|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
PHASE II BIOTRANSFORMATION OF XENOBIOTICS IN POLAR BEAR
(Lh-sus maritimus) AND CHANNEL CATFISH (Ictalurus punctatus)
JAMES C. SACCO
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
JAMES C. SACCO
This document is dedicated to Denise and my parents.
First and foremost, I would like to thank my mentor, Dr. Margaret O. James, for
her instruction, guidance, and support throughout my PhD program. Through her
excellent scientific and mentoring skills I not only managed to complete several
interesting studies but also rekindled my scientific curiosity with regards to
biotransformation and biochemistry in general. I greatly appreciate the advice and
instruction of Ms. Laura Faux, our laboratory manager, on enzyme assays, HPLC, and
fish dissection. The assistance and advice of Dr. David S. Barber, Mr. Alex McNally,
and Mr. Jason Blum at the Center for Human and Environmental Toxicology in walking
me through the complexities of molecular cloning are much appreciated. Academic
discussions with Dr. Liquan Wang, Dr. Ken Sloan, Dr. Joe Griffitt and Dr. Nancy
Denslow also helped me to interpret my results and design better experiments
Last but not least, I would like to thank my fiancee, Denise, and my parents, for
their support and encouragement throughout my doctoral studies.
TABLE OF CONTENTS
ACKNOWLEDGMENT S .............. .................... iv
LI ST OF T ABLE S ................. ................. vii........ ....
LIST OF FIGURES .............. .................... ix
AB STRAC T ................ .............. xii
1 BIOTRANSFORMATION AND ITS IMPORTANCE IN THE
DETOXIFICATION OF XENOBIOTICS ................. ...............1............ ....
2 PHASE II CONJUGATION: GLUCURONIDATION AND SULFONATION .........4
UJDP-Glucuronosyltransferases (UGTs) .....__.....___ ..........._ ............7
Sulfotransferases (SULTs) ............ ..... ._ ...............11....
3 SULFONATION OF XENOBIOTICS BY POLAR BEAR LIVER .........................14
Hypothesis .............. ...............17....
M ethodology ................. ...............17.................
Re sults ................ ...............23.................
Discussion ................. ...............32.................
4 GLUCURONIDATION OF POLYCHLORINATED BIPHENYLOLS BY
CHANNEL CATFISH LIVER AND INTESTINE ......____ ........_ ...............38
Hypothesis .............. ...............41....
M ethodology ................. ...............41....... .....
Re sults ....._ _................ ........_ _..........4
Discussion .........._ ......... ...... ...............53....
Conclusions and Recommendations .....__................. ........_ .........5
5 CLONING OF UDP-GLUCURONOSYLTRANSFERASES FROM CHANNEL
CATFISH LIVER AND INTESTINE............... ...............6
Piscine UGT Gene Structure and Isoforms ............. ..............60.....
Hypothesis ................ ...............62...
Method ology (part 1) ............... ...............62...
Results and discussion (part 1) .............. ...............74....
Methodology (part 2) .........__.. ..... ._ __ ...............79....
Overview of RLM-RACE ....._.. ................ ........_._ ......... 7
5' RLM-RACE procedure .............. ...............8 4....
3' RACE procedure ..................... ...............86.
PCR amplification of entire UGT gene ......___ ..... .._.. ......_._.........8
Results (part 2)............... ... ...............8
Nucleotide sequence analysis ................ ...............89........... ....
Protein sequence analysis ................ ...............100................
Cloning of entire UGT gene ................. ...............104........... ...
Discussion ................. ...............107................
Limitations ................. ............ ...............110......
Conclusions and recommendations ................. ......... ......... .............1
6 DETERMINATION OF PHYSIOLOGICAL UDPGA CONCENTRATIONS IN
CHANNEL CATFISH LIVER AND INTESTINE ......____ ........ __ ..............116
UDP-Glucuronic Acid (UDPGA) ................. ...............116................
Obj ective ................. ....._ __ ...............118......
M ethod Development ................. ...............118....... ......
Sample Digestion ....__ ................. .........__..........19
H PLC ............ _...... ...............1 1....
Final M ethod .............. ...............123....
Re sults............ ..... .. ...............125...
Discussion ............ __... ...._ ..... ._ ............12
Conclusions and Recommendations ....__ ......_____ .......___ .............2
A SEQUENCES OF UGT PARTIAL CLONES AND AMPLICONS ................... .....131
B SEQUENCES FOR UGT FULL-LENGTH CLONES FROM CATFISH LIVER..138
LIST OF REFERENCES ........._._ ...... ..... ...............144....
BIOGRAPHICAL SKETCH ........._... ...... .___ ...............157....
LIST OF TABLES
2-1 Expression of human UGT mRNA in various tissues ................. ........... ..........9
2-2 Tissue distribution of SULTs (cDNA and mRNA) in humans.............. ...... ........._12
3-1 Estimated kinetic parameters (Mean + SD) for (a) sulfonation and (b)
glucuronidation of 3-OH-B[a]P by polar bear liver cytosol and microsomes. ........24
3-2 Kinetic parameters (Mean + SD) for the sulfonation of various xenobiotics by
polar bear liver cytosol, listed in order of decreasing enzymatic efficiency. ...........27
4-1 Estimated kinetic parameters (mean & S.D.) for the co-substrate UDPGA in the
glucuronidation of three different OH-PCBs. ...........__......_ ................44
4-2 Kinetic parameters (Mean & S.D.) for the glucuronidation of 4-OHBP and OH-
PCBs. ........... ..... .. ...............48....
4-3 Comparison of the estimated kinetic parameters for OH-PCB glucuronidation in
catfish liver and proximal intestine .............. ...............48....
4-4 Comparison of kinetic parameters (Mean & SEM) for the glucuronidation of
OH- PCBs grouped according to the number of chlorine atoms flanking the
phenolic group ........._.__...... ..__ ...............49....
4-5 Results of regression analysis performed in order to investigate the relationship
between the glucuronidation of OH-PCBs by catfish proximal intestine and liver
and various estimated physical parameters. ............. ...............52.....
5-1 5' 3' Sequences of degenerate primers chosen. ......___ ... ...... ..............66
5-2 Primer pairs chosen, showing annealing temperature and estimated amplicon
length ................. ...............67.................
5-3 Results of BLASTn search of cloned putative partial UGT sequences ........._........78
5-4 Gene-specific primers used in initial 5'RLM-RACE study. ................ ................82
5-5 Gene-specific primers used in succeeding RLM-RACE study ............... .... ............83
5-6 Primers used for amplifying liver and intestinal UGT gene .............. ..................87
5-7 Results of blastn search for livUGTn (and intUGTn) ................ ......................92
5-8 Promoter prediction ................. ...............92........... ....
5-9 Results of blastp search for liv/intUGTp ................. .....__ ............. ......9
5-10 Results of blastn search for I35R C............... ...............99...
5-11 Potential antigenic sites on liv/intUGTp. ............. ...............101....
5-12 Results of blastp search for I3SRCp ................. ...............102........... .
5-13 Results of ClustalW multiple sequence alignment analysis of the cloned UGTs
and the original livUGTn .............. ...............106....
5-14 Conserved consecutive residues observed in catfish liver and mammalian UGTs
(sequences shown in Figure 5-13)............... ...............109.
6-1 UDPGA concentrations (C1M) in liver and intestine of various species ................. 117
6-2 Elution times of certain physiological substances (standards dissolved in mobile
phase) using the anion-exchange HPLC conditions described above...................124
6-3 UDPGA concentrations in CIM (duplicates for individual fish), in catfish liver
and intestine ................. ...............126................
LIST OF FIGURES
1-1 Schematic of select xenobiotic (represented by hydroxynaphthalene)
biotransformation pathways in the mammalian cell. ............. ....................3
2-1 Structure of the co-substrates PAPS and UDPGA (transferred moieties shown in
bold) and the formation of the polar sulfonate and glucuronide conjugates,
shown here competing for the same substrate ................. .............................6
2-2 Proposed structure of UGT, based on amino acid sequence .............. ..............7
2-3 Complete human UGT1 complex locus represented as an array of 13 linearly
arranged first exons. ............. ...............10.....
2-4 The human UGT2 family. ............. ...............10.....
3-1 Structures of sulfonation sub states investigated in thi s study ................. ...............1 5
3-2 Sulfonation of 3 -OH-B[a]P at PAPS = 0.02 mM ................. ................ ...._..25
3-3 Eadie-Hofstee plot for the glucuronidation of 10 CIM 3-OH-B[a]P, over a
UJDPGA concentration range of 5-3000 CIM. ............. ...............26.....
3-4 Sulfonation of 4'-OH-PCB79, PAPS = 0.02 mM. ................... ............... 2
3-5 Autoradiogram showing the reverse-phase TLC separation of sulfonation
products of OHMXC ................ ...............29........... ....
3-6 Autoradiogram showing the reverse-phase TLC separation of sulfonation
products from incubations with TCPM ................. ...............30........... ...
3-7 Autoradiogram showing the reverse-phase TLC separation of sulfonation
products of TCPM and the effect of sulfatase treatment............._. .. ........._._ ...31
3-8 Autoradiogram showing reverse-phase TLC separation of sulfonation products
from the study of PCP kinetics ....__. ................. ...............32. ...
4-1 Structure of sub states used in channel catfi sh glucuronidation study ................... ..42
4-2 UDPGA glucuronidation kinetics in 4 catfish............... ...............46
4-3 Representative kinetics of the glucuronidation of OH-PCB s in 4 catfi sh. ...............47
4-4 Decrease in Vmax with addition of second chlorine atom flanking the phenolic
group, while keeping the chlorine substitution pattern on the nonphenolic ring
constant ................. ...............50.................
4-5 Relationship between Vmax for OH-PCB glucuronidation in intestine and liver
and ovality .............. ...............53....
5-1 Summary of methods used to clone channel catfish UGT ..........__..................63
5-2 Products of PCR reaction. 1(from intestine), 2 and 3 (from liver) ................... ........75
5-3 Plasmid DNA obtained from cultures transformed with vector containing inserts
from liver and intestine. ............. ...............76.....
5-4 Product of ecoRI digest of purified plasmids containing liver inserts L1-L8..........77
5-5 5'- RLM-RACE and 3'- RACE ................. ........._._......... ................80
5-6 Primer positions for 5'- and 3'-RACE ......... ........ ................ ...............81
5-7 Full nucleotide sequence obtained for hepatic catfish UGT (livUGTn), derived
from 4 sequencing runs each. .............. ...............90....
5-8 Sizes and positions of partial UGT sequences (cross-hatched rectangles) from
intestine and liver, corresponding to two distinct isoforms, relative to complete
sequences for liver and intestinal UGT (solid rectangles). ............. ....................91
5-9 Identification of open reading frame using ORF Finder ................. ............... ....93
5-10 Predicted protein sequence liv/intUGTp from liv/intUGTn .............. ..................93
5-11 Comparison ofliv/intUGTp with homologous proteins in other fish, showing
scores and alignment of closely related sequences. ............. .....................9
5-12 Phylogram for fish UGT proteins homologous to liv/intUGTp .............. ..............96
5-13 Alignment of liv/intUGTp (excluding UTRs) with selected mammalian UGT
proteins, showing scores and multiple alignment of sequences, highlighting
important regions and residues (see discussion) .............. ...............97....
5-14 Phylogram for I.punctatus liv/intUGTp and selected mammalian UGT proteins ...98
5-15 Multiple sequence alignment between livUGTn and I35RC. ............. .................99
5-16 Results of NCBI conserved domain search ................. ...............100........... .
5-17 Kyte-Doolittle Hydrophobicity Plot for liv/intUGTp .............. .....................0
5-18 Results of NCBI conserved domain search for I35R Cp .............. ...................103
5-19 Alignment of predicted protein sequences from cloned catfish UGTs. Regions of
interest and the starting and ending residue of the mature product are
highlighted. ........... ..... ._ ...............104...
5-20 Cloning of livUGTn. ..........._.....__ .....__ .....__ ........... ....105
5-21 Cloning of intUGTn. ..........._.....__ .....__ .....__ ........... ....105
5-22 Multiple sequence alignment for fish sequences homologous to catfish UGT
isolated from liver and intestine, showing regions where substrate binding of
phenols is thought to occur for mammalian UGTI1A isozymes. ................... .........110
5-23 Results of 3' RACE performed on liver, showing multiple products obtained...... 111
5-24 3' RACE for I4. ................ ...............112..............
5-25 PCR amplification of UGT using degenerate primers. ........._._ ... ......_._.......1 14
6-1 Heat-induced degradation of UDPGA (boiling in 0.25 M H2PO4 buffer) ............120
6-2 Decomposition of UDPGA to UDP and UMP after boiling in 0.25 M H2PO4
buffer for 10 minutes ..........._ _..... .._ ...............120.
6-3 Effect of boiling liver tissue for 1 minute in two different concentrations of
buffer. A, 0.25 M H2PO4, pH 3.4; B, 0.30 M H2PO4, pH 4.3 ................ ............... 121
6-4 HPLC chromatogram for catfish AT17 liver. Center refers to region of liver
from which the sample was taken. ............. ...............122....
6-5 HPLC chromatogram for catfish AT18 intestine. Rep 2 refers to second sample
taken from AT18 intestine. .........__ _............ ...............123.
6-6 HPLC chromatogram of UDP, UDP-galacturonic acid (UDPGTA), and UDPGA
standards ................. ...............125................
6-7 Comparison of hepatic and intestinal [UDPGA] in 4 individual channel catfish. .126
Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
PHASE II BIOTRANSFORMATION OF XENOBIOTICS BY POLAR BEAR (Grsus
maritimus) AND CHANNEL CATFISH (Ictalurus punctatus)
James C. Sacco
Chair: Margaret O. James
Major Department: Medicinal Chemistry
Both polar bears and channel catfish are subject to bioaccumulation of persistent
toxic environmental pollutants including hydroxylated compounds, which are potential
substrates for detoxification via phase II conjugative processes such as sulfonation and
glucuronidation. The objectives of this dissertation were to (a) study the capability of
polar bear liver to sulfonate a structurally diverse group of environmental chemicals, and
to study the glucuronidation of 3-OH-B[a]P; (b) study the effects of chlorine substitution
pattern on the glucuronidation of polychlorinated biphenylols (OH-PCBs) by catfish liver
and proxi mal inte sti ne; (c) clone UDP -glucurono syltran sferase (UGT) from catfish liver
and intestine; (d) develop a method to determine physiological concentrations of UDP-
glucuronic acid (UDPGA) in catfish liver and intestine.
In the polar bear, the efficiency of sulfonation decreased in the order 3-OH-
>PCP, all of which produced detectable sulfate conjugates. Substrate inhibition was
observed for the sulfonation of 3-OH-B[a]P and 4'-OH-PCB79. The hexachlorinated
OH-PCBs, TCPM and PCP were poor substrates for sulfonation, suggesting that this may
be one reason why these substances and structurally similar xenobiotics persist in polar
OH-PCBs are glucuronidated with similar efficiency by channel catfish liver and
proximal intestine. There were differences in the UGT activity profile in both organs.
Both hepatic glucuronidation and intestinal glucuronidation were decreased with the
addition of a second chlorine atom flanking the phenolic group, which is an arrangement
typical of toxic OH-PCBs that persist in organisms.
One full length UGT from catfish liver, together with a full-length UGT (identical
to the liver UGT), and a partial sequence of a different UGT from catfish intestine were
cloned. The full-length catfish UGT clone appeared to be analogous to mammalian
UGTIAl or UGTIA6.
The anion-exchange HPLC method developed to determine UDPGA was sensitive,
reproducible and displayed good resolution for the co-substrate. The hepatic UDPGA
levels determined by this method were similar to those in other mammalian species and
higher than reported for two other fish species. This was the first time intestinal UDPGA
concentrations in any piscine species were determined; the values were similar to rat
intestine, but significantly higher than in human small intestine.
BIOTRANSFORMATION AND ITS IMPORTANCE INT THE DETOXIFICATION OF
The exposure of biological systems to environmental compounds which may be
potentially toxic to these systems has spurred the evolution of an elaborate, protective
biochemical system whereby these xenobiotics are eliminated from cells and whole
organisms, usually via chemical transformation (or biotransformation). This system is
composed of a multitude of enzymes, which while being distributed in many tissues and
organs, are principally located in organs such as liver, intestine and lungs. This is of
physiological significance since these tissues represent major routes of xenobiotic entry
into organisms. Within cells, biotransformation enzymes also display a level of
organization in that while some are soluble and found in the cytosol (e.g.
sulfotransferases (SULT), glutathione-S-transferases), others are relatively immobile and
membrane-bound (e.g. UDP-glucuronosyltransferases (UGT) and cytochrome P450s
(CYP) in the endoplasmic reticulum).
Since it is highly improbable that the organism has a substrate-specific enzyme for
metabolizing every potential xenobiotic, biotransformation enzymes are generally non-
specific, acting on a broad range of structurally unrelated substrates. In addition, several
isoforms of the same enzyme (or more than one enzyme) may catalyze product formation
from the same substrate, albeit at different rates and with different affinities. Enzymes in
the same superfamily as those that act upon xenobiotics can also biotransform
endogenous substances, indicating an equally important regulatory role for these
enzymes. This interrelationship between different enzymes and substrates can be
illustrated by the metabolism of P-estradiol in humans, which can be biotransformed both
via sulfonation (SULTIE1, which also acts on 7-hydroxymethyl-12-dimethylbenz-
anthracene, the product of CYP450-catalyzed hydroxylation of 7, 12-dimethyldibenz-
anthracene (Glatt et al., 1995)) and glucuronidation (UGTIA1, which can also conjugate
1-naphthol (Radominska-Pandya et al., 1999)).
While these enzymes mainly represent a cellular defense mechanism against
toxicity, occasionally procarcinogenic and protoxic xenobiotics are metabolized to active
metabolites that attack macromolecules such as DNA, proteins and lipids.
In exposed organisms, metabolism is an important factor in determining the
bioaccumulation, fate, toxicokinetics, and toxicity of contaminants. The majority of the
compounds of interest to this study are derived from Phase I metabolism of
environmental pollutants. These metabolites have been shown to have toxic effects both
in vitro and in vivo, effects that can be eliminated by Phase II biotransformation (Chapter
2). In addition, contaminant exposure can result in the induction or inhibition of both
Phase I and Phase II enzymes. For example, induction of CYP 1A (e.g., by polyaromatic
hydrocarbons (PAHs) or co-planar polychlorinated biphenyls (PCBs)), CYP 2B and
CYP3A (e.g., by o-chlorine substituted PCBs) will lead to increased formation of
hydroxylated metabolites. Thus, a balance between the CYP and conjugative Phase II
enzymes, sometimes directly mediated by the xenobiotic substrates and/or their
metabolites, is responsible for either the detoxification or the accumulation of toxic
metabolites in the body. The final removal of these metabolites from the cell is brought
about by several different groups of membrane proteins (e.g., organic anion transport
protein (OATP), multidrug-resistance associated protein (MRP)), a process sometimes
referred to as Phase III biotransformation (Figure 1-1).
\ cytosol ER membrane ER lumen
Figure 1-1. Schematic of select xenobiotic (represented by hydroxynaphthalene)
biotransformation pathways in the mammalian cell. For abbreviations see text.
PHASE II CONJUGATION: GLUCURONIDATION AND SULFONATION
Biotransformation has been conveniently categorized into two distinct phases.
While the consecutive numbering of these processes implies a sequence, this is not
always the case and the extent of involvement of both phases in the metabolism of a
compound depends on both its chemical structure and physical properties. Phase I
biotransformation usually consists of oxidations carried out largely by CYP enzymes and
flavin monooxygenases and hydrolysis reactions executed by ester hydrolase, amidase
and epoxide hydrolase (EH). A variety of chemical moieties can be conjugated to suitable
acceptor groups on xenobiotics as part of Phase II biotransformation, including
glucuronic acid (UGT), sulfonic acid (SULT), glutathione (GST), amino acids, and an
acetyl group (N-acetyltransferase) .
With the exception of acetylation, methylation and fatty acid conjugation, the
strategy of Phase II biotransformation is to convert a xenobiotic to a more hydrophilic
form via the attachment of a chemical moiety which is ionizable at physiological pH. The
resulting anionic conjugate is then readily excreted in bile, feces, or urine, and is
generally unable to undergo passive penetration of cell membranes. This metabolic
transformation also results in reduced affinity of the compound for its cellular target.
Enterohepatic recycling may result in the hydrolysis of biliary excreted conjugates and
the regeneration of the parent compound, which is then subj ect again to
biotransformation after being reabsorbed through the gut mucosa. In a few cases, the
conjugate is pharmacologically active, as in the case of morphine-6-glucuronide
(Yoshimura et al., 1973) and minoxidil sulfate (Buhl et al., 1990).
The moieties attached to the xenobiotic in the case of sulfonation and
glucuronidation are a sulfonate group (pKa 2) or glucuronic acid (pKa 4-5). The co-
substrates which supply these highly polar species are, respectively, 3'-phosphoadenosyl-
5'-phosphosulfate (PAPS) and uridine 5' -diphosphoglucuronic acid (UDPGA) (Figure 2-
1). The mechanism of both reactions, which occurs as a ternary complex, is a SN2
reaction, the deprotonated acceptor group of the substrate attacking the sulfur in the
phosphosulfate bond of PAPS, or the C1 of the pyranose ring to which UDP is attached in
an ot-glycosidic bond in the case of UDPGA. The resulting conjugates are then released.
PAP and UDP also leave the enzyme's active site and are subsequently regenerated.
There may be competition for the same acceptor group, especially for phenols.
Other acceptor groups that can be conjugated by both processes include alcohols,
aromatic amines and thiols. Glucuronidation is also active on other functional groups,
including carboxylic acids, hydroxylamines, aliphatic amines, sulfonamides and the C2 Of
1 ,3 -di carb onyl compounds. SULTs are generally high-affinity, low-capacity
biotransformation enzymes that operate effectively at low substrate concentrations. Thus,
typical Knas for the sulfonation of xenobiotic substrates are usually significantly lower
than Knas for the same substrates undergoing biotransformation by the low-affinity, high-
capacity UGTs. For example, kinetic parameters for the sulfonation and glucuronidation
of the antimicrobial agent triclosan in human liver are Km values of 8.5 and 107 C1M and
Vmax of 96 and 739 pmol/min/mg protein respectively (Wang et al., 2004).
o P P O
~UG T (ER)
Figure 2-1. Structure of the co-substrates PAPS and UDPGA (transferred moieties shown
in bold) and the formation of the polar sulfonate and glucuronide conjugates,
shown here competing for the same substrate.
The primary sequence of human UGTs ranges from 529 to 534 amino acids in
length (Tukey and Strassburg 2000). These 50-56 kDa proteins reside in the endoplasmic
reticulum, whereby the amino terminus and around 95% of the subsequent residues are
located in the lumen. A 17-amino acid-long transmembrane segment connects the
lumenal part of the enzyme with the short (19-24 residues) carboxyl-terminus located in
the cytosol (Figure 2-2). The active enzyme probably consists of dimers, linked together
at the C-terminus (Meech and Mackenzie 1997). The existence of tetramers for the
formation of the diglucuronide of B[a]P-3,6-diphenol has been suggested (Gschaidmeier
and Bock 1994).
Figure 2-2. Proposed structure of UGT, based on amino acid sequence
Based on evolutionary divergence, mammalian UGTs have been classified into four
distinct families (Mackenzie et al., 2005): family 1, which includes bilirubin, thyroxine
and phenol UGTs; family 2, which includes steroid UGTs; family 3, which includes
UGTs whose substrate specificity is, as yet, unknown (Mackenzie et al., 1997); family 8,
represented by UGT8Al which utilizes UDP-galactose as the sugar donor (Ichikawa et
al., 1996). Although the liver is the major site of glucuronidation in the living organism,
several other tissues have been shown to express UGTs. The small intestine appears to be
an equally important site of glucuronidation, particularly for ingested xenobiotics. In
addition, expression of some UGT isoforms is tissue-specific (Table 2-1).
The nine family 1 UGT isoforms (UGT1) are all encoded by one gene that has
multiple unique exons located upstream of four common exons on human chromosome
2q37 (Figure 2-3). The isoforms are generated by differential splicing of one unique first
exon (which encodes two-thirds of the lumenal domain, starting from the N-terminus,
288 amino acids long) to the four common exons (exons 2-5, which encode the remainder
of the lumenal domain, the transmembrane domain and the cytosolic tail, 246 amino
acids long). Due to this unusual gene structure and splicing mechanism, the UGT1
isoforms have variable amino-terminal halves and identical carboxyl-terminal halves.
While the first exon determines substrate specificity, the common exons specify the
interaction with UDPGA (Ritter et al., 1992; Gong et al., 2001). Thus, the major bilirubin
UGT (UGTIAl) of humans, rats and other species is encoded by exon 1 and the adjacent
4 common exons. The phenol UGT (UGTIA6) is encoded by exon 6 and the 4 common
The human UGT2 gene family includes three members of the UGT2A subfamily
and twelve members of the UGT2B subfamily (Mackenzie et al., 2005). The UGT2
proteins are encoded by separate genes consisting of six exons located on human
chromosome 4ql3. The region of the protein encoded by exons 1 and 2 is equivalent to
that encoded by the unique exons 1 of the UGT1 isoforms, and the subsequent
intron/exon boundaries are in corresponding positions in both gene families. Similar to
the UGTIA enzymes, the UGT2Al and 2A2 proteins have identical C-termini and
different N-termini that arise due to differential splicing of the first exon (Figure 2-4). By
contrast, the UGT2A3 gene comprises six exons that are not shared with the other two.
Table 2-1. Expression of human UGT mRNA in various tissues
UGT Liver Intestine Esophagus Kidney Brain Prostate Other tissues
1A3 J J Jb
1A6 J J Jb J testis, ovary
a Tukey and Strassburg 2000; King et al., 2000; Lin and Wong 2002; Wells et al., 2004
b only a third of the population expresses these isoforms in gastric epithelium (Strassburg
et al., 1998)
c expressed in bile ducts
Exons 1 C~*ommon
1A12plA11p 1A8 1A10 1A13p 1A9 1A7 1A6 1A51A41A3 1A~plA1
300 kb 218 kb 95 kb
Figure 2-3. Complete human UGT1 complex locus represented as an array of 13 linearly
arranged first exons.
Each first exon, except for the defective UGTIAl2p and UGTIAl3p pseudo
ones, contains a 5'proximal TATA box element (bent arrow) that allows for
the independent initiation of RNA polymerase activity that generates a series
of overlapping RNA transcripts (Adapted from Gong et al., 2001).
2B29p 2B17p 2815 2810 2A3 2B27p 2B26p 2B7 2811 2B28 2B25P 2B24P 2B4 2A1/2
I UI I UI I UI U U U U UI U U 1
2A1 2A2 2 3 4 5 6
1 1 11 I l
Figure 2-4. The human UGT2 family.
Each gene (not drawn to scale), consisting of six exons, is represented by a
white rectangle, except for '2A1/2', which represents seven exons (1 unique
first exon and shared exons 2-6). Adapted from Mackenzie et al. (2005).
Sulfotransferases can be either membrane-bound in the Golgi or in the cytosol.
While the membrane-bound SULTs sulfonate large molecules such as
glucosaminylglycans, the cytosolic enzymes are involved in the inactivation of
endogenous signal molecules (steroids, thyroid hormones, neurotransmitters) and the
Each cytosolic SULT is a single a/p globular protein with a characteristic five-
stranded parallel sheet, with a-helices flanking each sheet. The active enzyme is a
homodimer, with each polypeptide chain having a MW of about 35,000. Kakuta et al.
(1997) were the first group to solve the first X-ray structure for the SULT family. Mouse
estrogen sulfotransferase (mEST) was shown completed with PAP and the substrate
estradiol (E2). The binding of estradiol to human SULTIAl has also been demonstrated
(Gamage et al., 2005). Both PAPS- and substrate-binding sites are located deep in the
hydrophobic substrate pocket. The structures of four human cytosolic enzymes have also
been elucidated : SULT 1Al (Gamage et al., 2003), dop ami ne/c atechol ami ne
sulfotransferase (SULTIA3) (Bidwell et al., 1999; Dajani et al., 1999), hydroxysteroid
sulfotransferase (SULT2Al; hHST) (Pedersen et al., 2000), and estrogen sulfotransferase
(SULTIE1; hEST) (Pedersen et al., 2002).
Five SULT gene families have been identified in mammals (SULTsl-5). While
SULT enzymes have different substrate specificities, the repertoire of suitable substrates
is so broad that it is not uncommon that one substrate is biotransformed by more than one
enzyme. SULTs are distributed in a wide variety of tissues (Table 2-2). In humans, liver
cytosol has been shown to contain mostly SULTslA1, 2A1, and 1E1, with lesser amounts
a reviewed by Glatt 2002.
b mRNA of fetal tissues
Using 3 -hydroxy-benzo(a)pyrene (3-OH-B[a]P) and 9-OH-B[a]P, the existence of
multiple SULT isoforms in channel catfish liver and intestine, including a 3-
methylcholanthrene-inducible form of phenol-SULT in liver, has been established
(Gaworecki et al., 2004; James et al., 2001). The phenol-SULT in catfish liver and
of SULTs 1A2, 1B1, 1El and 2Al. While SULTIAl and SULTIE1 are responsible for
most of the phenol and estrogen SULT hepatic activity respectively, SULT2Al
(hydroxysteroid SULT) shows greater affinity for alcohols and benzylic alcohols (Mulder
and Jakoby, 1990; Glatt, 2002).
Table 2-2. Tissue distribution of SULTs (cDNA and mRNA) in humans
SULT Liver Intestine Esophagus Kidney Brain Lung Other tissues
J J Spleen, kidney,
J Jb Ovary, spinal
J J Thyroid gland,
Jb J Jb Endometrium,
J J Placenta,
intestine has been isolated as a 41,000 Da protein. A second protein with a molecular
weight of 31,000 Da, isolated from liver, has not been identified to date. Interestingly
enough, SULT activity with phenolic substrates is higher in intestine than liver (Tong and
James 2000). Other hepatic SULTs isolated and characterized from fish include
petromyzonol SULT from lamprey (Petromyzon marinus) larva (which displays 40%
homology with mammalian SULT2Bla, or cholesterol SULT) and a bile steroid SULT
from the shark Heterodontus portusjacksoni (Venkatachalam et al., 2004; Macrides et al.,
SULFONATION OF XENOBIOTICS BY POLAR BEAR LIVER
The lipophilicity and inherent chemical stability of persistent organic pollutants
(POPs) renders them excellent candidates for absorption through biological membranes
as well as accumulation in both organisms and their environment. Many POPs have been
shown to biomagnify in food webs to potentially toxic levels in top predators such as the
polar bear (Grsus maritimus), whose diet mainly consists of ringed seal (Phoca hispida)
blubber (Kucklick et al., 2002).
Since the sulfonation of xenobiotics has never been studied in the polar bear, the
obj ective of this study was to investigate the efficiency of this route of detoxification on a
select group of known environmental pollutants: 4'-hydroxy-3,3',4,5 '-
tetrachlorobiphenyl (4'OH-PCB79), 4'-hydroxy-2,3,3 ',4,5,5 '-hexachlorobiphenyl (4'-
OH-PCBl159), 4'-hydroxy-2,3,3 ',5,5 ',6-hexachlorobiphenyl (4'-OH-PCBl165),
pentachlorophenol (PCP), tris(4-chlorophenyl)-methanol (TCPM), 2-(4-methoxyphenyl)-
2-(4-hydroxyphenyl)- 1,1,1 -trichloroethane (OHMXC), 3 -hydroxybenzo(a)pyrene (3-OH-
B [a]P), triclosan (2,4,4'-trichloro-2 '-hydroxydiphenyl ether) (Figure 3-1). The OH-PCBs
were named as PCB metabolites, according to the convention suggested by Maervoet et
Polychlorinated biphenylols (OH-PCBs), major biotransformation products of
PCBs (James, 2001), have been shown to be present in relatively high concentrations in
polar bears (Sandau and Norstrom 1998; Sandau et al., 2000). The abundance of these
hydroxylated metabolites may be due to CYP induction (Letcher et al., 1996), inefficient
Figure 3-1. Structures of sulfonation substrates investigated in this study.
(1) 3-OH-B[a]P; (2) triclosan; (3) 4'-OH-PCB79; (4) 4'-OH-PCBl159; (5) 4'-OH-
PCBl65; (6) OHMXC; (7) TCPM; (8) PCP. Full names of each compound are
given in the text.
Phase II detoxication, and inhibition of their own biotransformation. The 4'-OH-PCB79
(an oxidation product of PCB congener 77) is a potent inhibitor of the sulfonation of
several substrates, including 3-OH-B[a]P in channel catfish intestine and human liver
(van den Hurk et al., 2002, Wang et al., 2005), 4-nitrophenol by human SULTIAl (Wang
et al., 2006), 3,5-diiodothyronine (T2) in rat liver (Schuur et al., 1998), and estradiol by
human SULTIE1 (Kester et al., 2000). Both 4'-OH-PCBl59 and 4'-OH-PCBl65 have
been shown to inhibit the sulfonation of 3-OH-B[a]P and 4-nitrophenol by human SULT
(Wang et al., 2005, 2006). Another compound detected in polar bears is PCP (Sandau and
Norstrom 1998), a commonly used wood preservative that has been implicated in thyroid
hormone disruption in Arctic Inuit populations (Sandau et al., 2002). TCPM is a globally
distributed organochlorine compound of uncertain origin, which was reported in human
adipose tissue (Minh et al., 2000). Polar bear liver contains 4000-6800 ng/g lipid weight
TCPM, the highest levels recorded for this compound in all species studied (Jarman et al.,
1992). TCPM is a potent androgen receptor antagonist in vitro (Schrader and Cooke
2002). OHMXC, formed by demethylation of the organochlorine pesticide methoxychlor,
is an estrogen receptor (ER) oc agonist, an ERP antagonist and an androgen receptor
antagonist (Gaido et al., 2000). The ubiquitous environmental pollutant benzo[a]pyrene is
mainly metabolized to 3-OH-B[a]P, a procarcinogen that can be eliminated via
sulfonation (Tong and James 2000). Together with its 7,8-dihydrodiol-9, 10-oxide and
7,8-oxide metabolites, 3-OH-B[a]P can form adducts with macromolecules and initiate
carcinogenesis (Ribeiro et al., 1986). Triclosan is an antimicrobial agent that has been
detected in human plasma and breast milk (Adolfsson-Erici et al., 2002). In vitro studies
have shown that triclosan inhibits various biotransformation enzymes, including SULT
and UDP-glucuronosyltransferases (UGT) (Wang et al., 2004).
The fact that 3-OH-B[a]P, triclosan, OHMXC, 4'-OH-PCB79, 4'-OH-PCBl59 and
4'-OH-PCBl165 have not been reported as environmental contaminants in polar bears to
date may be due to non-significant levels in the Arctic environment or efficient
metabolism via, for example, sulfonation. On the other hand, the presence of PCP and,
particularly, high amounts of TCPM in these Arctic carnivores, may indicate poor
sulfonation of these substrates. The polychlorobiphenylols 4'-OH-PCBl59 and 4'-OH-
PCBl165 are of interest since though they have not been detected in polar bears, they are
structurally similar to 4'-OH-PCBl72, one of the major OH-PCBs found in polar bear
plasma (Sandau et al., 2000). It is thus possible that these compounds are sulfonated with
similar efficiencies. The other major Phase II biotransformation pathway for the above-
mentioned compounds is glucuronidation. Polar bear liver efficiently glucuronidated 3-
OH-B[a]P and several OH-PCBs (Sacco and James 2004).
Sulfonation occurring in polar bear liver is an inefficient route of detoxification for
a structurally diverse group of environmental contaminants.
Unlabeled PAPS was purchased from the Dayton Research Institute (Dayton, OH).
Uridine 5' -diphosphoglucuronic acid (UDPGA) was obtained from Sigma (St.Louis,
MO). Radiolabeled [35S]PAPS (1.82 or 3.56 Ci/mmol) was obtained from Perkin-Elmer
Life Sciences, Inc. (Boston, MA). The benzo[a]pyrene metabolites 3-OH-B[a]P, B[a]P-3-
O-sulfate and B [a]P-3-O-glucuronide were supplied by the Midwest Research Institute
(Kansas City, MO), through contact with the Chemical Carcinogen Reference Standard
Repository of the National Cancer Institute. Dr. L.W.Robertson, U of Iowa, kindly
donated the 4'-OH-PCB79, and 4'-OH-PCBl59 and 4'-OH-PCBl65 were purchased
from AccuStandard, Inc. (New Haven, CT). PCP from Fluka Chemical (Milwaukee, WI)
was used to prepare the water-soluble sodium salt (Meerman et al., 1983). Triclosan and
sulfatase (Type VI from Aerobacter, S1629) were purchased from Sigma (St.Louis, MO),
while methoxychlor and TCPM were purchased from ICN Biomedical (Aurora, OH) and
Lancaster Synthesis, Inc. (Pelham, NH), respectively. The OHMXC was prepared by the
demethylation of methoxychlor and purified by recrystallization (Hu and Kupfer 2002).
Tetrabutyl ammonium hydrogen sulfate (PIC-A low UV reagent) was from Waters
Corporation, Milford, MA. Other reagents were the highest grade available from Fisher
Scientific (Atlanta, GA) and Sigma.
Animals. The samples used in this study were a kind donation from Dr. S. Bandiera (U
British Columbia) and Dr. R. Letcher (Environment Canada). They were derived from
the distal portion of the right lobe of livers of three adult male bears G, K and X. Bears G
and K were collected as part of a legally-controlled hunt by Inuit in the Canadian Arctic
in April 1993 near Resolute Bay, Northwest Territories, while bear X was collected in
November 1993 near Churchill, Manitoba, just after the fasting period. Liver samples
were removed within 10-15 minutes after death, cut into small pieces and frozen at -
196oC in liquid N2. The samples were subsequently stored at -80oC.
Cytosol and Microsomes Preparation. Prior to homogenization, the frozen polar
bear liver samples (~2g) were gradually thawed in a few ml of homogenizing buffer.
Homogenizing buffer consisted of 1.15% KC1, 0.05 M K3PO4 pH 7.4, and 0.2 mM
phenylmethylsulfonyl fluoride, added from concentrated ethanol solution just before use.
Resuspension buffer consisted of 0.25 M sucrose, 0.01 M Hepes pH 7.4, 5% glycerol, 0. 1
mM dithiothreitol, 0.1 mM ethylene diamine tetra-acetic acid and 0.1 mM phenylmethyl
sulfonyl fluoride. The liver was placed in a volume of fresh ice-cold buffer equal to 4
times the weight of the liver sample. The cytosol and microsomal fractions were obtained
using a procedure described previously (Wang et al., 2004). Microsomal and cytosolic
protein contents were measured by the Lowry assay, using bovine serum albumin (BSA)
A. Fluorometric method. The activity was measured on the basis that at alkaline pH, the
benzo[a]pyrene-3 -O-sulfate has different wavelength optima for fluorescence excitation
and emission (294/415 nm) from the benzo[a]pyrene-3 -O-phenolate anion (3 90/545 nm)
(James et al., 1997). Saturating concentrations of PAPS were determined by performing
the assay at 1 C1M 3-OH-B[a]P. The reaction mixture for detecting the sulfation of 3-OH-
BaP by polar bear liver cytosol consisted of 0.1 M Tris-Cl buffer (pH 7.6), 0.4% BSA,
PAPS (0.02 mM), 25 Clg polar bear hepatic cytosolic protein, and 3-OH-B[a]P (0.05-25
CIM) in a total reaction volume of 1.0 mL. SULT activity (pmol/min/mg) was calculated
from a standard curve prepared with B[a]P-3-O-sulfate standards. Substrate consumption
did not exceed 10%.
B. Radiochemical extraction method. This method, based on Wang and co-workers
(2004), was employed in the study of the sulfonation of 4'-OH-PCB79, 4'-OH-PCBl159,
4'-OH-PCBl65, triclosan, PCP, TCPM and OHMXC. Cytosolic protein concentrations
and incubation time were optimized for every test substrate to ensure that the reaction
was linear during the incubation period. Substrate consumption did not exceed 5%. The
incubation mixture consisted of 0.1 M Tris-Cl buffer (pH 7.0), 0.4% BSA in water, 20
CIM PAPS (10% labelled with 35S), 0.1 mg polar bear hepatic cytosolic protein, or 0.005
mg in the case of 4'-OH-PCB79 and triclosan, and substrate in a total reaction volume of
0.1 mL, or 0.5 mL in the case of TCPM. The OH-PCBs, triclosan and OHMXC were
added to tubes from methanol solutions, and the methanol was removed under N2 priOr to
addition of other components. The TCPM was dissolved in DMSO, the solvent being
present at a concentration not exceeding 1% in the final assay volume. Control
determinations utilizing 1% DMSO had no inhibitory effect on sulfonation. Aqueous
solutions of sodium pentachlorophenolate were utilized in the case of PCP. Tubes
containing all components except the co-substrate were placed in a water bath at 37oC
and PAPS was added to initiate the reaction. Incubation times were 5 min (TCPM), 20
min (4'-OH-PCB79, triclosan), 30 min (PCP) and 40 min (OHMXC, 4'-OH-PCBl59, 4'-
OH-PCBl165). The incubation was terminated by the addition of an equal volume of a 1:1
mixture of 2.5% acetic acid and PIC-A and water. The sulfated product was extracted
with 3.0 mL ethyl acetate as described previously (Wang et al., 2004) and the phases
were separated by centrifugation. Duplicate portions of the ethyl acetate phase were
counted for quantitation of sulfate conjugates.
C. Radiochemical TLC method. Since the ethyl acetate phase contains sulfate
conjugates formed from both the substrate of interest and substrates already present in
polar bear liver, TLC was used to quantify substrate sulfation in cases where SULT
activity was similar in samples and substrate blanks. After evaporating 2 ml of ethyl
acetate extract from the SULT assay under N2, the solutes were reconstituted in 40 C1L
methanol. For 4'-OH-PCBl59, 4'-OH-PCBl65, PCP and OHMXC, the substrate
conjugates were separated on RP-18F254s TOVeTSe phase TLC plates with fluorescent
indicator (Merck, Darmstadt, Germany) using methanol:water (80:20). For TCPM,
Whatman KClsF reverse phase 200 Clm TLC plates with fluorescent indicator in
conjunction with a developing solvent system consisting of methanol:water:0.28 M PIC-
A (40:60:1.9 by volume) were employed. Electronic autoradiography (Packard Instant
Imager, Meriden, CT) was used to identify and quantify the radioactive bands separated
on the TLC plate. The counts representing the substrate sulfate conjugate products were
expressed as a fraction of the total radioactivity determined by scintillation counting, thus
enabling the radioactivity due to the substrate conjugate to be accurately determined.
The identity of the conjugate of TCPM as a sulfate ester was verified by studying
its sensitivity to sulfatase. Polar bear cytosol (0.5 mg) was incubated for 75 minutes with
or without 200 CLM TCPM. The incubation was terminated, and the product extracted into
ethyl acetate as above. The ethyl acetate was evaporated to dryness and dissolved in 0.25
mL of Tris buffer, pH 7.5, containing 0 or 0.08 units of sulfatase. Following an overnight
incubation at 35oC, the reaction was stopped by the addition of methanol and the tubes
were centrifuged. The supernatants were evaporated to dryness, reconstituted in methanol
and analyzed by TLC as described above.
UDP-Glucuronosyltransferase Assay. The reaction mixture for detecting the
glucuronidation of 3-OH-B[a]P by polar bear liver microsomes consisted of 0. 1 M Tris-
HCI buffer (pH 7.6), 5 mM MgCl2, 0.5% Brij-58, UDPGA (4 mM), 5 Clg polar bear
hepatic microsomal protein, and 3-OH-B[a]P in a total reaction volume of 500 CIL. The
substrate, 3-OH-B[a]P in methanol, was blown dry under N2 in the dark in a tube to
which, after complete evaporation, a premixed solution of microsomal protein and Brij-
58 (in a 5:1 ratio) was added, vortexed, and left for 30 minutes on ice. Subsequently, the
buffer and water were added in that order and vortex-mixed. Immediately preceding a 20-
minute incubation at 37oC, UDPGA was added to initiate the reaction. The reaction was
terminated by the addition of 2 mL ice-cold methanol. Precipitated protein was pelleted
by centrifugation at 2000 rpm for 10 minutes. The supernatant, 2 mL, was then mixed
with 0.5 mL NaOH (lN) and the fluorescence of B[a]P-3-glucuronic acid measured at
excitation/emission wavelengths of 300/421 nm (Singh & Wiebel, 1979). The activity of
UGT (nmol/min/mg) was then determined.
Preliminary studies established the conditions for linearity of reaction with respect
to time, protein and detergent concentrations, at the same time ensuring that substrate
consumption did not exceed 10%. The apparent Km for UDPGA was determined by
performing experiments at a fixed concentration of 3-OH-B[a]P (10 C1M). Saturating
UDPGA concentrations were used in order to determine 3-OH-B[a]P glucuronidation
Kinetic Analysis. Duplicate values for the rate of conjugate formation at each substrate
concentration were used to calculate kinetic parameters using Prism v 4.0 (GraphPad
Software, Inc., San Diego, CA). Equations used to fit the data were the Michaelis-Menten
hyperbola for one-site binding (eq. 1), the Hill plot (eq. 2), substrate inhibition for one-
site binding (eq. 3) (Houston and Kenworthy 2000), and partial substrate inhibition due to
binding at an allosteric site (eq. 4) (Zhang et al., 1998).
v = Vmax[S] / (Km + [S]) (1)
v = Vmax[S]h / (S50h + [S]h) (2)
v = Vmax[S] / (Km + [S] + ([S]2/K,)) (3)
v = Vmaxy(1 + (Vmax2[S]/Vmax;K,)) / (1 + Km/[S] + [S]/K,) (4)
Values for K, and Vmax derived from equation 1 were used as initial values in the
fitting of data to equations 3 and 4. Eadie-Hofstee plots were used in order to analyze the
biphasic kinetics observed.
Sulfonation and glucuronidation of 3-OH-B[a]P
Optimum conditions for sulfonation were 10 minutes incubation time and 25 Cpg
cytosolic protein. A concentration of 0.02 mM PAPS provided saturating concentrations
of the co-substrate and enabled kinetic parameters at 1.0 CLM 3-OH-B [a]P to be calculated
by the application of eq. 1 (Table 3-la). The data for the sulfonation of 3-OH-B[a]P was
fit to a two-substrate model (eq. 3), whereby the binding of a second substrate to the
enzyme is responsible for the steep decline in enzyme activity at concentrations
exceeding 1 CLM (Figure 3-2a). Initial estimates of Vmaxi and Km were provided by the
initial data obtained at low [S] (non-inhibitory), while Vmax2 WAS constrained to 65 + 20
pmol/min/mg, which is slightly below the plateau in Figure 3-2a.
The kinetic scheme (Figure 3-2b) illustrates the proposed partial substrate
inhibition process, which assumes that substrate binding is at equilibrium, which is
probable due to the low turnover rate of SULT. The best fit of the data was provided by a
K, of 1.0 + 0.1 CLM. Binding of the second substrate molecule results in a tenfold
reduction in the rate of sulfonate formation.
a constrained variables to obtain best fit
b ValUeS for high-affinity component
c values for low-affinity component
Table 3-1. Estimated kinetic parameters (Mean f SD) for (a) sulfonation and (b) glucuronidation of 3-OH-B[a]P by polar bear liver
cytosol and microsomes. Values were calculated as described in the Methodology.
Substrate Vmax1 (app)
3-OH-B[a]P 500 f 8
PAPS 162 f 35
Sub state Vmax (app)
0.41 f 0.03
0.22 f 0.07
Vmaxl/Km Vmax2 (app) a
1220 f 70 65.0 f 20.0
1.01 & 0.10
66.2 f 26.8
3.00 f 1.18
1.53 f 0.56b, 1.47 f 0.48c
1.4 f 0.2
42.9 f 2.5b, 200 f 68c
1900 f 544
3 -OH-B [a]P
300- PB G
I PB K
liii v PB X
0 10 20 30
[3-OH-B(a)P] ( CMh)
PAPS K,, 10.4 IIM) PAPS V,,avy (471.8 pmol/min/mg)
1K, (1.2 p M)
E PPSV,,,, (45 .0 pmnol'mninl'm g)
Figure 3-2. Sulfonation of 3-OH-B[a]P at PAPS = 0.02 mM.
A. Each data point represents the average of duplicate assays for each bear,
while the error bars represent the standard deviation. The line represents the
best fit to the data of equation (3). B) Kinetic model for partial substrate
inhibition of SULT by 3-OH-B[a]P, after Zhang et al. (1998). E refers to
Optimum conditions for the glucuronidation of 3-OH-B[a]P by polar bear
microsomes were found to be 5 Cpg microsomal protein and a 20-minute incubation. A
concentration of 4 mM UDPGA was determined to be suitable for providing saturating
concentrations of the co-substrate. The binding of UDPGA to UGT at 10 CLM 3-OH-
B[a]P was shown to be biphasic, with a fivefold reduction in affinity at higher UDPGA
concentrations (Table 3-1b). The kinetic parameters for the co-substrate were calculated
by deconvoluting the curvilinear data in the Eadie-Hofstee plot (Figure 3-3). In the
presence of 4 mM UDPGA, the formation of B[a]P-3-O-glucuronide followed Michaelis-
Menten kinetics (Table 3-1b).
3- A high-affinity
0 10 20J 30 40
Figure 3-3. Eadie-Hofstee plot for the glucuronidation of 10 CLM 3-OH-B[a]P, over a
UJDPGA concentration range of 5-3000 CLM.
Each data point represents the average of duplicate assays for all bears, while
the error bars represent the standard deviation.
Sulfonation of other substrates
Triclosan sulfate was formed rapidly, with the overall kinetics conforming to a
hyperbolic curve (eq. 1) (Table 3-2). Substrate inhibition was observed for 4'-OH-PCB79
(Figure 3-4), with the data fitting equation (3). The value of K, that gave the best fit was
217 f 25 CLM (Table 3-2). Sulfate conjugation of 4'-OH-PCBl59 and 4'-OH-PCBl65,
which proceeded via Michaelis-Menten kinetics, was, respectively, 11 and 5 times less
efficient than the sulfonation of 4'-OH-PCB79 (Table 3-2). At a concentration of 10 CLM,
4'-OH-PCBl165 was observed to inhibit sulfonation of substrates already present in polar
bear liver cytosol by 60%.
Table 3-2. Kinetic parameters (Mean f SD) for the sulfonation of various xenobiotics by
polar bear liver cytosol, listed in order of decreasing enzymatic efficiency.
All data fit equation (1), except for 4'-OH-PCB79 and PCP, which fit
equations (3) and (2) respectively (see Methodology for equations).
Substrate Vmax Km Vmax / Km Ki
(pmol/min/mg) (yM) (pL/min/mg) (yM)
triclosan 1008 f 135 11 f 2
4'-OH-PCB79 372 f 38 123 f 20
OHMXC 51.1 f 7.8 67 f 4
4'-OH-PCBl165 8.6 f 2.0 17 f 7
TCPM 62.0 f 11.2 144 f 36
4'-OH-PCBl159 14.8 f 2.3 60 f 21
PCP 13.8 f 1.2 72 f 14b
aK, for bears G, K and X were 240, 220 and 190 LM
constrained to obtain the best fit for the data
bSso; h = 2.0 f 0.4
90.8 f 6.8
3.1 f 0.3 217 f 25a
0.8 f 0.1
0.56 f 0.17
0.44 f 0.06
0.28 f 0.12
0.20 f 0.05
respectively. These values were
0 100 200 300 4100 500 600 700 800
Figure 3-4. Sulfonation of 4'-OH-PCB79, PAPS = 0.02 mM.
Each data point represents the average of duplicate assays for each bear, while
the error bars represent the standard deviation. The line represents the best fit
to equation (4) for 4'-OH-PCB79.
Due to variable rates of sulfonation of these unknown substrates, autoradiographic
counts corresponding to the OHMXC-O-sulfate band were used to correct the activities
calculated from the scintillation counter data (Figure 3-5). This enabled the transformed
data to be fit into a Michaelis-Menten model (Table 3-2). The autoradiograms obtained
showed that increasing concentrations of OHMXC resulted in decreased counts for the
unknown sulfate conjugates (Figure 3-5). Sulfonation of the unknown substrates in polar
bear cytosol was reduced by half at OHMXC concentrations < 20 LM.
0 20 50 '100 200 300
Figure 3-5. Autoradiogram showing the reverse-phase TLC separation of sulfonation
products of OHMXC.
Incubations were carried out with the indicated concentrations of OHMXC.
The arrow indicates the sulfate conjugate of the OHMXC, while other bands
represent unidentified sulfate conjugates formed from endobiotics or other
xenobiotics in polar bear liver cytosol.
The total TCPM sulfate conjugate production formed after 5 minutes under initial
rate conditions did not exceed 30 pmol. TLC, followed by autoradiography, was thus
used to distinguish the TCPM-sulfate band (Rf 0.54) from other sulfate conjugates (Rf
0.05 and 0.72) originating from compounds in the polar bear liver cytosol (Figure 3-6).
The data obtained followed hyperbolic kinetics (Table 3-2). Even though the TLC
from the kinetic experiments showed a TCPM concentration-dependent increase of the
band corresponding to the purported TCPM-sulfate, and this band was absent in the
substrate blank, the fact remained that we were apparently looking at the only instance
ever reported of a successful sulfonation of an acyclic tertiary alcohol.
PC P100 CO C100 HO H100
Figure 3-6. Autoradiogram showing the reverse-phase TLC separation of sulfonation
products from incubations with TCPM using polar bear (P), channel catfish
(C), and human (H) liver cytosol in the absence of (0), and presence of 100
CLM TCPM (100).
The arrow indicates the sulfate conjugate of the substrate, while other bands
represent unidentified sulfate conjugates formed from endobiotics or other
xenobiotics in liver cytosol.
Thus, additional experiments were performed to verify the identity of this
conjugate. The purity of the TCPM was tested in the event that the additional band was
due to an impurity in the substrate. However, the substrate used was found to be free of
contaminants by HPLC (C18 reverse phase column, with detection at 268 and 220 nm,
using 90% methanol in water and a flow rate of 1 mL/min). A single peak was recorded
at 7.3 minutes. Another experiment involved a 60-minute incubation performed with 100
CLM TCPM and 0.1 mg cytosolic protein from polar bear, channel catfish and human
liver. For each of the three species, we detected a conjugate at Rf = 0.54. The substrate
blanks showed no band at the same position (Figure 3-6). The TCPM sulfate conjugate
from polar bear could be hydrolyzed by sulfatase (Figure 3-7), providing further evidence
of the sulfonation of this alcohol.
Figure 3-7. Autoradiogram showing the reverse-phase TLC separation of sulfonation
products of TCPM and the effect of sulfatase treatment.
A, incubation in the absence of TCPM (lane 1), and following treatment with
sulfatase (lane 2). B, incubation with 200 CLM TCPM (lane 3), and following
treatment with sulfatase (lane 4). The arrow indicates the sulfate conjugate of
the TCPM, while other bands represent unidentified sulfate conjugates formed
from endobiotics or other xenobiotics in polar bear liver cytosol.
Inhibition of sulfonation of substrates already present in the polar bear liver was
noted upon adding 1 CLM PCP (Figure 3-8). The data for PCP sulfonation fitted the
nonlinear Hill plot (eq. 2) (Table 3-2).
0 1 2 5 10 20 50 75 100
Figure 3-8. Autoradiogram showing reverse-phase TLC separation of sulfonation
products from the study of PCP kinetics.
The arrow indicates the sulfate conjugate of PCP, while other bands represent
unidentified sulfate conjugates formed from endobiotics or other xenobiotics
in polar bear liver cytosol.
The sulfonation of hydroxylated metabolites of benzo[a]pyrene has been reported
in various species, including fish (James et al., 2001) and humans (Wang et al., 2004).
Benzo[a]pyrene-3 -glucuronide has been shown to be produced by fish (James et al.,
1997), rats (Lilienblum et al., 1987) and humans (Wang et al., 2004). There are,
however, few studies investigating the kinetics of these conjugation reactions.
Glucuronidation of 3-OH-B[a]P was more efficient in polar bear liver than in human liver
or catfish intestine. On the other hand, the efficiency of sulfonation was similar to that
shown in human liver but around three times less than in catfish intestine (Wang et al.,
2004, James et al., 2001). From the limited comparative data available, it can be surmised
that, in general, polar bear liver is an important site of 3-OH-B[a]P detoxication,
particularly with respect to glucuronidation.
Substrate inhibition for the sulfonation of 3-OH-B[a]P has been observed at
relatively low concentrations of the xenobiotic in other species such as catfish and human
(Tong and James 2000, Wang et al., 2005). Data from the polar bear sulfonation assay
fitted a two-substrate model developed for the sulfonation of 17P-estradiol by SULTIE
(Zhang et al., 1998). This model was also used to explain the sulfonation profie observed
for the biotransformation of 1-hydroxypyrene, a compound structurally similar to 3-OH-
B[a]P, by SULTs 1Al and 1A3 (Ma et al., 2003). In the original model, SULTIE1 was
saturated with PAPS, and each of the estradiol substrate molecules bound independently
to the enzyme. The estradiol binding sites were proposed to consist of a catalytic site, and
an allosteric site that regulates turnover of the substrate (Zhang et al., 1998). The
substrate inhibition observed with polar bear liver cytosol at higher 3-OH-B[a]P
concentrations (>0.75 pLM) can thus be explained by the binding of a second substrate
molecule to an allosteric site, which leads to a two-fold decrease in affinity and an
eightfold decrease in V;;;a.
SULTs are generally high-affinity, low-capacity biotransformation enzymes that
operate effectively at low substrate concentrations. Thus, typical K;;s for the sulfonation
of xenobiotic substrates are usually significantly lower than K;;s for the same substrates
undergoing biotransformation by low-affinity, high-capacity glucuronosyltransferases
(UGTs). In polar bear liver, both pathways showed similar apparent affinities for 3-OH-
B[a]P, with K;;s of 0.4 and 1.4 CLM for sulfonation and glucuronidation respectively,
suggesting these two pathways of Phase II metabolism compete at similar 3-OH-B[a]P
concentrations. However, the apparent maximal rate of sulfonation was about 7.5 times
lower than the rate of glucuronidation.
It was previously reported that the maximum rate of glucuronidation of 3-OH-
B[a]P by polar bear liver was 1.26 nmol/min/mg, or around half the V;;;a value obtained
in this study (Sacco and James 2004). However, the preceding study utilized 0.2 mM
UJDPGA, which, as seen from Table 3-2a, is equivalent to the K,; (for UDPGA) of the
low-affinity enzyme, and thus does not represent saturating concentrations of the co-
substrate. The affinity of the enzyme for 3-OH-B[a]P did not change significantly with a
20-fold increase in UDPGA concentrations, suggesting that substrate binding is
independent of the binding of co-substrate. The binding of UDPGA was biphasic,
indicating that multiple hepatic UGTs may be responsible for the biotransformation.
Biphasic UDPGA kinetics have also been demonstrated in human liver and kidney for 1-
naphthol, morphine, and 4-methylumbelliferone (Miners et al., 1988a,b; Tsoutsikos et al.,
2004). While V;;;a was similar for both components, there was a fivefold decrease in
enzyme affinity for UDPGA as the co-substrate concentration was increased. The
involvement of at least two enzymes can be physiologically advantageous since it enables
the maintenance of a high turnover rate even as UDPGA is consumed. Although
physiological UDPGA concentrations in polar bear liver are unknown, mammalian
hepatic UDPGA has been determined to be around 200-400 CLM (Zhivkov et al., 1975,
Cappiello et al., 1991), implying that the observed nonlinear kinetics in the polar bear
may operate in vivo.
The rate of triclosan sulfonation was the highest of all the substrates studied;
apparent V;;; was twice as high as for 3-OH-B[a]P. However, the overall efficiency of
sulfonation of the hydroxylated PAH was still 13 times higher than for triclosan
sulfonation. The presence of three chlorine substituents (though none flanking the phenol
group) does not hinder the sulfonation of triclosan when compared to the 'chlorine-free'
3-OH-B[a]P. Triclosan sulfonation in polar bear liver was similar to human liver with
respect to enzyme affinity; however the maximum rate was tenfold higher in polar bears
than in humans (Wang et al., 2004). This may be one reason why triclosan has not been
detected in polar bear plasma or liver to date.
Our data fitted a model that indicates the substrate inhibition observed for 4'-OH-
PCB79 may be due to a second substrate molecule interacting with the enzyme-substrate
complex at the active site rather than an allosteric site, resulting in a dead-end complex.
Unlike 3-OH-B[a]P, sulfonation can only proceed via the single substrate-SULT
complex. Models of SULTIAl and 1A3, with two molecules of p-nitrophenol or
dopamine at the active site respectively, have been proposed as a mechanism of substrate
inhibition (Gamage et al., 2003, Barnett et al., 2004), while the crystal structure of human
EST containing bound 4,4'-OH-3,3',5,5 '-tetrachlorobiphenyl at the active site has not
provided any evidence of an allosteric site (Shevtsov et al., 2003). The slower sulfonation
of 4'-OH-PCB79 compared with 3-OH-B[a]P may result from the inductive effect of the
chlorines flanking the phenolic group rather than steric hindrance (Duffel and Jakoby,
1981). However, polar bear liver sulfonated 4'-OH-PCB79 more rapidly than the other
OH-PCB substrates studied.
The inclusion of two additional chlorine substituents on the non-phenol ring (with
respect to 4'-OH-PCB79) resulted in both 4'-OH-PCBl59 and 4'-OH-PCBl65 being
very poor substrates. Ineffieient sulfonation may be one reason why the related
compound 4'-OH-PCBl72 accumulates in polar bears. Some degree of substrate
inhibition may also be expected to contribute to this accumulation, as was observed with
Sulfonation was not an efficient pathway of OHMXC detoxification. The rate of
OHMXC-sulfonate formation was around 7 times lower than for 4'-OH-PCB79. Since
resonance delocalization of negative charge on the phenolic oxygen by the flanking
chlorines in chlorophenols may decrease Vmax by increasing the energy of the transition
state of the reaction (Duffel and Jakoby, 1981), it is possible that in the case of OHMXC
(with no chlorines flanking the phenolic group), product release, rather than sulfonate
transfer, may have been the rate-limiting step.
TCPM was a poor substrate for sulfonation, and this may be one reason why it has
been measured in such high amounts in polar bear liver. To our knowledge, sulfonation
of acyclic tertiary alcohols has not been reported in the literature. Despite the
considerable steric hindrance of three phenyl groups, the alcohol group could be
sulfonated. Although the alcohol in TCPM is not of the benzylic type, the presence of
three proximal phenyl groups may give this group some benzylic character, rendering
sulfonation of the alcohol possible. Both SULT 1El and SULT 2Al have been shown to
sulfonate benzylic alcohol groups attached to large molecules (Glatt, 2000). Sulfation of
the benzylic hydroxyl group leads to an unstable sulfate conjugate that readily degrades
to the reactive carbocation or spontaneously hydrolyzes back to the alcohol. Attempts to
recover TCPM-O-sulfonate from TLC plates resulted in recovery of TCPM from the
conjugate band, perhaps because of the conjugate' s instability.
A study of the sulfonation of PCP was complicated by the fact that it is a known
SULT inhibitor, often with K,s in the submicromolar range. In our experiments, this was
seen as a 74% decrease in formation of the unidentified sulfonate conjugates (band
shown at the solvent front in Figure 3-8) upon addition of 1 CLM PCP. Although PCP was
a strong inhibitor of SULTI1E1 (Kester et al., 2000), and has been postulated to be a dead-
end inhibitor for phenol sulfotransferases (Duffel and Jakoby, 1981), it was possible that
polar bear SULT 1A isoforms were not completely inhibited by PCP, or that other SULT
isoform(s) were responsible for the limited sulfonation activity observed. Thus, we have
shown that, in vitro at least, one mammalian species is capable of limited PCP
sulfonation. Even though the tertiary alcohol of TCPM was a poor candidate for
sulfonation, it was metabolized at twice the efficiency of PCP, which has a phenolic
group that is usually more susceptible to sulfonation. This demonstrates the extent of the
decreased nucleophilicity on the phenolic oxygen due to the resonance delocalization
afforded by the five chlorine substituents.
In summary, this study demonstrated that, in polar bear liver, 3-OH-B[a]P was a
good substrate for sulfonation and glucuronidation. Other, chlorinated, substrates were
biotransformed with less efficiency, implying that reduced rates of sulfonation may
contribute to the persistence of compounds such as hexachlorinated OH-PCBs, TCPM
and PCP in polar bear tissues.
GLUCURONIDATION OF POLYCHLORINATED BIPHENYLOLS BY CHANNEL
CATFISH LIVER AND INTESTINE
Polychlorinated biphenyls (PCBs) were extensively used as dielectrics in the mid-
twentieth century. Despite a ban on their use in the US, Europe and Japan since the mid
1970s, the chemical stability of PCBs has resulted in their persistence at all trophic levels
around the globe. Enzyme-mediated biotransformation is an important influence on PCB
persistence, and its significance in PCB toxicokinetics is dependent on congener structure
and the metabolic capacity of the organism.
Polychlorinated biphenylols (OH-PCBs) are products of CYP-dependent oxidation
of PCBs (James 2001). While OH-PCBs are more polar than their parent molecules, they
are still lipophilic enough to be orally absorbed, and distribute to several tissues (Sinjari
et al., 1998). Thus, not only have these compounds been detected in the plasma (which
represents recent dietary exposure, biotransformation, and remobilization into the
circulation) of a variety of animal species, such as polar bear (Sandau et al., 2004),
bowhead whale (Hoekstra et al., 2003), catfish (Li et al., 2003), and humans (Fangstroim
et al., 2002; Hovander et al., 2002), but also significantly, from a developmental
toxicology aspect, in fetuses and breast milk (Sandau et al., 2002; Guvenius et al., 2003).
OH-PCBs may contribute significantly to the recognized toxic effects of PCBs such
as endocrine disruption (Safe 2001; Shiraishi et al., 2003), tumor promotion (VondrBkek
et al., 2005) and neurological dysfunction (Sharma and Kodavanti 2002; Meerts et al.,
Elimination of these toxic metabolites via Phase II conjugation reactions, such as
glucuronidation and sulfonation, are thus important routes of detoxification. In view of
the persistence of certain OH-PCBs, it is surprising that only a few studies have
attempted to investigate the biotransformation of these compounds in animals or humans,
particularly by glucuronidation (Tampal et al., 2002; Sacco and James 2004; Daidoji et
al., 2005), which is normally a higher-capacity pathway than sulfonation.
Glucuronidation is catalyzed by a family of endoplasmic reticular membrane-bound
enzymes, the UDP-glucuronosyltransferases (UGTs), which transfer a D-glucuronic acid
moiety from the co-substrate UDP-glucuronic acid (UDPGA) to a xenobiotic containing
a suitable nucleophilic atom such as oxygen, nitrogen and sulfur. UGTs are mainly found
in the liver, but also in extrahepatic tissues, such as the small intestine and kidney (Wells
et al., 2004).
The various chlorine and hydroxyl substitution patterns possible on the biphenyl
structure may lead to significant differences in glucuronidation kinetics. One explanation
for the retention of certain OH-PCBs may thus be that they are poor substrates for
glucuronidation. Tampal and co-workers (2002) studied the glucuronidation of a series of
OH-PCBs by rat liver microsomes. Efficiency of glucuronidation varied widely, and
substitution of chlorine atoms at the m- and p-positions on the nonphenolic ring greatly
lowered Vmax. Weak relationships were observed between the dihedral angle, pKa, log D
and enzyme activity. The experimentally determined kinetic parameters determined in the
Tampal et al study were subsequently related to the physicochemical properties and
structural features of the OH-PCBs by means of a quantitative structure-activity
relationship (QSAR) study. Hydrophobic and electronic aspects of OH-PCBs were shown
to be important in their glucuronidation (Wang, 2005).
Most of the persistent OH-PCBs found in human plasma are hydroxylated at the p-
position, in addition to being meta-chlorinated on either side of the phenolic group. The
remaining substitution pattern on both rings is highly variable (Bergman et al., 1994;
Sjoidin et al., 2000). An OH group in the para position, with two flanking chlorine atoms
was associated with estrogen and thyroid hormone sulfotransferase inhibitory activity
(Kester et al., 2000; Schuur et al., 1998), and exhibited the highest affinity for
transthyretin (TTR) (Lans et al., 1993), the major transport protein in non-mammalian
species (Cheek et al., 1999). Such OH-PCBs were potent inhibitors of the sulfonation of
3 -hydroxybenzo[a]pyrene (Wang et al., 2005). In contrast, the OH-PCBs having an
unhindered hydroxyl group substituted at the para position (relative to the biphenyl bond)
have exhibited the strongest binding to the rodent estrogen receptor (ER), although the
competitive ER binding affinities were <100-fold lower than that observed for estradiol
(Korach et al., 1988; Arulmozhiraja et al., 2005).
In the channel catfish, individual OH-PCBs have been shown to inhibit the in vitro
intestinal glucuronidation of several hydroxylated metabolites of benzo[a]pyrene (BaP)
(van der Hurk et al., 2002; James and Rowland-Faux 2003). The in situ hepatic
glucuronidation of a procarcinogenic BaP metabolite, the (-)benzo[a]pyrene-7,8-dihydro-
diol, was also inhibited by a mixture of OH-PCBs, consequently increasing the formation
of DNA adducts (James et al., 2004). It is possible that these compounds inhibit their own
glucuronidation. The OH-PCB metabolites of 3,3,4,4-tetrachlorobiphenyl (CB-77), one
of the most toxic PCBs known, were poor substrates for catfish intestinal glucuronidation
(James and Rowland-Faux, 2003). This may help to explain the persistence of these
The glucuronidation kinetics of a series of potentially toxic p-OH-PCBs by channel
catfish liver and proximal intestine is influenced by the structural arrangement of the
chlorine substituents around the biphenyl ring.
Chemicals. A total of 14 substrates were used in this study (Figure 4-1). The
nomenclature of the OH-PCBs is based on the recommendations of Maervoet and co-
The following substrates (Catalog no. in parentheses) were purchased from
Accustandard (New Haven, CT): 4-OHCB2 (1003N), 4-OHCBl4 (2004N), 4'-OHCB69
(4008N), 4'-OHCB72 (4009N), 4'-OHCB106 (5005N), 4'-OHCBll2 (5006N), 4'-
OHCBl21 (5007N), 4'-OHCBl59 (6001N), and 4'-OHCBl65 (6002N). The compounds
4'-OHCB35, 4-OHCB39, 4'OHCB68, 4'-OHCB79 were synthesized by Suzuki-coupling
(Lehmler and Robertson, 2001; Bauer et al., 1995). The 4-hydroxy biphenyl (4-OHBP)
was purchased from Sigma (St.Louis, MO). 14C-UDPGA (196 CICi/Clmol) was obtained
from PerkinElmer Life and Analytical Sciences (Boston, MA). The 14C-UDPGA was
diluted with unlabelled UDPGA to a specific activity of 1.5-5 CICi/Clmol for use in
enzyme assays. PIC-A (tetrabutylammonium hydrogen sulfate) was obtained from
Waters Corp. (Milford, MA). Other reagents were the highest grade available from Fisher
Scientific (Atlanta, GA) and Sigma.
4' -OHCB3 5
4-OHCB39 4' -OHCB6 8
Cl ClI Cl' Cl
4' -OHCB79 4' -OHCB 106
4' -OHCB 112
4' -OHCB l59 4' -OHCBl165
Figure 4-1. Structure of substrates used in channel catfish glucuronidation study.
Animals. Channel catfish (Ictalurus punctatus), with weights ranging from 2.1 -
3.7 kg, were used for this study. All fish were kept in flowing well water and fed a fish
chow diet (Silvercup, Murray, UT). Care and treatment of the animals was conducted as
per the guidelines of the University of Florida Institutional Animal Care and Use
Committee. The microsomal fractions were obtained from liver and intestinal mucosa
4' -OHCB l21
using a procedure described previously (James et al., 1997). Only the proximal portion of
the intestine was used in the study. Protein determination was carried out by the method
of Lowry and co-workers (195 1) using bovine serum albumin as protein standard.
Glucuronidation assay. A radiochemical ion-pair extraction method was
employed to investigate the glucuronidation of the 4-OHPCBs and 4-OHBP. Substrate
consumption did not exceed 10%. Initial experiments determined the saturating
concentrations of UDPGA to be employed. The incubation mixture consisted of 0.1 M
Tris-Cl buffer (pH 7.6), 5 mM MgCl2, 0.5% Brij-58, 200 CIM or 1500 CIM [14C]UDPGA
(intestine and liver, respectively), 100 Clg catfish intestinal or hepatic microsomal protein,
and substrate in a total reaction volume of 0. 1 mL. Initially, the OH-PCBs were added to
tubes from methanol solutions and evaporated under nitrogen. In all cases, the protein
and Brij-58 were added to the dried substrate, thoroughly vortexed and left on ice for 30
minutes. Subsequently, the buffer, MgCl2, and water were added in that order and vortex-
mixed. After a pre-incubation of 3 minutes at 35oC, UDPGA was added to initiate the
reaction, which was terminated after 30 minutes incubation by the addition of a 1:1
mixture of 2.5% acetic acid and PIC-A in water, such that the final volume was 0.5 mL.
The glucuronide product was extracted by two successive 1.5 mL portions of ethyl
acetate. The phases were separated by centrifugation, and duplicate portions of the ethyl
acetate phase were counted for quantitation of glucuronide conjugate.
Physicochemical parameters. The structural characteristics of the OH-PCBs were
calculated using ChemDraw 3D (CambridgeSoft Corp., Cambridge, MA). Parameters
used were: the Connolly Accessible Surface Area (CAA, the locus of the center of a
probe sphere, representing the solvent, as it is rolled over the molecular shape), the
Connolly Molecular Surface Area (CMA, the contact surface created when a probe
sphere (radius = 1.4 A+, the size of H20), representing the solvent, is rolled over the
molecular shape), the Connolly Solvent-Excluded Volume (CSV, the volume contained
within the contact molecular surface, or that volume of space that the probe is excluded
from by collisions with the atoms of the molecule), the ovality (the ratio of the Molecular
Surface Area to the Minimum Surface Area, which is the surface area of a sphere having
a volume equal to CSV of the molecule), and dihedral angle (the angle formed between
the planes of the two rings, which is related to the extent of coplanarity of the molecule).
ACD/ILab software (Advanced Chemistry Development, Ontario, Canada) was used to
predict log P, log D (at pH 7.0), and the pKa (of the phenolic group).
Kinetic analysis. Duplicate values were employed for the rate of conjugate
formation at each substrate concentration to calculate kinetic parameters using Prism v4.0
(GraphPad Software, Inc., San Diego, CA). Equations used to fit the data were the
Michaelis-Menten hyperbola for one-site binding and the Hill plot for positive
The kinetics for UDPGA were analyzed for the glucuronidation of three
representative OH-PCBs (Table 4-1). Saturating concentrations of UDPGA were higher
in liver than in intestine (Figure 4-2). The glucuronidation of most of the OH-PCBs tested
followed Michaelis-Menten kinetics (Figure 4-3A). In the case of the glucuronidation of
4'OHCB35 by liver and 4'OHCBll2 by proximal intestine, the data fitted the Hill plot
Table 4-1. Estimated kinetic parameters (mean & S.D.) for the co-substrate UDPGA in
the glucuronidation of three different OH-PCBs.
Substrate Substrate Vmax (app) Km (app)
Concentration (CIM) (nmol/min/mg) (CLM)
4'-OHCB-3 5 500 0.87 & 0.20 697 & 246
4'-OHCB-72 250 0.32 & 0.14 247 & 162
4'-OHCB-69 200 0.20 + 0. 11 27 & 14
The estimated apparent maximal rate of glucuronidation of polychlorinated
biphenylols by channel catfish ranged from 124-784 pmol/min/mg for proximal intestine
and 404-2838 pmol/min/mg for the liver (Table 4-2). The Kms for individual OH-PCBs
tended to be different in the two organs, with a few exceptions (40HCB2, 4'OHCBl65).
Vmax was significantly higher in liver than in intestine. Conversely, the affinity of
intestinal catfish UGTs (Km range: 42-572 C1M) for the OH-PCBs tested was higher than
for liver UGTs (Km range: 111-1643 CIM). These contrasting differences are reflected in
the lack of any difference in the efficiency of glucuronidation in both organs when all the
OH-PCB substrates were considered (Table 4-3). Vmax for OH-PCB glucuronidation in
both organs were strongly correlated with each other (R2=0.74). This relationship did not
exist for Km (R2=0.003).
0 25 50 75 100
Figure 4-2. UDPGA glucuronidation kinetics in 4 catfish.
A) in liver, using 500 CLM 4'-OH-CB35. B) in proximal intestine, using 200
CLM 4'-OH CB69
0 250 500 750 1000 1250 1500
[4'-OH CB 159] (CLM)
0 100 200 300 400 500 600 700
[4'-OH CB 112] (CIM)
Figure 4-3. Representative kinetics of the glucuronidation of OH-PCBs in 4 catfish.
A) Michaelis-Menten plot for 4'-OHCB-159 by liver. B) Hill plot for 4'-
OHCB-112 by proximal intestine
Units for Km and Vmax are CLM and pmol/min/mg protein, respectively. Bold indicates Sso
in place of Km. ND, not done.
Table 4-3. Comparison of the estimated kinetic parameters for OH-PCB glucuronidation
in catfish liver and proximal intestine
Parameter Liver Intestine p-value
Vmax (app) 1370 & 275 364 & 70 0.002
Km (app) 567 & 128 210 & 46 0.016
Vmax/Km 3.7 & 0.6 3.4 & 1.8 0.857
(Mean & SEM for all OH-PCB substrates)
Table 4-2. Kinetic parameters (Mean & S.D.) for the glucuronidation of 4-OHBP and OH-
Substrate Vmax (app) Km (app)
4-OHCB 3 9
43 A 10
417 & 57
255 & 59
784 & 348
220 + 90
213 & 91
751 & 253
401 & 236
124 & 36
431 & 60
401 & 67
220 & 39
188 & 66
163 & 26
599 & 110
572 & 47
387 & 65
265 & 85
134 & 36
119 & 75
42 & 21
183 A 126
87 & 21
183 & 58
163 & 24
130 & 21
213 A 136
137 & 44
182 & 78
2277 & 849
2022 & 936
2838 & 1456
1716 & 536
2774 & 1153
869 & 318
1579 & 645
2144 & 1007
1046 & 408
404 & 116
502 & 235
583 & 95
614 & 202
455 & 89
242 & 76
1071 & 410
476 & 201
798 & 122
1643 & 545
207 & 97
318 & 91
111 & 28
The Vmax for glucuronidation in both proximal intestine and liver was significantly
decreased upon addition of a second chlorine substituent flanking the phenolic moiety,
while keeping the chlorine substitution pattern in the rest of the molecule constant (Table
4-4, Figure 4-4). The affinity of hepatic UGTs for the OH-PCBs appeared to increase
with the addition of a second flanking chlorine atom; however, this relationship did not
achieve statistical significance.
Table 4-4. Comparison of kinetic parameters (Mean & SEM) for the glucuronidation of
OH- PCBs grouped according to the number of chlorine atoms flanking the
Parameter Flanking chlorines p-value
Vmax (app), pmol/min/mg 2247 & 204 1002 & 274 0.007
Km (app), CIM 856 & 209 342 & 88 0.053
Vmax (app), pmol/min/mg 560 + 85 190 & 23 0.003
Km (app), CIM 274 & 97 191 & 53 0.473
The effect of chlorine substituents on the nonphenolic ring on glucuronidation of
OH-PCBs was also investigated. No significant differences on Km and Vmax could be
observed between the absence or presence of specific chlorine substituents on the
nonphenolic ring. The only exception was that the presence of an ortho-chlorine
significantly (p=0.03) decreased the Km in the proximal intestine.
I one flanking Cl
EC: two flanking Cl
0 3,4 2,4,6 2,3,4,5 2,3,5,6
Substitution pattern on nonphenol ring
Sone flanking Cl
..I~ two flanking Cl
Substitution pattern on
Figure 4-4. Decrease in Vmax with addition of second chlorine atom flanking the phenolic
group, while keeping the chlorine substitution pattern on the nonphenolic ring
A) proximal intestine. B) liver.
Regression analysis was performed between the kinetic parameters for the
glucuronidation of OH-PCBs and several physical parameters for these substrates (Table
4-5). The data for 40HBP was not used since this compound is not a OH-PCB. The
affinity of intestinal UGTs was negatively correlated with the Connolly solvent-
accessible surface area, the molecular surface area, solvent-excluded volume, ovality,
dihedral angle, log P, and positively correlated with pKa. The maximum rate of hepatic
glucuronidation was negatively correlated with the Connolly solvent-accessible surface
area, the molecular surface area, solvent-excluded volume, ovality, and log P, and
positively correlated with pKa (which showed a similar relationship with intestinal Vmax).
Ovality was also significantly negatively correlated with the maximum rate of intestinal
glucuronidation of the OH-PCBs studied (Figure 4-5).
A paired t-test performed in order to investigate the physicochemical parameters
involved in the significant decrease in Vmax observed for the glucuronidation of OH-
PCBs with two chlorine atoms flanking the phenolic group revealed that, for OH-PCBs
with this structural arrangement, pKa was decreased (p=0.02), while log P, and
parameters indicating molecular size (CAA, CMA, CSEV, ovality) were all increased (all
Sign in parentheses indicates type of correlation where it achieved significance.
Table 4-5. Results of regression analysis performed in order to investigate the
relationship between the glucuronidation of OH-PCBs by catfish proximal
intestine and liver and various estimated physical parameters.
'-- a Vmax (int)
g a Vmax (liv)
1.360 1.385 1.410 1.435
Figure 4-5. Relationship between Vmax for OH-PCB glucuronidation in intestine and liver
In comparison to catfish intestine, catfish liver displayed higher rates of
glucuronidation of OH-PCBs, however both organs collectively biotransform the OH-
PCBs studied with similar efficiency. This occurred because while the glucuronidation
Vmax in the intestine was lower than in the liver, the affinity of intestinal UGTs for the
OH-PCBs was higher than liver UGTs. However, the efficiency of glucuronidation of 4'-
OHCB69 was seven times higher in the proximal intestine; when the data for this
substrate was excluded, the efficiency of glucuronidation was significantly higher
(p=0.01) in liver.
The total UGT capacity in the liver is much greater than in intestine when the total
content of microsomal protein in these two organs is taken into consideration. In fact, the
levels of microsomal protein from liver were always higher than in the intestine of each
individual fish studied, possibly because of the decreased amount of endoplasmic
reticulum in enterocytes relative to hepatocytes (DePierre et al., 1987). Thus, the intestine
appears to compensate for the lower glucuronidation capacity by expressing UGTs with a
No relationship was established between Kms for the glucuronidation of OH-PCBs
in liver and intestine. When individual OH-PCBs were considered, there were significant
differences in efficiency. These results suggest that these two organs have different UGT
isoform profiles, with the intestine possessing one or more isoforms that display greater
specificity for OH-PCBs. Possible UGT isoforms responsible may be catfish enzymes
analogous to rat UGTIA1, UGTIA6 and UGT2B1 (Daidoji et al., 2005), and to plaice
hepatic UGTIBl, which has been shown to conjugate planar phenols (Clarke et al.,
The substrates 4'-OHCB69 and 4-OHCB39 were glucuronidated with the highest
efficiency in the intestine and liver respectively. 4'-OHCB35 showed the highest rates of
glucuronidation in both liver and intestine. The poorest substrates were 4-OHCBl4 in the
intestine and 4'-OHCBll2 in the liver. In contrast, rat liver glucuronidates 4-OHCBl4
with the highest efficiency, relative to other OH-PCBs studied (Tampal et al., 2002).
Overall, the efficiency of glucuronidation of the OH-PCBs by rat liver is higher than in
catfish liver. While these dissimilarities may be ascribed to differences in UGT isoform
type and expression due to the different species and tissues in the two studies, it may also
indicate an increased susceptibility of catfish to the toxic effects of OH-PCBs due to an
Compared to the OH-PCBs, 4-OHBP was the poorest substrate for glucuronidation.
This compound had the lowest Vmax in both liver and proximal intestine. The affinity for
4-OHBP in the intestine was also the lowest. In the liver however, the Km was
comparable to other OH-PCBs. These results are surprising in view of the fact that 4-
OHBP has been shown to be a good substrate for glucuronidation using rat, guinea pig,
beagle dog and rhesus monkey liver microsomes (Yoshimura et al., 1992), and human
expressed UGTs (King et al., 2000; Ethell et al., 2002). In isolated rat hepatocytes, 4-
OHBP is a cytotoxic major metabolite of biphenyl, impairing oxidative phosphorylation
(Nakagawa et al., 1993). These results suggest that this compound may be potentially
more toxic to catfish than to mammals, unless cleared by another pathway such as
While the decreased glucuronidation of 4-OHBP may be due to the lack of a
specific phenol UGT isoform in catfish, the known broad substrate specificity of phenol
UGTs, together with the observed higher rates of glucuronidation for the OH-PCBs, leads
us to hypothesize that this compound may be such a poor substrate due to its lower
lipophilicity, as has been observed for other substituted phenols (Kim 1991). In fact,
addition of a single chlorine atom flanking the phenolic group (as represented by
40HCB2) resulted in at least a tenfold increase in Vmax in both liver and intestine, with no
significant change in Km (with respect to 4-OHBP). This increased lipophilicity
(represented by an estimated log P increase from 3.2 to 3.8) appeared to impact the
formation of the glucuronide and not the initial binding of substrate to UGT. Good UGT
substrates tend to be lipophilic compounds which are thought to diffuse through the
endoplasmic reticular bilayer and reach the substrate-binding site in the lumenal N-
terminal part of the enzyme, which contains a region of strong interaction with the
membrane (Radominska-Pandya et al., 2005). For all the OH-PCBs studied, we only
observed weak inverse correlations (R2<0.3) between log P and intestinal Km and liver
Vmax. No significant relationship could be observed between parameters of lipophilicity
and intestinal Vmax. The absence and weakness of such relationships may reflect the need
for OH-PCBs with additional structural variation to be included in studies of this type.
Another explanation may be the perturbation of the lipid bilayer of the microsomes,
resulting in rate-limiting partitioning, which would not be present in vivo (Tampal et al.,
As the estimated pKa of the OH-PCBs increased, so did hepatic and intestinal Vmax
for glucuronidation. These results are in agreement with a previous OH-PCB
glucuronidation study in rats (Tampal et al., 2002). Thus, a greater proportion of ionized
OH-PCB molecules appear to have an adverse effect on glucuronidation. Such charged
molecules present at the active site of UGT may interfere with the charge-relay system
that relies on a basic negatively charged residue to deprotonate the phenolic group, prior
to transfer of glucuronic acid (Yin et al., 1994).
Since the use of microsomal systems to elucidate structure-activity relationships
involves incubations of substrate with a heterogeneous population of UGTs exhibiting
different levels of expression and activity, it was not the intention of this study to attempt
to predict the effect of molecular structure and physicochemical parameters on the
glucuronidation of OH-PCBs, which is better achieved using individual isoforms.
However, if any such effects can be observed at a microsomal level, then it is likely that
such processes are occurring in the organism, whose detoxification route depends on
various UGTs metabolizing substrate simultaneously and not in isolation. This may help
to further delineate the different toxicokinetics of OH-PCBs.
The p-OH-PCBs used in this study all had one or two chlorine atoms flanking the
phenolic group. This structural motif is of interest since it imparts several toxic properties
to these compounds. OH-PCBs with two flanking chlorines were found to be poorer
substrates than compounds with one flanking chlorine atom, in both liver and intestine.
Thus, for example, while 4'-OHCB35 was a very good substrate for glucuronidation,
addition of a second flanking chlorine (as in 4'-OHCB79) resulted in a greater decrease in
Vmax than the addition of two adjacent chlorine substituents on the aphenolic ring (as in
4'-OHCB 106). A comparison of the physicochemical parameters of the two different
structural arrangements suggests that lipophilicity, pKa, and molecular size may all be
contributing to this effect on Vmax.
The addition of a second chlorine atom imparts additional lipophilicity to the
molecule and may increase positive charge on the phenolic carbon atom, which results in
stronger binding to the active site (Wang 2005). This study did show a non-significant
decrease in Km with the addition of the second chlorine atom for both organs. On the
other hand, the 3,5-chlorine substitution pattern may interfere with the mechanism of
glucuronidation because of steric hindrance, although this has been disputed (Mulder and
Van Doorn 1975; Tampal et al., 2002).
The estimated pKas for OH-PCBs with two flanking chlorine substituents were
significantly lower than similar molecules with one flanking chlorine atom. This is
supported by limited experimental data showing that OH-PCBs with two flanking
chlorine atoms have pKa values as low as 6.4 (for 4'-OHCB39, Miller 1978). The
population of OH-PCB molecules which are ionized at physiological pH is significantly
more than OH-PCBs with one flanking chlorine atom, resulting in the adverse effect on
the enzymatically-catalyzed charged relay system described above. In studies conducted
with rat liver microsomes, a decreased maximal rate of glucuronidation was also
observed amongst OH-PCBs differing only in the number of chlorines flanking the
phenolic group (1 pair of OH-PCBs in Tampal et al., 2002; 2 pairs of OH-PCBs in
Daidoji et al., 2005). According to Daidoji and co-workers (2005), UGT2B1 is the
primary rat hepatic UGT isoform responsible for metabolizing OH-PCBs with one
flanking chlorine atom. UGTIAl appears to metabolize both, though with a preference
for structures with two flanking chlorines
These results are significant from a toxicological standpoint since almost all the
major OH-PCBs found in human plasma incorporate a 4'-hydroxy-3',5' -dichloro
structure (Sandau et al., 2002; Fangstrom et al., 2002; Hovander et al., 2002). It is
possible that one reason for the persistence of these OH-PCBs may be a reduced rate of
glucuronidation due to this structural arrangement.
Two or more chlorine substituents that are ortho to the biphenyl bond cause the
molecule to twist and assume a non-coplanar conformation. In the parent PCBs this leads
to toxicological differences, such as loss of AhR agonist activity. The estimated dihedral
angles for the compounds investigated in this study ranged from 360-760. The affinity of
intestinal, but not hepatic, UGTs appeared to increase with the degree of twisting,
suggesting that the predominant isoform(s) in catfish intestine binds more strongly to the
more twisted OH-PCBs. While this may be additional evidence of differences with
respect to isoform profiles between liver and intestine, the weakness of the relationship
(R2~0.3) precludes using this result to solidly support this hypothesis.
Similar to what has been reported for the glucuronidation of OH-PCBs in rats
(Tampal et al., 2002) and simple phenols by human UGTIA6 (Ethell et al., 2002), the
maximal rate of hepatic glucuronidation decreased with increased steric bulk. In the case
of intestinal glucuronidation this relationship was weaker. The enzyme affinity of
intestinal UGTs increased with increasing molecular size, perhaps because the bulkier
molecules tended to be more lipophilic. However, in contrast, the affinity of the liver
UGTs was not affected as much by the molecular size, at least within the restricted size
range offered by the OH-PCBs studied. At this point, no explanation for this discrepancy
between these two tissues is forthcoming.
Conclusions and Recommendations
OH-PCBs are glucuronidated with similar efficiency by channel catfish liver and
proximal intestine. There appear to be differences in the UGT isozyme profile in both
organs. The Vmax for both hepatic and intestinal glucuronidation was decreased with the
addition of a second chlorine atom flanking the phenolic group, which is an arrangement
typical of OH-PCBs that persist in organisms. Future research may be directed towards
cloning, sequencing and characterizing these catfish UGTs, in order to have a better
understanding of the specificity of individual UGT isoforms for particular chlorine
substitution patterns in OH-PCBs.
CLONING OF UDP-GLUCURONOSYLTRANSFERASES FROM CHANNEL
CATFISH LIVER AND INTESTINE
Piscine UGT Gene Structure and Isoforms
Fish are the most ancient vertebrate phylum, and account for over 40% of all living
vertebrate species (Clarke et al. 1992a). Clarke and co-workers (1992b) compared the
hepatic glucuronidation of several xenobiotics and endobiotics in plaice (Pleuronectes
platessa) and rat (Rattus norvegicus), species that are separated by more than 350 million
years of evolutionary divergence. Despite the fact that the plaice showed reduced
glucuronidation activity towards substrates such as morphine, bilirubin and steroids,
weak immunological cross-reactivity was obtained when anti-rat UGT antibodies were
used, indicating the presence of conserved common structural motifs between the two
Characterization of plaice UGTIB 1 (Accession number (AN): X741 16), an isoform
which conjugates planar phenols and is inducible by polyaromatic hydrocarbons (PAH),
confirmed the strong degree of conservation in gross exon structure and amino acid
character (signal peptide, membrane insertion, and stop sequences) between fish and
mammals. The greatest degree of similarity in amino acid sequence was found with
UGT1 rather than UGT2 (Clarke et al., 1992b, George et al., 1998). Allelic variations in
this UGTIB1 gene are presumed to be functionally silent (George and Leaver 2002).
While there is strong evidence for other distinct isoforms conjugating bilirubin, estrogen
and androgens, to date these have not been characterized. At least six distinct UGTs
exhibited tissue-specific expression in plaice (Clarke et al., 1992c). UGTIB2 mRNA has
recently been sequenced from marbled sole (Pleuronectes yokohamnae) liver (AN:
ABl20133), and a partial sequence of an unidentified UGT isoform has been obtained
from the orange-spotted grouper (Epinephehis coioides) (AN: AY735003). The existence
of a number of partial length sequences of UGT homologues from zebrafish (Dario rerio)
EST projects in GenBank provide evidence for the cDNA of 10 distinct UGTs. The
absence of cDNAs with the same 3'sequence and dissimilar 5-exon 1 coding sequence
suggests the absence of alternative splicing of UGTIA genes as seen in mammals. Thus,
George and Taylor (2002) have suggested the existence of three family 1-related UGTs
and another two related to the UGT 2 family in the zebrafish. In general, however, it
appears that fish possess multiple UGTs with similar functional and structural properties
to mammalian UGT.
Toxicologically, it is important to know whether xenobiotic pollutants such as
PAHs compete with steroids or bilirubin for the same active site on UGT, resulting in
physiological perturbations in reproductive and/or liver function. For example, Atlantic
salmon (Salmo salar) suffering from a multiple pollutant-induced j aundice were shown to
have decreased bilirubin UGT activity (George et al. 1992). Channel catfish are also
exposed to pollutants (such as PAHs and PCBs) which accumulate in sediments. Thus,
this organism may be a useful indicator of the bioavailability of these pollutants in such
sedimentary environments. In addition, the use of this Hish in aquaculture makes it
essential to understand every aspect of its detoxification mechanisms, since these will
ultimately impact human health.
While no UGTs have yet been cloned and characterized from channel catfish, this
species shows glucuronidation activity towards a variety of toxic xenobiotics, including
mono- and di-hydroxy metabolites of benzo[a]pyrene and OH-PCBs (James et al., 2001;
van den Hurk and James 2001; Gaworecki et al., 2004). As with other aquatic species,
pollutants which are direct substrates for glucuronidation, such as pentachlorophenol,
several OH-PCBs, 4-OH-heptachlorostyrene, and which have been shown to be
estrogenic and thyroidogenic, have been detected in channel catfish (Li et al., 2003).
Kinetic differences have been observed between hepatic and intestinal UGT activities,
suggesting expression of different isozymes in these two organs. Thus, knowing more
about the identity and substrate specificity of catfish UGTs will assist our understanding
of the effect of glucuronidation on the contributions of such metabolites to toxicity. Since
the absence of cDNAs with the same 3'sequence and dissimilar 5'exon 1 coding
sequence in fish suggests the absence of alternative splicing of UGTIA genes as seen in
mammals (Gong et al., 2001), additional information on piscine UGT gene structure is
also important from a phylogenetic perspective.
Multiple UGT isoforms are present in channel catfish liver and intestine
Methodology (part 1)
For convenience, a flowchart summarizing the various steps involved in the cloning
process is shown in Figure 5-1. Because the study utilizing the gene specific primers was
dependent on an initial study which utilized degenerate primers and led to the cloning of
partial sequences of UGT, the methodology and results sections are split correspondingly
mn two parts.
. _ _
Sequencing BLAST search
1. Des n GS primers
length UGTs 2. 5' and 3' RLMf-RACE
3. Clone and sequence
1. Des n GS primers
2. PCR with Super Taq Plus
3. Clone and sequence
Full-length cDNA clone
Figure 5-1. Summary of methods used to clone channel catfish UGT
Animals. A single female adult catfish was sacrificed. Total weights of liver and
intestinal mucosa were recorded. Tissues were immediately processed for RNA isolation.
RNA isolation. Approximately 0.1Ig of tissue from the liver and proximal intestinal
mucosa were homogenized in separate tubes with 1 mL Trizol@ reagent and placed on
ice. The homogenates were incubated for 5 min at room temperature (15-300C) to enable
complete dissociation of nucleoprotein complexes. Chloroform, 0.2 mL, was added and
1. RNA isolation
2. RT-PCR, using degenerate primers based on consensus sequences
U GT / ~a ion
cDNA f O transformation
pGEM T-Easy vector recombinant plasmid
UGT clones Ecoi ,MO
the tubes were shaken vigorously by hand for 15 seconds and then incubated at room
temperature for 2-3 min. The samples were then centrifuged at 12,000g for 15 min at 2-
80C. This separated the solution into an aqueous phase containing the RNA and an
organic phase containing DNA. The colorless upper aqueous phase was transferred to an
RNase-free tube. The RNA was precipitated by the addition of 0.5 mL propan-2-ol. The
samples were then incubated at room temperature for 10 min, followed by centrifugation
at 12,000g for 10 min at 2-80C. The RNA precipitate was now visible as a gel-like pellet
on the side and bottom of the tube. The supernatant was removed and the RNA pellet was
washed once with 1 mL 75% ethanol. The sample was vortex-mixed and centrifuged at
7,500g for 5 minutes at 2-80C. The RNA pellet was left to air-dry for a few minutes
following decantation of the ethanol. The RNA was dissolved in 100 CIL RNase-free
water for intestine, and 200 CIL RNase-free water for liver (since the solution in this case
appeared to be more concentrated), by passing the solution a few times through a pipette
tip. The solution was then incubated for 10 minutes at 55-600C. The samples were stored
at -800C. The purity of the RNA was checked by running the sample on 1% agarose gel
(with 9.5% formaldehyde) and 10x MOPS buffer. Bands corresponding to the 28S and
18S ribosomal subunits were observed. The purity of the RNA was also checked by
diluting the sample in 10mM Tris HC1, pH 7.5 and measuring the A260/A280 absorbance
ratio (ideally should be between 1.8 and 2.1i).
DNase treatment of RNA samples. This procedure was done in order to remove
contaminating DNA from RNA preparations, and to subsequently remove the DNase and
divalent cations from the sample. Portions of the RNA solutions were diluted to 100
Clg/mL with RNase-free water. The Ambion 9 (Austin, TX) DNA-removal kit was used.
The reaction mix, consisting of 25 CIL RNA, 2.5 CIL 10xDNasel buffer, and 3 CIL DNase I
was incubated at 370C for 1 hour. DNase inactivation reagent, 5 CIL, was added by means
of a wide pipette tip (due to the thick consistency of this reagent). The tubes were then
incubated for 2 min at room temperature, with gentle flicking. The tubes were then
centrifuged at 10,000g for ~1 min to pellet the DNase inactivation reagent. The
supernatant containing the RNA was transferred to a new RNase-free tube and stored at -
Generation of cDNA library. The Retroscript@ reagent kit manufactured by
Eppendorf (Westbury, NY) was employed in order to heat-denature the RNA. To each of
the two tubes were added 10 CIL liver or intestinal RNA (equivalent to 1 Gig) and 2 C1L
random decamers. The tubes were mixed, centrifuged briefly and heated for 3 min at 70-
850C in the thermocycler. Tubes were removed and left on ice for 1 minute. They were
centrifuged and put on ice again. The following components were added to each tube: 2
CIL 10xRT buffer, 1 C1L dNTP mix (10mM), 0.5 C1L RNase inhibitor, 1 C1L reverse
transcriptase, and RNase-free water to 20 C1L. The tubes were gently mixed and
centrifuged briefly. They were placed in the thermocycler for 1 hr at 42-440C, followed
by 920C for 10 min. The resulting cDNA was either stored at -200C or subjected to a
second round of PCR (liver, see below; for the intestine this procedure was performed a
few days after cDNA generation).
Degenerate primer design. A characteristic 'signature sequence', 44-amino acids
long, probably corresponding to the UDPGA binding site, has been shown to be highly
conserved amongst mammals and other vertebrates (Mackenzie et al., 1997). The relevant
amino acid and nucleotide sequences were compared in 4 species of fish using ClustalW.
The species investigated were Pleuronectes platessa UGTIBl, P.platessa UGT,
Pleuronectes yokohamnae UGT IB2, Epinephelus coiodes UGT and Danio rerio UGT.
Five primers were designed which could hypothetically bind to this sequence. The
application of exclusion criteria (degeneracy- <100-fold, poor or no matches with fish
sequences resulting from BLASTn searches, %GC content <40%, potential to self-
dimerize < -20 kcal/mol) resulted in the selection of two primers, designated as UGTR3
and UGTR4, and chosen to be reverse primers (Table 5-1). An additional reverse primer
(UGT RS) was chosen due to its low degeneracy (4-fold) and its complementarity to the
highly conserved N-terminal domain downstream of the signature sequence. Five
additional primers (UGTF3-7) were also chosen based on these same criteria. Since
these primers were complementary to sequences upstream of the signature sequence, they
were selected to be forward primers (Table 5-1).
Table 5-1. 5' 3' Sequences of degenerate primers chosen.
ID Sequence Direction
UGT F3 GTGGTSCTGGT SCCYGAAASYAGY Forward
UGT F4 CTTACWGAYCCMTTCYTKCC STGYGGC Forward
UGT F5 AAC AT GGTYYWWATYGGRGGYAT CAAC TGT Forward
UGT F6 ATYGGRGGYATCAACTGTGCA Forward
UGT F7 GAGT TTGT SVAHGGC TCW GGA Forward
UGT R3 AAAC AGHGGRAACAT CAVC AT Reverse
UGT R4 YC CYT GS TCK SCAAAC AGHGG Reverse
UGT R5 GT GRTAC TGRAT C CAGTT CAG Reverse
The primer pairs were selected in such a way that their melting temperatures did
not vary by more than 60C and their potential to heterodimerize was more than -20
kcal/mol (Table 5-2)
Table 5-2. Primer pairs chosen, showing annealing temperature and estimated amplicon
Pair Forward Reverse Amplicon length (bp)l T (oC)
3 UGT F6 UGT R5 618 52.6
4 UGT F7 UGT R3 288 53.8
5 UGT F6 UGT R3 339 53.8
6 UGT F7 UGT R5 567 52.6
7 UGT F5 UGT R4 363 61.1
8 UGT F3 UGT R4 1003 61.1
9 UGT F4 UGT R4 729 61.1
SBased on Danio rerio UGT sequence (Accession number NP_998587. 1)
PCR amplification of UGT cDNA. A 10 CIM solution of each primer in nuclease-
free water was made up. Each PCR tube consisted of 2 CIL DNA template (from catfish),
2 CIL forward primer, 2 C1L reverse primer, 0.5C1L Taq DNA polymerase (5U/C1L, in Mg
10x buffer), 1 CIL dNTP mix (10mM), 5 C1L 10xPCR buffer, and nuclease-free water up
to 50 CIL. Prior to the initiation of the PCR reaction, with the tubes in place, the
thermocycler lid was heated for two minutes at 1100C to prevent sample evaporation.
Thermocycler parameters (utilizing a gradient PCR program to adjust for the different
optimal annealing temperatures required by the various primer pairs) were as follows:
Stage Temp /oC Duration/min
Initial Denaturation 94 2
Denaturation 94 0.5
Annealing 5715 (L); 5515 (I)' 0.5
Extension 72 1.0
Final extension 72 7.0
annealing temperatures used for: L, liver; I, intestinal cDNA
The program consisted of 35 cycles of denaturation, annealing and extension.
The PCR products were subjected to electrophoresis on 1% agarose gel at 100V (in
lx TAE buffer (40 mM Tris base, 5 mM sodium acetate, 1 mM EDTA, pH 8.0)) using 30
CIL of PCR product; a 100bp DNA ladder was used for size estimates. The DNA bands
were visualized by placing on a UV transilluminator and recorded by photography.
Recovery of PCR product from gel and purification. The desired DNA band
was excised from the gel using a clean scalpel and transferred to a pre-weighed 1.5mL
microcentrifuge tube. The Wizard 9 SV Gel Clean Up system (Promega, Madison, WI)
was used to purify the PCR product by centrifugation. Membrane binding solution (4.5 M
guanidine isothiocyanate, 0.5 M potassium acetate, pH 5.0), 10 C1L per 10 mg gel, was
added to the gel slice. The mixture was vortexed and incubated at 510C for 10 min in
order to dissolve the gel slice. The tube was then briefly centrifuged at room temperature.
For every solution (derived from the cut gel slices), the following procedure was adopted.
One SV Minicolumn was placed in a collection tube. The dissolved gel mixture was
transferred to the SV Minicolumn assembly and incubated for 1 min at room temperature.
The assembly was centrifuged in a microcentrifuge at 16,000g for 1 minute and the liquid
in the collection tube was discarded. The column was washed by 700 C1L of Membrane
Wash Solution (10 mM potassium acetate, pH 5.0, 16.7 CIM EDTA, pH 8.0, 80%
ethanol). The assembly was centrifuged for 1 min at 16,000g, and the collection tube was
emptied. Another 500 CIL Membrane Wash Solution was added to the assembly, followed
by centrifugation for 5 minutes at 16,000g. The collection tube was emptied, and the
collection tube was recentrifuged for 1 minute to dry the column. The SV Minicolumn
was transferred to a clean 1.5 mL microcentrifuge tube and 50 CIL nuclease-free water
was applied to the column and incubated for 1 minute at room temperature. The
Minicolumn/micro-centrifuge tube was centrifuged for 1 minute at 16,000g. The
Minicolumn was discarded and the tube containing the eluted DNA was stored at -200C.
A portion of this DNA was diluted with 10 mM Tris-HC1, 1 mM EDTA, pH 8.0, and
used to calculate the DNA concentration by its absorbance at 260nm.
Ligation and transformation of E.coli. LB plates with ampicillin were first
prepared. LB Agar, 8.75 g, was weighed and dissolved in 250 mL, and the pH was
adjusted to 7.2 with NaOH. The solution was autoclaved for 30 min at 1200C. After the
medium cooled to around 500C, ampicillin was added to a final concentration of 100
Clg/mL. Some of the medium, 30-35 mL, was poured into 85-mm Petri dishes and the
agar left to harden. The plates were left overnight at room temperature and subsequently
stored in an inverted position at 40C
The ligation was performed using the p-GEM T-Easy Vector System@ supplied by
Promega. The volume of PCR product to be used in the ligation reaction could not exceed
3 CIL. The amount required was calculated from the following equation, which assumes
that the optimal insert:vector molar ratio is 3:1:
50 ng vector x a kb insert x 3 = ng insert required
3.0kb vector 1
where a is the approximate size of amplified insert
Because of this limit in sample volume, the amount of insert actually used was less
than that recommended since the concentration of purified DNA was relatively low. The
ligation reactions were setup as follows (all volumes in CIL) in 0.5 mL tubes:
Standard Positive Background
Component Reaction Control Control
2x Rapid Ligation Buffer, T4DNA ligase 5 5 5
pGEM T-Easy Vector (50 ng) 1 1 1
PCR-product (9 ng) 3----
Control insert DNA --- 2--
T4 DNA ligase (3 Weiss units/C1L) 1 1 1
DNase-free water 0 1 3
The ligation buffer was mixed vigorously before use. The reactions were mixed by
pipetting and incubated for 1 hour at room temperature, followed by storing overnight at
JM109 high-efficiency competent cells (>1x10s ofu/Clg DNA; Promega) were used
for transformation. The following procedure was performed using aseptic technique
(sterile tips and tubes, use of Bunsen flame to create upward convection in work area).
The tubes containing the ligation reactions were centrifuged for 1 minute at 10,000 rpm
and placed on ice. Another tube (transformation control, TC) was set up on ice; this
contained 0. 1 ng uncut plasmid (0. 1 CIL of 0. 1 mg/C1L solution used) in order to determine
the transformation efficiency of the competent cells. Tubes containing frozen aliquots of
JM109 cells were removed from -800C storage and placed on ice until thawed (~5 min).
The cells were mixed by gentle flicking of the tubes. Each ligation reaction, 2C1L, was
added to a 1.5 mL microcentrifuge tube on ice, followed by 50 CIL of cells (100 CIL were
added to the TC). The tubes were mixed by gentle flicking and left on ice for 20 minutes.
The cells were heat-shocked by placing in a 420C water bath for 45-50 sec. The tubes
were then returned to ice for 2 minutes. S.O.C medium (Invitrogen Corp., Carslbad, CA),
950 C1L, was added to each tube (900 CIL was added to the TC). The tubes were then
incubated for 1.5 h at 370C with shaking (~150 rpm). The ampicillin/LB plates were
removed from 40C storage, 100 CIL of 100 mM isopropylthiogalactoside (IPTG, a P-
galactosidase inducer) and 20 CIL of 50 mg/mL 5-bromo-4-chloro-3 -indolyl-P-D-
galactoside (X-Gal, hydrolyzed by P-galactosidase to yield a blue product) were added,
and the mixture spread on each plate. The agar was allowed to absorb these compounds
for 30 min at 370C. Samples, 100 CIL, of each transformation culture were transferred to,
and streaked on, duplicate LB/ampicillin/IPTG/ X-Gal plates; for the TC, 20C1L of tube
culture was diluted with 180 CIL of S.O.C. medium, and 100 C1L of this dilution was
applied to the agar plates. The plates were incubated overnight (~16h) at 370C. Plates
were then stored at 40C for 30 minutes to facilitate color development. The white
colonies should contain plasmids with the insert, while the blue colonies do not contain
the insert since the protein-encoding sequence of the lac Z gene in the vector is not
interrupted by the insert and hence can lead to P-galactosidase synthesis and catalysis of
the X-Gal reaction.
Colony PCR and culturing E.coli with insert of interest. Two white and one
blue colony from each plate were picked by a sterile wooden toothpick, which was
inserted in a PCR tube containing 5 CIL 10x PCR buffer, 5 CIL 10mM dNTP mix, 1 CIL
PUC/M13 forward primer, 1 CIL PUC/M13 reverse primer, 0.5 CIL Taq DNA polymerase,
and DNase-free water to 50 CIL. The pGEM T-Easy Vector contains binding sites for the
PUC/M13 primers. Thermocycler parameters were as shown previously, but with an
annealing temperature of 550C (no temperature gradient). PCR products were run on 1%
agarose gel at 100V, using 15 CIL of the PCR product.
Samples from colonies which showed the presence of insert on the gel were
extracted by an ethanol-flame sterilized metal hoop and dispensed into 14 mL sterile,
round-bottomed Falcon tubes containing 4 mL of LB medium with ampicillin by
swirling. The tubes were incubated with shaking for 16-20 h at 370C.
Purification of plasmid DNA. A sample, 850 CIL, of each culture medium was
diluted up to 1000 CIL with sterile glycerol and stored at -800C. The rest of the culture
medium was dispensed in 1.5 mL microcentrifuge tubes, which were centrifuged for 2
min at 10,000g. The supernatant was poured off and the tubes were blotted upside-down
on a paper towel to remove excess media. For plasmid purification, the Promega Wizard
Plus Minipreps 9 DNA Purification System was used. The cell pellets were resuspended
in 200 CIL of cell resuspension solution (50 mM Tris-HCI (pH 7.5), 10 mM EDTA, 100
Clg/mL RNase A). Cell lysis (0.2 M NaOH, 1% SDS) solution, 200 C1L, was added and
the tubes inverted 4 times to clarify the solution. Neutralization solution (1.32 M
potassium acetate, pH 4.8), 200 CIL, was added and mixed by inverting the tubes for 4
times, resulting in a white precipitate. The lysate was centrifuged at 10,000g for 5-20
minutes, depending on whether a cell pellet was clearly visible.
One Wizard@ Minicolumn was prepared for every Miniprep. A plunger was
removed from a 3 mL disposable syringe and set aside. The syringe barrel was attached
to the Luer-Lok@ extension of the Minicolumn. DNA purification resin (7 M guanidine
HC1), 1 mL, was pipetted in the barrel, followed by the cell lysate. The syringe plunger
was inserted in the barrel and used to push the slurry through the Minicolumn. The
syringe was detached from the Minicolumn and the plunger removed from the syringe
barrel. The barrel was then reattached to the Minicolumn. Column wash solution (80 mM
potassium acetate, 8.3 mM Tris-HCI (pH 7.5), 40 C1M EDTA, 55% ethanol), 2 mL, were
pipetted into the barrel of the Minicolumn/syringe assembly, and the solution was pushed
through the Minicolumn by the plunger. The syringe was removed, and the Minicolumn
was transferred to a 1.5 mL microcentrifuge tube, which was centrifuged at 10,000g for 2
min to dry the resin. The Minicolumn was transferred to a new 1.5 mL microcentrifuge
tube and 50 CIL nuclease-free water was added to the column and left for 1 minute. The
DNA was eluted by centrifuging at 10,000g for 20 sec. The Minicolumn was removed
and discarded, and the DNA solution stored at -200C. Products were visualized by 1%
agarose gel electrophoresis run at 100V, using 3 CIL of purified DNA and 10 CIL
quantitative 1kb plus DNA ladder (0.5 Gig).
Digestion with ecoRI. To ensure that the two DNA bands seen in the purified
plasmid DNA run on agarose gel were due to supercoiling of the DNA and not
contamination, the plasmid DNA was digested with the restriction enzyme ecoRI (pGEM
T-Easy Vector has restriction sites on either side of the insert). Plasmid DNA, 3 CIL, was
added to a tube containing 0.2 CIL acetylated BSA, 2 CIL 10x buffer, and 14.3 CIL DNase-
free water, and mixed by pipetting. ecoRI restriction enzyme (12 U/CIL), 0.5 CIL, was
added and the solution mixed by pipetting. The tubes were briefly centrifuged and
incubated for 2h at 370C. The products were run in 1% agarose at 100V, using all the
DNA sequencing and data processing. The concentration of the purified plasmid
DNA was determined prior to submission for sequencing. The DNA sequencing core
requires 1.5 Clg DNA for adequate processing. Cloned DNA sequences obtained were
then compared with nucleotide sequences in GenBank using the BLASTn tool provided
online (http://www.ncbi .nlm.nih.gov/BLAST). Multiple sequence comparisons were
done with SeqWeb, while two-sequence comparisons were done with the BLASTn 2.2.12
Results and discussion (part 1)
Primer pairs 4 and 6 successfully amplified cDNA from catfish liver; while primer
pair 4 amplified cDNA from proximal intestine. The size of the amplicons were
approximately 300bp (pair 4) and 600bp (pair 6) in size, with the gel-clean up system
effectively removing primer dimers and other contamination (Figure 5-2). The controls
indicated that ligation and transformation of the plasmid into E.coli were successful.
Purified plasmid DNA was obtained from several colonies (Figure 5-3), which were
denoted as L1-L8 for the liver and 11-14 for the proximal intestine. The two bands seen in
these gels, did not represent contamination, as verified by the restriction digest of the
plasmid, which resulted in a band corresponding to the vector and one corresponding to
the smaller insert (Figure 5-4).
3 2 1 ladder ladder 3 2 1 ladder
Figure 5-2. Products of PCR reaction. 1(from intestine), 2 and 3 (from liver)
A. pre-cleanup; B. post-cleanup with the gel clean-up system. 100kb ladder
shown for size estimation.
L8 L7 L6 L5 L4 L3 L2 L1 1kb ladder
14 13 12 I11
Figure 5-3. Plasmid DNA obtained from cultures transformed with vector containing
inserts from liver and intestine.
100bp ladder L8 L7 L6 L5 L4 L3 L2 L1 1kb ladder
Figure 5-4. Product of ecoRI digest of purified plasmids containing liver inserts L1-L8
The DNA sequences obtained are detailed in Appendix A. The results of the
BLASTn search (best five sequences for each insert) are summarized below (Table 5-3).
Very good matches were obtained with the Tetraodon nigroviridis cDNA, as well as
Pleuronectes yokohamnae UGT I B2, several Danio rerio sequences and
Strongylocentrotus purpuratus (sea-urchin) UGT2B sequences. There was also a good
similarity between the longer insert and the mammalian UGTIA sequences.
Better matches were obtained with the longer cDNA insert obtained from the liver
(95 sequences with score >50) than with the shorter insert from liver or intestine (9
sequences with score >50).
Table 5-3. Results of BLASTn search of cloned putative partial UGT sequences
Accession no. Short description Insert Score (bits) E-value
CNSOEYO6 Tetraodon nigroviridis, full length cDNA L1 123 3E-25
L4 115 8E-23
L7 113 6E-22
Il 107 2E-20
CNSOEVYF Tetraodon nigroviridis, full length cDNA L1 123 3E-25
L4 115 8E-23
L7 113 6E-22
Il 107 2E-20
AB120133.1 Pleuronectes yokohamae UGT1B2 mRNA L7 85.7 1E-13
L1 67.9 2E-08
L4 67.9 2E-08
Il 60.0 4E-06
AF104339 Macaca fascicularis UGT1A01, mRNA L7 63.9 5E-07
BC109404.1 Danio rerio cDNA clone L7 61.9 2E-06
BC100055.1 Danio rerio cDNA clone L1 60.0 4E-06
Il 52.0 1E-03
BX005348.9 Danio rerio DNA sequence from clone L1 60.0 4E-06
Il 52.0 1E-03
XM 792456.1 Strongylocentrotus purpuratus, UGT2B34 L4 54.0 3E-04
XM792428. 1 Strongylocentrotus purpuratus, UGT2B 17 L4 54.0 3E-04
SeqWeb analysis of the sequences showed that the short inserts were almost
identical with almost all the differences being located in the primer regions and thus may
be attributed to the degenerate nature of the primers. Sequence L1 was found to be 98%
similar to both sequences L7 and II. While this implied that all these sequences are
derived from the same isozyme, this could not be ascertained since most of the sequence
differences between UGT isoforms arise from the N-terminal (substrate-binding) domain
and only that part of the gene which codes for the highly conserved C-terminal domain
Methodology (part 2)
The next step in the cloning study was thus to design GSPs in order to extend the
partial UGT sequences obtained so far to the full-length gene.
Overview of RL~M-RACE
RNA-ligase mediated rapid amplification of cDNA ends, or RLM-RACE is a
procedure used to extend a known DNA sequence towards its 5'- and its 3'- ends
(Maruyama and Sugamo, 1994; Shaefer 1995).
In 5'-RACE, total RNA is treated with calf intestinal phosphatase (CIP) to remove
free 5'-phosphates from molecules such as ribosomal RNA, fragmented mRNA, tRNA,
and contaminating genomic DNA. The cap structure found on intact 5'-ends of mRNA is
not affected by CIP. The RNA is then treated with tobacco acid pyrophosphatase (TAP)
to remove the cap structure from full-length mRNA, leaving a 5'-monophosphate. A 45
base RNA Adapter oligonucleotide is ligated to the RNA population using T4 RNA
ligase. The adapter cannot ligate to dephosphorylated RNA because these molecules lack
the 5'-phosphate necessary for ligation. During the ligation reaction, the majority of the
full-length decapped mRNA acquires the adapter sequence as its 5'-end. A random-
primed reverse transcription reaction and nested PCR then amplifies the 5'-end of a
specific transcript (Figure 5-5). The Ambion kit used in this study provided two nested
primers corresponding to the 5'-RACE Adapter sequence, while two nested antisense
primers were designed to be specific to the target gene.
In 3'-RACE, first-strand cDNA is synthesized from total RNA using the supplied
3'-RACE Adapter. The cDNA is then subjected to PCR using one of the 3'-RACE
primers which are complimentary to the anchored adapter, and a user-supplied primer for
the gene under study (Figure 5-5). Although 3'-RACE may not require a nested PCR
reaction, this may also be performed if no significant amplicons are detected after the
CIP treatment to remove 5'PO4
from degraded mRNA, rRNA,
tRNA, and DNA
TAP treatment to remove cap
from full-length mRNA
reverse transcription with
3' RACE Adapter
G--P-PI-P~ AAAA ~NVTTTTT-adapter
G--P- P AAAAG---P--P--P AAAAA
5' RACE Adapter Ligation to
5'-RACE adapter AAAAA
5'-RACE adapter AAAAA
Figure 5-5. 5'- RLM-RACE and 3'- RACE
Design of gene-specific primers (GSPs) for initial 5'-RACE study. The initial
primers used for RACE were designed to be 20-24 bases in length, with 50% G:C
content, and with no secondary structure. Primers contained less than 3G or C residues in
the 3'-most 5 bases, and did not have a terminal G at the 3'-end. An online
oligonucleotide analyzer (www.idtdna.com) was used to determine whether potential
primers self-hybridized or hybridized to the primers supplied with the RLM-RACE Kit.
Figure 5-6 shows where the gene-specific primers and the primers supplied with the kit
should be positioned with respect to the DNA template.
5' RACE 5' RACE UGT-specific
outer inner 5'primer
5' RACE Adapter ~10bp
5'RACE UGT- 5'RACE UGT-
specific inner specific outer
3'RACE UGT- 3'RACE UGT-
specific outer specific inner
3' RACE Adiapter
Figure 5-6. Primer positions for 5'- and 3'-RACE.
The DNA templates selected were those identified from the previous study, that is,
the sequence isolated from liver (L6) and intestine (14). The primers used in this initial
study are shown in Table 5-4.
Table 5-4. Gene-specific primers used in initial 5'RLM-RACE study.
GSP ID Sequence (5' 3') Start positions PCR step
GSP OUT TGCTCTGAGGTCAGGTCGAA 397 Outer
GSP INN ACAGATACCCTCGTAGATGCCA 280 Inner
From 5' end of sense strand of partial sequence L6
Based on homology with the complete sequences of Pleuronectes platessa UGT1
(PPL249081) and M\'acaca fascicularis UGTIAl (AF104339) it was estimated that, for
the UGT sequence isolated from catfish liver, this sequence needed to be extend by ~920
bp to the 5' -end, and ~183 bp to the 3'-end. Unfortunately, the use of these primers led
to sequences which still lacked the 5'-end (L15R, L25R, L35R, Il5R, I25R, I35R; see
Appendix A). In addition, a high degree of non-specific binding was noted.
Design of GSPs for succeeding 5'- and 3'-RACE study. A new batch of GSPs
was designed (Table 5-5) using different criteria than the ones mentioned above in an
attempt to improve sensitivity. Primer 3.0 Software (http://frodo.wi .mit.edu/cgibin/
primer3/primer3_wYww.cgi) was used to design primers, based on the following criteria:
a. For 5'-RACE, GSPs with a GC-clamp at the 3'-end in order to reduce non-
specific binding were used,
b. The outer and inner primer melting temperatures for the GSPs were within a
degree of the RACE kit supplied primers,
c. For 5'-RACE, the inner primer was long (~27 bp) in order to reduce nonspecific
d. The primers were designed to anneal close (50-75 bp) to the existing 5'-end to
avoid large overlaps.
Different sets of primers were designed based on the cDNAs obtained by the study
involving the degenerate primers (I4) and the initial 5'-RACE study (I3 5R and L25R).
Table 5-5. Gene-specific primers used in succeeding RLM-RACE study
RACE Start' PCR step
(a) Liver L25R
(b) Intestine I4
(b) Intestine I35R
SStart position from partial DNA sequences obtained so far
5' RLM-RACE procedure
Calf intestinal phosphatase (CIP) treatment. Total RNA (not DNase treated) (2
CIL for liver; 1 CIL for intestine), 10 Gig, as well as 10 Clg of control RNA (mouse thymus)
were gently mixed with CIP buffer, CIP, and nuclease-free water in a total volume of 10
CIL. The mixture was incubated at 370C for 1 hour, and terminated by the addition of 15
CIL ammonium acetate solution. A 115 CIL volume of nuclease-free water was added,
followed by 150 CIL acid phenol-chloroform. The mixture was then vortexed thoroughly
and centrifuged for 5 minutes at room temperature and at > 10,000g. The aqueous phase
was transferred to a new tube, 150 CIL chloroform were added, and the mixture was
thoroughly vortexed and centrifuged for 5 minutes at > 10,000g. The top aqueous layer
was transferred to a new tube, 150 C1L isopropanol were added, followed by thorough
vortexing and chilling on ice for 10 minutes. The mixture was then centrifuged at
maximum speed (16,000g) for 20 minutes. The pellet was rinsed with 0.5 mL cold 70%
ethanol and centrifuged for 5 minutes at 16,000g. The ethanol was carefully removed and
discarded, and the pellet was allowed to air dry (but not completely). The pellet was
resuspended in 11 CIL nuclease-free water and placed on ice. At this point 1 CIL of the
CIP-treated RNA was reserved for the "minus-TAP" control reaction. This RNA was
carried through adapter ligation, reverse transcription and PCR in order to demonstrate
that the products generated by RLM-RACE were specific to the 5'-ends of decapped
Tobacco Acid Pyrophosphatase (TAP) treatment. CIP'd RNA, 5 C1L, was gently
mixed with TAP, 10XTAP buffer and nuclease-free water in a total volume of 10 CL.
The mixture was incubated at 370C for 1 hour.
5'RACE Adapter Ligation. CIP/TAP-treated RNA, 2 CIL, and 2 CIL of CIP-treated
RNA (minus-TAP control) was gently mixed with 1 CL 5'RACE adapter (5'-
CIL 10XRNA Ligase buffer (before use, the buffer was quickly warmed by rolling it
between gloved hands to resuspend any precipitate), T4 DNA Ligase (2.5 U/CIL), and
nuclease-free water in a total volume of 10 C1L. The mixture was incubated at 370C for 1
hour, after which it was stored at -200C.
Reverse transcription (RT). Ligated RNA, 2C1L, or minus-TAP control were
gently mixed with 4 CIL dNTP mix, 2 CIL random decamers, 2 CIL 10XRT buffer, 1 C1L
RNase inhibitor, 1 CIL M-MLV reverse transcriptase, and nuclease-free water in a total
volume of 20 CIL. The mixture was incubated at 420C (or 500C, see results) for 1 hour.
The reactions were stored at -200C.
Outer PCR. Each tube contained: 1 CIL RT reaction, 5 C1L 10XPCR buffer, 4 C1L
dNTPmix (4 mM), 2 CIL gene-specific or outer control (reverse) primer (10 CIM), 2 C1L
outer (forward) p ri mer ( 10 CM) (5'- GC TGAT GGC GAT GAAT GAAC AC TG-3'), 0.2 5
CIL Taq DNA polymerase (5 U/C1L), and nuclease-free water in a total volume of 50 CL.
A minus-template control was also included to ensure that one or more of the PCR
reagents was not contaminated with DNA.
Thermocycler parameters were as follows (Lid heating at 1100C):
Step Stage Temp/oC Duration/min
1 Initial denaturation 94 3
2 Denaturation 94 0.5
3 Annealing 59 + 21 0.5
4 Extension 722 13
5 Final extension 72 7
35 cycles of steps 2 4 were performed
1,2,3 These parameters were frequently changed to optimize the PCR. The values given
above are representative of parameters used with the GSP_OUT and control primers.
Inner nested PCR. A mixture was prepared, identical to the one for outer PCR,
except that the DNA template was now 1 C1L of the outer PCR, and 2 C1L each of both
inner primers. The sequence for the inner 5'RACE primer supplied with the kit was 5'-
CGC GGATCCGAACACTGCGTTTGC TGGCTTTGATG -3'. The thermocycler
parameters were similar except for the annealing temperature, which was typically higher
than the one used for the outer PCR.
3' RACE procedure
Reverse transcription. The following components were assembled in a nuclease-
free microfuge tube: 1 Clg total RNA (DNase-treated) from intestine or liver or control
(mouse thymus RNA), 4 CIL dNTP mix, 2 CIL 3'RACE Adapter (5'
GCGAGCACAGAATTAATACGACTCACTATAGG T12VN 3'), 2 CIL 10XRT buffer,
1 CIL RNase inhibitor, 1 C1L M-MLV reverse transcriptase, and nuclease-free water to 20
CIL. The reaction was mixed gently and incubated at 420C or 500C for 1 hour.
PCR. The procedure for the outer and inner PCR was similar to the one performed
for 5'-RACE, the only difference being the GSP and the kit-supplied primers used. The
sequences for the letters were as follows: Outer 5' -GC GAGCACAGAATTAATACGA
CT-3', Inner 5' -CGCGGATCCGAATTAATACGACTCACTATAGG-3'`
PCR amplification of entire UGT gene
Elucidation of the complete gene sequence for liver UGT from catfish by RLM-
RACE (via partial sequence overlap) enabled the design of gene specific primers which
are complementary to the gene itself as well as the untranslated region. The primers used
are shown in Table 5-6. All primers complementary to the untranslated region (UTR)
were designed with the help of Primer3 software, except for the pair of primers that were
complementary to the exact start and end of the gene (LIVUGTF1 and LIVUGTR1
Table 5-6. Primers used for amplifying liver and intestinal UGT gene
GSP ID Sequence (5' 3') Start position
UTR Fl CTGCTTCCTCTAGACGTAATTAGAAAC 40
UTR F2 CTCACATTCCTCCTCCTTCTTTTT 76
UTR R1 GAAC GT GGT GAT GAGAACACTATAACT + 121
UTR R2 TAGTGACATCATAACAACCGTAACTGC + 190
LIVUGT Fl ATGCCTCGTCTTCTTGCAGCTCTCTGT 1
LIVUGT R1 TCACTCCTTTTTGCTCTTCTGAGCCCT 1568
Due to the length of the amplicon (~1.6kb), Super Taq Plus polymerase (Ambion
Inc) was employed. This enzyme results in higher yields with amplicons >1kb. In
addition, this enzyme mixture has a proof-reading ability, which will be important for
future expression of the gene, as well as providing greater fidelity and processivity than
ordinary thermos Taq DNA polymerase. An extension temperature of 680C and an
extension time of 1.75 min were used for this PCR. Different combinations of UTR