in human 2B1b). Another 3-PAPS binding region, designated as 3PB (Yoshinari, et al ., 2001), is present at 121TMRNPKDVFTSS132 for catfish SULT2, and contains critical residues Arg123 and Ser131, which directly interact with the oxygen atoms of the 3-phosphate group (Kakuta, et al ., 1997). A comparison of this regi on in SULT2 sequences, both mammalian and fish, that share highest identity with catfis h SULT2 reveals a slightly different conserved motif from the originally de signated 3PB (shown in Figure 2-11), beginning three residues upstream from the original motif. Although catfish SULT2 shares less than 45% identity with either human SULT2 (43% with humanSULT2B1 and 39% with humanSULT2A1 ), a comparison of substrate binding sites and enzyme activities shows a greater similari ty to humanSULT2B1 than to humanSULT2A1. Falany et al (1989) found purified human liver SULT2A1 to be active with the 3-hydroxy group of all hydroxysteroids tested as well as the 17-hydroxy group of testosterone and the 3-phenolic

PAGE 126

126 hydroxy of -estradiol and estrone. Catfish SULT2, like humanSULT2B1, di splays no activity with the phenolic substrate 3 OH-BaP or with the steroid 17 -estradiol. HumanSULT2A1 has a broader substrate rang e than humanSULT2B1 and is not stereospecific, able to sulfonate both and -hydroxy steroids, while humanSULT2B1 is specific for -hydroxysteroids due primarily to residue Ile51 (Park et al ., 1999). Mutation of that residue to Asn in guinea pig SULT2 altered the stereo-specificity of the enzyme from -hydroxysteroids to -hydroxysteroids. The catfish SULT2, as well as the fish SULT2s sharing the highest identity, have Gln in this location, which, like Asn, is a polar, hydrophilic amino acid. While this might suggest that the catfish SULT2 (and the related fish SULT2s) may be stereo-specific for hydroxy substrates, it shows activity with DHEA, a -hydroxysteroid. Amino acids in the substrate bi nding pocket theorized to stabil ize steroid rings (Lee, et al 2003; Chang, et al ., 2004) are conserved in the human SUTLT2 enzymes and channel catfish SULT2. Steroidal A and B rings are stabilized by Trp77 in human SU LT2A1, which correlates to Trp103 in human SULT2B1 and Trp78 in catfish SULT2. The B and C rings are stabilized in SULT2B1b by Tyr257, which correlates to Tyr31 in SULT2A1 and Tyr233 in catfish SULT2. A second amino acid in this region, Leu260 for SULT2B1b and Leu234 for SULT2A1, is also similar in catfish as Val236. A conserved tryp tophan lining the binding pocket opening is at Trp98 for SULT2B1b, Trp72 for SULT2A1, and Trp73 for catfish SULT2. Other binding pocket residues are not the same between the en zymes; however, not enough is known to predict a substrate selectivity based on the sequences. HumanSULT2A1 showed activity with 4-OH-PCB that had chlorine atoms flanking either side of the hydroxy group, and i nhibition with a 4-OH-PCB with no vicinal chlorines (Liu, et al ., 2006). The 4-OH-PCB structures se lected for the assays with catfish SULT2 had at least one,

PAGE 127

127 and usually two, chlorines vicinal to the hydroxy group, but none showed activity or inhibition. In addition, catfish SULT2 show ed good activity with DHEA as substrate, up to 200 M, with no inhibition. It is interesting that the catfish SULT2 exhibi ts no activity with either testosterone or estradiol. Human SULT2B1 has no activity with either substrate, but that can be attributed to its specificity for -hydroxysteroids. The thre e zebrafish Family 2 sulfotransferases discovered and characterized exhibit a range of substrate prefer ences, but all sulfonate es tradiol (Yasuda, et al ., 2006). Only one of the three (SULT2 st2) shows activity with estrone, and only SULT2 st1 shows activity with 17 -hydroxypregnenolone and a compou nd the authors identified as 4androsterone-3, 17-dione. SULT2st1 also has a ten-fold or greater specific activity with pregnenolone than st2 or st3. Continued explorat ion into the substrate range of catfish SULT2, including the above mentioned compounds, is needed to shed light on the part icular role of this isozyme. Channel catfish are farmed worldwide, and wh ile US-raised catfish are raised under strict rules of allowed chemicals acco rding to FDA regulations, imported catfish have been found to contain drugs banned in this country for aquaculture farming (von Eschenbach, 2008). Aquacultured catfish from the Peoples Republic of China are currently on Import Alert due to the presence of new animal drugs and/or unsafe food additives. Testing of imported aquacu ltured species from 2004 to 2007 revealed that products imported from Asia and South American countries have been the primary source of violative product, as illegal chemical c ontent is termed. Of the co mpounds tested, malachite green, gentian crystal violet (also known as crystal violet) and fluoroqui nolones have been detected in imported aquacultured catfish (Plakas, et al., 1996; Thompson, et al ., 1999; von Eschenbach,

PAGE 128

128 2008). Malachite green and gentian violet are administered as an tiparasitic agents, as well as treatment of skin/gill di seases or fungal infections. A lthough neither chemical is a known substrate for sulfotransferase, their presence demons trates the reality of unwanted and potentially dangerous additives in imported catfish, and the poten tial for the presence of other substrates not currently being tested for. For example, the FDA sampling program does not generally test for drugs that some countries and the European Union (EU) have approved for use in aquaculture (GAO Report, 2011). Additionally, in fiscal year 2009, FDA tested about 0.1% of all imported seafood products for drug residues (GAO Report, 2011), increasing the probability that unknown and undetected xenobiotics exist in commercially available catfish. The role of sulfotransferase has been found to extend well beyond simply metabolic detoxification. The enzymes are important in hor mone regulation as well, particularly of the endocrine system. Evidence indicates an important role in endocrine disr uption due to inhibition of the enzyme by industrial, agricultural, a nd some food chemicals. Catfish farming is worldwide and commercially important, but the exposure of catfish to approved and unapproved chemicals is a concern. An understanding of the different sulfotransferases in the species is important. This work provides an important step in that knowl edge, with a demonstration of different sulfotransferase sequences of three di fferent families as well as characterization of a SULT2 isozyme.

PAGE 129

129 LIST OF REFERENCES Alnouti Y and Klaassen CD (2006) Tissue distributi on and ontogeny of sulfotransferase enzym es in mice. Toxicological Sciences 93(2), 242-255 Arukwe A (2001) Cellular and molecular response s to endocrine-modulators and the impact on fish reproduction. Marine Pollution Bulletin 42(8), 643-644 Assem FL, Kirk CJ, Chipman JK (2006) Subs trate characterization of a recombinant sulfotransferase SULT1 and mRNA expression in chub ( Leuciscus cephalus) tissues. Biochemical and Biophysical Research Communications 349, 900-905 Banoglu E (2000) Current status of the cytosolic su lfotransferases in the metabolic activation of promutagens and procarcinogens. Current Drug Metabolism 1, 1-30 Blanchard RL, Freimuth RR, Buck J, Wein shilboum RM, Coughtrie MWH (2004) A proposed nomenclature system for the cytosolic sulfotransferase (SULT) superfamily. Pharmacogenetics 14, 199-211 Bonefeld-Jorgensen EC (2010) Biomonitoring in Greenland: human biomarkers of exposure and effects a short review. Rural Remote Health 10 (2), Article #1362 Brian JV, Harris CA, Scholze M, Backhaus T, Booy P, Lamoree M, Pojana G, Jonkers N, Runnalls T, Bonf A, Marcomini A, Sumpte r JP (2005) Accurate prediction of the response of freshwater fish to a mixture of estrogenic chemicals. Environmental Health Perspectives 113(6), 721-728 Carlson DB, Williams DE (2001) 4-hydroxy-2-4 -6-trichlorobiphenyl and 4-hydroxy-25tetrachlorobiphenyl are estr ogenic in rainbow trout. Environmental Toxicology and Chemistry 20 (2), 351-358 Chang HJ, Shi R, Rehse P, Lin SX (2004) Identify ing androsterone (ADT) as a cognate substrate for human dehydroepiandrosterone sulfotransferase (DHEA-ST) important for steroid homeostasis: structure of the enzyme-ADT complex. Journal of Biological Chemistry 279(4), 2689-2696 Chen F, Lee Y, Jiang Y, Wang S, Peatman E, Abernathy J, Liu H, Liu S, Kucuktas H, Ke C, Liu Z (2010) Identification and characterization of full-length cDNAs in channel catfish ( Ictalurus punctatus ) and blue catfish ( Ictalurus furcatus ). PLoS ONE 5(7), e11546 Chen G, Battaglia E, Senay C, Falany CN, Radominska-Pandya A (1999) Photoaffinity labeling probe for the substrate binding site of human phenol sulfotransferase (SULT1A1): 7Azido-4-methylcoumarin. Protein Science 8, 2151-2157

PAGE 130

130 Chou HC, Lang NP, Kadlubar FF (1995) Metabolic ac tivation of the N-hydroxy derivative of the carcinogen 4-aminobiphenyl by human tissue sulfotransferases. Carcinogenesis 13, 413417 Colborn T, Dumanoski D, Meyers JP (1996) Ou r Stolen Future: How We Are Threatening Our Fertility, Intelligence and Survival. Pengui n Group Publishing, New York, New York Coughtrie MWH (2002) Sulfation through the looking glassrecent advances in sulfotransferase research for the curious. The Pharmacogenomics Journal 2,297-308 Dajani R, Hood AM, Coughtrie MWH (1998) A single amino acid, Glu146, governs the substrate specificity of a human dopamine sulfotransferase, SULT1A3. Molecular Pharmacology 54, 942-948 Duanmu Z, Weckle A, Koukouritaki SB, Hines RN, Falany JL, Falany CN, Kocarek TA, RungeMorris M (2005) Developmental expressi on of aryl, estrogen and hydroxysteroid sulfotransferases in preand postnatal human liver. The Journal of Pharmacology and Experimental Therapeutics 316(3), 1310-1317 Facemire CF and Gross TS (1995) Reproductive impa irment in the Florida panther: nature or nurture? Environmental Health Pe rspectives Supplements 103 Supplement 4 Falany CN, Vazquez ME, Kalb JM (1989) Purifi cation and characterization of human liver dehydroepiandrosterone sulphotransferase. Biochemical Journal 260, 641-646 Falany CN (1997) Enzymology of human cytosolic sulfotransferases. The FASEB Journal 11, 206-216 Falany CN, Xie X, Wang J, Ferrer J, Falany JL (2000) Molecular cloning and expression of novel sulphotransferase-like cDNAs from human and rat brain. The Biochemical Journal 346, 857-864 Fuda H, Lee YC, Shimizu C, Javitt NB, Strott CA (2002) Mutational analysis of human hydroxysteroid sulfotransferase SULT2B1 is oforms reveals that exon 1B of the SULT2B1 gene produces cholesterol sulfotransferase s, whereas exon 1A yields pregnenolone sulfotransferase. The Journal of Biological Chemistry 277 (39), 36161-36166 GAO: United States Government Accountability Office (2011) Seafood Safety: FDA needs to improve oversight of imported seafood and be tter leverage limited resources. (Accessed June 2011) Ga mage NU, Dubbleby RG, Barnett AC, Tresil lian M, Latham CF, Liyou NE, McManus ME, Martin JL (2003) Structure of a human carcinogen-converting enzyme, SULT1A1. Structural and kinetic implicat ions of substrate inhibition. The Journal of Biological Chemistry 278 (9), 7655-7662

PAGE 131

131 Gamage N, Barnett A, Hempel N, Duggleby RG Windmill KF, Martin JL, McManus ME (2006) Human sulfotransferases and their role in chemical metabolism. Toxicological Sciences 90(1), 5-22 Glatt H, Pauly K, Frank H, Seidel A, Oesc h F, Harvey RG, Werle-Schneider G (1994) Substance-dependent sex differences in the activation of benzylic alcohols to mutagens by hepatic sulfotransferases of the rat. Carcinogenesis 15(11), 2605-2611 Glatt H, Pauly K, Czich A, Falany JL, Falany CN (1995) Activation of benzylic alcohols to mutagens by rat and human sulfotransferases expressed in Escherichia coli European Journal of Pharmacology 293, 173-181 Glatt H, Boeing H, Engelke CEH, Ma L, Kuhlow A, Pabel U, Pomplun D, Teubner W, Meinl W (2001) Human cytosolic sulphot ransferases: genetics, char acteristics, toxicological aspects. Mutation Research 482, 27-40 Harshbarger JC, Coffey MJ, Young MY (2000) Intersexes in Mississi ppi river shovelnose sturgeon sampled below Saint Louis, Missouri, USA. Marine Environmental Research 50, 247-250 He D and Falany CN (2006) Charac terization of proline-serine-rich carboxyl terminus in human sulfotransferase 2B1b: Immunogenicity, sub cellular localization, ki netic properties, and phosphorylation. Drug Metabolism and Disposition 34(10), 1749-1755 Heiden TK, Hutz RJ, Carvan III, MJ (2005) A ccumulation, tissue distribution, and maternal transfer of dietary 2,3,7,8-tetrachlorodibenzop -dioxin: impacts on reproductive success of zebrafish. Toxicological Sciences 87(2), 497-507 Jobling S, Nolan M, Tyler CR, Brighty G, Sump ter JP (1998) Widespread sexual disruption in wild fish. Environmental Science & Technology 32(17), 2498-2506 Jones KC and de Voogt P (1999) Persistent organic pollutants (P OPs): state of the science. Environmental Pollution 100, 209-221 Jurgella GF, Marwah A, Malison JA, Peterson R, Barry TP (2006) Effects of xenobiotics and steroids on renal and hepatic estr ogen metabolism in lake trout. General and Comparative Endocrinology 148(2), 273-281 Kakuta Y, Pedersen LG, Carter CW, Negishi M, Pedersen LC (1997) Crystal structure of estrogen sulphotransferase. Nature Structural Biology 4(11), 904-908 Kauffman FC (2004) Sulfonation in pharmacology and toxicology. Drug Metabolism Reviews 36(3&4), 823-843

PAGE 132

132 Kester MHA, Bulduk S, van Toor H, Tibboel D, Meinl W, Glatt H, Falany CN, Coughtrie MWH, Schuur AG, Brouwer A, Visser TJ (2002) Potent inhibition of estrogen sulfotransferase by hydroxylated metabolites of polyhalogenated aromatic hydrocarbons reveals alternative mechanism for estrogenic activity of endocrine disrupters. The Journal of Clinical Endocrinology & Metabolism 87(3), 1142-1150 Kim SY, Suzuki N, Laxmi YRS, Shibutani S (2 004) Genotoxic mechanism of Tamoxifen in developing endometrial cancer. Drug Metabolism Reviews 36(2), 199-218 Kim SY, Laxmi YRS, Suzuki N, Ogura K, Watabe Ta, Duffel MW, Shibutani S (2005) Formation of tamoxifen-DNA adducts via o-sulfonation, not o -acetylation, of hydroxytamoxifen in rat and human livers. Drug Metabolism and Disposition 33(11), 1673-1678 Kohjitani A, Fuda H, Hanyu O, Strott CA (2006) Cloning, characterization and tissue expression of rat SULT2B1a and SULT2B1b st eroid/sterol sulfotransferase isoforms: Divergence of the rat SULT2B1 gene structure from orthol ogous human and mouse genes. Gene 367, 66-73 Komatsu K, Driscoll WJ, Koh YC, Stro tt CA (1994) A P-loop related motif (G xxgxxk) highly conserved in sulfotransferase s is required for binding the activated sulfate donor. Biochemical and Biophysical Research Communications 204, 149-157 Krimsky S (2001) Hormone disruptors: a clue to understanding the environmental causes of disease. Environment 43 (5), 22-31 Lee KA, Fuda H, Lee YC, Negishi M, Strott CA, Pedersen LC (2003) Crystal structure of human cholesterol sulfotransferase (SULT2B1b) in the presen ce of pregnenolone and 3phosphoadenosine 5-phosphate. The Journal of Biological Chemistry 278 (45), 4459344599 Leong JS, Jantzen SG, von Schalburg KR, Cooper GA, Messmer AM, Liao NY, Munro S, Moore R, Holt RA, Jones SJM, Davidson WS, Koop BF (2010) Salmo salar and Esox lucius full-length cDNA sequences reveal chan ges in evolutionary pressures on a posttetraploidization genome. BMC Genomics 11, 279 Liu Y, Apak TI, Lehmier HJ, Robertson LW, Du ffel MW (2006) Hydroxylated polychlorinated biphenyls are substrates and inhibitors of human hydroxys teroid sulfotransferase SULT2A1. Chemical Research in Toxicology 19 (11), 1420-1425 Lu LY, Hsieh YC, Liu MY, Lin YH, Chen CJ, Yang YS (2008) Identification and characterization of two am ino acids critic al for the substrate inhibition of human dehydroepiandrosterone sulf otransferase (SULTA1). Molecular Pharmacology 73(3), 660-668

PAGE 133

133 Matz MV (2003) Amplification of representative cDNA pools from microscopic amounts of animal tissue. Methods in Molecular Biology 221 103-116 Matz MV, Alieva NO, Chenchik A, Lukyanov S ( 2003) Amplification of cDNA ends using PCR suppression effect and step-out PCR. Methods in Molecular Biology 221, 41-49 Mellon SY, Griffin LD, Compa gnone NA (2001) Biosynthesis and action of neurosteroids. Brain Research Reviews 37, 3-12 Meloche CA and Falany CN (2001) Expressi on and characterization of the human 3 hydroxysteroid sulfotransferases (SULT2B1a and SULT2B1b). Journal of Steroid Biochemistry & Molecular Biology 77 261-269 Mills LJ and Chichester C (2005) Review of evidence: are endocri ne-disrupting ch emicals in the aquatic environment impacting fish populations? Science of the Total Environment 343 1-34 Nagler JJ, Bouma J, Thorgaard GH, Dauble DD (2001) High incidence of a male-specific genetic marker in phenotypic female Ch inook salmon from the Columbia river. Environmental Health Perspectives 109(1), 67-69 Park BC, Lee YC, Strott CA (1999) Identifica tion by chimera formation and site-selected mutagenesis of a key amino acid residue invol ved in determining stereospecificity of guinea pig 3-hydroxysteroid sulf otransferase isoforms. Journal of Biological Chemistry 274, 21562-21568 Pedersen LC, Petrotchenko EV, Negishi M (2 000) Crystal structure of SULT2A3, human hydroxysteroid sulfotransferase. FEBS Letters 475 61-64 Pedersen LC, Petrotchenko E, Shevtsov S, Negi shi M (2002) Crystal structure of the human estrogen sulfotransferase-PAPS complex. The Journal of Biological Chemistry 277 (20), 17928-17932 Petrotchenko EV, Doerflein ME, Kakuta Y, Pede rsen LC, Negishi M (1999) Substrate gating confers steroid specificity to estrogen sulfotransferase. Journal of Biological Chemistry 274, 30019-30022 Petrotchenko EV, Pedersen LC, Borchers CH, Tomer KB, Negishi M (2001) The dimerization motif of cytosolic sulfotransferases. FEBS Letters 490, 39-43 Purinton SC, Wood CE (1999) Ovine fetal estrogen sulfotransferase in brain regions important for hypothalamus-pituitary-adrenal axis control. Reproductive Neuroendocrinology 71 237-242 Plakas SM, El Said KR, Stehly GR, Gingerich WH, Allen JL (1996) Upta ke, tissue distribution, and m etabolish of malachite gr een in the channel catfish ( Ictalurus punctatus. Canadian Journal of Fisheries and Aquatic Sciences 53, 1427-1433

PAGE 134

134 Qian YM, Sun XJ, Tong MH, Li XP, Richa J, Song W-C (2001) Targeted disruption of the mouse estrogen sulfotransferase gene reveals a role of estrogen metabolism in intracrine and paracrine estrogen regulation. Endocrinology 142 (12), 5342-5350 Rauschenberger RH (2009) Nutrient and organoch lorine pesticide concentrations in American alligator eggs and their associa tions with clutch viability. Journal of Aquatic Animal Health 21, 249-261 Rehse PH, Zhou M, Lin S-X ( 2002) Crystal structure of hum an dehydroepiandrosterone sulphotransferase in complex with substrate. The Biochemical Journal 364 165-171 Richard K, Hume R, Kaptein E, Stanley EL, Visser TJ, C oughtrie MWH (2001) Sulfation of thyroid hormone and dopamine during human development: ontogeny of phenol sulfotransferases and arylsulf atase in liver, lung, and brain. The Journal of Clinical Endocrinology & Metabolism 86(6), 2734-2742 Sakakibara Y Takami Y, Nakayama T, Suiko M, Liu M-C (1998) Localization and functional analysis of the substrate sp ecificity/catalytic domains of human M-form and P-form phenol sulfotransferases. The Journal of Biological Chemistry 273(11), 6242-6247 Sakakibara Y, Suiko M, Pai TG, Nakayama T, Takami Y, Katafuchi J, Liu M-C (2002) Highly conserved mouse and human brain sulfotrans ferases: molecular cl oning, expression, and functional characterization. Gene 285, 39-47 Saraste M, Sibbald PR, Wittinghofer A (1990) Th e P-loop a common motif in Atp-binding and Gtp-binding proteins. Trends in Biochemical Sciences 15, 430-434 Seah VMY and Wong KP (1994) 2,6-Dichloro-4-n itrophenol (DCNP), an alternate-substrate inhibitor of phenolsulfotransferase. Biochemical Pharmacology 47(10), 1743-1749 Seplveda MS, Del Perio F, Wiebe JJ, Rausche nberger HR, Gross TS (2006) Necropsy findings in American alligator late-stage embryos a nd hatchlings from northcentral Florida lakes contaminated with organochlorine pesticides. Journal of Wildlife Diseases 42(1), 56-73 Strott CA (2002) Sulfonation and molecular action. Endocrine Reviews 23(5), 703-732 Sugimura T (1997) Overview of carcinogenic heterocyclic amines. Mutation Research 376, 211219 Thibaut R, Porte C (2004) Effects of endocri ne disrupters on sex st eroid synthesis and metabolism pathways in fish. Journal of Steroid Biochemistry & Molecular Biology 92, 485-494

PAGE 135

135 Thompson Jr. H, Rushing L, Gehring T, Lockmann R (1999) Persistence of gentian violet in channel catfish muscle ( ictalurus punctatus ) after water-borne exposure. Journal of Chromatography B 723, 287-291 Tong Z and James MO (2000) Purification and char acterization of hepatic and intestinal phenol sulfotransferase with high affinity for benzo[a]pyrene phenols from channel catfish, Ictalurus punctatus Archives of Biochemistry and Biophysics 376(2), 409-419 Turesky RJ (2004) The role of genetic polymorphisms in metabolism of carcinogenic heterocyclic aromatic amines. Current Drug Metabolism 5, 169-180 van den Hurk P, Kubiczak GA, Lehmler HJ, James MO (2002) Hydroxylated polychlorinated biphenyls as inhibitors of th e sulfation and glucuronidati on of 3-hydroxy-benzo[a]pyrene. Environmental Health Perspectives 110(4), 343-348 Van Lipzig MMH, Commandeur JN de Kanter FJJ, Damsten MC, Vermeulen NPE, Maat E, Groot EJ, Brouwer A, Kester MHA, Visser TJ, Meerman J HN (2005) Bioactivation of dibrominated biphenyls by cytochrome P450 activity to metabolites with estrogenic activity and estrogen sulfotra nsferase inhibition capacity. Chemical Research in Toxicology 18, 1691-1700 Verslycke TA, Vethaak AD, Arijs K, Janssen CR (2005) Flame retardants, surfactants and organotins in sediment and mysid shrimp of the Scheldt estuary (the Netherlands). Environmental Pollution 136, 19-31 von Eschenbach AC (2008) Enhanced aquaculture a nd seafood inspection Re port to Congress. ation/Seafood/Seaf oodRegulatoryProgram/ucm150954.htm (Accessed February 2011) Walker MK and Peterson RE (1991) Potencies of polychlorinated dibenzop-dioxins, debenzofurans and biphenyl congeners, relative to 2,3,7,8-tetrachlorodibenzop-dioxin, for producing early life stage mo rtality in rainbow trout ( Oncorhynchus mykiss). Aquatic Toxicology 21, 219-238 Wang LQ and James MO (2006) Inhibition of sulfotransferases by xenobiotics. Current Drug Metabolism 7(1), 83-104 Wang LQ, James MO (2007) Sulfonation of 17 -estradiol and inhibiti on of sulfotransferase activity by polychlorobiphenylols and celecoxib in channel catfish, Ictalurus punctatus Aquatic Toxicology 81, 286-292 Wellborn TL; Southern Regional Aquaculture Center (SRAC) pub lication No. 180, from website, http:// (accessed 11-06)

PAGE 136

136 Wood CE, Gridley KE, Keller-Wood M (2003) Biological activity of 17 -estradiol-3-sulfate in ovine fetal plasma and uptake in fetal brain. Endocrinology 144 (2), 599-604 Yasuda S, Liu MY, Yang YS, Snow R, Takaha shi S, Liu MC (2006) Identification of novel hydroxysteroid-sulfating cytosolic SULTs, SULT2 ST2 and SULT2 ST3, from zebrafish: Cloning, expression, charac terization, and developmental expression. Archives of Biochemistry and Biophysics 455, 1-9 Yoshinari K, Petrotchenko EV, Pedersen LC, Negishi M (2001) Crystal st ructure-based studies of cytosolic sulfotransferases. The Journal of Biochemical and Molecular Biology 15(2), 67-75 Zhang H, Varmalova O, Vargas FM, Falany CN, Le yh TS (1998) Sulfuryl tr ansfer: the catalytic mechanism of human estrogen sulfotransferase. The Journal of Biological Chemistry 273(18), 10888-10892

PAGE 137

137 BIOGRAPHICAL SKETCH Kristen Be lcher Merritt was born in Connecticut but raised in the small fishing village of Cedar Key, Florida. The daughter of a marine biologist father and ar tist mother, she blended these influences into a scientific examinati on of the marine world. Kristen was awarded a Bachelor of Science degree in chemistry, with highest honors, from the University of Florida (Gainesville). She subsequently received a Master of Science in food science and human nutrition, with a focus on toxicology, also from the University of Florida; that research project involved the study of methylmercury in green algae.