Cloning and Characterization of Rhodotorula glutinis Thymine
Jonathan W. Neidigh, Agus Darwanto, Adides A. Williams, Nathan R. Wall, and
Lawrence C. Sowers*
Department of Basic Sciences, Loma Linda UniVersity School of Medicine, Alumni Hall for Basic Science, Room
101, 11021 Campus Street, Loma Linda, California 92350
ReceiVed NoVember 24, 2008
Thymine hydroxylase (TH) is a member of the R-ketoglutarate-dependent nonheme iron dioxygenase
family that includes a series of DNA repair proteins including alkB. Substantial interest in this family of
enzymes derives from their capacity to modify DNA bases and precursors by oxidation. Previously, a
sequence has been published for cloned Rhodotorula glutinis TH. However, the minimal reported activity
of this enzyme, coupled with inconsistencies with previously published mass spectrometry data, compelled
us to reexamine TH. The sequence reported here differs from the previously reported sequence at two
amino acid positions and is consistent with previously reported mass spectrometry data. The cloned enzyme
characterized in this report displayed substantial activity, indicating that the sequence differences are
critical for activity. The substrate selectivity of TH against a series of pyrimidine analogues is consistent
with that reported for the wild-type enzyme and, in part, explains the mode of selection of uracil analogues.
A preliminary model of the active site has been constructed for the purposes of comparing TH with
other members of this family. TH and alkB share in common the capacity to oxidize N-methyl groups.
However, TH has the added capacity to oxidize the 5-methyl group of thymine, a property that is potentially
important for enzymes that could act on DNA and modify DNA-protein interactions.
The oxidation of thymine to 5-hydroxymethyluracil (HmU)1
was initially identified in rat liver and leukocytes (1, 2).
Subsequent investigation (3-5) in Neurospora crassa identified
and characterized an enzyme, thymine hydroxylase (TH), that
catalyzes the sequential oxidation of thymine to 5-carboxyuracil
(CarboxyU) via the intermediate oxidation products HmU and
5-formyluracil (FoU). TH was found to require iron and
R-ketoglutarate, making it a member of the nonheme iron
R-ketoglutarate-dependent dioxygenases (6, 7). This is an
important class of metabolic and catabolic enzymes found in a
wide range of organisms including bacteria, yeast, plants, and
In addition to studies of TH isolated from N. crassa, TH
isolated from Aspergillus nidulans (9, 10) and Rhodotorula
glutinis (11-13) has been characterized. In 1993, Stubbe and
co-workers reported on studies of a partially purified fraction
from the red mold R. glutinis (14, 15). In their studies, they
provided biochemical characterization of the TH activity, as well
as sequence information of the protein. An N-terminal sequence,
VSSGIVPPINFEPFLSGTPEDKLATA, was identified by Ed-
man degradation. An internal sequence was also identified in
the protein catalytic center by reaction with a pyrimidine
substrate analogue and mass spectrometry. The sequence of the
internal trypsin-generated peptide was NSIAF?SNPSLR, where
the unknown residue was proposed to be tyrosine (14). In
subsequent studies, the unknown amino acid was identified as
In 2005, Smiley and co-workers created a degenerate DNA
primer based upon the N-terminal peptide identified by Stubbe
and co-workers and used this primer to identify the TH sequence
from a cDNA library derived from R. glutinis (17). The
N-terminal amino acid sequence derived from the cloned TH
was MVSSGIVPVINFEPFLSGTP, consistent with the peptide
identified by Stubbe and co-workers, except that a valine
replaces proline at position 9. An internal predicted peptide,
NSIAFFSNPSLR, is also found within the DNA sequence of
cloned TH, consistent with the internal peptide identified by
Stubbe and co-workers. However, the internal peptide predicted
by the sequence of Smiley et al. (17) is preceded by cysteine,
not an arginine or lysine, and should therefore not give rise to
the tryptic fragment observed by Stubbe and co-workers.
Additionally, the enzyme cloned by Smiley et al. (17) had
negligible activity for the conversion of thymine to HmU.
Upon the basis of their cloned sequence, Smiley et al. (17)
were able to compare TH with the sequences of other known
proteins. The DNA sequences of the related species N. crassa
and A. nidulans are known (18-20), and both are reported to
have TH activities (4-7, 9, 10, 21). Furthermore, the cloned
TH sequence was sufficient to establish that the R. glutinis TH
was a member of the R-ketoglutarate dioxygenase family,
consistent with its known substrate properties. Enzymes within
this dioxygenase family can utilize either an RxS or an RxN
sequence motif for binding R-ketoglutarate. Although the THs
from N. crassa and A. nidulans are found within the RxS motif
* To whom correspondence should be addressed. Tel: 909-558-4480.
Fax: 909-558-4035. E-mail: firstname.lastname@example.org.
1Abbreviations: TH, thymine hydroxylase; ATCC, American Type
Culture Collection; MTBSTFA, N-methyl-N-[tert-butyldimethylsilyl]trif-
luoroacetamide; TBDMCS, tert-butyldimethylsilylchlorosilane; BCA, bicin-
choninic acid; FPLC, fast protein liquid chromatography; PCR, polymerase
chain reaction; GST, glutathion-S-transferase; BLAST, Basic Local Align-
ment Search Tool; NCBI, National Center for Biological Information; HmU,
5-hydroxymethyluracil; FoU, 5-formyluracil; CarboxyU, 5-carboxyuracil;
1MeU, N1-methyluracil; 1MeThy, N1-methylthymine; 1MeHmU, N1-
methyl-5-hydroxymethyluracil; 5MeC, 5-methylcytosine; 3MeU, N3-me-
thyluracil; 6MeU, 6-methyluracil.
Chem. Res. Toxicol. 2009, 22, 885–893
10.1021/tx8004482 CCC: $40.75
2009 American Chemical Society
Published on Web 04/03/2009
subgroup, the RxN sequence of Smiley et al. (17) surprisingly
placed the TH from R. glutinis within a different subgroup that
contained the alkB family of DNA repair enzymes (17).
Recently, there has been substantial interest in the Escherichia
coli protein, alkB, which repairs alkylation damage on DNA
(22). Bioinformatic analysis of the alkB amino acid sequence
by Aravind and Koonin predicted that alkB is a member of the
nonheme iron R-ketoglutarate-dependent dioxygenases; subse-
quent verification of this prediction demonstrated that alkB
oxidizes N-methyl groups spontaneously, regenerating the parent
DNA base by release of formaldehyde (23-26).
This class of proteins is of interest for further study as it might
reveal additional proteins involved in the modification of nucleic
acids that impact the control of gene expression (Figure 1). A
“demethylation” pathway for 5-methylcytosine (5MeC) was
suggested based on biological observations, but no clear data
have emerged to provide a definitive mechanism (27-30). One
model suggests that a dioxygenase of the alkB family might
act on 5MeC; however, no family member with this activity
has yet been identified (31).
Upon the basis of apparent inconsistencies in data previously
reported for R. glutinis TH (14, 17), in conjunction with mass
spectrometry data presented in this manuscript, we sought to
independently determine the sequence of TH. The derived
protein sequence reported here shows differences with both
previous reports but allows reconciliation of most of the data.
The cloned enzyme with this new sequence demonstrates
significant enzymatic activity, allowing the construction and
testing of several site-directed mutants as well as examination
of substrate preferences. A preliminary structural model has
allowed comparison of the active site residues of R. glutinis
TH with the alkB family of DNA repair enzymes. A previously
unknown activity that converts thymine to HmU in telomere
DNA regions of trypanosomatids has been recently identified
(32) that shares several key active site residues, suggesting that
further proteins that oxidize DNA bases are likely to be found.
Materials and Methods
Materials. The same strain (ATCC 2527) examined by Stubbe
and co-workers and Smiley et al. (14, 15, 17) of R. glutinis was
obtained from the American Type Culture Collection (ATCC,
Manassas, VA) and was grown in bulk by Encore Technologies
(Minnetonka, MN). Sodium ascorbate, R-ketoglutarate, ferrous
sulfate, and HEPES buffer were obtained from Fisher Scientific
(Tustin, CA). The silylating agent, N-methyl-N-[tert-butyldimeth-
ylsilyl]trifluoroacetamide (MTBSTFA) with 1% tert-butyldimeth-
ylsilylchlorosilane (TBDMCS), and the bicinchoninic acid (BCA)
protein concentration assay kit were obtained from Pierce (Rock-
ford, IL). Deuterium-enriched thymine-R,R,R,6-d4 (>98% enrich-
ment) was obtained from Cambridge Isotope Laboratories (Andover,
MA). Synthetic2H2-5-hydroxymethyluracil was prepared from2H2-
formaldehyde and uracil as previously described (33, 34). Sequenc-
ing grade trypsin, uracil, thymine, FoU, HmU, CarboxyU, and all
other laboratory reagents were obtained from Sigma-Aldrich (St.
Isolation of TH from R. glutinis. R. glutinis was grown in media
containing thymine as the only nitrogen source, essentially as
described previously by Thornburg et al. (14) without significant
modification. The primary difference is that we substituted a GC/
MS method for the measurement of TH activity as described below.
The purification steps involved ammonium sulfate precipitation,
Sephadex G-25, DE-52, and Sephadex G-100 chromatography. The
final monoQ fast protein liquid chromatography (FPLC) purification
resulted in a loss of enzyme activity as previously observed (14).
Protein concentrations were measured with the BCA kit using a
Cary 100 spectrophotometer (Varian, Walnut Creek, CA).
Peptide Mass Spectrometry Analysis. Purified TH was reduced
and alkylated prior to enzymatic digestion using 9.5 mM dithio-
threitol and 32 mM iodoacetamide, respectively. The reduced and
alkylated protein was digested with trypsin (1:50 ratio of protease
to substrate) overnight at 37 °C in 0.6 M urea and 10 mM Tris
buffer at pH 7.8. Tryptic peptides were analyzed with a Ther-
moFinnigan LCQ Deca XP mass spectrometer (Thermo, Waltham,
MA) equipped with a PicoView 500 nanospray apparatus and a 10
cm × 75 µM internal diameter C18 Biobasic bead column (New
Objective, Woburn, MA). Mobile phase A consisted of 2% aqueous
acetonitrile with 0.1% formic acid, while mobile phase B consisted
of 98% acetonitrile, 2% water, and 0.1% formic acid. Injected
peptides were eluted with a linear gradient from 20 to 60% mobile
phase B following a 5 min delay. Instrumental parameters were
previously optimized using a 5 mM solution of angiotensin infused
at a flow rate of 250 nL/min. The LC/MS/MS data for the purified
TH were analyzed using the de novo sequencing software Mas-
sAnalyzer 1.02 (Amgen, Thousand Oaks, CA) and PEAKS (Bio-
informatic Solutions Inc., Waterloo, Ontario, Canada). The software
program TurboSEQUEST implemented in Bioworks (Thermo) was
also used to analyze the LC/MS/MS data using an appropriate
database containing a TH protein sequence and the SEQUEST
Cloning and Isolation of Recombinant TH. The RNA from
R. glutinis was isolated using TRIzol reagent according to the
instructions from the manufacturer (Invitrogen, Grand Island, NY).
The mRNA obtained was reverse transcribed using the SuperScript
first-strand standard protocol (Invitrogen). First-strand cDNA
synthesis was performed by priming with 20 pmol of oligo-dT in
a 20 µL reaction mix containing 10 mM each dATP, dCTP, dGTP,
and dTTP, 40 U/µL RNaseOUT recombinant ribonuclease inhibitor,
and 50 U/µL SuperScript II reverse transcriptase. The reverse
transcriptase reaction was stopped by cooling to 4 °C for 10 min.
The resulting cDNA was then amplified with Phusion high-
fidelity DNA polymerase (New England Biolabs, Beverley, MA)
according to the manufacturer’s instructions. Thirty-six cycles of
Figure 1. Reactions catalyzed by alkB (left) and TH (right). Each
enzyme-catalyzed reaction also consumes one molecule of oxygen and
R-ketoglutarate and produces one molecule of succinate and carbon
Chem. Res. Toxicol., Vol. 22, No. 5, 2009 Neidigh et al.
polymerase chain reaction (PCR) (10 s at 98 °C, 30 s at 61 °C,
and 32 s at 72 °C) were performed. The following sequences of
the oligonucleotides used for PCR amplification were based upon
the cDNA sequence of TH [GenBank accession number AY622311,
National Center for Biological Information (NCBI)] reported by
Smiley et al. (17): 5′-CGG CGG GGA TCC ATG GTC TCG TCT
GGC ATC GTC-3′ (sense, carries a BamH1 restriction site) and
5′-CTT TTC CTT TTG CGG CCG CTT GAG GGC ACT GCT
GCA TTA C-3′ (antisense, carries a Not1 restriction site).
After resolution of the products by gel electrophoresis on a 1%
agarose gel, the expected 1035 base pair product was extracted using
the QIAquick gel extraction kit (Qiagen, Valencia, CA) and ligated
into the glutathion-S-transferase (GST) fusion expression vector
pGEX-4T-1 previously digested with BamH1 and Not1. Ligated
products were electroporated into E. coli BL21 Star DE3 (Invit-
rogen). The plasmid was isolated and purified using a QIAprep
Spin Miniprep kit (Qiagen), and both strands of the insert were
sequenced (Davis Sequencing, Inc., Davis, CA) using the primers
5′-TTG GTG GTG GCG ACC ATC CTC CAA-3′ (pGEXmcs5′)
and 5′-CTG CAT GTG TCA GAG GTT TTC ACC-3′ (pGEXmcs3′).
Purification of Recombinant TH. To purify TH expressed as a
recombinant protein in E. coli, 2 L of E. coli BL21 Star DE3
carrying the GST-TH construct were grown in LB broth with 50
µg/mL ampicillin at 37 °C until A600) 0.6-0.7 and then induced
overnight with 0.2 mM isopropyl-?-D-thiogalactopyranoside at 30
°C. The cells were harvested by centrifugation, resuspended in lysis
buffer (50 mM Tris, pH 7.4, 150 mM NaCl, 5 mM MgCl2, and
0.01% Triton-X) supplemented with 1 mM dithiothreitol and 1 mM
phenylmethylsulfonyl fluoride, and incubated at room temperature
for 30 min. Lysis was then completed by sonicating the suspension
on ice using a Branson Sonifier Cell Disruptor 200 and six bursts
of 10 s each, with a 90 s interval between pulses. The lysate was
clarified by centrifugation (12000 rpm for 30 min at 4 °C). The
supernatant was then mixed with swelled glutathion-agarose beads
(Sigma-Aldrich) and incubated at 4 °C overnight with gentle
agitation. The suspensions were centrifuged at 3000 rpm for 5 min
at 4 °C, and the beads were washed twice with lysis buffer and
then washed twice with thrombin buffer (50 mM Tris, pH 7.4, 150
mM NaCl, 5 mM MgCl2, and 1 mM dithiothreitol).
The recombinant protein GST-TH was resuspended in 15 mL
of thrombin buffer, cleaved with 100 U/mL thrombin (Sigma-
Aldrich) at 37 °C for 1 h, and then purified by FPLC using a
SuperDex 75 column (GE Healthcare, Waukesha, WI). The protein
was concentrated using centricon YM-10 membranes (Millipore,
Billerica, MA). The protein concentration was determined using
the BCA protein assay reagent kit (Pierce Chemical Co.). The
protein was analyzed on NuPAGE 4-12% Bis Tris Gels (Invitro-
gen, Carlsbad, CA), stained with Simply Blue (Invitrogen) and
confirmed by mass spectrometry. TH was digested with trypsin
following the protocol of Matsudaira (36) and analyzed by MALDI-
TOF-MS (Bruker Daltonics, Billerica, MA) and LC/MS/MS as
Bioinformatics Analysis and Comparative Modeling. A Basic
Local Alignment Search Tool (BLAST) search using the deduced
amino acid sequence of TH was performed to identify TH
homologues in other microorganisms with an E value of less than
1 × 10-4using the NCBI Web site with default parameters. A
multiple sequence alignment was generated using clustalX as
implemented in BioEdit (37).
The multiple sequence alignments were manually edited for
comparative modeling to ensure alignment of residues critical for
catalytic activity (see the Results and Discussion). Potential
modeling templates for TH were obtained by using a BLAST search
against the database of sequences corresponding to proteins with
known structure, which suggested anthocyanin synthase as the best
template model. The locations of gaps between the TH and the
anthocyanidin synthase amino acid sequences were manually
modified to maximize alignment of conserved residues and to ensure
that gaps were located in loops on the surface of the structural
model. The secondary structure of all sequences was predicted using
Profsec (http://cubic.bioc.Columbia.edu/predictprotein/) with the
default options (38). The predicted secondary structure of TH, when
compared with the observed secondary structure of anthocyanidin
synthase, was helpful when determining the best sequence alignment.
The TH model was built using InsightII (Accelrys, San Diego,
CA) to assign coordinates for the structurally conserved residues
and to predict the coordinates of loops randomly or based on a
database search of loop conformations within the protein databank.
An alignment of TH and anthocyanidin synthase detailing the
template coordinates used to build our model of the TH active site
is included in the Supporting Information.
Site-Directed Mutagenesis. Site-directed mutagenesis was used
to mutate Arg-285 to the Cys-285 observed by Smiley et al. (17),
generating a TH R285C mutant. The pGEX4T-1ADTH plasmid
containing the TH sequence was mutagenized using the Quick-
Change site-directed mutagenesis kit (Stratagene, La Jolla, CA)
following the manufacturer’s protocol using the primers 5′-GAT
GAC GCC CCG CTG CAA CTC GAT CGC-3′ (sense) and 5′-
GCG ATC GAG TTG CAG CGG GGC GTC ATC-3′ (antisense)
(Sigma-Genosys, Woodland, TX). The expected mutation was
confirmed by DNA sequencing (Davis Sequencing). Furthermore,
the protein was expressed and purified as above, and the mutant
TH R285C was confirmed by MALDI-TOF-MS using peptide
mapping analysis after trypsin digestion. A similar strategy was
used to generate the mutants N187D, N293D, and N250D.
Steady-State Enzyme Kinetics. The activity and kinetics of TH,
also known as thymine dioxygenase (EC 22.214.171.124), isolated from
R. glutinis or a recombinant protein and its mutants described above
were examined. The purity of TH samples was assessed by PAGE
stained with a coomassie blue stain and was g95% in all cases.
The standard enzyme reaction and buffer conditions defined by
Stubbe and co-workers (14) and used throughout this report, unless
explicitly stated otherwise, contained 0.45 mM R-ketoglutarate, 11
µM ferrous sulfate, 2.3 mM sodium ascorbate, and 50 mM HEPES
at a pH of 7.5 in a total volume of 200 µL. No enzyme activity
was observed when R-ketoglutarate or iron was omitted from the
enzyme reaction. The pH was varied from 6.5 to 8.0 to determine
that a pH of 7.5 was optimum for TH activity. All steady-state
kinetic assays were performed with 5-10 µg of protein in a 200
µL reaction volume. The assays were performed in 2 mL vials
placed in an aluminum block heated to 30 °C to control the reaction
temperature in all cases. When measuring the specific enzyme
activity of TH or comparing the relative activity of TH and its
mutants, the reaction was initiated by adding the substrate to obtain
a final concentration of 0.9 mM thymine and stopped 1 min later.
When testing the TH activity of fractions during purification of
the wild-type enzyme, deuterium-enriched thymine-R,R,R,6-d4 was
used as the substrate to eliminate interference from unlabeled
thymine present in R. glutinis extracts. The enzyme reaction was
stopped with the addition of hot absolute ethanol (200 µL) followed
by heating at 60 °C for 5 min. No residual TH activity was observed
following ethanol addition, and ethanol did not interfere in
subsequent mass spectrometry measurement of the pyrimidine
The kinetics for pyrimidine substrates were determined by
substituting thymine for the appropriate pyrimidine substrate and
varying its concentration. The concentrations of substrate ranged
from 10 µM to 2 mM. The standard reaction conditions defined
above and a 1 min reaction time were used. After addition of
internal standard, the GC/MS method described below was used
to quantitate product formation. Kinetics data were analyzed using
Prism 4 (GraphPad, La Jolla, CA) to perform nonlinear least-squares
analysis for the determination of Kmand kcatvalues along with the
calculated standard error. We note that steady-state kinetic param-
eters determined for TH must be interpreted with caution as the
pyrimidine concentrations needed to observe maximum enzyme
velocity approach the aqueous solubility of the pyrimidines
examined here. The high apparent Km value of FoU (near its
solubility limit) prevented accurate observation of a velocity
maximum and limited confidence in the nonlinear analysis; the
traditional Lineweaver-Burk, Eadie-Hofstee, and Hanes-Woolf
plots were also used in this case to derive Kmand kcat.
Characterization of Recombinant Thymine HydroxylaseChem. Res. Toxicol., Vol. 22, No. 5, 2009 887
Analysis of TH Reaction Products by GC/MS. Thymine and
its oxidation damage products can be separated, identified, and
quantified by GC/MS methods (39). After the enzyme reaction was
stopped, 20 µL of an internal standard solution containing 2-hy-
droxy-4-methylpyrimidine or2H2-hydroxymethyluracil was added.
The internal standard
examining the differences in activity for TH mutants, and 2-hy-
droxy-4-methylpyrimidine was used when determining the kinetics
of oxidation for a series of pyrimidine analogues. The internal
interference by R-ketoglutarate, which has the same mass following
derivatization (431 m/z). The resulting mixture was dried under
reduced pressure. A silylating mixture of acetonitrile (100 µL) and
silylating agent (MTBSTFA + 1% TBDMCS, 100 µL) was added,
and the sealed vial was heated at 140 °C for 30 min.
Pyrimidine products from the TH assay were measured with a
Hewlett-Packard 6890 gas chromatograph containing an Ultra-2
column interfaced with a Hewlett-Packard 5973 mass spectrometer.
An aliquot (1 µL) of the silylation reaction mixture was injected
onto the GC/MS with an inlet temperature of 250 °C. The oven
temperature was held constant at 100 °C for 2 min, increased at a
rate of 30 °C/min for 2.7 min to 180 °C, then increased at a rate of
16 °C/min for 6.3 min to 280 °C, and held at 280 °C from 2 min.
The amount of thymine and corresponding oxidation products was
determined by comparing the integrated area of the specific analyte
at a selected mass with that of the internal standard. The limit of
detection for all pyrimidines examined was less than 1 pmol injected
onto the column. The retention times for all compounds examined
and the ions used for the quantitation of oxidation products are
listed in the Supporting Information. Relative peak areas were
converted to relative concentrations based upon comparison of the
experimental areas with standard curves (R2g 0.99) constructed
for each of the analytes vs the internal standard.
Preparation and Characterization of N1-Methyl-5-hydroxym-
ethyluracil (1MeHmU). The 5-hydroxymethyl derivative of N1-
methylthymine (1MeThy) was synthesized using a method previ-
ously used to convert 2′-deoxyuridine to 5-hydroxymethyl-2′-
deoxyuridine (40). The synthesis of 1MeHmU is outlined in Scheme
1. To a solution of formaldehyde (CH2O, 20% w/w, 2.7 mL) and
triethylamine (TEA, 4 mL, 2.9 × 10-2mol) was added 1-methy-
luracil (1MeU, 481 mg, 3.8 × 10-3mol). The reaction mixture
was then heated at 60 °C for 7 days. Triethylamine was then
removed using a rotary evaporator, and the resulting residue was
dissolved in methanol (MeOH) and purified by normal phase
column chromatography (Silica Gel, Sigma-Aldrich). Unreacted
1MeU was eluted using a solvent mixture of 4% MeOH in
dichloromethane (CH2Cl2). 1MeHmU was then recovered using a
solvent mixture of 9% MeOH in CH2Cl2. The final product of 98.5
mg (17% yield) was obtained as a white powder: GC/MS C6H8N2O3
156.1H NMR (DMSO-d6): δ 3.24 (s, 3H, N1-CH3), δ 4.11 (d, 2H,
C5-H), δ 7.51 (s, 1H, C6-H), δ 11.22 (s, 1H, NH).
2H2-hydroxymethyluracil was used when
15N2,2H2-5-hydroxymethyluracil was not used due to
TH from R. glutinis Was Isolated and Examined by
Mass Spectrometry. Wild-type TH was isolated from R.
glutinis according to previously published methods (14). In the
procedure reported here, the use of radiotracers was substituted
by a GC/MS method to monitor the conversion of thymine to
HmU. To use this nonradioactive procedure to follow enzyme
activity during isolation from the host organism that contained
endogenous thymine and HmU, we used deuterium-enriched
thymine as the substrate. Substrate requirements and the apparent
molecular weight of the isolated enzyme were consistent with
previously published accounts (11, 14).
The isolated enzyme was digested with trypsin, and the
corresponding fragments were characterized by mass spectrom-
etry. Although insufficient data were obtained to determine the
entire protein sequence, one significant peptide was identified
as NSIAFFSNPSLR (Figure 2). This peptide is consistent with
one identified by Stubbe and co-workers as a critical sequence
within the substrate-binding domain (14). The identification of
this peptide, although consistent with the previous results of
Stubbe and co-workers (14), is inconsistent with the sequence
reported by Smiley et al. (17). Although the DNA sequence
provided by Smiley et al. (17) would encode this peptide
sequence, it is preceded by a cysteine residue and therefore
would not allow formation of the tryptic fragment described
TH Was Cloned and the Sequence Obtained Here Is
Different from the Previously Reported Sequence (17). TH
was cloned from the same R. glutinis strain used by the Stubbe
laboratory and Smiley et al. (14, 17). The PCR primer sequence
used here for cloning (Materials and Methods) was also based
upon the N-terminal peptide sequence reported by Stubbe and
co-workers (14) but differed from the primers used by Smiley
et al. (17). The DNA sequence of TH obtained here is presented
in Figures S1-S4 and Table S1 of the Supporting Information
and has been deposited in GenBank as accession number
The DNA sequence reported here predicted a protein of 332
amino acids, consistent with Smiley et al. (17), and a predicted
molecular mass of 37 kDa, consistent with the value reported
by Stubbe and co-workers (14) for the wild-type enzyme.
Substrate requirements and pH optimum for the cloned enzyme
were indistinguishable from the isolated enzyme (11, 14). In
accord with Smiley et al. (17), the DNA sequence predicted
valine at position 9, which differs from Edman sequencing data
from Stubbe and co-workers (14), who observed proline at this
The DNA sequence reported here differs from the previously
reported DNA sequence (17) at two positions. At nucleotide
position 551, we observe guanine, whereas adenine was
previously reported, and at position 853, we observe cytosine,
whereas thymine was previously reported. The corresponding
Scheme 1. Synthesis of 1Me5HmU
Figure 2. Tryptic peptides of TH purified from R. glutinis were
examined by LC/MS/MS. One highly significant peptide, NSIAFFS-
NPSLR, was identified. The mass spectrum shows the fragment ions
when the peptide corresponding to the indicated sequence is fragmented
by collision-induced dissociation. The masses expected for the y1-y11
ions are 175.12, 288.20, 375.23, 472.29, 586.33, 673.36, 820.43, 967.50,
1038.54, 1151.62, and 1238.65 m/z. The masses expected for the
b2-b11 ions are 202.08, 315.17, 386.20, 533.27, 680.34, 767.37,
881.41, 978.47, 1065.50, and 1178.58 m/z. The precursor mass and
the indicated fragments are consistent with the indicated sequence.
Chem. Res. Toxicol., Vol. 22, No. 5, 2009 Neidigh et al.
codons for our protein sequence at positions 184 and 285 would
be CGG and CGC, both encoding arginine, whereas the previous
study reported CAG and TGC, encoding glutamine and cysteine,
respectively. Our translated protein sequence contains arginine
at position 285, which would generate the tryptic fragments
observed by both our group and Stubbe and co-workers (14).
The TH Cloned Here Has Significant Activity against a
Series of Pyrimidines and Appears to Interact with the
Hydrogen-Bonding Face. Using the GC/MS method reported
here, the steady-state kinetic parameters for cloned TH were
measured. An example of the kinetic results for the conversion
of thymine to HmU is shown in Figure 3, and steady-state
kinetic parameters for additional pyrimidine analogues are
reported in Table 1. The corresponding structures are shown in
Figure 4. We observed that the cloned TH converted thymine
to HmU, FoU, and CarboxyU. Previous studies (14) were unable
to demonstrate that one protein activity could carry out all of
the sequential reactions, as the wild-type protein cannot be
purified to homogeneity and still retain enzymatic activity.
Within this series, Kmvaries more among the analogues than
does kcat, and the relative reactivity is Thy > HmU > FoU,
consistent with previously reported trends (14).
We also examine the activity of TH toward 1MeU and
1MeThy. The methyl group of 1MeU can be oxidized and
spontaneously lost, generating uracil and formaldehyde (Figure
4). Therefore, a methyl group in either the C5-position (Thy)
or the N1-position (1MeU) is oxidized by TH. The methyl
groups of 1MeThy at either the C5- or the N1-position could
be oxidized, depending upon the orientation with which it binds
to the active site. Oxidation of the N1-methyl group would
generate thymine, whereas oxidation at the 5-position would
generate 1MeHmU. Previously, 1MeHmU was proposed as a
product from TH-mediated oxidation of 1MeThy (14). We report
here that 1MeHmU was prepared by chemical synthesis and
characterized by NMR spectroscopy and high-resolution mass
spectrometry, allowing confirmation of the reactivity of 1MeThy.
As shown in Figure 4, 1MeThy could generate HmU via two
pathways. We also observed that TH does not convert 1MeHmU
to HmU, indicating that HmU must arise from 1MeThy by
oxidation first at the N1-position, generating thymine, followed
by oxidation in the 5-position, generating HmU. We also
observed that TH does not act on the nucleoside thymidine.
TH from R. glutinis does not act on 3MeU or 5MeC. A
methyl group in the N3-position would interfere with interaction
whereas a methyl group in the N1-position would not. The
apparent discrimination for thymine over 5MeC suggests that
hydrogen-bonding interactions with the O4 and N3H groups of
thymine are likely important for recognition.
We considered the possibility that the activity of TH might
be modulated by self-oxidation of one of the active site aromatic
amino acid residues, as has been reported for other enzymes
with similar substrate requirements (41-43). We also considered
the possibility that self-oxidation could account for earlier
ambiguity with respect to either phenylalanine or tyrosine at
position 291 (15, 16). Cloned TH was purified and studied by
MS prior to and after exposure to iron and R-ketoglutarate with
or without thymine. The tryptic peptide with phenylalanine at
position 285 was observed; however, no peptides containing
tyrosine or other oxidized phenylalanine derivatives at position
291 were observed, although the enzyme was incubated under
conditions that resulted in significant conversion of thymine to
HmU. This result indicates that the chemical oxidation of
phenylalanine to tyrosine is not involved in modulating the
activity of R. glutinis TH.
TH from R. glutinis Shows Substantial Homology with
TH from Related Yeast Species. The identity of the amino
acid at position 285 distinguishes our sequence from the
previously reported sequence and is significant with respect to
comparison among the yeast TH homologues. As shown in
Figure 5, the sequence reported here for this region of the protein
shows considerable homology with protein sequences derived
from the genome sequences of N. crassa and A. nidulans
(19, 20), whereas substantially less overlap is observed with
the previously reported sequence (17).
TH from R. glutinis Is a Member of the Nonheme Iron
r-Ketoglutarate Dependent Dioxygenase Family. Upon the
basis of substrate requirements, TH could be placed within the
family of dioxygenases. TH can now be placed within this
family upon the basis of sequence as well. Members of the
dioxygenase family contain sequence motifs to bind iron and
R-ketoglutarate. Iron is bound by the highly conserved HxD-
xn-H sequence motif, where x is any amino acid and xnis a
variable number of amino acids from 40 to 153 (8). The C5
carboxylate of R-ketoglutarate is stabilized by either the RxN
or the RxS sequence motif. The previously reported sequence
(17) contained RCNS, which allowed TH to be placed within
the overall dioxygenase family; however, as it had the RxN
motif, it was placed within the alkB subfamily as opposed to
the subfamily containing the other yeast TH homologues. In
contrast, we report that the corresponding sequence is RRNS,
allowing R. glutinis TH to be placed within the same subfamily
as the other yeast homologues.
Figure 3. Kinetics of recombinant R. glutinis TH oxidation of thymine.
All reactions used the standard conditions given in the Materials and
Methods and were performed in triplicate. The error bars indicate the
standard deviation of observed values. One minute reaction times result
in Kmand kcatvalues of 145 µM and 67 min-1, respectively.
Table 1. Steady-State Kinetic Parameters for the Oxidation
of Pyrimidine Substrates by THa
1. thymine67 ( 7145 ( 15
2. HmU25 ( 3 230 ( 43
3. FoU68 ( 21700 ( 200
5. 1MeU 16 ( 2110 ( 38
7. 1MeThy 26 ( 21300 ( 220
aThe structures of the substrates examined are found in Figure 4. The
sensitivity of the GC/MS assay corresponds to a detection limit for kcat
values of less than 0.01 min-1given the quantity of enzyme used. The
Kmand kcatvalues are reported with the standard error.
Characterization of Recombinant Thymine Hydroxylase Chem. Res. Toxicol., Vol. 22, No. 5, 2009 889
The Proposed Active Site of R. glutinis TH Can Be
Modeled Based upon Homology with Similar Enzymes. The
nonheme iron R-ketoglutarate-dependent dioxgenases comprise
a large and varied group of metabolic enzymes. A BLAST
search against a database of proteins with known structure
yielded isopenicillin N synthase and anthocyanidin synthase,
whose sequences are significantly (BLAST scores of 63 and
52, respectively) similar to that of R. glutinis TH and could
therefore serve as potential template structures. Anthocyanidin
synthase (PDB file 1GP5, 1GP6) was chosen as a template
structure because structures with bound R-ketoglutarate (or
succinate and oxygen) and iron were available (44). The
structural model was generated as described in the Materials
and Methods and is shown in Figure 6. When Arg-285 is
modeled as the residue that binds R-ketoglutarate, the TH
sequence RRNSIAFFSN forms a ?-strand with the underlined
residues oriented toward the active site.
The Homology Model Led to a Series of Mutant
Proteins That Were Expressed and Examined. To test the
importance of specific amino acids to the activity of R. glutinis
TH, a series of mutants were constructed. The first mutant was
R285C, which recreated the sequence of Smiley et al. (17) at
one of the two positions where our respective sequences differed.
Figure 4. Proposed reaction scheme for the pyrimidines examined in this study. The compounds shown are (1) thymine, (2) HmU, (3) FoU, (4)
CarboxyU, (5) N1-methyluracil (1MeU), (6) uracil, (7) 1MeThy, (8) 1MeHmU, (9) 5MeC, (10) N3-methyluracil (3MeU), and (11) 6-methyluracil
Figure 5. Sequence alignment of TH enzymes from R. glutinis (RGTH),
N. crassa (NSTH), and A. nidulans (ANTH). The conserved arginine
that binds R-ketoglutarate is highlighted. The alignment of the Smiley
et al. (17) sequence tests the proposition that the conserved arginine
that binds the R-ketoglutarate is Arg-284 instead of Arg-285.
Figure 6. Active site model of recombinant R. glutinis TH. The active
site was modeled using the published coordinates of anthocyanadin
synthase; the alignment used during modeling is included in the
Supporting Information. The active site residues discussed in the text
are shown, and the labels indicate the residue numbering of R. glutinis
Chem. Res. Toxicol., Vol. 22, No. 5, 2009Neidigh et al.
Arg-285 is within the active site as proposed by Stubbe and
co-workers (14), and the assignment of arginine at this position
allows close alignment among the yeast homologues. In our
hands, the R285C mutant had no detectable activity for the
conversion of thymine to HmU. This result suggests that the
sequence reported here is likely correct and confirms
the importance of Arg-285 as a critical active site residue.
We also prepared a series of mutants in which asparagine
was converted to aspartate to provide additional information
on the mechanism by which TH selects thymine analogues over
5MeC analogues. We note that in previous studies with uracil
DNA glycosylase and thymidylate synthase, discrimination
between thymine and cytosine analogues could be reversed by
substituting asparagine for aspartate (45-47). We therefore
mutated several asparagine residues near the active site to
aspartate residues. The kcat values of the mutants N187D,
N250D, and N293D, respectively, for thymine were found to
be 14.9 ( 0.3, 0.67 ( 0.07, and 3.2 ( 0.3 min-1, whereas the
corresponding Kmvalues were observed to be 67 ( 8, 1150 (
280, and 2100 ( 450 µM. While the first substitution had only
a modest impact on the enzyme activity, the latter two
significantly diminished enzyme activity through increasing Km.
Neither wild-type TH nor any of the mutants examined here
had any measurable activity against 5MeC. The data reported
here suggest that discrimination between thymine and 5MeC
cannot be attributed to a single amino acid.
The DNA bases in all living organisms are persistently
damaged by oxidation and require continuous repair (48, 49).
Damage can occur at multiple sites, including pyrimidine methyl
groups found on both thymine and 5MeC (50, 51). Although
the corresponding 5-hydroxymethyl pyrimidines are not mis-
coding, repair glycosylases have been identified in mammals
(52, 53). The 5-hydroxymethyl analogues of both thymine and
5MeC are known to interfere with sequence-specific DNA
protein interactions that could influence epigenetic programming
of gene activity (54-57). It has been suggested that cytosine
methylation patterns could be modulated by a “demethylase”
activity that acts by oxidative demethylation; however, no such
activity has yet been identified (27-30).
TH is an enzymatic activity that similarly oxidizes thymine
to HmU. TH could therefore serve as a model for other enzymes
that oxidize pyrimidine methyl groups and further assist in the
identification of other as yet unidentified enzymes that catalyze
similar chemistry. Interestingly, a related enzyme has recently
been identified in trypanosomatids that converts thymine
residues to HmU in telomeric DNA (32). TH is also related by
substrate requirements to the alkB family of R-ketoglutarate-
dependent dioxygenases that remove N-methyl groups from
DNA bases damaged by alkylating agents.
Although TH activity has been known for some time,
challenges with protein purification have limited the amount of
protein available for study. The absence of DNA sequence
information in the yeasts with known TH activity prevented
the use of genome-wide homology searches to identify candidate
gene sequences. We therefore began our studies by attempting
to determine the protein sequence of R. glutinis TH. While this
work was in progress, another group (17) reported a sequence
for TH while searching for a related gene involved in pyrimidine
metabolism in yeast using partial sequence information obtained
from tryptic peptides generated by Thornburg et al. (14). The
reported sequence (17) would not generate a tryptic peptide that
had been identified (14) as part of the catalytic pocket, and the
cloned TH had minimal activity.
We therefore recloned R. glutinis TH from the same strain
used by previous investigators. The DNA sequence obtained
here predicted a protein of similar molecular weight and amino
acid sequence differing in only two nucleotides from the
previously published DNA sequence (17). Both nucleotide
changes would result in amino acid substitutions. Importantly,
our sequence would generate a tryptic peptide from the catalytic
center previously observed (14) from the isolated enzyme.
Upon the basis of the sequence reported here, TH was cloned
and purified. The cloned enzyme showed substantial activity
and allowed us to demonstrate that one protein sequentially
converted thymine to HmU to FoU and CarboxyU. TH was
able to oxidize methyl groups in either the N1- or the
C5-positions of uracil analogues, indicating that the pyrimidine
can enter and bind in the active site in two potential orientations.
Oxidation of the N-methyl group resulted in spontaneous
demethylation, as has been reported for the alkB series of DNA
proteins. In contrast to alkB, which oxidizes methyl groups at
the N3-position of pyrimidines bases, TH oxidizes methyl
groups at the N1-position.
The alkB family of DNA repair proteins has been known for
some time to protect organisms from alkyating agent-induced
toxicity (23). Its mechanism of action and substrate targets were
identified only after results of homology comparisons were
published (25, 26). Avarind and Koonin (24) conducted homol-
ogy searches, which placed alkB in the larger family of nonheme
iron R-ketoglutarate-dependent dioxygenases, revealing its likely
mechanism of action as well as its likely DNA damage targets.
Subsequent studies have revealed additional homologues in other
organisms including human highlighting the potential strength
of genomic approaches, which rely upon knowledge of the DNA
coding sequence. We report here a sequence for R. glutinis TH
that generates an active protein and places TH by sequence
within the larger family of dioxygenases and correctly places it
within the subfamily of TH enzymes from other yeast species.
The dioxygenase family comprises a large group of enzymes
involved in numerous biochemical metabolic pathways, gener-
ally involving oxidative demethylation reactions. The active site
structure of nonheme, R-ketoglutarate-dependent dioxygenases
with known structures is highly conserved. The distances
between the active site iron, the CR atom of the two histidine
and one aspartate residue that bind the iron atom, and the CR
atom of the R-ketoglutarate binding arginine residue vary no
more than 0.5 Å when comparing the crystal structures for a
biochemically disparate group of enzymes including anthocya-
nidin synthase (1GP5), alkB (2FD8, 3BIE, 3BKZ, and 3BI3),
ABH2 (3BU0 and 3BUC), and ABH3 (2IUW). Homology
modeling within this family has been used to generate a
proposed structure of the active site of TH. This preliminary
structure indicated the probable importance of Arg-285 in the
binding of R-ketoglutarate and suggested that the R285C
mutation in the previously reported sequence might explain its
apparent lack of activity (17). This mutant was constructed,
tested, and found to be without activity, as suggested by the
Another important structural question is how TH can selec-
tively act on thymine but not 5MeC. In other enzymes that act
on pyrimidines, including thymidylate synthase and uracil DNA
glycosylase, specificity for thymine over cytosine analogues is
afforded by a specific asparagine residue that hydrogen bonds
to the N3H and O4 carbonyl of thymine analogues. Conversion
of asparagine to aspartate alters these hydrogen-bonding interac-
Characterization of Recombinant Thymine HydroxylaseChem. Res. Toxicol., Vol. 22, No. 5, 2009 891
tions and changes the specificity from thymine to cytosine
(45-47). A series of TH mutants were constructed here in which
asparagine residues near the catalytic site were substituted with
aspartate. Activity toward thymine was diminished for two of
the three mutants, while neither the wild-type enzyme nor any
of the mutants demonstrated any activity in the conversion of
5MeC to 5-hydroxymethylcytosine. The selectivity toward
thymine therefore likely results from interactions with specif-
ically oriented side chain residues or multiple amino acids.
The active sites of TH and the alkB family share considerable
similarity within the active site with respect to the binding of
R-ketoglutarate, iron, and oxygen. Whereas TH acts on a free
base, alkB acts on alkylated thymine residues within the context
of DNA. The capacity of alkB to bind duplex nucleic acids is
found on the amino terminal side of the oxidation center. The
active site in TH that allows selective binding of thymine
analogs is as yet unknown but likely exists on the carboxyl
terminal side of the catalytic center. Further structural studies
will be needed to define the substrate interactions with TH;
however, the apparent large number of members of the
dioxygenase family that can act on diverse substrates suggests
that additional enzymes might be found that could modify DNA
bases and thus have important biological functions.
Supporting Information Available: Table of GC/MS data
for derivatized products of TH reactions; DNA sequence for
the TH gene from R. glutinis; sequence alignment of TH
enzymes from R. glutinis (RGTH), N. crassa (NSTH), and A.
nidulans (ANTH); model of the TH active site; and sequence
alignment of nonheme dioxygenases that oxidize methylated
pyrimidines. This material is available free of charge via the
Internet at http://pubs.acs.org.
(1) Fink, K., Cline, R. E., Henderson, R. B., and Fink, R. M. (1956)
Metabolism of thymine (methyl-C14 or -2-C14) by rat liver in vitro.
J. Biol. Chem. 221, 425–433.
(2) Schandl, E. K. (1973) Thymine 7-hydroxylase activity in normal and
leukemic leukocytes. Biochem. Biophys. Res. Commun. 52, 524–529.
(3) Abbott, M. T., Kadner, R. J., and Fink, R. M. (1964) Conversion of
thymine to 5-hydroxymethyluracil in a cell-free system. J. Biol. Chem.
(4) Liu, C. K., Shaffer, P. M., Slaughter, R. S., McCroskey, R. P., and
Abbott, M. T. (1972) Stoichiometry of the pyrimidine deoxyribo-
nucleoside 2′-hydroxylase reaction and of the conversions of 5-hy-
droxymethyluracil to 5-formyluracil and of the latter to uracil-5-
carboxylic acid. Biochemistry 11, 2172–2176.
(5) Liu, C. K., Hsu, C. A., and Abbott, M. T. (1973) Catalysis of three
sequential dioxygenase reactions by thymine 7-hydroxylase. Arch.
Biochem. Biophys. 159, 180–187.
(6) Abbott, M. T., Schandl, E. K., Lee, R. F., Parker, T. S., and Midgett,
R. J. (1967) Cofactor requirements of thymine 7-hydroxylase. Biochim.
Biophys. Acta 132, 525–528.
(7) McCroskey, R. P., Griswold, W. R., Sokoloff, R. L., Sevier, E. D.,
Lin, S., Liu, K., Shaffer, P. M., Palmatier, R. D., Parker, T. S., and
Abbott, M. T. (1971) Studies pertaining to the purification and
properties of thymine 7-hydroxylase. Biochim. Biophys. Acta 227, 264–
(8) Hausinger, R. P. (2004) FeII/alpha-ketoglutarate-dependent hydroxy-
lases and related enzymes. Crit. ReV. Biochem. Mol. Biol. 39, 21–68.
(9) Shaffer, P. M., and Arst, H. N., Jr. (1984) Regulation of pyrimidine
salvage in Aspergillus nidulans: A role for the major regulatory gene
are A mediating nitrogen metabolite repression. Mol. Gen. Genet. 198,
(10) Shaffer, P. M., Cairns, J. R., Dorman, D. C., and Lott, A. M. (1984)
Separation and characterization of pyrimidine deoxyribonucleoside 2′-
hydroxylase and thymine 7-hydroxylase from Aspergillus nidulans.
Int. J. Biochem. 16, 429–434.
(11) Wondrack, L. M., Hsu, C. A., and Abbott, M. T. (1978) Thymine
7-hydroxylase and pyrimidine deoxyribonucleoside 2′-hydroxylase
activities in Rhodotorula glutinis. J. Biol. Chem. 253, 6511–6515.
(12) Wondrack, L. M., Warn, B. J., Saewert, M. D., and Abbott, M. T.
(1979) Substitution of nucleoside triphosphates for ascorbate in the
thymine 7-hydroxylase reaction of Rhodotorula glutinis. J. Biol. Chem.
(13) Warn-Cramer, B. J., Macrander, L. A., and Abbott, M. T. (1983)
Markedly different ascorbate dependencies of the sequential alpha-
ketoglutarate dioxygenase reactions catalyzed by an essentially
homogeneous thymine 7-hydroxylase from Rhodotorula glutinis.
J. Biol. Chem. 258, 10551–10557.
(14) Thornburg, L. D., Lai, M. T., Wishnok, J. S., and Stubbe, J. (1993) A
non-heme iron protein with heme tendencies: An investigation of the
substrate specificity of thymine hydroxylase. Biochemistry 32, 14023–
(15) Thornburg, L. D., and Stubbe, J. (1993) Mechanism-based inactivation
of thymine hydroxylase, an alpha-ketoglutarate-dependent dioxygenase,
by 5-ethynyluracil. Biochemistry 32, 14034–14042.
(16) Lai, M., Wu, W., and Stubbe, J. (1995) Characterization of a novel,
stable norcaradiene adduct resulting from the inactivation of thymine
hydroxylase by 5-ethynyluracil. J. Am. Chem. Soc. 117, 5023–5030.
(17) Smiley, J. A., Kundracik, M., Landfried, D. A., Barnes, V. R., Sr.,
and and Axhemi, A. A. (2005) Genes of the thymidine salvage
pathway: Thymine-7-hydroxylase from a Rhodotorula glutinis cDNA
library and iso-orotate decarboxylase from Neurospora crassa. Bio-
chim. Biophys. Acta 1723, 256–264.
(18) Braun, E. L., Halpern, A. L., Nelson, M. A., and Natvig, D. O. (2000)
Large-scale comparison of fungal sequence information: mechanisms
of innovation in Neurospora crassa and gene loss in Saccharomyces
cereVisiae. Genome Res. 10, 416–430.
(19) Galagan, J. E., Calvo, S. E., Borkovich, K. A., Selker, E. U., Read,
N. D., Jaffe, D., FitzHugh, W., Ma, L. J., Smirnov, S., Purcell, S.,
Rehman, B., Elkins, T., Engels, R., Wang, S., Nielsen, C. B., Butler,
J., Endrizzi, M., Qui, D., Ianakiev, P., Bell-Pedersen, D., Nelson,
M. A., Werner-Washburne, M., Selitrennikoff, C. P., Kinsey, J. A.,
Braun, E. L., Zelter, A., Schulte, U., Kothe, G. O., Jedd, G., Mewes,
W., Staben, C., Marcotte, E., Greenberg, D., Roy, A., Foley, K.,
Naylor, J., Stange-Thomann, N., Barrett, R., Gnerre, S., Kamal, M.,
Kamvysselis, M., Mauceli, E., Bielke, C., Rudd, S., Frishman, D.,
Krystofova, S., Rasmussen, C., Metzenberg, R. L., Perkins, D. D.,
Kroken, S., Cogoni, C., Macino, G., Catcheside, D., Li, W., Pratt,
R. J., Osmani, S. A., DeSouza, C. P., Glass, L., Orbach, M. J.,
Berglund, J. A., Voelker, R., Yarden, O., Plamann, M., Seiler, S.,
Dunlap, J., Radford, A., Aramayo, R., Natvig, D. O., Alex, L. A.,
Mannhaupt, G., Ebbole, D. J., Freitag, M., Paulsen, I., Sachs, M. S.,
Lander, E. S., Nusbaum, C., and Birren, B. (2003) The genome
sequence of the filamentous fungus Neurospora crassa. Nature
(London) 422, 859–868.
(20) Galagan, J. E., Calvo, S. E., Cuomo, C., Ma, L. J., Wortman, J. R.,
Batzoglou, S., Lee, S. I., Basturkmen, M., Spevak, C. C., Clutterbuck,
J., Kapitonov, V., Jurka, J., Scazzocchio, C., Farman, M., Butler, J.,
Purcell, S., Harris, S., Braus, G. H., Draht, O., Busch, S., D’Enfert,
C., Bouchier, C., Goldman, G. H., Bell-Pedersen, D., Griffiths-Jones,
S., Doonan, J. H., Yu, J., Vienken, K., Pain, A., Freitag, M., Selker,
E. U., Archer, D. B., Penalva, M. A., Oakley, B. R., Momany, M.,
Tanaka, T., Kumagai, T., Asai, K., Machida, M., Nierman, W. C.,
Denning, D. W., Caddick, M., Hynes, M., Paoletti, M., Fischer, R.,
Miller, B., Dyer, P., Sachs, M. S., Osmani, S. A., and Birren, B. W.
(2005) Sequencing of Aspergillus nidulans and comparative analysis
with A. fumigatus and A. oryzae. Nature 438, 1105–1115.
(21) Bankel, L., Lindstedt, G., and Lindstedt, S. (1977) Thymine 7-hy-
droxylase from Neurospora crassa. Substrate specificity studies.
Biochim. Biophys. Acta 481, 431–437.
(22) Sedgwick, B. (2004) Repairing DNA-methylation damage. Nat. ReV.
Mol. Cell Biol. 5, 148–157.
(23) Kondo, H., Nakabeppu, Y., Kataoka, H., Kuhara, S., Kawabata, S.,
and Sekiguchi, M. (1986) Structure and expression of the alkB gene
of Escherichia coli related to the repair of alkylated DNA. J. Biol.
Chem. 261, 15772–15777.
(24) Aravind, L., and Koonin, E. V. (2001) The DNA-repair protein AlkB,
EGL-9, and leprecan define new families of 2-oxoglutarate- and iron-
dependent dioxygenases. Genome Biol. 2, research0007.10007.8.
(25) Trewick, S. C., Henshaw, T. F., Hausinger, R. P., Lindahl, T., and
Sedgwick, B. (2002) Oxidative demethylation by Escherichia coli
AlkB directly reverts DNA base damage. Nature (London) 419, 174–
(26) Falnes, P. O., Johansen, R. F., and Seeberg, E. (2002) AlkB-mediated
oxidative demethylation reverses DNA damage in Escherichia coli.
Nature (London) 419, 178–182.
(27) Bhattacharya, S. K., Ramchandani, S., Cervoni, N., and Szyf, M. (1999)
A mammalian protein with specific demethylase activity for mCpG
DNA. Nature (London) 397, 579–583.
(28) Cedar, H., and Verdine, G. L. (1999) Gene expression. The amazing
demethylase. Nature (London) 397, 568–569.
Chem. Res. Toxicol., Vol. 22, No. 5, 2009Neidigh et al.
(29) Wolffe, A. P., Jones, P. L., and Wade, P. A. (1999) DNA demethy-
lation. Proc. Natl. Acad. Sci. U.S.A. 96, 5894–5896.
(30) Mayer, W., Niveleau, A., Walter, J., Fundele, R., and Haaf, T. (2000)
Demethylation of the zygotic paternal genome. Nature (London) 403,
(31) Koivisto, P., Robins, P., Lindahl, T., and Sedgwick, B. (2004)
Demethylation of 3-methylthymine in DNA by bacterial and human
DNA dioxygenases. J. Biol. Chem. 279, 40470–40474.
(32) Yu, Z., Genest, P. A., ter Riet, B., Sweeney, K., DiPaolo, C., Kieft,
R., Christodoulou, E., Perrakis, A., Simmons, J. M., Hausinger, R. P.,
van Luenen, H. G., Rigden, D. J., Sabatini, R., and Borst, P. (2007)
The protein that binds to DNA base J in trypanosomatids has features
of a thymidine hydroxylase. Nucleic Acids Res. 35, 2107–2115.
(33) LaFrancois, C. J., Fujimoto, J., and Sowers, L. C. (1998) Synthesis
and characterization of isotopically enriched pyrimidine deoxynucleo-
side oxidation damage products. Chem. Res. Toxicol. 11, 75–83.
(34) LaFrancois, C. J., Yu, K., and Sowers, L. C. (1998) Quantification of
5-(hydroxymethyl)uracil in DNA by gas chromatography/mass spec-
trometry: Problems and solutions. Chem. Res. Toxicol. 11, 786–793.
(35) Eng, J. K., McCormack, A. L., and Yates, J. R. (1994) An approach
to correlate tandem mass-spectral data of peptides with amino-acid-
sequences in a protein database. J. Am. Soc. Mass Spectrom. 5, 976–
(36) Matsudaira, P. (1990) Limited N-terminal sequence analysis. Methods
Enzymol. 182, 602–613.
(37) Sauder, J. M., Arthur, J. W., and Dunbrack, R. L., Jr. (2000) Large-
scale comparison of protein sequence alignment algorithms with
structure alignments. Proteins 40, 6–22.
(38) Ward, J. J., McGuffin, L. J., Buxton, B. F., and Jones, D. T. (2003)
Secondary structure prediction with support vector machines. Bioin-
formatics 19, 1650–1655.
(39) Privat, E., and Sowers, L. C. (1996) Photochemical deamination and
demethylation of 5-methylcytosine. Chem. Res. Toxicol. 9, 745–750.
(40) Burdzy, A., Noyes, K. T., Valinluck, V., and Sowers, L. C. (2002)
Synthesis of stable-isotope enriched 5-methylpyrimidines and their use
as probes of base reactivity in DNA. Nucleic Acids Res. 30, 4068–
(41) Kinzie, S. D., Thevis, M., Ngo, K., Whitelegge, J., Loo, J. A., and
Abu-Omar, M. M. (2003) Posttranslational hydroxylation of human
phenylalanine hydroxylase is a novel example of enzyme self-repair
within the second coordination sphere of catalytic iron. J. Am. Chem.
Soc. 125, 4710–4711.
(42) Muller, I., Stuckl, C., Wakeley, J., Kertesz, M., and Uson, I. (2005)
Succinate complex crystal structures of the alpha-ketoglutarate-
dependent dioxygenase AtsK: Steric aspects of enzyme self-hydroxy-
lation. J. Biol. Chem. 280, 5716–5723.
(43) Ryle, M. J., Koehntop, K. D., Liu, A., Que, L., Jr., and and Hausinger,
R. P. (2003) Interconversion of two oxidized forms of taurine/alpha-
ketoglutarate dioxygenase, a non-heme iron hydroxylase: Evidence
for bicarbonate binding. Proc. Natl. Acad. Sci. U.S.A. 100, 3790–3795.
(44) Wilmouth, R. C., Turnbull, J. J., Welford, R. W., Clifton, I. J., Prescott,
A. G., and Schofield, C. J. (2002) Structure and mechanism of
anthocyanidin synthase from Arabidopsis thaliana. Structure 10, 93–
(45) Kavli, B., Slupphaug, G., Mol, C. D., Arvai, A. S., Peterson, S. B.,
Tainer, J. A., and Krokan, H. E. (1996) Excision of cytosine and
thymine from DNA by mutants of human uracil-DNA glycosylase.
EMBO J. 15, 3442–3447.
(46) Agarwalla, S., LaPorte, S., Liu, L., Finer-Moore, J., Stroud, R. M.,
and Santi, D. V. (1997) A novel dCMP methylase by engineering
thymidylate synthase. Biochemistry 36, 15909–15917.
(47) Handa, P., Acharya, N., and Varshney, U. (2002) Effects of mutations
at tyrosine 66 and asparagine 123 in the active site pocket of
Escherichia coli uracil DNA glycosylase on uracil excision from
synthetic DNA oligomers: Evidence for the occurrence of long-range
interactions between the enzyme and substrate. Nucleic Acids Res.
(48) Lindahl, T. (1993) Instability and decay of the primary structure of
DNA. Nature (London) 362, 709–715.
(49) Mullaart, E., Lohman, P. H., Berends, F., and Vijg, J. (1990) DNA
damage metabolism and aging. Mutat. Res. 237, 189–210.
(50) Zuo, S., Boorstein, R. J., and Teebor, G. W. (1995) Oxidative damage
to 5-methylcytosine in DNA. Nucleic Acids Res. 23, 3239–3243.
(51) Burdzy, A., Noyes, K. T., Valinluck, V., and Sowers, L. C. (2002)
Synthesis of stable-isotope enriched 5-methylpyrimidines and their use
as probes of base reactivity in DNA. Nucleic Acids Res. 30, 4068–
(52) Haushalter, K. A., Todd Stukenberg, M. W., Kirschner, M. W., and
Verdine, G. L. (1999) Identification of a new uracil-DNA glycosylase
family by expression cloning using synthetic inhibitors. Curr. Biol.
(53) Baker, D., Liu, P., Burdzy, A., and Sowers, L. C. (2002) Characteriza-
tion of the substrate specificity of a human 5-hydroxymethyluracil
glycosylase activity. Chem. Res. Toxicol. 15, 33–39.
(54) Cannon-Carlson, S. V., Gokhale, H., and Teebor, G. W. (1989)
Purification and characterization of 5-hydroxymethyluracil-DNA gly-
cosylase from calf thymus. Its possible role in the maintenance of
methylated cytosine residues. J. Biol. Chem. 264, 13306–13312.
(55) Boorstein, R. J., Chiu, L. N., and Teebor, G. W. (1989) Phylogenetic
evidence of a role for 5-hydroxymethyluracil-DNA glycosylase in the
maintenance of 5-methylcytosine in DNA. Nucleic Acids Res. 17,
(56) Rogstad, D. K., Liu, P., Burdzy, A., Lin, S. S., and Sowers, L. C.
(2002) Endogenous DNA lesions can inhibit the binding of the AP-1
(c-Jun) transcription factor. Biochemistry 41, 8093–8102.
(57) Valinluck, V., Tsai, H. H., Rogstad, D. K., Burdzy, A., Bird, A., and
Sowers, L. C. (2004) Oxidative damage to methyl-CpG sequences
inhibits the binding of the methyl-CpG binding domain (MBD) of
methyl-CpG binding protein 2 (MeCP2). Nucleic Acids Res. 32,
Characterization of Recombinant Thymine HydroxylaseChem. Res. Toxicol., Vol. 22, No. 5, 2009 893